Garry Kasparov On Modern Chess Part Three - Kasparov Vs Karpov 1986-1987 (Garry Kasparov)
Garry Kasparov On Modern Chess Part Three - Kasparov Vs Karpov 1986-1987 (Garry Kasparov)
Garry Kasparov On Modern Chess Part Three - Kasparov Vs Karpov 1986-1987 (Garry Kasparov)
PART THREE
VS
1986-1987
5
WUII
{
j\.
m
\P
't:'
I'lRr,~\\O~
PI\E~~
55
I
• _ .. - .
*.
....... , _ _ _ u . ., _
EVERYMAN CHESS
GARRY KASPAROV 0 ODERN CHESS
PAROV
vs
RPOV
1 -1
The right of Garry Kasparov to be identified as the author of this work has been asserted in
accordance with the Copyrights, Designs and Patents Act 1988.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system or transmitted in any form or by any means, electronic, electrostatic, magnetic tape,
photocopying, recording or otherwise, without prior permission of the publisher.
ISBN: 9781857446258
Distributed in North America by The Globe Pequot Press, p.o Box 480,
246 Goose Lane, Guilford, CT 06437-0480.
All other sales enquiries should be directed to Everyman Chess, Northburgh House,
10 Northburgh Street, London EC1V OAT
tel: 020 7253 7887 fax: 020 7490 3708
email: info@everymanchess.com
website: www.everymanchess.com
Everyman is the registered trade mark of Random House Inc. and is used in this work under
license from Random House Inc.
Foreword 5
This volume - the second in the trilogy of my games against Anatoly Karpov - is mainly
devoted to our third and fourth matches (1986 and 1987). In both of them there was an
incredibly dramatic finish, and for a long time afterwards the arguments about these
events excited the chess world.
After winning the world crown in 1985 I had very little time in which to enjoy the taste
of victory: the return match started the very next summer.
By the FIDE rules approved in 1949, the champion was obliged to defend his title every
three years in a match with the winner of the qualifying series. But in my case this process
was violated. Back in 1977, to satisfy Karpov, FIDE reinstated the regulation concerning a
return match, which had operated in only two cycles and had been abolished in the early
1960s. If the champion lost, he would automatically, without any qualifying process, have
the right to playa return match the following year. This right was used with great benefit
to himself by Botvinnik, who won return matches against Smyslov (1958) and Tal (1961).
He lost to them in matches for the world championship, but he twice regained the crown,
restricting the reign of his' offenders' to just one year.
In order to avoid such a fate, I also had to defeat Karpov in a return match (Lon-
don/Leningrad 1986). Moreover, we had already played 72 games - the longest match in
history (1984/85) and the second match (1985), in which I had wrested the title from Kar-
pov ... In a fierce struggle I managed to win the return match.
But the trials did not end there! Despite all our additional matches, the FIDE three-year
qualifying cycle remained unchanged, and a year later I had to playa match against the
next challenger. Not surprisingly, my opponent was again Karpov. For him the path to the
summit was far shorter than usual. FIDE relieved him of the need to play in the series of
Candidates matches and allowed him directly into a 'super-final', where the ex-champion
crushed Andrey Sokolov, the winner of the qualifying series.
On the eve of our fourth match (Seville 1987) Karpov stated that he had no doubts he
would be successful. I also said that I did not see any reason why I should lose my title. I
was firmly convinced that, when I needed to, I could win against Karpov. But Botvinnik
5
Kasparov vs. Karpov: 1986-1987
warned me: 'You are playing more strongly, but there is one psychological danger which
you must overcome: you must realise that it is possible to lose the match! Such a possibility
exists, and if you remember about it, this will mobilise your strength.' But during the
encounter I periodically ceased to realise that I was playing a match for the world champi-
onship: the psychological fatigue, accumulated over the three years, made itself felt.
It was the encounter in Seville that became the real return match for me. The first three
matches had merged in my consciousness into one gigantic battle, and only after my
victory in 1986 did I fully feel that I was the champion, having demonstrated that my
success in 1985 was no accident. This was also recognised by the Soviet authorities, who
awarded me the Order of the Red Banner of Labour (by the irony of fate, the Order was
presented to me by Politburo Candidate Member Pyotr Demichev - the chairman of the
organising committee of the first two matches). I began travelling about and appearing in
public, and there were also unaccustomed duties ... But the next year in Seville I had to
solve those psychological problems which in their time had proved insuperable for Smys-
lov and Tal. Again a match with the same tenacious opponent, and losing was something I
could not contemplate: now it would be a long time before I would have a chance to re-
coup my losses! This severe trial nearly ended in disaster for me. But, by winning the last,
24th game, I levelled the score and retained my title.
Whereas this is the first time I have commented on the fourth match (apart from a few
games which I annotated for Informator), the second and third matches have already been
covered in detail in myoId book Dva matcha (1987).1 By the standards of the time this was
quite a high-quality work, but by no means all the comments have stood the test of time -
and of the computer! Many chess evaluations have undergone changes, and this has forced
me to make substantial corrections to a seemingly harmonious 'exemplary text'. The
elimination of analytical myths, passed on from one generation to the next, is one of the
main aims of my entire project.
Four matches for the world championship in the space of three years is an unprece-
dented occurrence in history! Regarding this, Botvinnik wrote: 'The most important part of
chess culture is created by great players, and their talent should be preserved; they should
not be forced to work until they are worn out, until they are traumatised - this is the
sacred duty of FIDE. At present the FIDE leaders are acting in the opposite direction. This
has provoked a natural reaction on the part of those who create chess culture, on the part
of grandmasters - they have united, to repulse tyranny.' The story of the creation of the
legendary Grandmasters Association (GMA) is another topic of this book.
For their help in the preparation of this manuscript for publication I should like to thank my
former trainers Alexander Nikitin and Alexander Shakarov, as well as Sergey Voronkov, editor of
the books 'Dva matcha' (1987) and 'Bezlimitny poedinok' (1989), and Viktor Chepizhny, for many
years editor-in-chief of chess literature published by 'Fizkultura i sport'.
1 An English translation was published in two separate volumes: New World Chess Champion (Per-
gamon 1986) and London-Leningrad Championship Games (Pergamon 1987) - Translator's note.
6
Chapter One I
Difficult negotiations
The victory over Karpov on 9 November 1985 was one of the main landmarks of my life.
World champion! Initially the weight of this title had a paralysing effect, which kept me in
a state of euphoria. The waves of joy which engulfed me made a level-headed assessment
of the situation impossible. I should remind you that I was just 22 years old.
A week after the end of the match I gave a talk at the progressive Sovremennik Theatre
in Moscow. I spoke for several hours. It was late, but people did not leave until deep into
the night. The knowledgeable and emotional audience was enthralled by my simple tale of
adventures, misadventures, and the traditional happy ending. It seemed that my victory
over Karpov had buried all the grief and conflicts of the past and that I could now begin
writing my biography anew. But I should have known that the past would not disappear
without a trace.
On becoming world champion, I deliberately adopted a conciliatory attitude towards
the Chess Federation, since I was hoping that our domestic chess situation would improve
and I was afraid of harming this by any incautious statements. 'There is a big difference
between Garry Kasparov the challenger and Garry Kasparov the world champion!' I
declared at a press conference after the match in Moscow. The western journalists were
quick to notice this change. Possibly they had been hoping for some sensational revelations
with which to titillate their readers. If so, they were disappointed.
'Chess hero Kasparov buries the Hatchet', announced one headline. 'No more scandals, says
Kasparov', said another. 'Kasparov pointedly refuses to be drawn into fresh attacks against those
he once accused of trying to rob him of victory', wrote The Times.
And it was true. Now that I'd won the world title, I was determined not to abuse my
position in the way that Karpov had done for many years, which was what, strictly speak-
ing, I had accused him of. The fact that to a significant extent his power had been directed
personally against me made me indignant, of course, but I was also sickened by the
thought that in chess a future dictatorship was possible! The chess world was in serious
need of rebuilding on a democratic basis, and this meant that the champion himself should
not exceed his powers.
Not surprisingly, my honeymoon period with the chess authorities did not last long.
7
Kasparov vs. Karpov: 1986-1987
Grounds for a conclusive split were provided by the question of a return match.
Return match ... The very words sounded strange - twenty-five years had passed since
the time of the last such match. What, in general, was the point of granting a world cham-
pion the right to a return match? It was said that so there should not be an accidental
champion. But could a champion be called accidental, after he had won the title after a
marathon contest of 72 games?! In the given situation a return match was against common
sense, elementary logic, and also the chess regulations. FIDE officials put themselves above
the law, arbitrarily interpreting the regulations and changing them in the course of the
world championship contest.
The Congress in Graz, which concluded three days before the start of our second match
(1985), approved Campomanes's decision to terminate the first match (1984/85) and ap-
proved new rules - or, strictly speaking, additions to the rules - for the world champion-
ship. The idea of an unlimited match, which had been roundly condemned back in 1927
after the Alekhine-Capablanca match and re-emerged in 1975 at the insistence of Fischer
(supported in 1977 by Karpov for his battle with Korchnoi), once again gave way to the
traditional contest of 24 games.
But 'for some reason' the abolition of the unlimited match did not lead to the abolition
of the return match. This baby fathered by the unlimited contest was successfully left on
the doorstep of the match with the new regulations. Karpov was granted the exclusive
right to a return match. Exclusive, because it was not extended to future champions. This
meant that, while retaining the champion's advantage of an unlimited match, Karpov
acquired another one, which was traditional for limited matches: in the event of a draw, he
would remain champion. But even this wasn't enough! In the event of a double failure,
Karpov reserved another privilege for himself: to begin the fight for the world title not
from the start of the Candidates series, as Smyslov, Tal, Petrosian and Spassky had done,
but directly from a super-final- just one match with the winner of the qualifying series!
Averbakh: 'And yet the rules for the Candidates matches stated something quite different: an
ex-champion should join the Candidates event at the quarter-final stage. It is probable that, feeling
guilty towards Karpov for changing the conditions for the termination of the first match, in Graz
Campomanes as though "repaid the debt", but the honourable congressmen could not have failed to
realise that their decision would grant Karpov excessive privileges. It was unfair both to Kasparov,
and to the Candidates - Vaganian, Sokolov, Timman and Yusupov.'
This triple defensive rampart erected around Karpov was incomparably greater than
the notorious dollar rampart that champions sometimes sheltered behind in the old pre-
FIDE times!
All these privileges were enshrined in law at the Graz Congress in the form of 'Special
rules for the 1985 world championship match'. Also stated with cynical frankness was the
'special price' of these rules: 'Point 6. The match organisers will pay to FIDE a sum equivalent to
24 per cent of the prize fund, i.e. a sum corresponding to 24 drawn games'.
Thus the Soviet Union, as organiser of the 1985 match, paid FIDE 24 per cent of the
prize fund instead of 16 per cent, which corresponded to the actual number of draws in the
match and was laid down in the basic rules. The difference came to 128,000 Swiss francs!
Such was the fee that FIDE obtained from the USSR Sports Committee for its new-fangled
special rules.
It was not long before grounds for justifying the return match were found. As if noth-
8
Thirsting for Revenge
ing had happened, Karpov suddenly awarded himself victory in the terminated match: 'I
ICon then with a score of 5-3, and Kasparov has done the same now. The score is equal: 8-8J' This
'argument' began to be actively used, but it didn't stand up to criticism. Karpov had won
four of the first nine games out of 72. So by the end of our battle the question of who was
the stronger had been settled pretty conclusively!
In one of his interviews Karpov said that I was a meteor in chess, like Tal. Why in par-
ticular did he compare me with Tal? Was it because he had been champion for only one
year? In this subtle way the idea was suggested to the public that Karpov had lost the title
only temporarily. In his statements he kept harking back to the decisive 24th game of our
1985 match. He simply couldn't get used to the idea that he'd lost it, and he claimed that a
win for him in that game (for which read: in the match!) would have been more logical -
and 'logically' he still considered himself the champion.
On 5 December 1985, less than a month after losing the title, Karpov claimed his right to
a return match, which under the new rules had to be implemented between 10 February
and 21 April 1986. Thus FIDE was giving Karpov the right to a return match within just
three months - an unprecedented occurrence in the history of chess! This was the third
arbitrary change to the rules over the past few years, and all of them to Karpov's advan-
tage. In all, he was allowed 96 games to defend his title, whereas I - the new champion-
was given no privileges at all! Moreover, if I were to lose the return match, I faced the
dubious honour of also becoming the youngest ex-champion.
A new storm was brewing, which I should have anticipated. But in the euphoria pro-
voked by my November triumph, I had failed to see that defeating Karpov at the chess
board was a necessary but insufficient condition for resolving all the problems. Above all, I
had underestimated Campomanes. In a sense my victory had done him a good service:
now he could claim that the high standard of the games in the second match vindicated his
decision to terminate the first one. In addition, by forcing me to keep playing chess, he
thereby prevented me from playing politics against him. Even after going through hell and
high water, the Filipino had still landed on his feet.
I wanted no less than Karpov to be an active champion, but I hadn't signed up to play
exclusively against Karpov for the rest of my life! It should not be forgotten that our end-
less encounters had disrupted the schedule of international tournaments, to the detriment
of many players. But FIDE seemed to have forgotten that its primary concern was sup-
posed to be the welfare of chess and chess players. It appeared that at that moment the
FIDE officials were only concerned with two questions: returning the champion's title to
Karpov and the state of their finances, since world championship matches were their main
source of revenue.
But at the same time, my matches with Karpov had greatly stimulated interest in chess
around the world, and I felt that I didn't have the right to deceive the hopes of so many
chess lovers.
After losing the match, Karpov immediately went off to the first World Team Champi-
onship (Luzern 15-28 November 1985), where he scored 5 out of 7 (+3=4) on the top board
and the USSR team gained a convincing victory. For my part, at the end of the year I
played a match of six games with Timman, who was then considered the No.3 player in
the world (Hilversum, 15-22 December 1985). For a champion who had only just won the
title a month earlier, this was an unusual thing to do: I was risking my reputation and
9
Kasparov vs. Karpov: 1986-1987
simply could not afford to lose. The prize fund was quite modest, but money wasn't the
aim of my trip to Holland: I wanted the Western chess public to have the opportunity of
seeing for themselves the new world champion. And the match was a success! A spectacu-
lar win in the final game also brought me overall victory: 4-2 (+3-1=2).
After the match I gave a press conference in which I spoke out openly against the stag-
ing of the return match. 'This match should not take place, because no one should have so
many privileges,' I said. 'The world champion keeps his title in the event of a drawn
match, and that is enough. If I lose, I will give up my right to a return match. According to
the decision adopted by the FIDE Congress, I have become world champion for one year -
1986. I was informed about the right to a return match granted to Karpov just a few days
before the start of the second match in Moscow. I had no time to make a protest.'
I also took the opportunity to consider a possible replacement for Campomanes and
found one in the person of the 38-year-old Brazilian, Lincoln Lucena, who had agreed to
stand in 1986 for the post of FIDE President in tandem with Raymond Keene. The English
grandmaster had previously been one of Campo's allies, but now he was pinning his hopes
on me, expecting that the Soviet Federation, which played an almost decisive role in the
FIDE Presidential elections, would speak out against the Filipino. Western Europe - in other
words, the active chess world! - rejected Campomanes, but he held on thanks to the USSR,
its satellites and the developing countries. And Lucena performed the role of representative
of the 'third world', challenging the dictatorship. A part of this pre-election campaign was
the staging of the return match in London, although in late 1985 Keene had said that we
might be able to convene an extraordinary FIDE congress and abolish the return match ...
'In the past few years', I said at a press conference,' there have been a lot of problems in
the chess world. During my stay here I have met with people who are prepared to work
for more democracy in chess life. They want chess players to know what is going on, they
want the rules to be honest and the opinions of players to be respected. I also want this. As
world champion I consider it my duty to support people working for these aims.' Unfor-
tunately, this press conference proved to be a mistake, which later was to cost me dearly. It
was going to need more than goodwill and good intentions to beat Campomanes.
On returning home I received a message of support from the newly-formed European
Chess Union, which had written to the USSR Chess Federation suggesting that the return
match should be cancelled:
The Council of the European Chess Union held a meeting in Hilversum, Holland, on 22 Decem-
ber 1985 to discuss, among other things, the question of the planned return match between Karpov
and Kasparov.
We wish to make the following points against the idea of a return match:
1. It could have a ruinous effect on the players' health (as Campomanes noted when he stopped
the 1984185 match).
2. Such a large number of games between the same players will lower the importance of chess in
the eyes of the public.
3. Neither of the players can bring any benefit to chess by participating in other competitions,
because they are spending all their time playing against each other.
4. No world champion has ever held the title for less than a year before being required to defend
it, whereas in this instance the period is just three months.
10
Thirsting for Revenge
5. The idea of a return match was not put to the Council of Grandmasters for discussion.
6. The FIDE Congress in Graz only examined the terms of the return match, and did not ap-
prove them by a majority of two-thirds of the votes (as required by the FIDE statutes, clause 4.14).
7. Kasparov was only notified of the return match two days before the start of the 1985 match.
He thus had no opportunity to comment on, or protest against, the decision. Moreover, the FIDE
executive council did not inform the congress in Graz of Kasparov's previous protests against a
return match.
In view of the above, the Council of the European Chess Union is of the opinion that the return
match should not take place, and respectfully asks the USSR Chess Federation to do everything in
its power to cancel it.
R. LITTORIN, President
This request was supported by Timman, Larsen and others. Their help could not have
corne at a better time. 'I am very happy to hear that the European Chess Union has been
formed to look after the interests of the strongest chess-playing nations,' I said in my reply.
'The New Year is normally a time of peace, but sometimes you have to fight to gain that
peace. I hope that in the next year we will restore genuine peace in the chess world, and
that chess players will then be able to concentrate their efforts on creating beautiful games
under normal conditions.'
It has to be said that this statement of support had been extracted with some difficulty
from the European grandmasters. Firstly, they were apprehensive of the despotic power of
FIDE. One of them said to me: 'If that's the way they treat the world champion, what will
they do to me?' Secondly, at that time many saw my fight against the system of dictator-
ship and tyranny as a fight for my own interests, and considered the problems of the
return match to be an internal affair between me and Karpov. So, despite the support I
received from a number of players - Ljubojevic, Seirawan, Najdorf and Short (who said
'the world championship cycle stinks'), the grandmasters did not unite firmly enough
against the staging of the return match.
In a letter of 4 January 1986 I called on the USSR Chess Federation to state its position
regarding the return match. In particular, I said:
'The tense situation which has arisen over the return match is the result of the arbitrary inter-
pretation and frequent alteration of the match rules by the FIDE President. Campomanes's high-
handed actions are the subject of strong and well-justified criticism in the chess world. Thus the
European Chess Union has written to the USSR Chess Federation calling on it to use all its influ-
ence to cancel the forthcoming return match. I still do not know our Federation's answer to that
appeal. I therefore consider that we will have the unprecedented situation whereby the rules are
being interpreted in two different ways: one way for Karpov, and another way for everyone else. The
only rules that can be considered binding are the standing rules which give no one any special
privileges. No one person has the sole right to lay down the rules - neither the FIDE President, nor
the world champion. There is now an obvious need to put matters in order and create proper rules
which are fair and permanently in force, and which will be adopted by FIDE as soon as possible.
'In view of the above, I consider Karpov's personal entitlement to a return match to be illegal,
and I refuse to take part in such a contest. The right to a return match in a limited competition must
be accorded to all world champions, or to none.'
11
Kasparov vs. Karpov: 1986-1987
However, the position of the USSR Chess Federation had been decided long before. As
had the position of Karpov, who declared in an interview in 1986: 'Irrespective of the outcome
of the present contest, I consider the return match to be an historic right of world champions, and I
see no need to give up this right.'
In the end I had to agree to play. This was, of course, a concession on my part, but the
threat to strip me of the world title by administrative measures was by no means an empty
one. By then I was well aware that it is easier to demonstrate your right at the chess board,
than in casuistic arguments with chess officials. Just a short while before, they had been
saying that the unlimited match was stopped because of their concern for the health of the
world's two leading players. Now they were pushing just as vigorously for a return match
to begin as soon as possible.
The FIDE headquarters in Luzern began issuing menacing statements that I would be
stripped of my title if I refused to play on the stipulated date. Campomanes said that the
rules required me to give a formal undertaking by 7 January, or else I would be disqualified.
As the deadline approached, tension in the chess world mounted. But when the day
came, nothing happened. In a terse message from Luzern, Campomanes was forced to
admit that his ultimatum had been based on a misreading of the rules. The contestants did
indeed have two weeks in which to confirm their readiness to play - but only after the
match venue had been announced, and in fact it had not yet been decided.
Meanwhile, bids had already been received from Leningrad and London. The former
had offered one million Swiss francs, and the latter 800,000 more. However, Leningrad's
million was partly symbolic, since it was of value only to FIDE, which would receive its
share of the prize fund in hard currency. The players themselves, in accordance with
Soviet rules, could expect only a small part of the prize fund, and, moreover, in roubles -
72,000 between the two of them.
The English were very keen to host the championship, because it would coincide with
the centenary of the first world championship match. They had failed in a bid for the 1985
match, after Marseilles had offered more money - 1,600,000 Swiss francs (but in the end
the match went to Moscow, which offered the same amount).
The fate of the return match was of particular concern to Keene, who was afraid that
London would again be denied the opportunity to stage a world championship match, on
the organisation of which he and his colleagues had spent so much time and energy. The
days leading up to 13 January, when Campomanes was due to announce the match venue,
were very tense for all of us. It became known that Campomanes had flown to Vienna,
where Karpov was playing (and where he shared 2nd-8th places in an unusually strong
Swiss tournament, scoring 6 out of 9). The FIDE President's actions - first to consult with
the challenger, and only then with the champion - looked strange (although natural for
FIDE President Campomanes). After a series of fruitless discussions, Campomanes an-
nounced that nothing had been decided.
The situation had become intolerable, and on 18 January I publicly announced that I
was refusing to play the match. Apart from the issues of principle involved, all these off-
the-board intrigues had made it impossible to make any serious chess preparations. Of
course, I realised that Karpov wouldn't be ready to play in February either, so why was he
pressing for an early start to the match? Simply he knew that in February I would be
without the support of two of my trainers, Dorfman and Timoshchenko, who would be
12
Thirsting for Revenge
unable to go abroad at that time: a month before the proposed start of the return match,
they had both become 'non-travellers'. What's more, the Army Sports Committee had not
even given Timoshchenko and Vladimirov permission to join me for a training session. In
fact, from 1983 onwards, whenever they wanted to travel to training sessions, they had to
obtain permission from the highest authorities - right up to the USSR Ministry of Defence!
On 19 January, with the situation at boiling-point, the following telex from the British
Chess Federation was received in Moscow:
'It is our view that the reinstatement of a return match would be inadvisable in a two-year world
championship cycle. Such a reintroduction would lead to congestion of international tournament
schedules. We stress, however, that it is ideally a matter which the players concerned should decide
themselves. We take this opportunity to repeat our offer that if the return world championship
match of the current cycle goes ahead, then London, the British Chess Federation and the Greater
London Council are very willing to stage the first twelve games here with the second twelve in
Leningrad. It is in our view essential, though, that any problems concerning the staging of the
return match should be amicably settled to the mutual satisfaction of both world champion Kasparov
and challenger Karpov. It is of extreme importance that no public problems damage the interna-
tional chess movement and its reputation worldwide. In particular, we would wish to avoid the
unseemly default situations which occurred in 1975 and 1983.'
This message was intended to clear the air, offer an alternative solution, and involve the
players themselves in the discussion. I had always declared my readiness to negotiate and
listen to other points of view, but in this case I considered that a return match was unjust.
At the same time it was now part of the championship rules and hence a fact I had to learn
to live with. I knew I would have to defend my title sooner or later, but it was vital now to
have as long a respite as possible.
The search for a compromise was helped by a political factor. The 27th Congress of the
USSR Communist Party was due to begin in February, so the last thing the sports authori-
ties wanted was a public scandal, and disqualifying me was certain to cause a scandal that
would attract worldwide publicity. In the circumstances Karpov couldn't afford to be seen
as the one who rocked the boat. It ceased to be a matter of whether the match would take
place, but when.
Of course, Karpov's patrons were very pleased when I announced that I was refusing to
playa return match: once again, as in the summer of 1985, they had a pretext for trying to
have me disqualified.
At an urgent session of the USSR Chess Federation on 21 January, Sevastyanov had
said as much: either I agreed to play in February, or March at the latest, or Karpov would
be declared the champion. I kept insisting: why such haste? Why should I have to play two
matches in one year, something that had never happened before? It was explained to me
that the FIDE schedule of regional and international tournaments would be disrupted. It
was the old ping-pong principle at work again: our federation says we must do something
because FIDE demands it, then FIDE hits the ball back, saying our federation requested
this, and we have to respect its wishes because it is the strongest federation in the world,
and so on ad infinitum.
They were carrying on as if nothing in the country had changed. But it should be re-
membered that this was now 1986, not 1985 and especially not 1983. The time of cynical
reprisals was receding into the past. And besides, I was not the only one affected by the
13
Kasparov vs. Karpov: 1986-1987
situation: because my battles with Karpov had dragged on for so long, two world champi-
onship cycles were overlapping. The other contestants had to be considered - Sokolov,
Vaganian, Timman and Yusupov. Therefore instead of a return match with Karpov I
suggested holding a three-player match-tournament (an old idea of Botvinnik: this FIDE
regulation was in force for the matches in 1951 and 1954) involving me, Karpov and the
winner of the Candidates matches, thereby concluding the two cycles at one go and restor-
ing the traditional system for contesting the world championship.
But Karpov flatly rejected this idea, declaring: 'What right do you have to suggest this? The
world championship title - is it your own personal property? It's the property of FIDE. You can
only take up your position - that's all. And in general I don't see that there is anything to be dis-
cussed, because here is a document stating that in the event of Kasparov defeating the world cham-
pion he is obliged to playa return match. Here is his signature. What is there to talk about?'
But before the 1985 match I had been practically forced to accept this unjust decision
taken at the congress in Graz: at that moment my protest would have threatened the start of
the match.
Averbakh: 'At a meeting of the federation Sevastyanov tried to put pressure on the world cham-
pion, but there was a clear division of the presidium members into two groups: one was for Kasparov
and the other for Karpov. In this difficult situation an unexpected solution was found by the neutral
Vera Tikhomirova: she suggested putting Kasparov and Karpov in a separate room, so that they
should come to a joint decision. A break was announced. The two rivals went off and a short time
later they returned with an agreed draft document ... '
As a result of these difficult discussions, on the following day, 22 January, the final text
of a joint agreement was hammered out, presenting Campomanes with a fait accompli. For
me the key point was that the match would not take place earlier than the end of the
summer! Here is the text of the agreement:
World champion Garry Kasparov and ex-world champion Anatoly Karpov, having examined the
complicated situation that was created during the world championship competition and wanting to
eliminate differences of opinion and avoid a situation where, in the absence of an agreement, they
would have to accept any extraordinary measures decided by FIDE, have agreed the following:
1. The return match between them as foreseen by 1985 FIDE rules will take place.
2. It is to begin in July or August 1986. This postponement from the date set up earlier is neces-
sary for both chess players, in order to rest and restore their strength after having played 72 games
against each other in 14 months.
3. Hoping that their wish will meet with the understanding of the leadership of FIDE and the
whole chess world, they think a super-final (the match between the winner of the Candidates and the
ex-world champion) will take place in February 1987 and the next world championship match in
July or August 1987.
4. Kasparov and Karpov have undertaken the following obligations towards each other:
a) neither of them will playa match with the winner of the Candidates cycle until the return
match between them is over;
b) the winner of the return match guarantees the loser that he will not playa world champion-
ship match with the winner of the Candidates cycle until that player has a played a match against
the ex-world champion;
c) under all circumstances our position as stated in a) and b) will stay permanent.
14
Thirsting for Revenge
5. After examining the bids for the organisation of the return match from the cities of Leningrad
and London received by the President of FIDE dated 16 December 1985, Kasparov and Karpov
express their wish to play the match in Leningrad.
If however, the organisers of either of them withdraw their bids in connection with the changed
dates of the match, the participants are ready to consider other bids which are made in accordance
with FIDE regulations by 1 April 1986, so that the venue, dates of return match and the team of
arbiters can be announced a month later.
The document reflected the shaky balance of power that now existed: I had been unable
to avoid the return match that was being foisted on me, but Karpov's absolute rule had
also come to an end.
The agreement was signed by Karpov, Sevastyanov and myself, and it had the backing of
the USSR Chess Federation. In February Karpov and I travelled to the FIDE headquarters in
Luzern to secure the acceptance of our agreement and to decide on the match venue.
At the time I sincerely believed in the importance of our mission, thinking that, having
formed a united front with Karpov, I could strike a strong blow against Campomanes.
Indeed, Campomanes looked very gloomy and he kept complaining about the intolerable
high-handedness of the Soviet Chess Federation, which he had to submit to if he was to
avoid a split in the chess world. The other leading FIDE officials starred in their usual role
as extras and unanimously agreed with the President, joining him in loudly expressing
displeasure at the Soviet ultimatum. I was enormously proud of my diplomatic victory
and felt that a serious crack had appeared in my enemies' alliance. True, Gavrilin, vice-
chairman of the USSR Sports Committee, who flew in to Luzern shortly after us, gave
cautious support to Campomanes's attempt to save face by suggesting that the match
should begin in May, but, inspired by my success, I turned this down flat.
Only later did I realise that I'd been rather cleverly bamboozled. The clash between
FIDE and the USSR Chess Federation was more apparent than real, and had arisen because
the change in the political situation in our country had wrecked the plan of action against
me that they had cooked up together. All the details of the forthcoming return match had
in fact been decided at a secret meeting in Zurich airport involving Campomanes, Gavrilin,
and Anderton, an official of the British Chess Federation.
In the end the match was shared between London and Leningrad. Originally the inten-
tion had been to hold the entire match in the Soviet Union, but just in the nick of time FIDE
thought up a new rule, according to which no chess federation - in this case the Soviet
Federation - could be the organiser of two successive world championship matches. At an
executive meeting of FIDE it was soon confirmed that the contest would begin on 28 July
in London, where the first 12 games would take place, and the interval between the Lon-
don and Leningrad parts of the match would be not more than 6 days. The highly-
experienced West German Lothar Schmid was confirmed as the chief arbiter.
The days I spent in Luzern were also memorable for the remarkably smooth relations I
had with Karpov. An outsider might well have taken us for bosom buddies. We spent all
our spare time playing cards (together with my friend Frederic Friedel, a journalist and
computer expert). Even Campomanes, who had seen a few odd things in his time, was
flabbergasted by the idyllic picture of us locked in battle over the card table in my hotel
room. But the illusions soon vanished when they came up against reality.
15
Kasparov vs. Karpov: 1986-1987
Blitz Preparation
On returning home in the middle of February, I learned that Dorfman in Lvov and Ti-
moshchenko in Novosibirsk had been refused permission to travel abroad, and the latter
even to my training session. I spent a lot of nervous energy sorting out the problem of my
trainers, which even involved flying to Novosibirsk to negotiate with the military authori-
ties. But it was only much later that I found out what was going on at that time with my
team ...
Alexander Nikitin remembers:
'The reasons why someone was aiming to reduce to the minimum the interval between matches
some became clear. It proved possible to "rehabilitate" Timoshchenko only a month later, and with
great difficulty. Also in February, Sevastyanov and Karpov unexpectedly paid a visit to Lvov, where
grandmaster Dorfman was living. They appeared not so much in large auditoriums, as in the offices
of the local authorities. It may be a coincidence, but after their departure the processing of travel
documents for our trainer was halted. Garry had to appeal for help to the trades union leaders, and
only then was this "accidental" problem resolved.
'The one who felt most comfortable was our third trainer - Vladimirov, who was also serving in
the army, as a warrant officer. He was living in Alma Ata, but he continued to travel freely around
the country, sometimes on personal affairs and without obtaining the necessary documents. For a
person in the armed forces this involved a serious disciplinary risk, and we were surprised that he
continually got away it. His unauthorised travels cannot have gone unnoticed, and I think that at
some time early in 1986 our opponents were able to find a weak point in Kasparov's training group.
'Beginning in March, for three months and with only small breaks, the training base in Zagulba
became the home for our team. In the middle of the preparation period it was accidentally discovered
that Vladimirov was secretly copying out certain sections of our joint analyses. Instead of immedi-
ately asking him to explain the reason for such a flagrant violation, it was decided to pretend that
nothing had happened, not to say anything to Kasparov, and to travel to London with the same
team, after taking measures against a leak of information. The program of work had to be changed,
reducing joint analyses to the minimum. Thus a deep internal split in the training group was
discovered ... '
In the 1980s KGB Colonel Viktor Litvinov accompanied me on trips aboard, and at
matches he was the deputy leader of my delegation. This is what he recalls:
'Early in the summer of 1986 Dorfman summoned me [from Baku] to a training session in
Zagulba. "Viktor Petrovich, Vladimirov keeps writing down our analyses!" I discussed this with
Nikitin: what could this signify? By the unwritten rules, this behaviour was improper. It was
agreed with Nikitin that he and Dorfman would analyse each game of the match and try to deter-
mine whether or not Karpov knew anything about our preparations. Having decided this, we trav-
elled to London for the first half of the return match.
'In London Garry said to me several times that there was a leak of information from our team.
But we were playing abroad and everyone was particularly concerned that on no account should
there be a scandal, and so no one had the intention of investigating, especially since it was not
known how Vladimirov would react. And so it all dragged on, now in Leningrad, until the 19th
game ... Later we learned that Zaitsev, Karpov's chief trainer, was constantly surprised why Tolya
played one way, rather than another: after preparing one strategy, he would follow a completely
different one.
'I cannot say for sure that it was Vladimirov who was passing on information, but later we
16
Thirsting for Revenge
learned that Kasparov's entire entourage had been investigated. And it was probably no accident
that Dorfman was always with Vladimirov. They seemed to be inseparable friends. But ... it was
Dorfman who pointed the finger at Vladimirov.'
That spring, despite being so anxious and busy, with the help of Trades Union officials
I resurrected the Botvinnik school, in which at one time I had begun my way into top-class
chess. Now it became known as the Botvinnik-Kasparov School, and in April 1986, in
Petsov on the outskirts of Moscow, Mikhail Moiseevich and I held its first session, in which
13 young talents from the far corners of the country took part.
In the meantime Karpov confidently won the first SWIFT Tournament in Brussels - 9
out of 11 (+7=4), and in early June a double-round super-tournament in Bugojno - 8V2 out
of 14 (+4-1=9). For my part, prior to a battle for the world title I was not eager to play in
any tournaments and I followed my usual mode of preparation: why change a successful
habit?
In May I made a short tour of Europe. I first flew to Basel, where I played another train-
ing match, already the fourth that year, with one of the top Western grandmasters - this
time the Englishman Tony Miles. When I won SV2-V2 (of course, not without some good
fortune) he lamented: 'I thought I was playing the world champion, not some monster
with a hundred eyes!' I then travelled to Barcelona to collect my third chess Oscar and
went on for a familiarisation visit to London.
After returning to Baku, I immediately shut myself away in my out-of-town base in
Zagulba and concentrated entirely on preparing for the match. There was little time left,
but the experience acquired in the previous matches helped me to highlight what was
essential and solve the main problems. My team of trainers worked most productively,
and I had not even any suspicion that this was to be our last session together ...
With Black against 1 d4 I prepared the Grunfeld Defence - it should have been a surprise
to Karpov. Initially we analysed the main variation -1 d4ctJf6 2 c4 g6 3ctJc3 dS 4 cxdSctJxdS
S e4 ctJxc3 6 bxc3 iLg7 7 iLc4 cS 8 ctJe2 ctJc6 9 iLe3 0-0 10 0-0 iLg4 11 f3 ctJaS (d. Game No.32,
note to White's 12th move), then also the variations with 4ctJf3 iLg7 S 'iVb3, S iLgS or S iLf4,
and also 4 iLf4. Those were the first serious attempts to investigate the intricacies of this very
complicated opening, and as yet many of our analyses lacked depth and accuracy.
In reserve was the Queen's Gambit, which we looked at far more for White.
Of course, I also continued to prepare for 1 e4 (who could have known that Karpov had
given up this move for ever?). We analysed the Najdorf Variation a great deal, including
the sharp variations with 6 iLgS e6 7 f4 - and not only 7... iVb6, but also the main, 'Fischer'
line 7 ... iLe7 8 'iVf3 'iVc7 9 0-0-0 ctJbd7.
We carried out a large-scale opening program for White. It was planned to alternate the
moves 1 d4 and 1 e4, and after 1 d4 against the Nimzo-Indian Defence - 1...ctJf6 2 c4 e6 3
ctJc3 iLb4 - to hammer away with 4ctJf3, which had justified itself in the 1985 match. How-
ever, here Karpov was able to find an acceptable defence, and towards the end of the
match I switched to 3ctJf3 (and once I played 3 g3).
In the Queen's Gambit we made a thorough study of the variation 1 d4 ctJf6 2 c4 e6 3
ctJf3 dS 4 ctJc3 iLe7 S iLgS h6 6 iLxf6 iLxf6 7 e3 0-0 8 .l::!.c1 c6 9 iLd3 ctJd7 10 0-0 dxc4 11 iLxc4,
which had created long-term problems for Karpov in the 23rd game of the 1985 match. In
the pre-computer era this comparatively harmless line proved quite dangerous for Black,
as was confirmed in the 1986 match.
17
Kasparov vs. Karpov: 1986-1987
After 1 e4 a difficult problem was posed by the Petroff Defence - 1...e5 2 ct:lf3 ct:lf6, and
using the old ideas with 3 ct:lxe5 we were unable to solve it (Game No.6). Towards the end
of the London half of the match Timoshchenko and I prepared the variation with 3 d4, but
in Leningrad Karpov played only for the seizure of the initiative, by employing his favour-
ite Zaitsev Variation in the Ruy Lopez - 2 ... ct:lc6 3 ..ib5 a6 4 ..ia4 ct:lf6 5 0-0 ..ie7 6 ~el b5 7
..ib3 d6 8 c3 0-0 9 h3 ..ib7 10 d4 ~e8 11 ct:lbd2 ..if8. Both in this and in other openings we
had some interesting novelties prepared, but not one of them was employed: all the time
Karpov kept one step ahead, finding ways of avoiding them! The nature of this clairvoy-
ance, unprecedented in chess history, will be studied in the chapter on the return match.
Just before the departure to London I saw the first signs of discord in my team of train-
ers. Timoshchenko, who had certain disagreements with his colleagues, suddenly de-
manded that I should appoint him my official second (as in the 1984/85 match). But I
declined and Gennady, offended, declared that he no longer considered himself under any
obligations and that after the London half of the match he would go off to an army event.
Which is what happened. But in Leningrad I obtained reinforcements - Mikhail Gurevich
and, a little later, Elmar Magerramov. As for Karpov, in London he was helped by Zaitsev,
Makarychev and Salov, and in Leningrad also by Beliavsky.
I think that my publishing of an analysis of the causes of Karpov's defeat in 1985 helped
him in his preparations for the return match. However, in any case it was clear that Karpov
would not repeat the mistakes generated by complacency or self-confidence, and that in a
match which was of crucial importance in both the chess and the non-chess sense he would
mobilise all his inner reserves. But I too, remembering the sad fate of Smyslov and Tal and
with Botvinnik's steely training behind me, had carried out an enormous amount of pre-
paratory work and I sensed that I had reached a qualitatively new level of chess under-
standing.
In general, a fascinating encounter, different from the two preceding matches, was ex-
pected. The concluding round of a duel stretching over three matches within the frame-
work of one cycle was due to finally resolve matters.
18
Chapter Two I
The Third Match: 1986
19
Kasparov vs. Karpov: 1986-1987
servative Party). The Prime Minister listed alpine peak; older men, with heavy East Euro-
the qualities needed to become a top chess pean faces that seemed to harbour a deep,
player: accuracy of thinking, imagination, serious knowledge of the game; and well-kempt
and good physical condition. She remarked secretaries and officials in pin-stripes, who had
that these qualities are also needed to come straight from work.'
become a good politician - 'but chess is The latest technology meant that every
limited in time', she added: 'we deal in move was instantly displayed on the moni-
unfinished business'. Then she and the tors, the pieces having been sensitised
chief arbiter Lothar Schmid carried out the electronically by a coil inside them, which
drawing of lots, which gave Karpov the was linked to a computer. This was the
white pieces in the first game. brainchild of David Levy and Kevin
In London I tried not to feed the gossip O'Connell of Intelligent Chess Software
about the hostility between me and Karpov: and it was revolutionising chess as a spec-
we were both now ambassadors for our tator sport. One commentator wrote: 'The
country. At a press conference I said: 'My main function of the champion and challenger
position remains unchanged, but in the conducting their private contest on the ballroom
months to come the issue between us will stage is to make a move. Instantaneously the
be decided not by words, but by how we move is relayed to thousands of television sets
play chess'. Some of the newspapers were scattered around the hotel. In bars, conference
categorical: 'This is war! There were no boxing rooms, press offices, lobbies, TV lounges, com-
gloves in sight, but that could not disguise the mentary positions and smoking rooms, the other
fact that this was a two-man war.' According players immediately begin to work through the
to The Spectator: 'If looks could kill, one of the zillions of new options that have opened to
greatest world title battles in chess history them... Chess is perhaps the only game in the
would have ended before it started yesterday world where there are no simple spectators: here
evening.' they are all participants!'
What most interested the public was the ... And thus began a new encounter at
contrast between our personalities. As one the chess board with Karpov. Almost
commentator wrote: 'Every competitive sport everything that was said about the first
needs heroes with whom the public can identify. game of the previous match also applied
Tennis had McEnroe and Borg, then Becker and here, with the difference that the time
Lendl; snooker had Alex Higgins and Steve between the matches had increased the
Davis. For ten years the chess world was ruled importance of every factor. The two players
by the quiet, calm and cautious Karpov. When now had not 48, but 72 games behind them,
he was dethroned by the brash and highly but the main thing was that the change in
temperamental Kasparov, chess, as a sport, our status was bound to change the charac-
suddenly came to life again. If the USSR wants ter of the play. The psychological novelty of
to create a new image for itself Kasparov must the situation - for me, the first match as
be one of its best ambassadors.' champion, and for Karpov the first meeting
We were all impressed by the level of with a new world champion (before this
public interest in chess in Britain. Every my opponent had only once played a
day there were long queues for tickets. current champion: Spassky in 1971) - was
Here, according to The Times, were 'bronzed bound to affect the overall strategic course.
young tourists in garish anoraks, looking for all Tal: 'On 28 July at 17.00 Greenwich Mean
the world as if they had just conquered some Time the chief arbiter of the match started the
20
The Third Match: 1986
21
Kasparov vs. Karpov: 1986-1987
22
The Third Match: 1986
'¥VcS as!), and here 16 ... '¥Vc3! 17 ~d1 ~fd8 London 1999);
(17 ... ~ab8!?) 18 f3 (18 '¥VcS ttJxe4) 18 ... aS 2) IS Jl.bS 'iYc2 16 Jl.xd7 (16 'iYe4?!
(Beliavsky) would have given Black excel- "iVxd2+!) 16 ... Jl.xd7 17 'i¥e4 'iVa4 18 f3 (18
lent chances. ~1 Jl.c6 19 ttJf3 'iYa6 is no better) 18 ... aS 19
Then in the 21st Century White never- ttJc4 Jl.fS 20 'iYf4 (Bareev-Kalka, Bundesliga
theless gained an advantage by 14 c6! ttJeS 2000) 20 ... axb4! 21 ttJb6 'iYc2 22 0-0 (22
IS ~S! '¥VxbS 16 Jl.xbS ~fb8! 17 a4 a6 18 c7 ttJxa8? Jl.d3) 22 ... ~xa3 23 ~ac1 'iYe2 24 oUfe1
~b7 19 Jl.e2 ~xc7 20 f4 ttJc6? 21 ~c1 ~cc8 22 ~S with double-edged play. Here 16 O-O!?
bS axbS 23 Jl.xbS ttJd4 24 ~xc8+ ~xc8 2S ~f2 is more dangerous: 16 ... ~ab8 (16 ... ~db8 17
(Ikonnikov-Schmitt, Selestat 2007). How- 'iYa6 "iVxd2 18 ~ad1 "iVc3 19 ~xd7) 17 'i¥c7
ever, 20 ... ttJg4 was necessary, with fair ~bc8 (17 ... ~dc8?! 18 'iYaS ttJeS 19 oUfd1 and
compensation for the pawn and real wins) 18 'iYxa7 oUa8 19 'iYc7 oUac8 20 'iYg3!
chances of a draw: 21 h3 ttJf6 22 g4 ~b8, or '¥Vxd2 21 Jl.a6 l:ta8 22 Jl.b7, and White's
21 0-0 ~c2 22 fS ~xd2 23 Jl.xg4 Jl.d7 24 ~f2 connected passed pawns are stronger than
~xf2 2S ~xf2 ~b8 26 bS axbS 27 as b4 28 the black knight.
~b1 b3 29 a6 ~a8 30 .ie2 Jl.c8. Not surprisingly, S... cS?! has almost
Apart from 13 e4, White can also con- gone out of serious use. What is strange is
sider a line which arose in the 1990s: 13 e3!? something else: Karpov 'took me at my
ttJd7 14 a3 ~fd8 (14 ... aS? IS Jl.bS). word', that after 7 cxdS! Black is alright,
and in the match he did not again employ
the variation 4 ttJf3 Jl.g7 S Jl.f4. This is
contrary to the logic of match play! Such
sharp lines are normally checked carefully,
with the aim of refuting the opponent's
idea. Perhaps in his analysis Karpov carne
to the same erroneous conclusion as I did:
that the complications are favourable for
Black? Well, this sometimes happens, but in
this unusual match we were to arrive at
identical conclusions with frightening
regularity ...
7 ~c1?!
Analysis diagram A solid, but rather inactive move - by
avoiding the complications, White essen-
In this position IS '¥Va6 or IS Jl.bS has tially gives up the fight for an opening
been played, for example: advantage. 'To follow the risky path with 7
1) IS '¥Va6 '¥Vc2 16 '¥Vd3 '¥Vxd3 (16 ... '¥Va4 17 cxdS without special analysis would at the
'¥Vc3! Van Wely-Kamsky, Groningen 1995) least have been imprudent, and Karpov
17 Jl.xd3 ttJxcS 18 Jl.c2 (little is promised by considered it advisable to take the game
18 Jl.xg6 hxg6 19 bxcS ~ac8 20 ~c1 ~dS) along quiet lines.' (Taimanov)
18 ... ttJd3+ 19 .ixd3 ~xd3 20 ~e2 (20 ttJf3 as! 7... dxC4!
21 bS a4 is equal, Riazantsev-Macieja, Of course!
Moscow 2002; 20 ttJe4!?) 20 ... ~dS!? 8 e3
(20 ... l:tad8 21 ~hd1), trying to neutralise 8 'iYa4+ 'iYxa4 9 ttJxa4 is also unpromis-
White's extra pawn by ... a7-aS (Mah-Lalic, ing. Here the press centre bulletin recom-
23
Kasparov vs. Karpov: 1986-1987
mended 9 ... ttJc6 10 e3 (10 lixc4 i..e6) with a win after 14 ... 'li'dS? IS lixcS or
10 ... i..d7 11 i..xc4 (I would add 11 ttJc3 ttJhS 14 .. :i\¥e6? IS lixcs+ i..fS 16 ttJxf6+ exf6 17
12 i..xc4 licS) l1...ttJaS etc. with compensa- lidS. And although Black is saved by
tion for the pawn. But the most accurate 14 ... ttJxdS! IS J:l.xcS+ i..fS 16 i..h6 ttJd7 17
and also the simplest is 9 ... i..d7! 10 .uxc4 .uxaS 'li'xb2 IS i..d3 ttJSb6 19 lidS f6, I do not
ttJa6 11 e3 i..xa4 (not 11...0-0? 12 c6!) 12 think that he should avoid 1O ... 'li'xc4.
lixa4 ttJxcS 13 i..bS+ (or 13 .uc4 ttJfd7 14 lic2 10 i..xc4
ttJe6 with equality) 13 ... ~fS 14 'ua3 (14 lic4?! Refraining for the moment from the ex-
ttJd3+; 14 lid4 ttJe6) 14 ... ttJdS IS i..eS i..xeS change of queens - 10 'li'xc4 'li'xc4 (now if
16 ttJxeS ~g7 with good play (Azmai- 10 ... 'li'b6 there is 11 'iVbS) 11 i..xc4 0-0 etc.
parashvili-Leko, Leon 1994). After 10 ttJeS i..d7 (10 ... 0-0 11 ttJxc6 i..d7 or
11 'li'xc4 'ilVb6 is also good) 11 ttJxd7 ttJxd7
12 'iVxc4 0-0 Black's lead in development at
least compensates for the opponent's two
bishops. 10 ttJbS is also no better: 10 ... 0-0 11
ttJc7 J::!.bS (l1...eS!?) 12 i..xc4 'li'hs with
equality.
10 ... 0-0 11 0-0 i..d7!
11...i..g4, which was recommended as
an alternative, was also not bad, but it was
far less logical. 'The problem of the opposi-
tion of the white rook and the black queen
is solved by Kasparov in the most active
and effective way, reminding his opponent
8.. :~xcS that on the c1-cS route there are two barri-
The path to equality is rather more ers, whereas the d7-a4 road is almost
complicated after S... O-O 9 i..xc4 'iYxcS, deserted.' (Tal)
transposing into positions from the 4 i..f4
variation, which at that moment seemed
dangerous to me because of 10 ttJbS, but
which later occurred in the 9th and 11th
games.
9 '¥Va4+ ttJc6
9 ... i..d7 10 '¥Vxc4 'iYxc4 11 i..xc4 ttJc6 12
0-00-013 lifd1 liacS 14 e4 lifdS would have
forced equal chances. Black can also try to
sharpen the play by 10 ... '¥i'b6, attacking the
b2-pawn and intending 11...0-0 and ... licS.
In the event of 11 i..c7 'iYxb2 12 lib1 '¥Va3 13
lixb7 0-0 or 11 ttJbS ttJa6 12 i..c7 'li'e6 this
idea would succeed. It is more interesting 12 'li'bs
to play 11 i..e2 (if 11 i..d3 0-0 12 0-0, then The unpleasant diagonal 'X-ray' com-
12 ... ttJa6) 11...0-0 12 0-0, and after 12 ... licS in pels White to go in for the exchange of
Injarmatar Karpov and Zaitsev suggested a queens. 'For roughly half an hour Karpov
queen sacrifice - 13 'li'xcS+!? i..xcS 14 ttJdS! looked for a way to retain the tension, but
24
The Third Match: 1986
without success. And the move made is 2) 19 ~xa6!? J:ta8! 20 ~b7 bxc3 21 J:txc3
essentially the start of peace negotiations.' ttJb4! 22 ~xa8 J:txa8 23 b3 with a position
(Taimanov) which is hard to evaluate. At the board it
The last chance of fighting for an advan- did not appeal to me, but after the possible
tage was 12 l:I.fdl ttJeS 13 ~bS, but here too 23 ... ~e6 Black's two minor pieces are
after 13 ... ttJxf3+ 14 gxf3 ~xbS IS 'iYxbS hardly weaker than White's rook and two
'~xbS 16 ttJxbS a6 and ... J:tac8 the draw pawns.
would be not far off. However, there was no need for a risky
12 ... 'iYxb5 13 ~xb5 J:tacS 14 J:tfdl J:tfdS sharpening of the play: why create addi-
White is very slightly more active, but tional weaknesses for yourself when you
within a few moves even this disappears. are a step away from a draw, especially in
15 h3 h6 the first game of a match?
Preparing to bring out the bishop to e6. is J:txdS+
The immediate IS ... ~e6 would have al- Things would not essentially have been
lowed White to break the symmetry to his changed by 18 ttJeS ttJxeS 19 ~xeSl:txdl+ 20
advantage: 16 .itxc6! l:I.xdl+ 17 l:I.xdl bxc6 l:I.xdl ttJe8 or 20 ... ttJd7 21 ~xg7 'it>xg7 22 ~f3
(17 ... J:txc6? 18 J:td8+) 18 ~eS. l:tc7.
16 'it'fl (or 16 e4 a6) 16 ... a6 17 ~e2 lS ... J:txdS 19 ttJe5
In the event of 17 ~a4?! Black could 'It is now White who is required to dis-
have activated his game by 17 ... bS (18 ~b3 playa certain accuracy.' (Tal)
:'LJaS). 19 ... ttJxe5 20 ~xe5l:td2
A show of activity. After 21 ttJbl I was
intending simply to return with my rook to
d8, and then play ... ttJf6-e8.
17 ... ~e6
Here too 17 ... bS was possible, leading
after 18 a4 b4 to interesting and unexpected
complications: 21 b3
1) 19 ttJbl (19 ttJa2 b3 20 ttJc3 ttJb4) After making this move, Karpov offered
19 ... ttJaS! 20 ~xa6 (20 J:txc8 J:txc8 21 ~xa6 a draw, which, of course, was accepted
.l¥..xa4) 20 ... J:txc1 21 J:txc1 ~xa4 with excel- (Yz-Yz). If 21...ttJdS, then 22 ~xg7 'it'xg7 23
lent play for Black, although after 22 ttJbd2 l:td1. Times: 1.47-1.48.
(22 ~c7 J:tdS) 22 ... ttJdS 23 J:tal ttJb3 24 J:txa4
ttJxd2+ 2S ttJxd2 ttJxf4 26 exf4 J:txd2 27 J:txb4 Naturally, this confident draw with
.l¥..xb2 it all ends in a draw; Black in the first game strengthened my
25
Kasparov vs. Karpov: 1986-1987
fighting spirit. Not yet knowing how well- analysed it and in general I did not know of
informed Karpov was about my opening it - at that time information spread in
preparation, I was optimistic about the different ways. I had to stop and think for a
prospect of further employing the Grunfeld long time ...
Defence. 'The move order employed by Karpov
Two days later in my first 'White' game looks very logical: the pressure on d4 does
I decided in the first instance to check what not allow White to make moves only on
Karpov would embark on in the Nimzo- general grounds. It was no accident that
Indian Defence, where he had so suffered Kasparov spent 35 minutes on his next
in the previous match. And I did not have move.' (Tal) I had to take an important
to wait long for a novelty to appear... decision, which would determine the
further course of the battle - find the best
reaction to the crisis in the centre.
Game 2
G.Kasparov-A.Karpov
World Championship Match,
2nd Game, London 30/31.07.1986
Nimzo-Indian Defence E20
26
The Third Match: 1986
1984), which went 1 ltJf3 d5 2 d4 ltJf6 3 c4 e6 his opponent liked to play shows that the
4 ltJc3 ~b4 5 g3 dxc4 6 ~g2 c5 7 0-0 ltJc6 8 first two matches with Karpov had made
dxc5, and after 8 .. .'Yi'xdl 9 ~xdl ~xc5 10 him into a genuinely universal player.'
ltJd2 ltJa5 11 ltJde4 White gained the initia- (Nikitin)
tive. However, neither in this old game, nor
in the modern encounter Radjabov-J.Polgar
(Biel 2007), did White manage to win.
But perhaps Karpov was intending to
improve Black's play somewhere - for
example, by avoiding the queen exchange
with 8 .. .'Yi'a5!? (if 8 .. .'Yi'e7, apart from 9 ~g5
or 9 ltJd4 there is the computer leap 9 ltJe5!,
promising White an advantage: 9 ... 0-0 10
'iVa4! etc.). This seemed unclear to me:
indeed, after 9 ltJd2 O-O! 10 ltJxc4 'iVxc5 11
ltJe3 ~d8 or 9 ltJa4 0-0 10 'iVc2 ~d7 Black
has no reason for complaint.
Later we analysed 9 ltJd4 ~d7 10 ltJdb5 9... ltJxd4
or 9 ... ltJxd4 10 'iVxd4 'iVxc5 11 ~e3 with Of course, Black cannot be satisfied with
some pressure for White. 9 ~g5 ~xc3 10 9 ... ltJxc3? 10 bxc3 ltJxd4 11 cxb4 e5 12 e3 -
bxc3 looks more energetic, and if 10 ... 'iVxc5 White's advantage is obvious. And after
11 ~xf6 gxf6, then 12 'iVd2! (12 ltJd2 is 9 ... ~xc3?! 10 bxc3 ltJde7 he replies simply
weaker because of 12 ... f5!, but not 12 ... 0-0?! 11 0-0 (or 11 ~f4!? 'iVa5 120-0, since 12 ... e5,
13 ltJe4 Krogius-Borisenko, Moscow 1966) given by me in the book Ova matcha, is
12 ... ~d7 13 'iVh6 'iVe7 14 ltJd2 with excellent weak on account of 13 ltJxc6 ltJxc6 14 ~xc6+
compensation for the pawn. However, after bxc6 15 'iVd6!), and the acceptance of the
10 ... ltJd5!? 11 ltJd4 0-0 (ll.. ..id7?! is worse: pawn sacrifice leaves Black facing difficult
12 ~xd5! exd5 13 Mbl) 12 ltJxc6 bxc6 the problems: 11...ltJxd4 12 cxd4 'iVxd4 13 ~bl
situation remains unclear, for example: 13 ltJd5 (it is weaker to play 13 ... 0-0 14 ~b4! or
iYd4 (13 .id2 e5) 13 ... ~a6 14 ~d2 ~fe8 13 ... ltJc6 14 'iVc2 0-0 15 ~fdl!) 14 'iVc2 0-0 15
(with the idea of ... e6-e5) 15 'iVe5 f6 16 'iVh5 e4 ltJb6 16 ~b4 ~e8 17 ~fdl with the initia-
~ad8, and Black holds on. tive.
If this variation had occurred again in 10 ltJxdS ~xd2+ 11 'YWxd2 ltJc6?!
the match, after calm analysis at home I Not the best of the three possible moves.
would have played 7 0-0. But in this game, Probably Black's position did not cause
on encountering a surprise, I decided to Karpov any particular fears, and he as-
satisfy myself with a minimal initiative - sumed that the choice here was not of
after all, the match was only just beginning. critical importance.
Besides, I thought that the position after the The other knight move - 11...ltJf5 would
simplification provoked by 7 cxd5 was still have forced the exchange of queens, but
by no means drawn. after 12 ltJe3!? (if 12 ltJf4 'iVxd2+ 13 ~xd2
7... ltJxdS 8 .id2 cxd4 9 ltJxd4 there is the defence 13 ... ~d7 14 ~hc1 ltJd6)
White is aiming for an albeit slight, but 12 .. .'Yi'xd2+ 13 ~xd2 ltJd6 14 ~ac1 ~d7 15
long-term advantage. 'The fact that Kas- ctJc4 or 12 ... ltJxe3 13 'iVxe3 'iVa5+ 14 'iVd2
parov went in for a position of the type that 'iYxd2+ 15 ~xd2 ~e7 16 ~ac1 ~d8+ 17 ~e3
27
Kasparov vs. Karpov: 1986-1987
.l:tb8 18 .l:tc7+ t!.d7 19 t!.hc1 ~d8 20 t!.7c4 the places Black in a difficult position after 14
endgame is permanently inferior for Black: "Yi'xd4 ctJxd4 15 O-O-O!. After 13 ... e5
it is difficult for him to avoid the creation of (Brodsky-Tarlev, Voronezh 2006) he has to
weaknesses of some kind in his position. reckon with 14 t!.dl or 14 .Jixc6. And in the
But why move the knight from d4? It is event of 13 ... .Jid7 14 0-0 "Yi'e7 extreme cau-
far more ambitious to support it with tion is demanded of Black, for example: 15
11...e5!, since if 12 e3 or 12 t!.dl there is the t!. fd 1 ~fd8 16 a3 .Jie8 17 t!.xd8 "Yi'xd8
tactical trick 12 ... .Jih3!. In the event of 12 0-0 (17 ... t!.xd8 18 ~xc6!) 18 t!.dl "Yi'e7 19 b4,
0-0 (Winants-Short, Brussels 1987) White retaining an enduring initiative.
can claim a small plus after 13 e3 .Jih3 14 f4,
but it is more accurate to play 12 ... .Jie6! 13
e4 0-0 or 13 'iVe3 "Yi'd6. Finally, the sharp
attempt 12 ctJc3 0-0 13 0-0-0 is parried by
13 ... "Yi'c7 14 e3 ctJb5. Thus 11...e5! would
have solved most simply the problem of
how to equalise, as was also confirmed by
later experience.
12 ... "Yi'xd2+
After 12 ... 0-0 White can suddenly
change his mind and avoid the exchange of
queens - 13 ctJd3 (Bacrot-Short, 3rd match
game, Albert 2000).
13 ~xd2 .Jid7
Let us sum up the result of the opening.
For the moment the white pieces are
12 ctJf4?! slightly more active, but the symmetry of
'Kasparov spent a considerable time on the pawn structure and the absence of
this move.' (Tal) But, in contrast to the plan weaknesses in Black's position should soon
beginning with 7 cxd5, this striving for nullify this symbolic advantage. Literally
simplification (just like 12 ctJc3?! "Yi'xd2+ 13 one or two accurate moves, and a draw can
~xd2 .Jid7, which was tried in the source be agreed. In the history of our encounters
game Vinuesa-Schwartman, Mar del Plata (I have in mind the unlimited match) there
1936) is not at all justified - after the ex- were numerous games where an attempt to
change of queens White remains with only gain an advantage without any risk met
illusory chances of success. with a precise reply, after which the players
Far more interesting play results from signed a peace agreement, to the obvious
12 "Yi'e3! 0-0 (12 ... "Yi'a5+ 13 ctJc3 "iVb6 also dissatisfaction of the spectators.
does not give clear equality, Bacrot-Schenk, The same result, but with far less moral
France 2004) 13 ctJc3. Now the persistent detriment for chess (this was only the 2nd
desire to exchange queens - 13 ... "Yi'd4?! game, and the London spectator does not
28
The Third Match: 1986
often witness events of this standard) could 14 ... 0-0-0?! IS ctJd3 ~b8 16 ~el ~e8 17 ctJcS!
easily have been predicted here - the (17 J::i.c3 is weaker because of 17 ... ctJd4 18
defending side's set-up was just too solid. Mac1 ~c6!), preventing ... b7-b6 and not
This is what Mikhail Tal did, commenting allowing the black knight to move. After
on this game on the radio. And indeed, at 17... J::i.c8 18 J::i.c3 Black's position is rather
that moment it was impossible to assume dangerous, as is illustrated by the amusing
that the return match would spring so variation 18 ... b6? 19 ctJa6+ ~b7 20 ctJb4 with
many surprises, radically changing the a forced win for White - 20 ... aS 21 ctJxc6
popular opinion about the character of ~xc6 22 J::i.xc6 J::i.xc6 23 J::i.c1 J::i.hc8 24 a4!. It is
chess encounters between the two 'Ka's. a paradox, but for a long time powerful
But it all began from this seemingly life- analytical programs do not see that Black
less position, in which Karpov did not loses because of the pin: White advances
display his customary accuracy, while I his kingside pawns and at a convenient
proved uncharacteristically persistent in moment he transposes into a won pawn
the 'squeezing out' of minimal chances. endgame. This is a striking example of
However, the further course of the match 'gaps in the education' of chess machines!
was to show that this development of 15 ctJd3 J::i.hcS
events was not accidental and that I was If IS ... J::i.hd8, then 16 J::i.c3(cS) and J::i.acl.
ready to play for a win in 'simple' endings. Assuming that White will not object to a
quick draw, Karpov prepares to make use
of the c-file for further simplification.
16 ctJC5
14 J::i.hcl
The most natural move. The immediate
14 ctJd3 was also possible, hoping for
14 ... b6? IS J::i.hc1 J::i.c8 16 J::i.c3 and J::i.ac1 with 16 ... J::i.abS?!
the seizure of the c-file. However, after As yet this hardly spoils anything, in
14 .. .'it>e7 it transpires that the threat of ctJcS contrast to the incautious 16 ... ~e8? 17 J::i.c3
is illusory: IS ctJcS J::i.ad8! 16 J::i.hc1 (16 Mac1 l:tab8 18 Mac1, when Black runs into diffi-
ctJeS!) 16 ... b6 with equality. culties (18 ... b6? 19 ctJa6; 18 ... ctJd8? 19
14... ~e7 ctJxb7!). After Tal's recommendation
The correct reply. In the centre the king 16 ... J::i.c7 there would also have followed 17
is far more safely placed than on the queen- J::i.c3 J::i.ac8 18 J::i.ac1 or 17 ... J::i.d8 18 ctJd3 with
side, where, although it defends the b7- some pressure.
pawn, it could itself become a target: However, the immediate 16 ... b6! would
29
Kasparov vs. Karpov: 1986-1987
have solved the problems more simply and l:!.b3 l:!.c7 24 l:!.b4 h6 (Georgadze-G.Garcia,
quickly - after 17 LiJxd7 (17 LiJa6 l:!.d8! 18 Pontevedra 1986) or 19 l:!.a3 a6 20 l:!.d1 'It>e7
'It>e1 l:!.ac8 with equality) 17... 'lt>xd7 18 l:!.c3 21 'It>e3 (21 l:!.b3 l:!.c5) 21...l:!.c2 22 l:!.b3 LiJc6
l:!.c7 19 l:!.ac1 (or 19 l:!.d1 l:!.ac8) 19 ... l:!.ac8 White's pressure would have evaporated.
White has no more resources for strength- lS ... ..tc6
ening his position, for example: 20 f4 LiJa5 Again avoiding ... b7-b6, the move that
21 l:!.xc7+ l:!.xc7 22 l:!.xc7+ 'It>xc7, and with suggests itself. '18 ... ..te8? 19 LiJxb7! is bad
four pawns against four on the kingside for Black. In this game too the ex-world
Black's defence is not so complicated as champion advances his pawns very care-
with three against three (d. the ending of fully, endeavouring to avoid the slightest
the 10th game), since the white king has no weakening of his position. But in the given
real possibility of breaking through. But, instance I think that Yuri Razuvaev was
leaving himself with a knight against the right, when on the pages of Sovietsky Sport
bishop on g2, Karpov did not want to he recommended the sharp 18 ... b6. In the
weaken the light squares. event of 19 LiJd3 Black's defence is ex-
17l:!.C3 tremely easy: 19 ... l:!.xc3 20 l:!.xc3l:!.c8. Things
The planned doubling of rooks. are rather more complicated after 19 LiJxd7
'It>xd7 20 l:!.d3+! 'It>e7 (20 ... WeS 21 l:!.cc3! -
G.K.) 21 l:!.a3 a5 (21...l:!.xc1? 22 l:!.xa7+) 22
l:!.xc8 l:!.xc8 23 l:!.b3 l:!.b8, and although the
endgame is certainly favourable for White,
he apparently should not have any real
opportunity to exploit the weakness of the
b5-square. Possibly Karpov underestimated
the clever knight manoeuvre carried out by
his opponent.' (Tal)
Black also experiences a certain discom-
fort after 18 ... l:!.c7 19 b3!? - a waiting move:
19 ... ..te8 20 f4, and if 19 ... ..tc6 there is 20
LiJa6 bxa6 21 ..txc6 LiJxc6 22 l:!.xc6 l:!.d8+ 23
17 ... LiJdS 'It>e3 l:!.xc6 24 .l:txc6 l:!.d6 25 l:!.c7+ l:td7 26 l:!.c5
'Black defends very accurately: after with a slightly better rook endgame.
safely covering the b7-point, he intends to In any event, White exerts light pres-
exchange rooks on the c-file.' (Suetin) And sure. In the book Ova matcha I followed the
he is ready for the exchange of bishops by other commentators in condemning
... £c6 or the exchange tZJxd7 without the 18 ... ..tc6, writing that ' soon Black encoun-
weakening ... b7-b6. Karpov apparently ters real difficulties'. However, analysis
thought that I was playing on purely carried out with the help of computer
through obstinacy, and he was aiming to programs has significantly changed the
deny White even any purely symbolic evaluation of the events which occurred in
advantages. the game. Specific variations do not con-
lSl:!.acl firm the version that Black fell back step by
Assuming that for the moment there is step and reached a bad position. In fact,
no need for 18 LiJxd7 'It>xd7: after 19 l:!.ac1 despite the errors committed, until the
l:!.xc3 20 l:!.xc3l:!.c8 21l:!.a3 a6 22l:!.d3+ 'It>e7 23 blunder 38 .. JH3?? at no point was he in
30
The Third Match: 1986
danger of losing. But Karpov's persistent bishops (after ... f7-f6 and ... .tc6), although
unwillingness to create any weaknesses in by 21...~cS he could have gone into a
his position undoubtedly prolonged the position similar to that examined in the
struggle and gave me additional chances. note to Black's lSth move, with only a
slight difference: after 22 ~a3 a6 23 ct:Jxd7
(23 ~b3? .tb5!) 23 ... 'it>xd7 24 .l:i.b3 l:tc7 the
placing of Black's queenside pawns on a6
and b7 - instead of a7(a5) and b6 - is nomi-
nally to White's advantage. However, here
Black's defence would not have been
difficult.
After the move in the game Black is still
within the equality zone, but... 'Defending
this position against an extremely persis-
tent opponent is an unpleasant business.
Perhaps that was why in the 4th game
Karpov employed a different system of
19 ct:Jd3 defence.' (Tal)
White changes the attacking front, 22 b4!
avoiding the exchange of bishops, but An important nuance. To carry out his
allowing the exchange of one pair of rooks. plan of exchanging bishops Black is now
19 e4!? was interesting - this move looks forced to play ... a7-a6 (since the immediate
rather artificial (it blocks the bishop on g2), 22 .. .f6? is bad in view of 23 ct:Jd3 .tc6 24
and therefore it was underestimated. But in .txc6 ct:Jxc6 25 b5 ct:Jd4 26 ~c7+ WfS 27 a4),
this way White retains the pressure on the and in the future the pawns on a6 and b7
c-file for the moment, whereas Black has to will inevitably come under attack not only
solve the problem of his knight on dS. If by the light-square bishop, but also by the
19 ... f6 with the idea of ... ct:Jf7 White has the knight, if it returns to c5.
reply 20 .th3 with an attack on the e6-
pawn. Black also does not have complete
equality after 19 ... a5 20 ct:Jb3 b6 21 e5, for
example: 21...Wd7 22 .tf1! .te4 23 .tb5+
..iLc6 24 ~xc6 ~xc6 25 ct:Jd4 ~bcS 26 ct:Jxc6
.:tJxc6 27 f4, or 21.. ..td7 22 ct:Jd4 ~xc3 23
~xc3 l::tcS 24 ~b3 ~bS 25 .te4 h6 26 f4 etc.
19 ... .td7
Possibly 19 ... f6, opposing White's plan,
should have been preferred. I was intend-
ing to press with 20 .te4 h6 21 b4 .td7 22
.:tJc5 .tc6 23 .td3, but after 23 ... ct:Jf7 24 ct:Ja6
(24 .tc4 e5) 24 ... ~aS nothing real for White
is apparent. 22 .•• a6
20 ct:Jes ~XC3 21 ~XC3 .te8 'During the last ten or so moves White
Again psychology: as a matter of princi- has made considerable progress: each of his
ple Karpov is aiming for the exchange of pieces is more active than the opponent's.
31
Kasparov vs. Karpov: 1986-1987
Nevertheless, Black's position, which has 2S ~xc6 ttJxc6 26 f4!? (if 26 ttJcS Black
no weaknesses, does not yet give grounds equalises by 26 ... aS!) 26 ... Md8 27 ~e3 Mc8!
for pessimistic forecasts. Was it necessary my chances of success would have been
to play 22 ... a6, as though urging White to slight (28 ttJcS ttJb8!).
transfer his knight to a peripheral post, but 23 ... h6
one that is very unpleasant for Black - the Each of the other pawn advances has
as-square? I think that things would have definite drawbacks. However, also after
been easier for Black after 22 ... ~d6 23 f4 f6 23 .. .£6 24 ttJd3 fS 2S ~f3 ~c6 there would
24 ttJc4+ ~e7 followed by the moving of not have been anything terrible for Black:
one of his minor pieces from the 8th rank.' 26 ..Itxc6 ttJxc6 27 a3 Md8 (27 ... eS 28 ttJcS as
(Tal) is also good) 28 ~e3 Mc8 29 f4 (29 ttJcS ttJeS
However, after 2S ttJaS he would still 30 ~d2 Mc7 with equality) 29 ... MC7 30 ttJeS
have been under some pressure: 2S ... b6?! 26 ~d6, or 26 ttJeS! (recommended by Tal)
ttJb3 ~d6 27 ttJd4 as (27 ... ~d7 28 Md3! ~e7 26 ... ~d6! (Tal gave only 26 ... ~xf3 27 exf3!
29 Ma3) 28 a3, or 2S ... ~d6 26 e4!? ttJc6 ~d6 28 f4 'with a threatening, if not deci-
(26 ...b6?! 27 eS+!) 27 eS+ fxeS 28 ttJc4+ ~e7 sive positional advantage for White') 27
29 ttJxeS ttJxb4 30 Mc7+ ~f8 31 a3 ttJc6 32 ttJxc6 ttJxc6 28 ~xc6 bxc6, maintaining the
~xc6 and 'it'c3. status quo (29 Md3+ ~c7 30 ~c3 MbS 31 ~c4
as).
24a3
Again Black faces a choice.
23~e4
At first sight very logical and consistent.
White intends to provoke the advance of
the black pawns on the kingside, with the 24 ... f6?!
aim of creating a further weakness, thus Karpov sticks to his plan, although he
widening the active front. But the position could have advantageously carried out the
of the bishop on e4 gives Black additional simplifying 24 ... aS!, for example: 2S f4 axb4
tactical possibilities. 26 aXb4 ttJc6 27 ~xc6! bxc6 (27 ... iLxc6? 28
The prophylactic 23 a3 would have re- Mxc6!) 28 ttJd3 Ma8! 29 'it'e3 Ma2 30 MCS fS 31
duced the effectiveness of the undermining MaS Mc2, extinguishing White's initiative,
23 ... aS? in view of 24 bxaS Ma8 2S MCS ~d6 or 2S bxaS Ma8! 26 f4!? (26 ttJc4 iLbS! or 26
26 ttJd3 ~c6 27 e4 f6 28 f4 ttJf7 29 eS+! fxeS MCS ~d6 27 ttJd3 iLc6 is no better for White)
30 ~xc6! bxc6 31 fxeS+ ~c7 32 ttJb4 and 26 ... MxaS (26 ... f6 27 ttJc4 iLbS will also do)
wins. But here too after 23 ... f6 24 ttJd3 ~c6 27 Mc7+ ~d6 28 Mc8 ~e7 29 ttJc4 Ma4 with
32
The Third Match: 1986
fairly simple equality (30 l:tc7+ 'it>f8 31 'it>c3 with the exchange of bishops. 2S ... 'it>d6 was
bS 32 ttJd6l:txa3+ 33 'it>b4l:ta1). perhaps preferable, although Black is
25 ttJd3 already experiencing difficulties.' (Suetin)
The best practical chance and the logical In fact things are by no means so
development of the oltg2-e4 idea was 2S gloomy, and Black's position is far more
ttJg6+!?, hoping for 2S ... 'it>d6 26l::td3+ 'it>c7 27 solid than it appeared at the time both to
ttJf4 when Black has more problems than in the commentators, and to the match par-
the game: 27 ... oltc6? 28 oltxc6 bxc6 29 ttJhS, ticipants themselves.
or 27 ... gS 28 l:tc3+ 'it>d6 29 ttJd3 oltc6? 30 26 oltxc6 ttJxc6
oltxc6 ttJxc6 31 ttJcS. Black can also defend
with 27 ... ..if7 (or 27 ... oltbS 28 l:tc3+ 'it>d6) 28
l:tc3+ 'it>d7 29 ttJd3 b6, but this too does not
yet guarantee complete equality (30 ttJb2!
eS 31 ttJa4 i!.e6 32 oltd3 as 33 oltbS+ 'it>d6 34
e4 etc.).
It would probably not have suited Kar-
pov to play 2S ... oltxg6 26 oltxg6 (since he did
not want to give his opponent the' eternal'
advantage of bishop over knight), but here
Black had a comparatively clear way to
draw: 26 ... ttJc6 27 olte4 'it>d6 28 f4 (28 ..ixc6?!
bxc6 29 l:tcSl:tbS!) 28 ... 'it>c7 29 'it>e3 l:td8 etc.
By retaining the knight, I was optimisti- 27 ttJC5
cally hoping that I would be able to achieve In the book Dva matcha I condemned
more - use it to attack the black pawns and, this 'hasty' move and recommended 27 f4
by tying down the opponent's pieces, play l:td8 28 'it>e3, 'aiming for the further activa-
for domination and the exploitation of the tion of the pieces'. However, after 28 ... l:tc8!
tactical features of the position. Black has no particular reason for concern
(a similar position was already reached in
the note to White's 23rd move).
27 ...ttJe5?!
27 ... aS! would have solved Black's de-
fensive problems more quickly, for exam-
ple: 28 bS ttJd4 29 a4 (29 l:td3 b6!) 29 ... b6 30
ttJd3 l:td8.
28 f4 ttJd7 29 ttJb3!
For the first time White has gained some
perceptible advantage: it is easier for his
more active pieces to attack the enemy
pawns. Black is condemned to passive
defence, but... the position is too close to a
25 ... oltc6 draw.
'The number of pieces is reduced by 29 ...'it>d6
one, but Black's difficulties are not re- I thought that this was' another serious
duced.' (Tal) 'It was hardly correct to hurry inaccuracy', and that 'it was essential to
33
Kasparov vs. Karpov: 1986-1987
play 29 ... fS, hindering White's possible tiJxc8 37 tiJb7+ and tiJcs with domination
activation in the centre'. Now 30 tiJd4 g6 31 for White (Ova matcha), but 3S ... 'it'd7 36 g4
l:i.e3 tiJf8 does not promise White anything .l:!.c8 37 .l:!.xc8 'it'xc8, and the draw is immi-
real, and he has to decide on more radical nent.
measures: 30 e4!? fxe4 (not 30 ... tiJf6?! 31 Even so, one can understand Karpov not
l:i.c7+ ~d6 32 l:i.xg7 tiJxe4+ 33 Wd3 l:i.c8 34 wanting to leave his pawn on g7.
tiJaS! b6 3S tiJc4+ 'it'c6 36 a4) 31 'it'e3 tiJf6?! 32 31We3
l:i.c7+ Wf8 33 We2! tiJdS 34l:i.d7 (threatening '31 tiJaS was also not bad.' (Tal)
tiJcS) 34 ... b6 3S tiJd4 l:i.e8 36 tiJxe6+ l:i.xe6 37
l:i.xdS l:i.c6 38 Wd2 ~f7 39 l:i.d4, obtaining a
rook endgame with an extra pawn. True,
the active 31...gS 32 fxgS hxgS 33 tiJd2 g4! 34
tiJxe4 l:i.f8 would have been quite sufficient
for a draw.
However, 29 ... Wd6 also does not place
Black under threat of defeat. It is another
matter that many of his moves in this stage
of the game, as Tal put it, 'bear the stamp of
a shortage of time'.
30 e4 (otherwise there is nothing to hope
for)
31 ... e5
31...b6? 32 tiJd4 was bad for Black, but
31...gxf4+!? 32 gxf4 eS came into considera-
tion, in order to activate the rook: 33 tiJaS
exf4+ 34 'it'xf4 tiJeS 3S 'it'fS (or 3S lIh3 b6 36
tiJb3 tiJf7 37 tiJd4 .l:!.c8) 3S ... b6 36 tiJc4+ tiJxc4
37 .l:!.xc4 .l:!.g8!, gaining a draw in the rook
endgame: 38 .l:!.c2 (38 lIc3 .l:!.g2) 38 ... .l:!.g1 39
Wxf6 .l:!.f1+ 40 Wg6 .l:!.f3 41 a4 .l:!.a3 42 as bxaS
43 bxaSl:i.h3 44 'it'f6 .l:!.f3+.
32 fxg5!
The best practical chance, despite the
fact that the f-file is now opened and in the
30 ... g5 future only Black can make use of it. This
Black is intending ... e6-eS and with this exchange, with the aim of weakening the
aim he avoids the possible blockade of his eS-pawn, creates a new and unexpected
g-pawn after 30 ... eS (30 ... b6? 31 tiJd4) 31 fS situation which unnerves Karpov.
(Tal recommended 31 We3, but after 31...b6 If 32 fS Black could have played 32 ... b6!,
there is nothing better than 32 fS) 31...b6 32 when there is no way of creating a break-
'it'd3. However, this would not have been through: 33 a4 hS 34 Wd3 (34 h3 bS) 34 ... g4
fatal: 32 ... l:i.b7 (intending ... l:i.c7) 33 l:i.c8 l:i.b8! 3S bS axbS 36 axbS .l:!.a8 37 .l:!.c6+ We7. And
or 32 ... .l:!.a8 33 tiJd2 bS (but not 33 ... aS? 34 32 tiJaS gxf4+ (32 ... hS 33 .l:!.d3+ Wc7 34 l:i.d2
tiJc4+ We7 3S tiJe3) 34 tiJb3 tiJb6 3S tiJaS, and exf4+ 3S gxf4 .l:!.d8 would also appear to be
now not immediately 3S ... lIc8? 36 .l:!.xc8 suitable) 33 gxf4 exf4+ 34 'it'xf4 tiJeS would
34
The Third Match: 1986
3S ... b6
Not a bad move, but a committal one:
for the sake of counterplay on the f-file,
33 ... g4! Karpov abandons his queenside pawns to
Black must urgently create counterplay, their fate. 35 ... et:Jf6 36 l:!.c5 et:Jd7 37 et:Jc4+ ~e7
by exploiting the f-file and the weakness of (37 ... ~e6? 38 l:!.c7! et:Jf6 39 et:Ja5 b6 40 et:Jc4
the e4-pawn. 'In all probability, this was and wins) 38l:!.d5 ~e6 39l:!.d6+ ~e7 40 l:!.h6
intuition, although at that moment it was l:!.c8 41 ~d3 was advantageous to White.
hard to imagine that it was the advance of However, things are by no means so
his kingside pawns which would save bad for Black: 35 ... ~e6!? was possible, and
Black.' (Tal) If 33 ... et:Jf6 White would have if 36 l:!.c7 (36 l:!.c3 ~d6) - 36 ... l:!.f8! 37 ~e2
gained an obvious advantage by 34 ~c5! et:Jf6, quickly creating counterplay: 38 et:Jxb7
et:Jg4+ 35 ~e2 et:Jxh2 36 et:Jc4+ or 34 ... et:Jd7 35 et:Jxe4 39 ~e3 (39 et:Jc5+ ~d5) 39 ... ~d5 40
et:Jc4+ ~e6 36 l:!.c7 (say, 36 ... g4 37 et:Ja5 b6 38 l:td7+ ~c6 41 l:!.d3 ~xb7 42 ~xe4 l:!.f2 43
et:Jc4 b5? 39 l:!.c6+ ~e7 40 et:Jd6 and wins). ~xe5l:!.xh2 44 ~f4 h4 with equality.
And 33 ... h5 could have transposed after 34 Thus the strictly waiting move by the
l:!.c2 g4. king (which is also the simplest) would
34l:!.c2 have demonstrated that Black's zugzwang
Inviting the opponent to 'declare his in- is in fact an illusion, whereas the unclear
tentions' (the alternative was 34 ~d3!? h5 35 ... b6, although it does not lose, forces him
35 l:!.c1), since Black is in an unusual to become anxious and devise something in
zugzwang position and sooner or later will the time scramble. Objectively, all the fears
be forced to allow the white pieces into his here are exaggerated and White's winning
territory. In time-trouble such a prospect is chances are more psychological than real.
35
Kasparov vs. Karpov: 1986-1987
36 ~c6+ rj;;e7 37 ttJC4 ~f8 38 rj;;e2! concluded the game immediately, but the
Of course! 'To judge by Kasparov's feverish haste of time-trouble continues to
lightning reaction to every move, he had affect matters ... 'Karpov's impulsive move
calculated all this well in advance.' (Tal) provoked an instantaneous reaction by
38 .•. ~f3?? White. Apparently this knight manoeuvre
Surprisingly, this time-trouble blunder was an idea that Kasparov had been plan-
was not severely criticised by the commen- ning long before.' (Tal)
tators, who thought that even after the 39 ... ttJf6!
'best' move 38 ... ~f6 (?) Black would have 39 ... rj;;d8 would have led to a dangerous
lost. And indeed, after 39 ~c8! (nothing real position after 40 ~h6 ttJf6 41 ttJc4 ttJxe4 42
is given by 39 ~c7?! rj;;d8! 40 ~a7 b5 41 ttJe3 ~xh5 ~c3 43 ttJxb6 ~xa3 44 ~xe5 ttJc3+ 45
~c6! 42 ttJd5 ttJf6) 39 ... ~e6 (39 ... b5 40 ttJe3 or rj;;d2 ~b3 46 ttJd5 ttJxd5 47 ~xd5+ rj;;e7 48
39 ... ~f8 40 ~c7 is no better) 40 ~a8! ~c6 41 ~d4 ~b2+ 49 rj;;c3 ~xh2 50 ~xg4.
rj;;d3 b5 42 ttJe3 ~d6+ 43 rj;;e2 he has a 40 ~xb6 ttJxe4 (the time control has passed,
difficult position: 43 ... rj;;f7 44 ~a7 and ttJd5, but ... )
or 43 ... rj;;f6 44 ttJd5+ rj;;g6 (44 ... rj;;g7 45 ~e8)
45 ~g8+ rj;;f7 (45 ... rj;;h6 46 ttJe7) 46 ~d8 rj;;e6
47 ~h8 ttJf6 48 ttJxf6 rj;;xf6 49 ~h6+ rj;;e7 50
~xh5 with an extra pawn and good win-
ning chances.
However, by 38 ... b5! (already the only
move!) Black would have created counter-
play sufficient for a draw: 39 ttJe3 (if 39
ttJd6, then 39 ... ~f6 40 ttJf5+ rj;;d8 41 ~c1
ttJb6) 39 ... ttJf6 40 ttJd5+ ttJxd5 41 exd5 ~f3 42
~e6+ rj;;d7 43 ~xe5 ~xa3 44 ~xh5 rj;;d6 or 40
~xa6 ttJxe4 41 ttJd5+ rj;;d7 42 ~h6 ~f5 43 ~h8
rj;;d6 44 ~d8+ rj;;e6 45 ttJc7+ rj;;e7 46 ~d3 ttJf6
47 ttJd5+ rj;;e6 48 ttJc3 e4. 41 ~xa6
'This move was also made very quickly
by Kasparov,' writes Tal. 'True, in contrast
to last year's encounter, he made it on the
board, and did not seal it openly' (as in the
19th game of the 1985 match). In this way
the turn to move was passed to my oppo-
nent.
Here the game was adjourned. Later I
was reproached for the fact that I not did
think of and did not give the 'winning'
check 41 ttJd5+!? (associating this with my
traditional impulsiveness and nervous-
ness). But it would have been better not to
39 ttJe3?? sink into thought, but to make the knight
Obvious - and unbelievable! 39 ~c7! ~c3 move equally quickly - then on his sealed
(39 ... rj;;e6 40 ~xd7) 40 rj;;d2 would have move Karpov would have had a difficult
36
The Third Match: 1986
choice, since only one of the two king Karpov's sealed move was obvious, and
moves gives Black a definite draw: a painstaking analysis was made of the
1) 41...Wf7?! 42 Mxa6 h4! (42 .. .'~Jc3+? 43 position. But, alas, a win could not be
ctJxc3 Mxc3 44 Wd2, and the b-pawn ad- found. In the event of 41...ctJc3+ 42 'it>d2
vances). I thought that here Black had ctJe4+ 43 Wd3 ctJc3! 44 Wxc3 l:he3+ 45 ~c4
sufficient counterplay, but years later I Me2 Black could also have counted on a
discovered the strong reply 43 gxh4! (43 draw, despite the activity of the white king.
ctJe3 ctJc3+!) 43 ... Mh3 44 h5! Mxh5 45 Wd3 42 'it'd3 ctJd6 43 Ma7+
Mh4! (there is nothing else) 46 Mc6 ctJg5 47 It was established that after 43 a4 Mxh2
ctJe3, and Black is balancing on the verge of 44 b5 J::th3! (not 44 ... h4?! 45 ~a7+ 'it'e6 46
defeat: 47 ... ctJf3 48 b5 ctJe1 + 49 Wd2 ctJf3+ 50 gxh4 g3 47 .l:.g7!) 45 b6 lIxg3 46 ~a7+ 'it'e6
WeI Mxh2 51 b6 ~h1 + 52 Wb2 .l:i.h2+ 53 Wc3 47 b7 ctJxb7 48 .l:txb7 .t:.g1 49 l:i.b6+ 'it'd7 the
ctJd4 54 Mc5 etc. extra piece does not give White a win: 50 a5
2) 41...Wd7! 42 Mxa6 ctJc3+! 43 ctJxc3 Mal 51 a6 g3 52 l::tg6 h4 53 a7 h3 or 50 Mg6
MXc3, and White cannot win the rook g3! (50 .. .l::ta1 is worse in view of 51 ctJc4! e4+
ending: the black king is close to his passed 52 'it'c3) and ... h5-h4 with a draw.
pawns, and the rook will attack the h2- 43 ... <;¥;>e6441::th7
pawn. A curious moment, especially in the
light of the mysterious and inexplicable
events which occurred in this match.
Analysis diagram
37
Kasparov vs. Karpov: 1986-1987
46 ... .l:ta2 47 tDd5+ 'it'd7 48 .l:txh5l:ta3+ 49 tDc3 2) 45 ... .l:txh2. Essential: according to Tal,
e4+ 50 'it'd4 and wins), to which we devoted 'Black's counterplay on the kingside should
our main attention before the resumption. save the game for him'. But if this is so,
White is unable to gain an advantage after then literally only by a miracle - 46 .l:th6+
either 46 a4 (Tal) 46 ... .l:ta2 47 .l:txh5 .l:ta3+, or (46 tDxg4? .l:td2+) 46 ... 'it'd 7 47 'it'c5 tDf7 48
46 tDxg4 e4+ (Tal gives only 46 ... .l:ta2? 47 .l:th7 'ite6 49 b5, for example:
tDxe5+), for example: a) 49 ... .l:te2 50 tDc4 Mc2 51 'it'b4! (after 51
1) 47 'it'e3 .l:ta2 48 tDe5+ (48 tDf6+ 'it'd8 'it'd4 ltg2 52 b6 Black is saved by 52 ....:xg3
with a draw) 48 ... 'it'e7! 49 .l:txh5 .l:txa3+ 50 53 a4 Mb3 and ... g4-g3) 51...llg2 52 b6 e3! 53
'it'f4 e3!? 51 tDc6+ 'it'd7 52 tDd4 .l:td3 53 .l:th7+ .l:txh5 with the idea of Mh1 or 53 tDxe3 Mxg3
'it'e8 54 tDc6 (54 tDe2 .l:tb3 or 54 'it'e5 tDb5! is 54 tDc2, and White retains winning chances;
no better) 54 ... Mc3 or 53 'it'e5 e2 54 .l:th7+ b) 49 .. .1la2! 50 b6! .l:txa3 51 tDc4 Mb3 52
tDf7+! 55 Mxf7+ 'it'e8 with a draw; tDd2 .l:tc3+ 53 'it'b4 Md3, and although
2) 47 'it'c3 .l:th3 48 tDf6+ (48 tDe5+ 'it'e7!) Black's position is dangerous, I have not
48 ... 'it'e7 49 tDxh5 tDf5! 50 .l:th7+ 'it'd6 or 49 found a win for White: 54 tDxe4 .l:td4+ 55
tDd5+ 'it'd7 50 .l:txh5 Mxg3+ 51 'it'd4 .l:txa3 52 'it'a5 .l:txe4 56 b7 .l:te1 with a forced draw (57
.l:th7+ 'it'c6 53 .l:tc7+ 'it'b5 54 .l:tc5+ 'it'a4 with a b8'iV .l:ta1+ 58 'it'b6 Mb1+ 59 'it'a7 Mxb8 60
draw. 'it'xb8 tDg5 61 lhh5 tDe4) or 54 tDb3 'it'f6
(this enables Black to save the game by a
single tempo) 55 b7 .l:td8 56 Mxh5 e3 57 tDc5
tDd6 58 .l:th6+ 'it'e5 59 .l:te6+ 'it'd5 60 Mxe3
.l:tb8 61 .l:te7 'it'c6 with a draw.
To this day I don't understand why I
didn't play 45 'it'd4 - to all appearances, I
was confused by the provocative move
44 ... e4+?!.
45 ... tDb5+ 46 'it'C4
45 'it'c3? (12)
Missing a good winning chance - 45
'it'd4! would have set Black a difficult
choice:
1) 45 ... .l:td2+? 46 'it'c5! .l:td3 47 Mh6+! (47
tDc4? .l:td5+! 48 'it'c6 tDf5) 47 ... 'it'd7 48 tDc4!
(Tal; 48 tDg2 .l:tc3+ 49 'it'd5 Md3+ 50 ~e5 e3 is
unclear) 48 ... tDxc4 49 'it'xc4 Mxa3 50 .l:txh5
'it'e6 (50 ... e3 51 Me5 'it'd6 52 .l:te4 and wins)
51 .l:tg5! (in the book Ova matcha only 51 46 ... tDxa3+
'it'd4? .l:tb3! is given) 51...e3 52 .l:tg8! with a Now the game quickly ends in a draw.
won rook endgame: 52 ... 'it'f6 53 Me8 'it'f7 54 47 'it'd4 (also nothing is promised by 47 'itc5
.l:te5! 'it'f6 55 Me4 'it'f5 56 .l:te8 'it'f6 57 b5 .l:ta4+ .l:te2!) 47 ... .l:txh2 48 .l:th6+ 'it'd7
58 'it'c5 Ma3 59 'it'b4. Simpler than 48 ... 'it'f7 49 'it'e5 tDc2! 50
38
The Third Match: 1986
.i::i.f6+ 'it'g8 51 tiJf5 e3 with a draw. easy life in quiet positions! I think that
49 tiJd5 games such as this one had a definite
A last flicker of activity - a vain attempt psychological effect: they induced Karpov
to restore the lost domination. to play more energetically. As we will see,
in both this and our later matches he would
cease 'without fail' exchanging the queens
and would aim for a complicated fight.
In the 3rd game the Griinfeld Defence
again enabled me to gain a draw with Black
without any particular problems.
Game 3
A.Karpov-G.Kasparov
World Championship Match,
3rd Game, London 01.08.1986
Grunfeld Defence 079
49 ... h4 (the most radical) 50 .i::i.xh4 (50 gxh4
g3) 50 .. Jhh4 51 gxh4 g3 52 tiJf4
52 tiJe3 t"Llc2+! would not have changed 1 d4 tiJf6 2 c4 g6 3 tiJf3 (3 g3 c5 - Game
anything. No.25; my opponent most often played 3
52 ... tiJC2+ Yz-Yz tiJc3) 3... .Itg7 4 g3
Draw agreed on White's proposal. After After the stormy events of the 2nd game
the game, in an interview with the English Karpov wants to gain a slight plus, without
press, Karpov stated that in the adjourned the traditional risk inherent in the main
position the chances of White winning and variations of the Griinfeld Defence. Natu-
of Black drawing were equal. At the time I rally, we expected this variation: here
regarded this as merely a psychological Black's chances of counterplay are limited.
ploy, but now I think that my opponent 4 ... c6
possibly did not in fact see the clear draw The main drawback to the g2-g3 varia-
after 44 ... Mxh2!. tion is that Black can choose a symmetric
Times: 3.22-3.26. set-up, where White does not have very
many possibilities. Such a rather passive
On the whole I was happy with this but solid course for Black is typical of
game, since, although Karpov solved his match play. Later in tournaments I also
opening problems, for a long time he defended the 'pure' Griinfeld against
suffered in an ending and just before the Karpov - 4 ... d5 5 cxd5 tiJxd5 6 .Itg2 tiJb6
time control he could have lost. For him (Amsterdam 1988) or 5 .Itg2 0-0 6 0-0 dxc4 7
this was a new, uncustomary situation - a tiJa3 c3 8 bxc3 c5 (Frankfurt (rapid) 1999;
depressing fight for a draw in a position Linares 2001).
where there was hardly any material left 5 .Itg2 d5 6 cxd5
(usually it was his opponents who had to Nowadays 6 'iYb3 is fashionable, or, with
do this). Such a turn of events must have the inclusion of castling by both sides, 7
given him serious food for thought. It 'iYb3. But for a long time it was thought that
turned out that he would no longer have an the exchange on d5 gave White the best
39
Kasparov vs. Karpov: 1986-1987
40
The Third Match: 1986
9 ... ttJfd7 10 ttJf3 Openings (1976) gave the old game Pod-
10 f4 (Game Nos.13, 28, 30) is more logi- gaets-Marszalek (1974), which after
cal, and after the 13th game for a long time 12 ... 'iVe7 13 'iVd2 favoured White. But in
it became the main line. By retreating his Romanishin-Andersson (Indonesia 1983)
knight, Karpov probably wanted to retain Black equalised with 12 ... ttJe4 13 ttJeS?!
some initiative and probe his opponent: ttJxc3, although in the event of 13 ttJxe4
how would he defend in a slightly inferior dxe4 14 ttJeS White would nevertheless
position without active counterplay? How- have retained a slight plus. 12 ... ~d7 is
ever, this plan did not produce any real quieter (this is probably what I would have
effect. played): 13 ~d6 ~e8 14 e3 with light pres-
10... ttJc6 11 ~f4 sure for White and ... a solid position for
Black!
12 ... ~d7 13 'iVd2
11 ... ttJf6
A rather solid continuation. This posi-
tion has also arisen many times via a differ- 13 ... ttJxes
ent move order - for example, 8 0-0 e6 In the variation 13 ... ttJhS 14 ~gS f6 IS
(instead of 8 ... ttJe4!) 9 ~f4 ttJc6. ttJxc6 ~xc6 16 ~h6 I did not like my knight
11...~6!? also comes into consideration. on the edge of the board. Another attempt
Only, after 12 ttJa4 Black should not reply to put pressure on the centre is more inter-
12 ... 'iVaS 13 ~c1 bS (Akhmilovskaya- esting - 13 ... ~6!? 14 ~e3! ~fc8 IS ~ac1
Chiburdanidze, 12th match game, Bor- ~e8 16 f4 (16 f3 ttJd7), 'and Black cannot
zami/Sofia 1986) because of 14 ~xc6 bxa4 IS solve the problem of the knight on eS
'ii'd3 ~b7 16 ~d2 with a slight advantage to without detriment to his position' (Dva
White, but 12 ... ~S! 13 ~c1 ttJb6 with matcha), although after 16 ... ttJe7 17 g4 ttJd7
complicated play: White can win the queen 18 ~f2 f6 19 ttJf3 ~f7 this detriment is
- 14 ttJc3 'iVxb2 IS ~d6 ~d8 16 ~c2, but insignificant: 20 gS ttJfS 21 gxf6 ttJxf6 22
after 16 ... 'iVxc2 17 'iVxc2 ~xd6 18 e3 ~d8 ttJeS ~e8 with complicated play. White
Black gains full compensation for it and the may be slightly better, but if he wants to
play becomes double-edged. play for a win, he also cannot manage
12 ttJes without creating weaknesses.
Of course, immediately occupying the 1411..xes
central square. 12 ~c1!? is also not bad - the It is not possible to gain an advantage
first edition of the Encyclopaedia of Chess by 14 dxeS?! ttJg4 IS e4 on account of the
41
Kasparov vs. Karpov: 1986-1987
strong reply 15 ... d4! 16 'ii'xd4 .ic6 17 promise any advantage after 18 .. :~8 19
'ii'c5(d6) 'ii'b6. 'ii'e5 'ii'xe5 20 dxe5 CiJd7 21 f4 lbc5. There-
14 ... .ic6 fore Karpov continues manoeuvring.
17 ... lbf6 18 'ii'f4
If 18 f3, then 18 ... ~d6! with the same
idea of 19 e4 dxe4 20 fxe4 e5.
18 ...'ii'b8 19 'iixb8
After 19 'ii'e5 'ii'xe5 20 dxe5lbg4 21 f4, as
recommended by the commentators, there
is simple equality after 21...lbe3 22 l:td4
lbxg2 23 'it>xg2 f6 24 exf6+ 'it>xf6.
19 ... l:taxb8 20 f3
Beginning to threaten the advance of the
pawns - both e2-e4, and b2-b4-b5. But
Black's position is solid.
20 ... l:tfd8 21 'it>f2
15 ~fdl The immediate 21 e4 is parried by the
15 l:tac1!? was more natural and more standard 21...dxe4 22 fxe4 e5! 23 d5 (23
useful, with the idea of 15 ... l:tc8 16 l:tc2 and dxe5lbg4) 23 ... .id7 and ... lbe8-d6.
l:tfc1, although here too after 15 .. .tiJe4 16 21 ... l:tbc8 22 e3lbe8 23 l:td2 lbd6 (a success-
lbxe4 .ixe5 17 dxe5 dxe4 18 'ii'f4 'ii'd4 19 ful piece arrangement) 24l:tdc2
l:tc2 (19 h4?! is incorrect in view of
19 ... 'ii'xb2 20 h5 'ii'xe2! 21 h6 'iVh5 22 g4 g5)
19 .. .f6 20 exf6 l:tad8! Black can hope for
equality: 21 e3 'ii'xf6 22 'ii'xf6l:txf6 23 b4l:tf7
24 h4 h6 25 l:tfc1l:tfd7 etc.
The move in the game, seemingly multi-
purpose and 'purely Karpovian' (it ex-
cludes 15 ... lbe4, and gradually prepares f2-
f3 and e3-e4), proves on verification to be
unnecessary.
ls ... lbd7
Immediately exchanging the dark-
square bishops. 15 ... 'ii'e7 16 l:tac1 l:tfc8 17
l:tc2 lbe8 was also acceptable - White's 24 ... 'it>f8?!
slight pressure is perfectly bearable: 18 A routine move, playing the king to the
.ixg7 'it>xg7 19 l:tdc1 lbd6 20 'ii'f4 l:td8 21 centre (although there is nothing for it to do
'ii'e5+ f6 22 'ii'e3 lbf5 23 'ii'd2 l:td7 24 .ih3 at e7). The immediate 24 ... f5! was more
lbd6 25 b3 ~e8 etc. accurate, not fearing the weakening of the
16 .ixg7 'it>xg7 17 l:tacl e5-square, killing the idea of e3-e4 and
If 17 f3 (the only genuinely active plan) beginning to seize space: 25 .ifl (25 h4 h6,
there would also have followed 17 ... lbf6, while after 25 f4 White creates a weak
when 18 e4 runs into 18 ... dxe4 19 fxe4 e5! 20 square for himself - e4) 25 ... g5 26 .id3 'it>f6
d5 (20 dxe5 'ii'b6+ and ... lbg4) 20 ... .id7 with with comfortable play and imminent ex-
a good game for Black. 18 'ii'f4 also does not changes on the c-file.
42
The Third Match: 1986
43
Kasparov vs. Karpov: 1986-1987
he would have to play more sharply ... the solidity of Black's defences would not
In the 4th game I gained my first win, of yet be obvious either to me, or to Karpov.
which I had every right to feel very proud - 4 ... c5 5 g3 cxd4
an unusual decision in the middlegame and For the first time Black chooses the main
the confident conversion of an advantage in line of the variation, instead of 5 ... ttJe4
the ending enabled a complete game to be (Game No.53 in Kasparov vs. Karpov 1975-
produced. 1985, 5 ... ttJc6 6 ~g2 ttJe4 (Game Nos.65, 69 in
Incidentally, this game saw the start of a Kasparov vs. Karpov 1975-1985) or 6 ... d5!?
unique opening duel: the two players (Game No.2). To judge by Karpov's deter-
constantly duplicated each other's analysis, mined appearance, there could be no doubt
including analytical mistakes! As a result I that here he had done some serious prepa-
gained the strange feeling that for the ration.
positions reached on the board Karpov and 6 ttJxd4 0-0 (Black is not enticed by the
I had prepared together. Of course, fore- fashionable 6 ... ttJe4 7 'ii'd3 'ii'a5) 7 ~g2 d5
sight and successful 'spot-on' preparation Whereas in our second match Karpov
happen to everyone, but here we were did his utmost to exploit the weaknesses on
talking about the almost complete coinci- White's queenside, in the third one, as we
dence of all analyses and assessments. see, he plays completely differently, dis-
Karpov suffered from the same 'chess carding the largely mythical idea of win-
blindness' as I did, although sometimes he ning the white c-pawns (which possessed
nevertheless found 'holes' in my analyses, him during the entire second half of our
with sad consequences for me. We will previous match) in favour of free develop-
examine this 'insight phenomenon' on the ment.
basis of numerous specific examples.
Game4
G.Kasparov-A.Karpov
World Championship Match,
4th Game, London 04.08.1986
Nimzo-Indian Defence f20
44
The Third Match: 1986
Analysis diagram
45
Kasparov vs. Karpov: 1986-1987
ctJd2 White has some advantage (Ivanchuk- regains his pawn, retaining some pressure
Topalov, Wijk aan Zee 2006). But after (Ubilava-Groszpeter, Sochi 1984);
10 ... d4 (10 ... dxc4 is also suitable) 11 ~aS 2) 10 ... ctJbd7 11 ctJbS ttJb6 12 0-0 .td7 13
'iVe8! 12 .tgS ctJc6 13 ~c7 ~d7! (a typical l:tdl ttJfdS 14 ctJd6 "iVf6 IS l:td4 .tc6 16 e4
defence) 14 ~xd7 ctJxd7 IS l:tc1 as! Black with a complicated fight (Karpov-Portisch,
has no reason for complaint (Kamsky- Luzern 1985), which occurred after our
Christiansen, USA Championship 1993). second match: evidently the ex-champion
was trying to adapt to positions of this
type;
3) 10 ... ttJdS 11 0-0 .td7 12l:tdl ~c8 13 e4
ttJb6 14 .tf4 l:te8 IS .td6 eS 16 ctJfS ctJc6 17
ctJe3 .th3 18 .tcS, regaining the pawn with
a comfortable game (Lerner-Suba, Moscow
1986).
The character of the play in these games
appealed to me and could hardly satisfy
Black. Therefore in our preparations for the
match we looked for other ways of defend-
ing. Black's main problem in the original
position is the development of his queen-
I will explain the story of how Black's side pieces, and an attempt to solve it
next move came into being. Initially 9 ... eS suggested to us the idea of exploiting the
10 ctJbS dxc4 was played here, and it tran- development of the queen at b3 by 9 ... ctJc6!
spired that after 11 'YWxc4 a6 12 ctJc7 l:ta7 13 10 cxdS ctJaS!.
ctJdS ctJxdS 14 ~xdS ~c7 (Holzl-Portisch, But my opponent too was not idle, espe-
Luzern Olympiad 1982) or even 14 .. .'iVxdS cially since one of his trainers, Ubilava, was
IS .txdS bS the chances are equal. Then the 'godfather' of the 8 'iVb3 variation.
Georgian players introduced 11 ~a3!, and 9 ... ctJc6!
after the game Georgadze-Polugayevsky
(Moscow 1983), which went 11...ctJc6 12
.te3! .te6 13 l:tdl ~8 14 0-0 ~d8 IS ~cS
l:txdl?! 16l:txdl a6 17 ctJd6 ctJd718 ~a3 ~c7
19 "iVa4!, the evaluation of the variation
changed in favour of White.
Black had to seek a new way - 9 ... dxc4,
but here too 10 ~a3! (inviting 10 ... eS 11
ctJbS) is possible. It was around this gambit
idea that the main discussion developed.
For the pawn White has sufficient compen-
sation, and the games played confirmed
this assessment:
1) 1O ... ~6 11 O-O! (11 .te3 "iVa6! Ubi- Karpov made this move with lighting
lava-Polovodin, Tallinn 1983) 11 ... ttJbd7 12 speed - apparently, with the aim of stun-
.te3 ctJdS 13 l:Iabl 'ifcs 14 'iVxcs ttJxcS IS ning me. It was evident that the creative
.txdS exdS 16 ctJbS ctJe6 17 l:tfdl, and White process of seeking the strongest continua-
46
The Third Match: 1986
hon for Black had proceeded in parallel... there is 14 ... eS IS fxeS tLlac4 with equality.
10 cxdS It was this that gave rise to the unusual
There is nothing else: after 10 tLlxc6? move 12 1i'd3, which sets Black certain
bxc6 11 0-0 .i.a6 only White may have tactical problems. However, a mistake had
problems - he has reinforced the oppo- crept into our home analysis - a phenome-
nent's pawn centre, while he himself still non which was not unusual in the pre-
has weaknesses and is behind in develop- computer era.
ment.
10 ... tLlas!
The crux of Black's idea. 10 ... exdS is
weaker on account of 11 O-O! (but not 11
l,gS tLlxd4 12 cxd4 1i'aS+ Szekely-Hazai,
Kecskemet 1983) 11...l::!.e8 12 .i.gS (and if
12 ... tLlxd4 13 cxd4 l::!.xe2, then 14 l::!.fel) or
1l ... h6 12 .i.f4 (Jobava-Sturua, Tbilisi 2001)
with appreciable pressure for White.
12 ... .i.d7?
It is paradoxical, but true: this seem-
ingly natural move, which in our prepara-
tions we considered to be best, turned out
to be second-rate, and the genuinely best
reply 12 ... 1i'c7!, which was subsequently
recommended by the commentators, was
not seriously examined by us! After briefly
assessing 13 tLlbS 1i'c4? 14 1i'xc4 tLlxc4 IS
ll1i'c2 tLlxdS .i.xdS exdS 16 tLlc7, we did not return again
'With the c-file open and his pieces able to this line, but later it transpired that
to develop normally, the weak c3-pawn 13 ...1i'c6! is far stronger, and if 14 .i.a3, then
illone should be sufficient for Black to gain 14 ... l::!.d8 IS e4 a6 or IS ... tLlc4 (Izeta-Smagin,
good counterplay', I wrote in Revolution in Novi Sad 1986) with easy equality. Nothing
:Ize 70s. Of course, we realised that this is also given by 14 a4 a6! (Shchukin-Aseev,
position looked very promising for Black, St. Petersburg 1998) or 14 0-0 .i.d7 IS a4 a6!
ilnd yet we had found a possibility of 16 e4 axbS 17 exdS1i'c4.
fighting for the initiative. Soon in a game with Suba (Dubai
121i'd3!? Olympiad 1986) I again chose this varia-
It was clear that after 12 0-0 Black had tion, hoping that my opponent would not
two good replies: 12 ... .i.d7 13 ~a3l::!.e8 14 c4 know the latest recommendations, but he in
:c8 (Piskov-Zakharov, Lvov 1986) or 13 e4 fact replied 12 ...1i'c7!. I decided to avoid 13
:':Jb6 14 eS 1i'c7 (Adianto-Tisdall, Norway tLlbS 1i'c6! and take a risk (how else could I
1992), as well as 12 ...1i'c7!? 13 e4 tLlb6 (Bar- play for a win?): 13 0-0 .i.d7 14 e4?! (14
lov-Stone, New York 1988), and if 14 f4 .i.xdS exdS IS .i.f4 was more solid) 14 ... tLlb6
47
Kasparov vs. Karpov: 1986-1987
15 f4, but after 15 ... e5 16 fxe5 tbac4 17 g4?! The commentators suggested 13 ... tbb6!?,
tbxe5 18 ~g3 tbbc4 19 .i.f4 ~c5 I ended up but then too after 14 c5 tbbc4 15 a-a! l:!.c8 16
in an inferior position and in a very diffi- tbb3 .i.c6 17 .i.xc6 lhc6 18 Md1 (18 ~c3 b6
cult struggle I gained a draw only by a 19 .i.f4 e5 20 l:!.fd1 ~e8 is unclear) White
miracle. would have had somewhat the better
This was a very important moment in chances, for example: 18 ... ~xd3 19 exd3
the match. 12 ... ~c7, solving all Black's Ma6 20 Mb1 tbxb3 21l:!.xb3 tba5 22 l:!.a3 tbc6
problems, was overlooked not only by me, 23 l:!.xa6 bxa6 24 .i.e3 etc.
but also by Karpov! We both considered
the best move to be 12 ... .i.d7, after which
White retains the initiative. I was unaware
of the improvement right up to the time
when we moved to Leningrad, when an
exchange of information occurred and one
of the trainers said to me: 'In that line there
is 12 ... ~c7 13 tbb5 ~c6!'. To all appear-
ances, Karpov too did not immediately
learn about this: right to the end of the
London half of the match - in the 8th, 10th
and 12th games - he avoided the Nimzo-
Indian Defence, answering 1 d4 only with
1...d5. In our home analysis we had planned 14
Thus we proceeded along parallel ana- .i.a3 here, in order after 14 ... l:!.c8 to play 15
lytical courses. Our 'unseen dispute' in this Mel 'with advantage', but at the board I
game concluded two moves later. .. saw that 14 ... l:!.e8 15 0-0 tbec6! was far
stronger - Black provokes favourable
simplification and easily equalises (16 tbf3
e5 17 ~c3 .i.e6). Karpov was clearly aiming
for this position! What was I to do? After
some thought, I devised a stronger move.
140-0!
For the present it is too early to bring
out the bishop to the vulnerable a3-square-
it may also be developed at g5 or at e3, and
it may also remain at el.
From this moment on both players be-
gan acting spontaneously. And Karpov also
sank into thought: Black has certain prob-
13 c4 lems to solve.
13 .i.a3?! is premature in view of 14 ...l:!.c8?!
13 ... l:!.e8 14 tbf5 (14 c4 tbxc4!) 14 ... exf5 15 This first move of his' own' is seemingly
.i.xd5 .i.c6 16 l:!.d1 (Ftacnik-Palkovi, the most consistent (it attacks the c4-pawn),
Bundesliga 1995) 16 ... ~d7, and if 17 c4, but it is not the best. If 14 ... tbec6 both 15
then 17 ... tbxc4!. tbf3 e5 16 tbg5 and 15 tbb3 are unpleasant,
13 ... tbe7 since after 15 ... tbe5 16 ~c3 tbexc4 17 tbxa5
48
The Third Match: 1986
~xa5 18 .tb2 f6 19 l:!.fdl White has strong the direction of his activity. The knight at c4
pressure for the pawn. is a strongpoint for Black, with which his
The best chance was 14... .tc6!? (a rec- hopes of counterplay are associated.
ommendation of Gufeld). It is psychologi- From this point it is interesting to follow
cally not easy to present the opponent with how with every move White confronts his
two bishops against two knights - 15 tiJxc6 opponent with direct threats.
~exc6, but here the knights coordinate well, 17 ... tiJes
whereas it is hard for the white bishops to The little tactical trick 17 ... ttJg6 (with the
expand their influence (say, 16 .itf4 'YWxd3 17 idea of 18 .txc4 ttJe5) runs into a strong
exd3l:!.fd8 18 l:!.fdl l:!.d7 and ... l:!.ad8). There- rejoinder - 18 f4!.
fore in my preparations for the 8th game I
was intending 15 .ta3! .txg2 16 ~xg2,
retaining the initiative (16 ... l:!.c8 17 tiJb5!).
'Everywhere White is a little better' -
this was the conclusion we arrived at in
London, not yet being aware of 12 ... 'YWc7!,
and we tirelessly analysed the position after
14 0-0. However, it was never to occur
again ...
is tiJb3!
A very strong move, forcing the oppo-
nent to agree to the exchange of the weak
c4-pawn for the pawn on b7. At the same
time White not only rids himself of a 18 'YWe3
chronic weakness, but he also evicts the 18 'YWd6 also came into consideration -
black knight from c4 with an unusual 18 ... .itc8? is bad because of 19 'YWxe5! .txa6
bishop manoeuvre. 20 .itb2 ttJf5 21 e4! .itxfl 22 exf5, and also
lS ... tiJxC4 16 .txb7 l:!.C7 after 18 ... l:!.c6?! 19 'YWxe5l:!.xa6 20 ttJc5l:!.a5 21
.td2 l:!.b5 22 a4 Black has a difficult posi-
tion. However, it is unclear whether White
achieves anything after 18 ... ttJ7g6, for
example: 19 .ite3 .tc8 20 'YWxd8l:!.xd8 21 .tb5
tiJe7 with a slightly inferior, but perfectly
tenable ending.
18 ...ttJC4?
A tempting, but in fact faulty move. It
was correct to play 18 ... f6! 19 .ta3, when in
the event of 19 ... l:!.f7?! 20 l:!.ac1 the white
pieces, up till now rather slow to develop,
would have become very active, but after
the accurate reply 19 ... .tc8! White's advan-
17.ta6! tage would have remained minimal: 20
The light-square bishop has fulfilled its l:!.fdl ttJd5 21 .txc8 'YWxc8 22 l:!.xd5 ttJc4! 23
mission on the long diagonal (17 .te4?! tiJg6 'YWc3 exd5 24 .txf8 'YWxf8 25 'YWd3 'YWd6 26 .l:!.dl
does not give anything real) and it changes ttJb6 etc.
49
Kasparov vs. Karpov: 1986-1987
23 ~fd1!?
A psychologically unpleasant move for
Black - after the forced exchange of queens
It seemed to me that Black was obliged the rhythm of the play changes. 23 'iVf3
to play 19 ...'iVa8 20 'iVxa8 ~xa8, and al- suggested itself, but after 23 ...'iYb6 24 ~fd1
though the evaluation of this endgame is ~d8! for the moment no direct gains are
clear - White's advantage is undisputed - evident. In the given specific instance it is
after 21 .i.g5 lbd5 (21...f6 is inferior: 22 with the queens off the board that the main
~ac1! fxg5 23 l:tfd1! 'it'f7 24 ~xc4) 22 ~fc1 drawback of Black's position becomes
lbd6 it is still possible to hold on. However, apparent - the insecure, poorly coordinated
21 .i.f4!? is stronger, and only if 21...e5 - 22 placing of his pieces. It is very important
i/..g5 f6 23 ~fc1 fxg5 (23 ... .i.e6? 24 .i.e3 leads that White is able to gain complete control
to loss of material) 24 ~xc4 ~xc4 25 .i.xc4+ of the d-file.
'it'f8 26 lbc5 with a great advantage. 23 .. :iVxd3 24l:txd3l:te8
Therefore it is hard to say which reply is After 24 ... ~a6 25 lbc5 ~c6 26 ~b1 ~e8 27
better - 19 ...'iVa8 or 19 ... lbd6. In both cases ~b7 or 24 ... l:tc2 25 l:td7 l:te8 26 lbd4 l:tc4 27
Black already has a difficult position. l:tc1 l:txc1 + 28 .i.xc1 lbd5 29 .i.d2, despite
20 'iYd3! the pawn symmetry and the limited mate-
The three-move dance has concluded to rial, White has considerable winning
White's clear advantage: the queen has chances.
returned with honour to its chosen square, 25 ~ad1 f6?
whereas the black knight is not destined to Again seemingly the most natural move
do the same. Apparently Karpov was (an escape square for the king), but it does
hoping to solve his problems thanks to the not take into account the concrete nuances
symmetry of the pawn structure, but he is of the position. For ... e6-e5 Black lacks just
let down by his 'hanging' knights. one tempo, and he loses.
50
The Third Match: 1986
2S ... h6?! 26 :dS! was also bad for Black, Black does not have time to capture the a2-
but 2S ... g6 would have enabled him to pawn: 29 ... ttJb6 30 .i.xb6 axb6 31 ~d7 g6 32
avoid loss of material - 26 e4 l:tc2 27 .l:l.dS ~c7 ttJb4 33 l:te4! ttJdS 34 .l:l.g7+ 'it>hS 3S l:!.h4
.l::.xdS 2S l:txdS+ ~g7 29 ~eS with the hope hS 36 .l:!.xg6 ~h7 37 .l:!.gg4 'it'h6 3S ttJd4 :l.eS
of exploiting the weakness of the dark 39 l:tgS and wins.
squares: 29 ...11c7 30 eS! .l:!.d7 31 ttJcS l:tc7 32 28 ttJbS! .l:!.e6
ttJe4 or 29 .. .fS!? 30 ttJd4 (30 ..Iixe7 'it'f7! is 2S ... .:xa2 would also not have saved
unclear) 30 ....:c4 31 ttJf3 fxe4 32 ttJgS, and Black in view of 29 ttJc7 ~c2 (29 ... J:tfS 30
although Black is far from lost, his position ttJxe6 .l:l.eS 31 ttJc71HS 32 ~e3) 30 ttJxeSl:!xcS
is very dangerous. That is also the case after 31 :l.dS 'litf7 32li!.ld7.
2S ... ttJdS!? 26 e4 ttJf6 (26 ... ttJc3? 27 l:tdS) 27 29..1ixe7!
ttJd4 (27 ~dS!?) 27 .. .11c4 2S f3 ttJb6 29 ttJbS Another non-routine decision, based on
l:tc2 30 .l:!.dS h6 31 .l:!.xeS+ ttJxeS 32 ':dS ':cS concrete calculation: now Black loses a
33 l:.xcS ttJxcS 34 eS etc. pawn in an unfavourable situation.
But here it would still have been possi- 29 ... ttJxe7 (29 .. Jhe7? 30 l:.dS+ 'it>f7 31 ~xcS)
ble to defend, whereas now the issue is 30l:i.d7
decided unexpectedly quickly.
26 ttJd4! ~b6
Or 26 ... .:a6 27 ttJbS. The sacrifice of the
e-pawn by 26 ... ~c4 27 ttJxe6 ttJc6 2S ttJd4
was no worse than the game continuation,
where the a-pawn was lost, but for the
moment it is not clear why Black should
give up material.
27..1ies
Unexpectedly it transpires that the rook
has nowhere to go.
51
Kasparov vs. Karpov: 1986-1987
ct:Jd6 Ma2 37 l:tb5 Md8 38 ct:Je4 and wins. 41...Md6 (41...ct:Jc5 42 l:tc7) 42 ct:Je7 Md1 43
34~g2 e5 Ma8 ~h7 44 a7 Mal 45 ct:Jc6 ct:Jg5 46 Me8.
The exchange of rooks should have been Times: 2.39-2.29.
delayed with 34 ... MC2!, although after 35
Mb1 all the same White would achieve his Some commentators wrote that 'Kas-
aim: 35 ... e5 36 a5 Mc5 37 ct:Jc3 Mxb1 38 ct:Jxb1 parov won an excellent game in Karpovian
ct:Je6 39 ct:Jd2 and wins. style'. There is no doubt that my style and
35 Md3! Karpov's are markedly different - I have
With the idea of Mc3 (35 Md6 Mc2 is less always preferred a complicated, dynamic
convincing). The exchange of a pair of battle. I was flattered by comparisons with
rooks deprives Black of his saving hopes. Karpov in technique, but for such a com-
35 ... ~h7 parison my play in the 2nd game would
After 35 ... ct:Je6 there would have fol- have been more appropriate. Here, in the
lowed 36 Md6 Mxd6 37 ct:Jxd6 - Black is fourth game, I made clear and strong
unable both to defend his weakened king- moves, in accordance with the demands of
side, and to prevent the advance of the a- the position - and in this, of course, Karpov
pawn: 37... Mb2 38 a5 l:ta2 39 a6 ~h7 40 Me7 does not have a monopoly. Even today I
ct:Jd8 41 Md7 ct:Jc6 42 ct:Je8 Mxa6 43 Mxg7+ consider this game, conducted consistently
~h6 44 Mf7. and energetically by White from beginning
36 MC3 Mbc8 (forced) 37 Mxc6 Mxc6 38 ct:JC7! to end, to be one of the best that I played in
ct:Je6 39 ct:Jd 5 our long-running duel.
White activates his pieces to the maxi-
mum, at the same time creating tactical First Disaster
threats. Alas, the scores were immediately lev-
39 ... ~h6 40 as e4 elled ... The character of the play in the 1st
Here the game was adjourned. and 3rd games had convinced me that my
choice of the Griinfeld Defence was correct.
But after Karpov's defeat in the 4th game it
could be expected that he would adopt a
more effective weapon against this defence.
And that is what happened: in the 5th
game, with a serious of strong and confi-
dent moves, my opponent refuted the
audacious plan prepared by me, which was
based on an analytical mistake.
GameS
A.Karpov-G.Kasparov
41a6! World Championship Match,
The sealed move. Accuracy to the end - 5th Game, London 06.08.1986
the attempt by Black to create threats to the Griinfeld Defence 082
white king are clearly too late. Karpov
resigned the game without resuming (1-0).
This decision was by no means premature: 1 d4 ct:Jf6 2 c4 g6 3 ct:JC3 d5 4..if4
52
The Third Match: 1986
In contrast to the 1st game, which went .l:!.cS as!, and 18 ..txaS? is bad because of
-± tLlf3 ..tg7 S ..tf4, Karpov immediately 18 ... ..tc3! (Vallejo-Mikhalevski, Calvi a
brings out his bishop to f4, intending e2-e3. 2007).
This old variation began to come into 6 dXc5 'ii'a5 7.l:!.C1
fashion after the 14th game of the Euwe- A variation which has been popular
Alekhine match (Groningen 1935), which since the time of the sensational 2nd game
continued 4 ... tLlhS? S ..teS f6 6 ..tg3 tLlxg3 7 of the Petrosian-Fischer match (Buenos
hxg3 c6 8 e3 (Game No. 13 in Volume II of My Aires 1971). A forced draw results from 7
Great Predecessors). Simple development cxdS (7 tLlf3 tLle4!) 7 ... tLlxdS 8 'ii'xdS ..txc3+ 9
seemed to be the soundest reaction to bxc3 'ii'xc3+ 10 <;t>e2 'ii'xal 11 ..teS 'ii'c1
Black's opening escapade, and this was (Black can also try fighting on with
played by Levenfish, Reshevsky, Flohr and 11...'iVbl!? 12 ..txh8 ..te6 13 'ii'd3 'ii'xa2+ 14
Capablanca. <;t>f3 f6) 12 ..txh8 ..te6 13 'ii'xb7 'ii'c2+ (Filip-
4 .....tg7 5 e3 Pachman, Bucharest 19S4).
7 'ii'a4+ 'ii'xa4 8 tLlxa4 has also occurred -
an attempt to gain an advantage in the
ending. I three times defended Black's
position against Timman - twice in Am-
sterdam (1988), and then in Belfort (1988),
and on the whole the opening duel con-
cluded in my favour: after 8 ... 0-0 9 tLlf3 tLle4
10 ..teS ..td7 11 tLlc3 tLlxc3 12 bxc3 dxc4 13
..txc4 .l:!.c8 Black regains the pawn and has
no problems.
5... c5
At that time the gambit line S... O-O 6
cxdS tLlxdS 7 tLlxdS 'ii'xdS 8 ..txc7 had been
widely played, but was considered favour-
able for White: 8 ... tLlc6?! 9 tLle2 ..tg4 10 f3
::'ac8 11 tLlc3 'ii'e6 12 ..tf4 (Timman-Schmidt,
Indonesia 1983).
The endgame after 8 ... tLla6! 9 ..txa6 (9
..tg3?! ..tfS! Levenfish-Ragozin, Leningrad
1936) 9 ... 'ii'xg2 10 'ii'f3 'ii'xf3 11 tLlxf3 bxa6
(Flohr-Botvinnik, A VRO Tournament, 7... tLle4
Holland 1938) was also thought to favour In contrast to the 1st game, here after
White. However, the 21st century has 7 ... dxc4 there is 8 ..txc4 - and yet this is the
shown that Black has sufficient counter- safest continuation for Black, which was to
play: 12 .l:!.gl f6 13 .l:!.c1 .l:!.f7 14 <;t>e2 ..td7 IS occur in the 9th and 11 th games, and later it
tLld2 eS 16 dS ..tbS+ and 17 .. .£S (Radjabov- became the main line.
Ivanchuk, Linares 2006), or 12 .l:!.c1 ..tb7 13 S cxd5 tLlxc3 9 'ii'd2 'ii'xa2 10 bxc3
<;t>e2 .l:!.ac8 14 .l:!.hdl f6 IS dS fS 16 b4 ..tb2 17 Later 10 .l:!.xc3 was also tried, but with-
53
Kasparov vs. Karpov: 1986-1987
out particular success: after 10 ... 0-0 only game published in Informator,
(10 ... ~xc3?! 11 'iiixc3 is too dangerous) 11 Schmidt-Gross (Naleczow 1984). We did
~c4 'iVaI + 12 l:i.c1 'iVxb2 Black repelled the not find a refutation of Black's idea, and in
attack. addition we were counting on the surprise
value and the unusual nature of the result-
ing positions. After the win in the previous
game such a risk seemed to me to be quite
justified. However, my opponent proved to
be fully prepared for the surprise.
10 .. :ihd2+?! (an objectively dubious move,
compared with 10 ... 'iVa5) 11 <;t>xd2 ttJd7 12
~bS 0-0 13 ~xd7
'If 13 c6 there follows 13 ... ttJc5', writes
Karpov (citing the little-known source
game Szymczak-Novak, Rimavska Sobota
1977). 'Now Black gains the advantage of
the two bishops, but, as it transpires, both
One of the tabiyas of the Grunfeld De- of them will be crippled.'
fence, where everyone automatically 13 ... ~xd7 14 e4
played 10 ... 'iVa5 11 ~c4 ttJd7 12 ttJe2 - Karpov played quite quickly, which is
following the Petrosian-Fischer game (Game not surprising: all these moves are natural.
No.29 in Volume III of My Great Predeces- After the seemingly strong 14 l:i.b1 there is
sors), when it transpired that 12 ... ttJxc5! 14 ... ~f5 (or 14 ... e5 15 ~g3 ~f5 Hubner-
gives Black comfortable play. In the mid- Staudt, Bochum 1990) 15 l:i.xb7 e5! 16 ~g3
1980s White switched to 12 ttJf3 ttJxc5 13 (16 dxe5 ~e4) 16 ... ~e4 with an excellent
~e5, but here too after 13 ... ~xe5 14 ttJxe5 f6 position.
15 ttJf3 b5! 16 ~a2 ttJe4 (or 16 ... ~f5) 17 'iiib2 14 ...fS! (the correct decision, in my opinion:
~d7 18 0-0 l:i.c8, as was shown both by later the weak white pawns must be attacked) 15
games and by our analyses, Black has eS
adequate counter-chances. And the subtle
13 0-0 0-0 14 ~e5, with the idea of 14 ... ~xe5
15 ttJxe5 f6 16 d6+!?, is parried by 16 ... e6 or
16 ... <;t>g717 dxe7l:i.e8.
In this game, and in the 9th, and in the
11th, was Karpov really prepared to go in
for the double-edged position after
10 ... 'iVa5, which is not at all in his style?!
Such play, with an element of strategic risk,
was to some extent unnatural for him.
However, in this match he often chose
continuations which did not correspond
with his concepts of correct play ...
But we, to avoid' superfluous' analytical lS ... e6
work, had prepared another, very rare A new, sharp idea, which was found the
move - 10 ... 'iVxd2+, familiar to us from the day before the game and became the main
54
The Third Match: 1986
line of our analysis. Gross played J5 ...!:i.ac8 l:!.xc3 22 'it'xc3 l:!.b8 23 .txa7 l:!.xb5 24 l:!.xb5
and after 16 c6?! bxc6 17 d6 exd6 18 exd6 .ixb5 or 20 dxe6 .txe6 21 cxb5l:td8+ 22l:!.d3
l:!.f6 he obtained a good position. But 16 c4! .ic4 23 l:!.xd8+ l:txd8+ 24 'it'c3 .txb5 25 .txa7
is far stronger: 16 ... l:!.xc5 17 .te3 l:!.c7 .tc6, and in both cases the most probable
(17 ... l:!.a5 does not change things) 18 CLjf3 result is a draw.
(Seirawan-Adorjan, New York (rapid) 1987) 17 ... bxc6 18 d6
or 18 f4 (Petursson-H.Olafsson, Reykjavik The critical moment of the game, but I
1988), and White's powerful centre guaran- did not notice this and I quickly made the
tees him some advantage (at any event, the prepared move.
statistics of this variation are unfavourable
for Black).
16 C4l:!.fc8
'After 16 ... g5 17 .txg5 .txe5 18 tiJf3 .tg7
19 l:!.bl and l:!.hel White would have com-
pleted his development, retaining strong
pressure.' (Karpov)
55
Kasparov vs. Karpov: 1986-1987
Popovic, Belgrade 1989) 23 ... .tLb2+! (pre- game White will be playing with an extra
venting .l:tc1-b1) 24 .l:tc2 .l:tb4, and it is not piece. It is true that the transfer of the
clear what White can extract from the knight to d3 will take three or four moves,
position: 25 ttJe5 (25 g3 as) 25 ... ..ixe5 26 fxe5 but can Black do anything during this time?
as or 25 lihc1!? .l:txf4 (25 ... a5!?) 26 .l:tb1 ~f7!
27 .l:tb7 ~e8.
It would appear that 18 ... g5! gives suffi-
cient counterplay. However, Black's crush-
ing defeat in this game created such a
strong impression that the variation with
10 .. :YWxd2+ did not in fact become estab-
lished in practice (which, however, also
aided the successful employment of
1O .. .'YWa5).
20 ... aS
Black has nothing apart from the ad-
vance of his a-pawn. 20 ... .l:tcb8 was recom-
mended, and it stands to reason that I
analysed this possibility during the game:
21 f3 .l:tb2+ 22 .l:tc2 .l:txc2+ 23 ~xc2 ..ia4+ 24
~d2 .l:tb8 25 .l:ta1 .l:tb2+ 26 ~e1 and wins.
21 f3 a4 22 .l:the1! (35)
An exceptionally important move, kill-
ing Black's counterplay. 'Over-protection of
19 h4! (Karpov made this move quickly the e5-pawn.' (Karpov) In the event of 22
and confidently) 19 ... h6 (06) ttJf2? Black would have had time for 22 ... g5!
Still not suspecting anything and hoping 23 hxg5 hxg5 24 ~h2 f4! 25 ttJd3 ..ie8!,
for 20 ttJf3 ..ic6. The immediate 19 ... a5 activating his bishop, for example: 26 ..ig1
would have changed little in view of 20 ..ig6 27 ttJxc5 ..ixe5 (Karpov) or 26 g3 fxg3
ttJh3! (but not 20 ttJf3 h6!) 20 ... a4 21 f3!, 27 ..ixg3 a3 with good counter-chances.
when Black cannot manage without 21...h6 Thus the surprise effect, on which I was
(in order to threaten ... g6-g5), since 21...a3? pinning my hopes, did not work. 'I man-
22 ttJf2 a2 23 ttJd3 and .l:ta1 is completely aged to find an almost mathematical solu-
hopeless for him. tion to the position,' Karpov writes in his
20 ttJh3! (12) book My Best Games. This is indeed a deep
A cold shower! This signals the end of plan (19 M!, 20 ttJh3!, 22 .l:the1), fully in
the plan with 18 ... c5: White prevents ... g6- Karpov's style, irrespective of whether it
g5, and after f2-f3 (also 'killing' the light- was found at home or at the board.
square bishop) he transfers his knight to d3, 22 ... a3 23 ttJf2
securely defending his e5-pawn and attack- With 22 .l:the1 Karpov lost a tempo, but
ing the c5-pawn. The bishop at g7 remains he solved the main problem - the switching
shut in, and thus for the remainder of the of his knight to the key square d3.
56
The Third Match: 1986
57
Kasparov vs. Karpov: 1986-1987
a 'mathematical solution to the position' is the 4th game: that my opponent was fol-
not easy. Black can resist, and it is clearly lowing a parallel analytical course, as if he
premature to assume that the outcome is knew beforehand what variation I was
already decided. intending to employ. I became nervous and
26 hxgs hxgs 27 il..xgs was tormented my doubts, since I trusted
my trainers completely and I did not yet
know of the events described in the chap-
ters 'Blitz Preparation' (p.16) and' A Trap
for the Champion' (p.262).
The following day I took my first time-
out, in order to analyse the character of the
play in the first few games and plan a new
strategic course. For Black we made inten-
sive repairs to the Grunfeld, and we de-
cided to begin the next white game with 1
e4, for a complete change of scene. Al-
though this also evoked memories of the
previous match, where after a loss in the
27 .. .'~f7? (28) 4th game I unsuccessfully 'switched sides'
The last act of the drama. Again 27 ... llb8 in the 5th, we had many fresh ideas in the
was more tenacious (even without the g- Ruy Lopez (we were expecting it, in analo-
pawn), since the tempting 28 lLlxc5? llb2+ gous fashion to the previous match), and I
29 'it>c1 llxg2 30 .td2 is bad, not in view of was in the mood for a large-scale battle.
Karpov's suggestion 30 ... .th6(?) 31 il..xh6(?) Despite the heavy defeat, I retained my
llc3+ 32 'it>d1 lld3+! 33 lLlxd3 il..a4+ 34 'it>c1 optimism and I felt far more confident than
llc2+ with 'a study-like draw by perpetual a year earlier in Moscow.
check' - this is refuted by 31 'it>b2! llxf3 32 However, Karpov avoided the Ruy Lo-
llgl llfg3 33 llxg2 and lLlxd7, winning, but pez, again demonstrating brilliant prepara-
because of 30 ... il..c6! 31 d7 (31 f4 il..h6!) tion, and after excessively active play in the
31...il..xd7 32 lLlxd7 .th6 with a simple opening I had defend carefully to save an
draw. inferior endgame.
White would again have had to find the
correct move 28 'it>e2! (with the idea of
28 ... llbb3 29 lLlxc5 llb2+ 30 'it>f1 - Karpov), Game 6
for example: 28 ... il..c6 29 il..c1 lla5 30 il..e3 G.Kasparov-A.Karpov
'it>f7 31 llec1 llb3 32 llc2 llba3 33 'it>d2 'it>e8 World Championship Match,
34 llc3 and wins. But now he doesn't have 6th Game, London 11.08.1986
to find anything. Petroff Defence C42
28 il..f4 (neutralising all the threats)
28 ...llb8 (too late!) 29 llecl il..c6 30 llC3 llas
31llc2llba832lLlcll-0 le4es
Times: 2.01-2.24. Until Seville 1987, my opponent did not
employ any other move in matches for the
This opening disaster strengthened an world championship.
unpleasant feeling, which first arose during 2 lLlf3 lLlf6
58
The Third Match: 1986
59
Kasparov vs. Karpov: 1986-1987
An attempt to fight for an opening ad- if 13 ...'iWc8, then 14 'iYe3! ctJe6 (14 ... ctJfS?! IS
vantage, which was new for that time. The 'iYeS is worse, as is 14 ... ctJc2?! IS 'iVe2 ctJxal?
line chosen by me a year earlier - 11 .l::tel+ 16 iLxf6 gxf6 17 ctJdS, winning) IS ctJe4!
iLe7 12 'iVdl ctJe6 13 cxdS ctJxdS 14 iLb5+ c6 (stronger than IS iLxf6 iLxf6 16 ctJdS or 16
IS ctJxdS cxbS 16 'iVb3 0-0 17 ctJxe7+ 'iVxe7 18 ctJe4, which has been seen in practice), and
'iYxbS a6 19 'iVb3 .l::tfd8 20 iLe3 .l::tac8 21 .l::tac1, White has an unpleasant initiative: IS ... 0-0
led after 21...h6 22 h3 ctJd4! to the simplifi- 16 ctJxf6+ iLxf6 17 iLxf6 gxf6 18 I!.adl or
cation of the position and a draw (Game IS ... ctJxe4 16 iLxe7 ctJd6 17 iLxd6 cxd6 18
No.67 in Kasparov vs. Karpov 1975-1985). 21 iLbS+.
h3 h6 22 .l::tadl! is slightly better for White It was because of 11 'iVh3! that the move
(Ivanchuk-Kramnik, Dortmund 2008). order 6 ... ctJc6 7 0-0 iLg4 almost went out of
Already in the 21st century Yuri Dok- use, and the variation 11 I!.el+ iLe7 12 'iVdl
hoi an and I analysed 16 iLf4!? (preventing would arise only after 7 ... iLe7 8 I:!.el iLg4 9
castling; if first 16 a4, then not 16 ... b4? 17 c4 ctJf6 10 ctJc3 iLxf3 11 'iVxf3 ctJxd4 12 'iYdl.
iLf4!, but 16 ... 0-0! 17 axbS iLcs with equal- But in 1986 the natural move 11 iVh3,
ity) 16 ... ctJxf4 (16 ... .l::tc8 17 iLeS!) 17 .l::txe7+ strangely enough, was overlooked by both
'it'f8 18 .l::teS 'iYd6 19 'iVd2!, but I was not able players! We concentrated our analysis on
to employ this bold idea, which a few years the variations with 11 'iYe3+, although at a
later overwhelmed Black in the game pre-match training session back in the
Naiditsch-Kramnik (Dortmund 2008): spring I expressed my scepticism about
19 ... ctJg6? 20 .l::teel f6 21 .l::tadl 'it'f7 22 'iVe3! White's chances: Black retains an extra
l:!he8 23 ctJe7! 'iVxe7 24 'iib3+ 'it'f8 2S I!.xe7 pawn and a solid enough position. But
.l::txe7 26 ~xbS and White won. Dorfman actively tried to persuade me of
19 ... 'iVxeS! was essential, although after the promising nature of White's position,
20 'iVb4+ 'it'e8 21 'iVxbS+ 'it'd8 22 .l::tdl ctJe2+ putting forward seemingly convincing
23 'it'hl! ctJd4 24 'iVxb7 I!.c8 2S h3! White's arguments: the black king is kept in the
attack is worth the sacrificed rook: 2S ... .l::tcS centre and very unusual play begins.
(2S ... gS!? 26 'iVxa7 'it'e8 or 26 'iYxf7 I!.cs - 11 ... ctJe6 12 cxdS itJxdS 13 itJxds 'iVxds 14
Landa) 26 ctJe3 (26 'iVa8+ \t>d7 27 'iVxa7+ iLe4 'iYbs
\t>d6 28 ctJe3 \t>e6 is no better) 26 ... I!.e8 27
ctJfS, when 27 ... I!.c7? 28 'iVb8+ \t>d7 29 ~xd4+
'it'c6 30 'iVb3! .l::te6 31 'iYf3+ \t>b6 32 .l::tb4+ \t>cS
33 I!.b8 is bad for Black (P.5mirnov-
Neelotpal, Pardubice 2008), but after
27 ... 'iVel+ 28 \t>h2 (28 .uxel?! I!.xel+ 29 \t>h2
ctJxfS) 28 ... 'iVxdl 29 '¥Wb8+ \t>d7 30 'iVd6+ ~c8
31 'iVxcS+ ~b8 32 ctJxd4 'iVa4 the position is
roughly equal.
However, before my fourth match with
Karpov it transpired that the greatest
problems are posed by 11 'iVh3! dxc4 12
iLxc4 iLe7 13 iLgS!, and after 13 ... 0-0 14
.l:tadl cS IS .l::tfel the storm clouds gather All this was performed quite quickly:
over the black king (Kupreichik-Yusupov, both players had aimed for this position. In
S4th USSR Championship, Minsk 1987). Or it my trainers and I had discovered some
60
The Third Match: 1986
interesting ways of developing White's two bishops (the light-square bishop being
initiative, giving full compensation for the especially strong, having no opponent and
sacrificed pawn. Alas, we didn't manage to operating on both wings), the possibility of
find anything more - neither at the training provoking weaknesses in Black's position
sessions, nor immediately before the game. with constant threats, and also of harassing
Everywhere Black has sufficient defensive the black king after kingside castling. The
resources, so that my scepticism was well- position of the black queen on the edge of
founded. Even so, attracted by the diversity the board creates additional motifs for
of White's threats, I decided to go in for this combinations, and if events develop rap-
variation, hoping that at the board it would idly it may prove to be out of play. Never-
not be easy for Karpov to solve the prob- theless, Black's game is based on a sound
lems arising. But that wasn't the case! positional foundation, and it is instructive
15 a4 to follow how, against all kinds of tactical
tricks by White, he finds replies which
maintain a dynamic balance.
15 ... 'iYa6!
The cowardly 15".'iYc5? gave White a
big advantage in the game Ivanchuk-Serper 16l:i.dl!
(Sochi 1986): 16 ~xb7 l:i.b8 17 b4! 'i¥b6 18 A useful, multi-purpose move. In the
'iYxb6 cxb619 ~c6+ ~d8 20 l:i.dl+ ~c7 21 b5. first instance White prevents queenside
And after 15".'i¥b6 16 'iYxb6 axb6 17 ~xb7 castling and takes control of the d-file. It
l:i.d8 18 ~c6+ ~e7 19 b3 the two bishops was less good to play 16 ~d3 ~6 17 'iVf3
promise White a somewhat better end- 0-0-0 or 16 'iYf3 ttJd4!, when after 17 'ii'e3
game: 19".'iii'f6 (19".f5!?) 20 a5! bxa5 21 there is no point in Black winning the
~b2+ 'iii'g6 22l:i.xa5 ~c5 23 ~e4+ f5 24 ~bl exchange by 17".ttJe2+?! 18 ~hl ttJg3+ 19
etc. hxg3 'iYxfl + 20 'iii'h2, but he should simply
By retreating his queen to a6, Black re- castle - 17".0-0-0!, remaining a pawn up
tains his extra pawn with the hope of with the better game.
calmly completing his development. In- 16 ..• ~e7!
deed, at first sight it is not altogether clear The most solid move, emphasising that
for what White has sacrificed a pawn. White has nothing special. Although in our
There are no obvious weaknesses in Black's preparations we devoted more time to two
position, and his king will soon be free to other continuations:
castle. But even so, White has his trumps: 1) 16".~c5 17 'iVf3, and it is hard for
61
Kasparov vs. Karpov: 1986-1987
Black to find the correct reply. White's rich bishop would have allowed White bound-
attacking potential is graphically illustrated less scope for creative thinking, which is
by the variation 17... c6? 18 1:!.d7!? Wxd7? 19 what induced me to go in for this position. I
iYxf7+ 'it'd6 (19 ... .i.e7? 20 .i.fS) 20 .i.f4+ ttJxf4 assumed that here White was not taking
21 'iVxf4+ ~e7 22 iYeS+ ~f7 23 iYfS+ ~e8 24 any particular risk, whereas Black had to
l:.el! .i.e7 2S ~e6 or 23 ... Wg8 24 .i.d3! play extremely carefully .
.i.xf2+ 2S ~hl bS 26 .i.xbS cxbS 27 iYdS+
with crushing threats. 18 ... 0-0 is more
tenacious, but here too after 19 b4! difficult
trials await Black: 19 ... .i.d4 20 .i.d3 'lWb6 21
.l:I.bl as 22 iVe4! g6 23 .i.c4! .i.xf2+ 24 ~hl
etc. The immediate 18 b4! is even better, for
example: 18 ... .i.xb4 19 .l:I.bl .i.cs 20 .l:I.d7! 0-0
21 .i.b2! with a decisive attack (21...ttJgS 22
'ikg4 'ikaS 23 M).
The exchange 17 ... .l:I.d8 18 .l:I.xd8+ ttJxd8
also favours White in view of 19 .i.d2! 0-0
20 .i.c3 (20 b4!? .i.d4 21 .i.xh7+) 20 .. :~h6 21
~dl ttJc6 22 g3 with excellent compensa-
tion, or 20 ... .l:I.e8 21 .i.xh7+ ~xh7 22 'YWhS+ 17 b4
'ltg8 23 ~xcS ttJe6 24 iVb4, regaining the The main line of our analysis. Now
pawn and retaining his positional pluses. White's idea begins to become clear -
17 ... .l:I.b8! 18 b4 .i.d4 19 bS iVb6 is combined play on both wings. The pawn
stronger, with unclear play, although here offensive on the queenside pursues the aim
too White has various possibilities: 20 of regaining the pawn, and at the same
.i.a3!? (a tempting exchange sacrifice) time the appearance of the dark-square
20 ....i.xal 21 .l:I.xal ttJgS 22 iVg4 ttJxe4 23 bishop on the long diagonal will cause the
iVxe4+ iVe6 24 iVc2(f4) with sufficient black king to feel uncomfortable even after
compensation, or 20 as iVxbS 21 ~bl iYxaS kingside castling. We also analysed other
22 .i.c6+ 'it'f8 23 .i.dS .i.f6 - Black is now moves:
three pawns up, but after 24 .i.d2 iVa4 2S 1) 17 ~3!? It all began with this idea -
.i.xe6 fxe6 26 .i.f4 White is out of the' dan- to keep the black king in the centre! But
ger zone' (26 ... ~g8 27 .i.xc7 .l:I.f8 28 ~xb7 after 17 ... .l:I.d8 18 .l:I.xd8+ .i.xd8 19 .i.e3 .i.f6
.i.d4 29 ~e2 .i.xf2+ 30 ~hl iYc6 31 ~b8 the weakness of White's queens ide pawns
.l:I.xb8 32 .i.xb8 .i.M 33 .i.d6 hS 34 iYxe6+ gives Black sufficient counterplay: 20 iVfS
with a draw). 'lte7 (20 ... .i.xb2? 21 .l:I.bl iYxa4 22 .i.c2! iVb4
2) 16 ... .i.d6 (to the defence of the c7- 23 iVdS! leads to the loss of a piece: 23 ... 'lte7
pawn) 17 ~f3 (17 .i.d3 "iVb6 18 .i.bS+, forc- 24 iVa2 b6 2S h3) 21 .l:I.el .l:I.d8 with ap-
ing 18 ... ~f8, is also unclear) 17 ... 0-0-0 18 proximate equality. However, the position
.i.e3 .l:I.hf8 19 .l:I.dS g6 20 g3 or even 19 b4!? after 17 ~3 is complicated and is still of
.i.xb4 20 ~dbl cS 21 .i.xcs ttJxcS 22 .l:I.xb4 interest.
ttJxe4 23 'iNxe4, and in both cases White's 2) 17 iVf3 (an attempted improvement)
initiative fully compensates for the pawn 17 ....l:I.d8 18 .i.d3 iVaS! 19 .i.d2 .i.b4 20 .i.e3
deficit. 0-0 21 iVxb7 ttJcS 22 .i.xcS .i.xcs 23 iVbs 1J2-1f2
In general, either development of the (Timman-Yusupov, 2nd match game,
62
The Third Match: 1986
Hilversum 1986). 18 l:!.xd8+!? ttJxd8 19 .i.e3 variation with 16 ... .i.c5), or 18 ... l:I.b8 19 .i.b2,
(Yusupov) is more promising, with excel- and it is hard for Black to complete his
lent compensation for the pawn: 19 ... 0-020 development. However, with 18 ... l:td8! he
'iVh3!? h6 (20 ... g6 21 .i.h6) 21 'iYf5 g6 22 'iVb5 parries the attack, by returning nearly all
iLf6 23 'Yi'xa6 bxa6 24 l:!.b1, and Black con- his extra pawns: 19 .i.b2 l:txd1+ 20 ltxd1
tinues to experience some discomfort. But ~xa4 21 .i.d5 .i.d6 22 .i.xg7 .l:tg8 23 .i.xb7
17... l:!.b8!? is more accurate, and if 18 b4, (23 .i.d4?! .l:tg6) 23 ...'iYf4 24 .i.c6+ We7 25
then not 18 ... 0-0 (Howell-Ivanchuk, Gran- 'iYxf4 ttJxf4.
ingen 1986/87) because of 19 b5 (Kuporo- We also studied 18 'li'b3!? c5 19 .i.d3
sov-Forintos, Budapest 1990), but 18 .. .'iVc4! 'iYb6 20 .i.b5+ 'it'f8 21 .i.b2 with an attack
- this move, found by us, had not yet been which fully compensates for the two-pawn
tried. deficit. But here there is the computer
I should add that 17 b3, depriving the resource 18 ... .i.e7!, which at the time we
black queen of the c4-square, is passive underestimated. Now White is not prom-
because of 17 ... .i.f6! (seizing the long di- ised anything by 19 'iYxb7 'iixb7 20 .i.xb7
agonal) 18 l:!.a2 0-0. But now 17 ... .i.f6 would l:!.b8 21 .i.c6+ Wf8, or 19 .i.xb7 ttJc5! 20 .i.xa6
have led to an unpleasant endgame for ttJxb3 21 .i.b5+ 'it'f8 22 .i.c6 l:!.c8. Perhaps 19
Black: 18 b5 ~6 19 'iYxb6 axb6 20 .i.xb7 .i.e3!? 0-0 20 'iYb1 l:!.ad8 21 .i.xh7+ 'i£th8 22
iLxa1 21 .i.xa8 (21 .i.a3 c5!) 21...0-0 22 .i.c6 .i.d3 is preferable, remaining a pawn down,
l:!.d8 23l:!.xd8+ ttJxd8 24 .i.d5. but retaining practical chances of fighting
What attracted me about the 17 b4 thrust for the initiative.
was its expansiveness and boldness. Such As we see, after 17 ... .i.xb4 Black would
moves are especially effective when they have had adequate defensive resources. But
occur for the first time and come as a sur- Karpov's move is undoubtedly the safest
prise to the opponent. But in this match I and most sensible.
was quite unable to surprise Karpov in the 18 'iYh3!
opening ... The 6th game was no exception. Events begin to develop rapidly and by
force, making a crisis imminent. An impor-
tant nuance is that, unfortunately, it is not
possible to cut off the black queen: 18 b5
l:!.ad8! (the weakness of the back rank!) 19
.i.d2 'iYb6 20 'iWxb6 cxb6 21 .i.xb7 .i.f6 22
l:!.a2 l:!.d4 23 l:!.e1 ttJc5 24 .i.c6 l:txa4, and
White is fighting only for a draw.
18 ... g6!
Accurately calculated: White is unable
to exploit the weakening of the long diago-
nal. If 18 ... h6, then 19 .i.xh6?! gxh6 20 'ii'xh6
is dubious in view of 20 ... ttJg5 21 'iYxa6 bxa6
22 .i.xa8 .l:txa8. Little is also promised by 19
17 ... 0-0! ~f5 (19 b5 Mad8!) 19 ... g6 20 ~f3 (20 'iVe5
Black coolly completes his development. Mad8! 21 .i.b2 'i£th7) 20 ... l:!.fd8 21 .i.xh6
If 17... .i.xb4, then 18 ~f3 looks tempting, Mxd1+ 22 l:!.xd1 'iYxa4 23 .i.xb7 l:!.d8. But
with the ideas 18 ... ttJc5? 19 Md4! (Timman), after 19 .i.d3! White's chances are neverthe-
18 ... c6? 19 l:!.b1 .i.c5 20 l:!.d7! (as in the less rather better: 19 .. .'iVc6 20 b5 'iYc5 21
63
Kasparov vs. Karpov: 1986-1987
ii.xh6, or 19 .. :~ib6 20 as 'iVxb4 21 ii.a3 'iVh4 19 ... cS!? 20 ii.b2 ltJd4 is not so clear, for
22 'iVxh4 ii.xh4 23 ii.xf8 'it'xf8 24 ii.c4 etc. example: 21 ~d3 'iVb6 22 as 'iWc7 23 bxcS
(Aseev-Ivanchuk, Irkutsk 1986) 23 ... Mfd8!
with the idea of ... ii.f6, or 21 'it'h1 ii.f6 22
bxcS Mad8 23 Mel, and White, after regain-
ing the pawn, pins his hopes on his two
bishops, but with accurate play Black
should be able to maintain the balance.
Nevertheless, only 19 'iVc3 would have
allowed White to fight for an advantage,
since the move in the game encountered a
strong reply ...
19 ...'iVC4!
The unexpected entry of the queen cre-
ates disharmony in the white ranks. White
19 ii.b2?! is a pawn down, determined measures are
The penultimate move of our home demanded of him, but now he is obliged to
analysis. It looks very strong: White intends lose time ... We clearly underestimated the
to set up the 'iV + ii. battery, and it appears strength of this move. Here I thought for a
to be not easy to defend against the threat long time - especially since there was much
of 20 'iVc3 without detriment to the posi- to think about.
tion: 19 ... ltJgS? 20 'iVh6 ii.f6 21 ii.xb7! or
19 ... ii.xb4? 20 'iVh4! (20 ii.xg6!? is also
decisive) 20 ... cS 21 ii.eS and wins. Black
also has a difficult position after 19 ... ltJf4?!
20 'iVf3 'iVe2 21 ii.xb7 'iVxb2 (21...Mab8 22
ii.d4) 22 ii.xa8 ltJe2+ 23 'it'h1 ltJc3 24 Mfl
ii.xb4 2S ii.c6. However, we failed to take
something into account, and in his 'parallel'
analysis Karpov was able to foresee some
veiled defensive resources.
Therefore the battery should have been
set up in reverse order - 19 'iVc3! (of course,
not 19 bS Mad8!). This was the best practical
chance, since in the event of my recom- An important moment: it will decide
mendation in Dva matcha 19 .. .tLigS (19 ... fS? what course the game takes.
20 ii.dS ii.f6 21 'iVb3 ~xa1 22 ii.xe6+ is bad 20Md7
for Black) 20 ii.xgS ii.xgS 21 'iVxc7 Mad8! This move, which had been planned at
Black loses a pawn - 22 ii.dS! (Huzman; the home, took me 69 minutes! Alas, it was
earlier 22 Mxd8 Mxd8 23 ii.xb7 'iVe2 24 ii.f3 only now that I saw the accurate reply
'iVb2 2S Mfl three times led to a draw) 20 ... Mae8!; I realised that our analysis had
22 .... 'it'g7 23 bS 'iVf6 24 'iVxb7, although after been incomplete and that things would not
24 ... Mfe8 2S 'iVc6 'iVeS 26 g3 Md6 27 'iVc4 turn out well for White.
ii.d2! 28 ii.f3 Md4 29 'iVc6 Me7 30 'iVc2 Med7 But what was I to do? If 20 ii.xb7, then
he can hope for a draw. 20 ... Mad8!, seizing the initiative: 21 Mdc1
64
The Third Match: 1986
'lixb4 22 ~a3 'li'xb7 23 ~xe7 (23 l:!.cbl 'li'e4 self with maintaining the balance. He is
24 l:!.el ctJf4) 23 ... ctJf4 24 'li'g4 l:!.fe8!, or 21 forced to agree to a slightly inferior end-
'lic3 l:!.xdl + 22 l:!.xdl 'iYxc3 23 ~xc3 l:!.b8 24 game, since everything else is unsuitable.
~d5 ~xb4 25 ~f6 ~d6. 20 'li'e3 is well met 21 ~xb7? is bad: 21...'li'xb4 22 ~a3 ~xa4
by 20 ... ~g5!, and the' cunning' 20 l:!.el?! by 23 ~xe7 Ilxe7 24 ~c3 (24 ~b2 ctJf4) 24 ... c5!
either 20 ... l:!.ad8 or 20 ... ~xb4 21l:!.e3 (threat- 25 ~b2 ctJd4 26 Ilxa4 ctJe2+ 27 Wfl ctJxc3 28
ening 'li'xh7+) 21...~d2 or 21...f5. ~xc3 Ilxb7 and wins.
In this situation it would perhaps have The commentators, who can readily sac-
been most appropriate to force a draw rifice other players' pawns, suggested 21
immediately by 20 ~d5! 'li'c2 21l:!.abl (more Ilel?! 'iVxb4 (21...~xb4? 22 ~xg6!) 22 ~c3.
accurate than 21 ~e5 l:!.ad8!) 21...l:!.ad8 22
':dc1! 'li'd2 23 l:!.dl. But I was nevertheless
tempted by the rook move, thinking that
the draw would not run away.
20 l:!.d7 does indeed look impressive, for
example: 20 ... ~d6(xb4)? 21 ~xg6!, or
20 ... 'li'xe4?! 21 'li'c3 f6 (21...ctJd4? is bad: 22
':xd4 'li'e5 23l:!.el 'li'f6 24 l:!.xe7'li'xe7 25 l:!.e4)
22l:!.xe71i'd5 23 'ifu3 with an attack: 23 ... ctJg5
24 ~e3 l:!.f7 25 h4, or 23 ... ctJg7 24 l:!.xc7 l:!.ad8
25 l:!.el (25 ... l:!.fe8 26 l:!.f1!). And the endgame
arising after 20 ... 'li'xb4?! 21 ~xg6 ctJg5 22
~xh7+ ctJxh7 23 ~a3 ~4 24 ~xe7 'li'xh3 25
gxh3 or 2l...1!Vh4 22 ~f5'li'xh3 23 ~xh3 ~fe8 Analysis diagram
24 ~xe6 fxe6 25 Ilxc7 b6 26 ~fl promises
Black nothing but problems. During the game I considered the posi-
But he has in reserve a simpler measure. tion after 22 ... ctJg5 (not 22 ... ~xa4? 23 ~xg6!)
20 ... Ilae8! (of course!) 23 'li'd3 (23 'li'g3 ctJxe4) 23 ... 'li'xa4 and was
afraid that the attack would not succeed,
for example: 24 Ild4 'li'a3 25 Ilal ~c5 26
Ild5 ctJxe4! 27 Ilxc5 ~xc5 with advantage to
Black. Later, when I was working on the
book Ova matcha, I thought that White
would nevertheless be saved by 24 ~xg6!?
If 24 ... ~f6? there is a pretty mate - 25
~xf7+! ctJxf7 26 'li'g3+ ctJg5 27 ~xg5+!! etc.
And in the event of 24 ... hxg6 25 Ildxe7! (25
Ilexe7 ctJe4!) 25 ... Ilxe7 26 Ilxe7 'li'f4 (not
26 ... l:!.e8? because of 27 'li'd4!) he finds a
drawing combination - 27 'li'xg6+ fxg6 28
Ilg7+ Wh8 29 Ilf7+.
21 ~d5 (15) Today, however, with the aid of the
Played already with the full realisation computer microscope, I have come to the
that White's spectacular idea has proved conclusion that here too after 24 ... ~d6!
ineffective, and that he must concern him- White would have faced a difficult struggle
65
Kasparov vs. Karpov: 1986-1987
for survival: 25 ~xe8 ~xe8 26 ~xf7 'iVa1+! 27 23 .i.xb4 (there is nothing to be done: 23
.i.xa1 .l:.e1+ 28 "iYfl ~xfl+ 29 'it'xfl ct:Jxf7 30 "iYf3? is fatal because of 23 ... iVxc3!)
.i.e4 b5 etc. 23 ...ct:Jxh3+ 24 gxh3 .i.xb4 25 ~xc7
Besides, apart from 21...'iYxb4, 21.. ..i.d6!? The lively tactical skirmish has con-
22 'iYf3 ct:Jd4! 23 iVe3 (23 iVc3? 'iYe2!) 23 ... ct:Jc6 cluded in a prosaic endgame. On his next
is also good for Black. move White will restore material equality,
21 .. :iWxb4 22 ~C3 and it may seem that a peace agreement is
imminent. But the point is that (with rooks
on the board) White's broken pawns on the
kingside may become a target. The f2-pawn
demands particular care - on its retention
depends the fate of the game. In addition,
the doubled h2- and h3-pawns are by no
means the same as pawns on f2 and f3: they
are not participating in the play, and Black
may have the possibility of attacking not
only the f2-pawn, but also the one on h2 (if
the white king should move too far away).
This is the type of position in which
Karpov was especially strong: he has no
22 ... ct:Jf4! specific threats, but his chances are some-
The clearest and most practical decision, what better, and without any risk he can
guaranteeing against any accidents: Black patiently manoeuvre for a long time, await-
takes play into an ending in which he even ing a mistake by the opponent (as a typical
has a slight positional advantage. example it is sufficient to recall the 20th
After 22 .. .'iVb6?! 23 a5 iVa6 24 ~e1 White game of the previous match, in which
would have retained chances of an attack: Karpov tried to 'squeeze' me right up to the
24 ... ~f6 25 .i.xf6 ct:Jg5 26 .i.xf7+ ct:Jxf7 27 .i.e7 85th move). Therefore I was fully aware
ct:Jd6 28 ~fl .l:.f7 29 ~xd6 cxd6 30 ~xf7 'it'xf7 that definite trials awaited me.
31 iVxh7+ 'it'f6 32 ~4+.
22 ... 'iYf4!? 23 .i.xe6 fxe6 24 'iYxe6+ ~f7 is
sharper, and now 25 ~e1? iVxf2+ 26 'it'h1
.i.f8! is bad for White, as is 25 .i.d4? because
of a surprising computer tactic - 25 .. :~Yf5!
26 iVxf5 gxf5 27 ~xc7 .i.f6! 28 ~c4 ~fe7!
(threatening mate on e1) 29 'it'fl ~e4 or 29
f3 ~c7! 30 ~b4 a5 and wins. Uncommonly
pretty play: in a succession of continuous
threats Black exploits the weakness of the
back rank, a pin on the 4th rank, and the
long diagonal! But after the accurate move
25 ~fl! White maintains the balance:
25 ... iYf5 26 'iYxf5 gxf5 27 ~xc7, or 25 ... b6 26 25 ... b6
iVe2! (with the idea of .l:.e1 or g2-g3) A somewhat unexpected decision -
26 ... iVf5 27 tl.xc7 .i.f8 28 iVxe8 tl.xc7 29 .i.b2. Black retains his b-pawn and secures the
66
The Third Match: 1986
cS-square for his bishop, which will set its White has to play very carefully (37 ..ItbS
sights on the f2-pawn. Many would have Mc2 followed by ... ..Itc5), to avoid ending up
preferred 25 ... Me5, aiming to activate the in zugzwang. Passivity and inertness in
pieces as quickly as possible, but after 26 such positions can prove catastrophic.
Md 1 Mg5+ 27 ~f1 ..Itd6 28 Mxb7 ..Itxh2 29 37 ...fs 38 Md3 ..Ites 39 Me3 ~f6 40 Me4
.ltb3 Black's advantage is short-lived. He The dubious 40 f4?! would have been
has more chances after 25 ... Me7!? 26 Mael unnecessarily weakening.
Md8 (nevertheless retaining the a-pawn) 27 40 .•. gs
..Itxb7 Md2, although here too with accurate Black has advanced all his kingside
play White should be able to draw. pawns, but for the moment he has not
26 Mxa7 ~g7 27 Md7! (an accurate move: in created any real chances. White constantly
view of the threat of l:tel-c7 Black has to keeps the bishop at cS under fire, so that if
exchange this excessively active rook) the black rook moves off the c-file he can
27 ... l:td8 play a4-aS. And otherwise he hopes to
If 27 ... Me5, then 28 Mel is good, for ex- exploit the weakness of the f5-pawn and by
ample: 28 ... Me7 29 Mxe7 kxe7 30 Mc7(c6), or attacking it to provoke .. .fS-f4.
28 ... MfS 29 ~g2 ..Itc5 30 as!. 41 Me2 ~es (here the game was adjourned)
28 l:txd8 Mxd8 29 Md1 (White is not afraid 42..1te4
of the pin, since he is intending to retreat The sealed move. 42 Md2 was possibly
his bishop to f3) 29 ...Md6 30 Md3 more accurate, for example: 42 ... ..Itd4 43 Mc2
While preventing the possible activation .i.c3 44 Me2+ ~f4 4S Me6 ..Itd4 46 Md6 ..Itc5
of the black king (30 ... ~f6 31 Mf3+), White 47 Md5 with a draw.
intends to wait, basing his play in accor-
dance with the opponent's actions.
30... hs 31 ~f1Md7 32 ~g2 ..Ites 33 ~f1 h4
67
Kasparov vs. Karpov: 1986-1987
The attempt to 'latch on' to the f5-pawn by not display the same persistence. Perhaps
playing the rook to f3 and bishop to d3 also Karpov considered his position in the
does not give a complete guarantee. Black match to be so favourable, that he did not
plays .. .f5-f4, penetrates with his rook onto see any point in playing flat out?
the second rank, and then, covered by his Times: 2.40-2.26.
bishop (from c3), his king also advances to
the 2nd rank, after which the bishop re- At any event, the initiative in the match
turns. This regrouping is not hard to carry had passed to Karpov. The initiative in the
out, but the question is whether or not match is a psychological phenomenon, but
White can be put in zugzwang and the e1- for the one who loses it, it often turns into a
square gained for the king. 'material' loss (an analogous situation arose
In the assessment of such positions, after the 3rd game of the 1985 match). And
general reasoning undoubtedly prevails, in the next game Karpov had an opportu-
and a concrete calculation of all the possi- nity to reinforce that psychological success
bilities is not easy. Here are some sample which had accompanied him in the 5th and
variations: 6th games, and to take the lead.
1) 42 .. .l:td7 43 ~b5l:td5 (otherwise there Indeed, before the 7th game I was in a
follows a4-a5) 44 lIe2+ 'it'f4(f6) 45 lIc2 or 43 slight state of panic, since after the 5th
~e2 ~b4 44 .l::tc4 ~a5 45 .l::tc2 lld4 46 ~b5 game I did not yet know an accurate reply
lId1 47 lIe2 'it'f4 48 l:te3, and White's de- to the 4 ~f4 variation in the Griinfeld
fences hold; Defence. We intensively analysed a new
2) 42 ... ~b4 43 Wf3 lld7 (43 ... Wd4 44 variation, but its critical position seemed to
~d3!) 44 l:te2+ 'it'd4 45 ~e6. It is probable me to be unclear, although there was
that 43 ~b3 lld7 44l:te2+ 'it'f4 45 ~e6 lle7 46 already the feeling that Black should have a
l:f.e3 is stronger, when it is not apparent good game (we only fully prepared this
how Black can strengthen his position variation for the 9th game - d. the note to
(46 ... g4 47 hxg4 fxg4 48 h3 ~c5 49lIe2). White's 10th move). And at the last mo-
In short, I was not absolutely clear about ment I decided to change opening, al-
the adjourned position - Black could have though I realised just how risky a step this
'tormented' me for quite a long time yet, was.
although a draw seemed most likely. Great Such decisions before the very start of a
was my surprise when the following morn- game are always most unpleasant. I re-
ing I learned that Karpov was offering a member how before the 4th game of my
draw without resuming (Yz-Yz). Even now Candidates match with Beliavsky (1983) I
such a step seems to me to be an inadmissi- agonised over whether or not to play the
ble luxury in a match for the world cham- Tarrasch Defence. And when I was already
pionship. on my way to the hall, I decided to play the
After all, usually my opponent endeav- Nimzo-Indian Defence. I lost that game,
oured to exploit the slightest chance! Thus although not because of the opening:
in our second match he played on in the 8th simply that day I was not prepared for a
game, which was also adjourned in a large-scale battle. In the 6th game I did
harmless position, and in the fourth match employ the Tarrasch - and I defended quite
he turned up for the resumption of the 17th easily ...
game in order to make just a few more In such situations you have to trust your
moves. But here, for some reason, he did intuition and not be afraid to follow the
68
The Third Match: 1986
path which you consider correct. The In contrast to 6 e3 i.f5 7 g4 (21st game of
betrayal of the Griinfeld in the 7th game the 1985 match), this move used to have a
could have put me in a difficult match harmless reputation, but in the game Petro-
situation, if Karpov, who was again armed sian-Beliavsky (Kislovodsk 1982), which
to the teeth in the opening, had exploited Karpov in fact follows, White demonstrated
all his chances. a new and interesting idea.
Game 7
A.Karpov-G.Kasparov
World Championship Match,
7th Game, London 13.08.1986
Queen's Gambit 031
1 d4d5
The Griinfeld Defence is given tempo-
rary retirement, and at last on to the stage
comes the Queen's Gambit, which at one
time was aptly called the opening of world 6 ... g6
championship matches. We had earlier Nowadays the best is deemed to be
prepared it for both sides, for Black as an 6 ... i.d6! (the source game: Beliavsky-Geller,
'emergency landing'. Moscow Interzonal 1982). I remember that
2 c4 e6 3 ct:JC3 i.e7 4 cxd5 exd5 5 i.f4 Geller played this after a long think, stood
In the previous match this variation oc- up from the board and, sorrowfully shak-
curred in the 20th, 21st and 22nd games. ing his head, whispered: 'I seem to have
Karpov played 5 ... c6, and I - 5 ... ct:Jf6. Of blundered a pawn ... ' Of course, this was a
course, we expected that the variation joke - in fact 7 CDxd5?! i.xf4 8 ~e4+ CDe7 9
would be repeated, and before the game CDxf4 ~a5+ 10 'it>dl i.f5 favours Black.
almost the entire team analysed both these Beliavsky chose 7 i.xd6 iYxd6 8 e3 CDf6 9
possibilities, including a controversial idea i.d3 and won the game, after which 6 ... i.d6
of Timoshchenko (with whom I had earlier almost went out of use. But later the idea of
looked at the Queen's Gambit separately). 8 ... CDe7 9 i.d3 b6! and ... i.a6 was found,
But, to all appearances, this idea did not with a comfortable game for Black (Harik-
come as a surprise to Karpov. rishna-Kasimdzhanov, Mumbai 2003). And
5... c6 if 7 i.g3, then 7... CDe7 8 e3 i.f5 9 iVb3 ct:Jc8!
Another change of role! If 5 ... ct:Jf6 Kar- is good, with the intention of ... i.xg3 and
pov could have gone 6 e3 0-0 7 ct:Jf3 i.f5 8 ... CDd6, equalising (Baburin-Vaganian, Los
h3 (22nd game of the 1985 match) or 7 i.d3, Angeles 1997), since 10 ~xb7? is not possi-
against which I had prepared 7... c5!. Inci- ble because of 10 ... CDb6.
dentally, in the next game he himself was In the end 6 iYc2 lost its attraction be-
to play 5 ... CDf6 6 e3 0-0, and after 7 i.d3 - cause of 6 ... i.d6!. Back then, however, the
7... c5! (yet another example of how our main move was 6 ... g6 - since the time of the
analytical excavations coincided). 6th game of the Korchnoi-Spassky match
6~C2 (Kiev 1968), where after 7 0-0-0 ct:Jf6 8 f3
69
Kasparov vs. Karpov: 1986-1987
ttJa6! 9 e4 ttJb4 10 iYb3 .i.e6! 11 eS ttJd7 12 a3 fully plays 10 .i.d3! (but not 10 ttJge2 M!)
as! Black developed a very strong attack. 10 ... .i.xd3 11 "iYxd3 ttJbd7 12 ttJge2, intend-
7 e3 (in 1989 Shirov introduced the double- ing e3-e4 and 0-0 (Ionescu-Geller, Sochi
edged 7 f3!?) 7 ... .i.fs 8 ~d2! 1986) or 0-0-0 (M.Gurevich-Geller, Moscow
This is Tigran Petrosian's unexpected 1987). 11...ttJa6 12 ttJge2 ttJc7 13 0-0 ttJe6 14
idea (instead of the simplifying 8 .i.d3). .i.eS is also no better for Black (Topalov-
White is not concerned about the loss of Karpov, Wijk aan Zee 1998).
time: he has a solid position in the centre The statistics of these variations are de-
and is hoping to exploit the weakening of pressing for Black, but I had not thought
the dark squares on the kingside (the move about the more solid 9 ... ttJbd7 (d. the note
... g7-g6), while the g6-pawn restricts the to my 9th move).
bishop on fS and soon allows White to 9 f3
drive if off its important diagonal. We looked at this position a few hours
before the game.
8... ttJd7
Black's main problem is the develop- 9... ttJb6?!
ment of his minor pieces, which can be Timoshchenko's aforementioned idea.
driven back after f2-f3 and e3-e4-eS (or g2- Objectively this manoeuvre is dubious (in
g4). In order to reduce the effectiveness of view of a possibility for White on the 11th
the pawn offensive, I decided to delay the move which was overlooked by the two
development of my knight from g8. players), but at the time it seemed to us to
After the energetic 8 ... ttJf6 9 f3 (we also make sense. Especially since the sharp
studied 9 ttJge2) 9 ... cS?! White retains the 9 ... gS 10 .i.g3 hS?! is dangerous in view of
initiative by both 10 .i.bS+ ttJc6 11 dxcS 11 e4! (Semkov-Geller, Plovdiv 1988;
.i.xcs 12 ttJa4 .i.e7 13 .i.h6 (Petrosian- Yakovich-Geller, Elista 1995).
Beliavsky), and 10 .i.h6! cxd4 11 exd4 a6 It is sounder to prevent e3-e4 by 9 ... ttJgf6
(11...ttJc6!? Gulko-Beliavsky, Linares 1989) (transposing into the 8 ... ttJf6 9 f3 ttJbd7
12 g4 .i.e6 13 ttJge2 ttJbd7?! 14 .i.g2 ttJb6 IS variation), although after 10 g4 (10 ttJge2!?)
b3 ~c8 16 0-0, when it is hard for Black to 10 ... ..te6 11 0-0-0 Black's position is rather
find a shelter for his king (Kasparov-Short, passive and White retains a slight but
Thessaloniki Olympiad 1988). enduring advantage, for example: 11 ... hS 12
And if 9 ... hS (at the time we considered gS ttJh7 13 M 0-0 14 ..td3 ~e8 IS ttJge2 ttJhf8
this move to be satisfactory) White success- 16 ~b1 as 17 ~c2 a4 18 ..tg3 a3 19 b3 ~c8 20
70
The Third Match: 1986
tDf4 .i.b4 21l:tc1 'ifVaS 22l:thdl etc. play either 12 tLlge2 dxe4 13 fxe4 tLlc4 14
10 e4 .i.e6 'ifVc2, with an advantage thanks to his
There is no sense in conceding the cen- strong pawn centre (if 14 ... tLlg4, then IS
tre by 10 ... dxe4?! 11 fxe4 .i.e6, since after 12 .i.xc4 .i.xc4 16 h3 tDf6 170-0), or 12 eS tLlhS
tDf3 White has a very comfortable position: 13 .i.h6 or 12 ... tLlfd7 13 b3, when the ma-
12 ... tDf6 13 .i.d3 0-0 140-0 .i.g4 IS .i.e3 .i.xf3 noeuvre of the knight to b6 proves to be a
16 gxf3 or 14 ... tDc4 IS 'ifVe2 bS 16l:tadl. blank shot.
After 10 ... .i.e6 we considered both 11
0-0-0 tDf6 12 eS tDfd7 and 11 .i.e3 tDf6 12 eS
tDfd7 to be unclear (for example, 13 b3 f6!
Graf-Sturua, Komotini 1993), while if 11
.i.d3 dxe4 White appears to have a pawn
hanging. Therefore we were most afraid of
11 eS, and ...
11 ... hS
Preparing a post for the bishop at fS. But
11...fS!? would have been sharper and
psychologically more unpleasant for Kar-
pov. This would have set him a difficult
choice - whether to retain the passed eS-
pawn, but allow Black counterplay (say, 12
11 eS tLlh3 h6 or 12 .i.d3 h6 13 g4 'ifVd7 14 'ifVc2
Karpov took this committal decision .i.b4! IS gxfS .i.xfS 16 .i.xfS .i.xc3+ 17 bxc3
confidently, without hesitation - the move gxfS etc.), or remain with an isolated d4-
is fully in his style. White relieves the pawn after 12 exf6 tLlxf6, which is fraught
'uneasy' tension in the centre, seizes space with risks for both sides: if 13 0-0-0 there is
and restricts the knight on g8; the variation a choice between 13 ... 0-0 (Karpov) and the
11...tDc4 12 .i.xc4 dxc4 13 tDe4 is in his immediate 13 ... tLlc4 (with the idea of 14
favour. In Gufeld's opinion, 'the further .i.xc4 dxc4 IS 'ifVe2 'ifVd7 16 l:tel 'it'f7), and if
plan is also clear - White should exchange 13 b3 - between 13 ... aS 14 .i.d3 a4 and
the dark-square bishops and play f3-f4-fS, 13 ... tLlhS 14 .i.h6 .i.d6 IS 0-0-0 ~f6 followed
when Black will have practically no way of by ... 0-0-0 or even ... 'it'd7.
opposing the attack on the kingside.' However, today I am not sure that
However, both Karpov and I over- 11...hS is objectively worse than 11...fS. It
looked that White could still have made the seems to me that both moves leave Black
useful move 11 .i.d3! (neutralising the idea roughly identical chances, but 11...hS leads
of 9 ... tDb6), since in fact the pawn is not to a more blocked position.
hanging: 11...dxe4?! 12 fxe4 'ifVxd4? is bad 12 .i.d3 (here too if 12 b3 there is the logical
because of 13 tDbS!. And if 11...tDf6 he can reply 12 ... aS!) 12 .. Ji'd7
71
Kasparov vs. Karpov: 1986-1987
Continuing to prepare the exchange of bilities and forces White to take unusual
bishops by ... ..if5. White stands better after decisions.
12 ... ltJc4?! 13 ..ixc4 dxc4 14ltJge2 'lWa5 150-0 14g3
0-0-0 16 'lWe3! ..if8 (16 ... h4 17 h3) 17 ltJe4 If 14 ..ig3, then 14 ... ..ie7! with the idea of
ltJh6 18 ..ig5 .i:td7 19 .i:tfd IltJf5 20 'lWf2. ... ltJh6 - say, 15 a4 as 16 ltJge2 ltJh6 with
13 b3 good play.
More subtle than 13 ltJge2, after which 14 ... ..ie7
there would have followed 13 ... ..if5 14 b3 Black has prevented ltJh3. 14 ... ..id8 was
..ixd3 15 'lWxd3 h4 (and if 16 .ie3, then no better - the place for the bishop is still at
16 ... ltJh6 17 ..ixh6 .i:txh6 18 f4 f5 with equal- e7.
ity). 15 ~f2 (the start of some creative, non-
But now if 13 ... ..if5 White can play 14 routine play) 15 ... .if5 16 .if1!
..ixf5 'lWxf5 15 ltJh3! - this position did not A change of plan in the changed situa-
appeal to me: 15 .. .f6?! 16 0-0 0-0-0 17 a4, or tion. 16 nel ..ixd3 17 'lWxd3 0-0-0 was
15 ... ltJd7 16 0-0 ltJf8, and here apart from unclear. Karpov's idea is more unpleasant:
the quiet move 17ltJdl the attempt to open White is threatening to drive back the
a second front is interesting - 17 b4!?, for bishop from f5 (h2-h3 and g3-g4) and
example: 17... .ixb4 18 .i:tabl ..ixc3 19 'lWxc3 'suffocate' his opponent.
'iVd7 20 ..ig5 ltJe6 21 f4 with the initiative
for the pawn. Nevertheless, 13 ... ..if5 was
not such a bad move - Black exchanges the
bishops while retaining a solid pawn
structure.
However, here another, at first sight
fanciful idea, occurred to me ...
16 ... ~f8?
A serious positional mistake, not men-
tioned in Ova matcha. Black kills his coun-
terplay against the white king, involving
... f7-f6. 16 ... a5!? (Karpov) came into consid-
eration, for example: 17 a4 ..ib4 18 ..ig5 f6!
19 exf6 0-0-0 20 f7 ltJe7, or 17 a3 f6! 18 ~g2
13 ... ..ih4+!? a4 19 b4 g5 20 exf6 gxf4 21 fxe7 ltJxe7 22
In the book Ova matcha I followed the "iVxf4 0-0-0 23 nel ltJg6 24 'ii'd2 h4 25 b5
other commentators in condemning the .i:tdf8 26 bxc6 'lWxc6 with a double-edged,
bishop check: 'The loss of time involved roughly equal game.
with this manoeuvre aggravates Black's Even the immediate 16 ... f6 was alright,
difficulties.' But now I evaluate it differ- and if 17 .i:tel, then 17 ... a6 (but not
ently: this move creates additional possi- 17 ... 0-0-0? 18ltJb5!). The undermining move
72
The Third Match: 1986
... f7-f6 was part of Black's compulsory to White's positional pluses are added the
program and would have justified his two bishops. White's advantage is gradu-
entire preceding play. Instead of this he ally increasing.
castles artificially and solves the problem of 19 ... kxh3+ 20 'iitxh3 ~g7 (I did not con-
the development of his knight at g8, but... sider the variations with 20 ... 'iVd7+ 21 'it>g2,
at too high a price! since I was planning ... ttJd7) 21 ~g2 ttJd7
17 ~g2 as is a3 The manoeuvre of the knight to e6 is a
Karpov is true to himself - he wants to sensible action, although only a slight
deny Black even a glimmer of counterplay. consolation.
The direct 18 h3 a4 19 g4 was possible, 22 kd3 ttJfS 23 i..e3 ttJe6 24 ttJe2 ttJh6
when 19 ... hxg4?! 20 hxg4 .l:i.xhl 21 ~xhl A critical moment. Here I had just 16
ke6 (the piece sacrifice 21...axb3 22 gxf5 minutes left for 16 moves ...
'iVxf5 is not a serious attempt) 22 .l:i.bl axb3
23 axb3 ~g7 24 kd3 favours White, but
after 19 ... ke6 20 .l:i.bl axb3 21 axb3 f5! (this
cannot be avoided) 22 exf6 kxf6 Black still
has a playable position. After 18 a3 it is
more difficult for him to carry out his
plans.
ls .. :iVdS
Black has in mind 19 h3 g5 (which is not
possible after 18 ... ~g7?! 19 h3! and g3-g4)
and at the same time he vacates the d7-
square for the future manoeuvre ... ttJb6-d7-
f8-e6. 18 ... a4 (Karpov) would not have
promised an easy life after 19 b4 h4 20 g4 In my view, Black's position is strategi-
ke6 21 ttJh3 or 19 ... ttJc4 20 kxc4 dxc4 21 h4 cally lost - he has no way of opposing the
~g7 22 ttJge2. advance of the kingside pawns (which
could have been begun a move earlier - 24
h3 etc.). Knowing Karpov's iron grip, I took
a pessimistic view of my chances. When
suddenly ...
25 b4?!
Without the slightest necessity White
opens a second front, when the game could
have been decided by direct action on the
kingside: 25 h3! followed by g3-g4 and f3-
f4-f5 (here 25 ... ttJf5 would not have helped
in view of 26 i..xf5 gxf5 27 'iVc2 etc.). This
move was in Karpov's style, and it is sur-
prising that he did not make it.
19 ttJh3! 'Relying on his bridgehead in the centre
A new correction to the plan. Since (the e5-pawn!), White should have attacked
Black cannot tolerate the appearance of this where the opponent is weaker: on the
knight at g5, he is forced to exchange it, and kingside: writes Gufeld, recommending
73
Kasparov vs. Karpov: 1986-1987
the 'obvious' 2Sl:i.hfl 'iYd7 26 ~xh6+ (in my 30 ~c4 l:i.c8 31 'iYd3 was no better in
opinion, 26 f4!? is better) 26 ... l:i.xh6 27 f4 fS view of 31...h4! and ... l:i.cS, driving away the
28 exf6+ ~xf6 29 fS - 'and what else could knight. It is possible that Karpov consid-
White want?' However, after 29 ... gxfS 30 ered the position after 30 ~e4 to be strate-
~xfS ~d6! 31 h4 l:i.f8 Black successfully gically favourable for him: powerful knight
defends (32l:i.ae1 ~d8 etc.). on dS, the black rooks disconnected ... But
Although White still retains the advan- after the next move the rook at h6 comes
tage after the strange move 2S b4, now into play.
Black is at least able to display some activ- 30... M!
ity. The situation has become sharper, and
2S .. :iWb6 26 bS? now it is no longer easy for White to find
Staggering! 26 l:i.ab1! axb4 27 axb4 fol- an accurate reply, especially with time-
lowed by b4-bS was far stronger, when trouble imminent.
there is no way for Black to equalise. If
27 ... l:i.a3, then first 28 l:i.hc1, so that in reply
to b4-bS there should not be ... c6-cS. Karpov
recommended 27 ... l:i.a4 28 bS 'iYaS (28 ... cS?!
29 ctJc3) with the evaluation 'unclear', but
after 29 'iYxaS! l:i.xaS 30 bxc6 bxc6 31 l:i.b7
White has clearly the better endgame, for
example: 31...l:i.a3 32 l:i.d1 ctJfS 33 ~xfS gxfS
34 ctJf4! ~a2+ 3S 'it>gl ctJxf4 36 ~xf4 1:te8 37
~c1 l:i.a6 38 l:i.c7 l:i.a4 39 l:i.bl. Black also
remains with problems after 27 ... ctJfS 28
~xfS gxfS 29 bS etc.
31l:i.hf1?!
A seemingly natural move, but not the
best. 31 g4?! h3+ was dubious, but 31l:i.hc1
or 31 f4 was better with interesting compli-
cations, for example:
1) 31l:i.hc1 ctJcS. In the book Ova matcha I
now gave 32 l:i.ab1 hxg3 33 hxg3 ctJxe4 34
fxe4 'iYd7 3S l:i.h1 (this was also considered
by Karpov in New in Chess) 3S ... l:i.ah8 36
'iYxd4 'iYg4 37 e6+ f6 38 l:i.xh6 ~e2+ 39 'iYf2
'iYxe4+ 40 'iYf3 'iYc2+ with a draw, but 3S
'it>f3! l:i.f8 36 ~xd4 f6 37 e6! gives White an
26 ... cs! advantage. Therefore it is correct to play
Of course, not 26 ... cxbS? 27 ctJc3 ctJc7 28 32 ... d3! 33 'iYe3 hxg3 34 hxg3, and now
ctJxbS ctJxbS 29 l:i.hbl. Now, however, the either 34 ... ctJxe4, or 34 ... l:i.c8 3S l:i.h1 t!.xh1 36
two sides' chances become equal. l:i.xh1 ~gS 37 f4 ctJxe4 38 'iYxe4 l:i.cs 39 ctJe3
27 ctJC3 (Black is also satisfied with 27 dxcS ~e7 40 'iYxb7 d2 with compensation for the
~xcS 28 ~xcS 'iYxcS) 27 ... cxd4 28 ~xh6+ pawn, or 34 ... 'iYd7 3S l:i.h1l:i.ah8 36 ctJf6 'iYd8
l:i.xh6 29 ctJxdS 'iYd8 30 ~e4 with a sharp fight.
74
The Third Match: 1986
32 ~xd4!? ttJb3 33 ~c3 is more interest- trouble!) 36 ~f3 (36 fxg5 ~xhl and ... l:'tc2)
ing, but in this case too after 33 ... hxg3! 34 36 ... ttJd4+ 37 ~e3 ttJc2+ 38 <;t>xd3 ttJxal 39
hxg3 ttJxal Black maintains the balance: 35 ~xal f5! 40 .if3 .ie7 and wins. However, 35
''iiUxal 'iYh8 36 e6+ f6 37 ~f2 l:1h2+ 38 ~e3 fxg5! ~xhl 36 ~xhl ~xd5 37 .ixd5 ~xd5+
'il!h6+ 39 f4 'il'h3 40 l:1c7! ~xg3+ 41 .if3l:1d8!, would have led to a drawn position which
forcing perpetual check, or 35 f4 f5 36 exf6+ also occurred in the game (38 ~gl! - d. the
(36 .if3 g5!) 36 ... .ixf6 37 ttJxf6 ~xf6 38 note to White's 38th move).
''iiUxf6+ ~xf6 39l:1xall:1h7; 34 ~xh1.ig5 35 f4 ~C5
2) 31 f4 ttJc5 (31..J:tc8!?) 32 'ijVxd4 ttJb3 33
'l'i'c3 ttJxal 34 Itdl!. This domination in the
centre looks unpleasant for Black (34 ... ~e8?
35 ttJxe7 ~xe7 36 ~d7! and wins), but after
34 ... ~f8 35 ~xal hxg3 36 hxg3 ~d7!
(36 ... ~c8?! is worse in view of 37 ttJf6!) he
forces the exchange 37 ttJxe7 ~xe7 and
easily gains a draw: 38 ~c3 ~d8! or 38 ~d4
g5! etc.
In playing 31 l::thfl, through inertia Kar-
pov apparently thought that the position
still favoured White: g3-g4 is threatened,
while after the exchange on g3 the rook
returns to hI and White continues to domi- 36 fxg5
nate in the centre ... But in fact now Black At this point, for the first time in the
has some advantage. game, Karpov caught up with me on the
31 ... hxg3?! clock and, apparently unsettled, decided to
Returning the favour; the immediate simplify the position as soon as possible.
31...~c8! was more accurate. I was afraid of White would also have held on after 36
32 g4 (in the event of 32 ... h3+ the hI-square ~a2 d3 37 ~f3 ttJd4+ 38 ~e3 ttJf5+ 39 ~f3,
has now been prepared for the king) but 37 ... ~c2! 38 'iVaI ~c4 would have
32 ... l:!.c5 33 ttJxe7 ~xe7 34 f4, not noticing caused him more problems.
the simple tactical stroke 34 ... g5!, which But 36 ~gl! was more cunning, with the
enables Black to seize the initiative: 35 fxg5 idea of 36 ... l:1xd5? 37 .ixd5 ~xd5 38 'iYh2!
(35 f5? ttJf4+ 36 ~xf4 ~xe5!) 35 ... ~xg5 36 ~f8 39 'iYh8+ ~e7 40 fxg5, when White has
'l'i'xg5+ ttJxg5 37 .ixb7 ~xb5 38 ~f5 ~g6. every chance of winning. The only defence
32 hxg3 ~c8 33 ~h1 is 36 ... f5!. A lively discussion revolved
Also after 33 ttJxe7 ~xe7 34 ~hl ~ch8 35 around 37 fxg5!, and it was established that
~xh6l:!.xh6 36 .id5 (36 ~xa5? ~g5 with an 37 .. .fxe4? 38 ttJf6! e3 (38 ... ttJf8 39 ~a2!) 39
attack) 36 ... ~h5 37 .i:ldl the chances would ~h7+ ~f8 40 ~a2 was bad for Black (40 ... e2
have been equal, for the reason that Black 41 ~xe2 l:!.c1 + does not help in view of 42
prematurely opened the h-file. ~h2 d3 43 'iVg4).
33 ... ~xh1?! However, Black is saved by 37 ... d3!! (the
Black's lack of time tells. 33 ... .ig5! 34 f4 endgame after 37 ... ~xg5!? 38 ~xg5 ttJxg5 39
~c5! was more cunning, pushing White .ig2 ttJe4 is also tenable). In the event of 38
towards the precipice - 35 ~a2? d3! (such a ~2?? ~c1+ and 38 ~h7+?? ~xh7 39 'iYh2+
move can easily be overlooked in time- ~g8 40 ttJf6+ 'iVxf6! White even loses. Only
75
Kasparov vs. Karpov: 1986-1987
76
The Third Match: 1986
before.
5... ttJf6
Game 8 'My' move from the previous match, in-
G.Kasparov-A.Karpov stead of the approved 5 ... c6 6 e3 (Game
World Championship Match, No.73 in Kasparov vs. Karpov 1975-1985) or 6
8th Game, London 15.08.1986 iVc2 (Game No.7).
Queen's Gambit 035 6 e3 0-0
In the words of Push kin, 'We all learned
little by little ... ' In the 22nd game of the
1 d4ds 1985 match Karpov had the white pieces
I reverted to 1 d4, in order once again to here, but now he is ready to defend this
test the solidity of the Nimzo-Indian De- same position with Black.
fence, but from that day Karpov switched The theory of the variation with 5 ... ttJf6
to his reserve opening (the improvement on 6 e3 was only just being developed, and
the 4th game, 12 ... ~c7!, had not yet come Ubilava's gambit line - 6... kfS (with the
within our field of view). idea of 7 iVb3 ttJc6 8 iVxb7 ttJb4 9 i.b5+ 'it'f8)
2 c4 e6 3 ttJc3 ke7 had not yet acquired its 'seal of approval'.
When it was necessary to hold out, the Karpov was to employ it only in the 12th
Queen's Gambit also served me faithfully, game of our next match, and after a quick
but, by the irony of fate, the overall result draw (Game No.39) it began to come into
was in favour of my opponent, whereas fashion. Although Timoshchenko and I
Karpov, who experienced greater problems were already looking at 6... it.fS in 1986, our
with Black, had until this game successfully team focussed on 6 ... 0-0. This move also
avoided defeat. attracted my opponent...
4 cxd5 exd5 5 kf4
7 kd3!
A position from the previous game has It is logical not to allow the develop-
been reached - the opening duel is contin- ment of the black bishop at fS. In the event
ued with change of colours! This occurred of 7 ttJf3 cS?! 8 dxcS i.xcs 9 i.e2 ttJc6 10 0-0
several times in our matches, but in the Black is left with an 'eternal' weakness at
given instance the two players quickly dS (10 ... d4? 11 ttJa4), but after 7 ... kfS! his
avoided 'repeating the past' and events position is quite acceptable, and it is
developed far more dynamically than unlikely that Karpov was concerned about
77
Kasparov vs. Karpov: 1986-1987
the plan with 8 h3 and g2-g4, which he h3 ~xe2 12 ctJxe2 (12 "iYxe2 d4!) 12 ... ~d6,
himself chose in the 22nd game of the when Black has a slightly inferior, but
previous match. And in the 20th game he acceptable position, since it is difficult for
prevented ... ~fS by 6 'iic2 0-0 7 e3, weaken- White to prevent ... dS-d4.
ing his control over d4, and I equalised We also analysed 8 ctJf3 ctJc6 9 0-0 c4
with the typical 7... cS! 8 dxcS ~xcS 9 ctJf3 with the evaluation 'unclear', or 9 ... ~g4 10
ctJc6 10 ~e2 d4 etc. dxcS ~xcS 11 h3 ~xf3 12 'iYxf3 d4 13 ctJe4
With 7 ~d3 White has also weakened ctJxe4 14 ~xe4 dxe3, 'and Black is alright,
his control over d4, which the opponent since IS 'iVhs after lS ... exf2+ 16 ~h1 fS 17
immediately exploits by initiating play in ~xfS g6 18 ~xg6 hxg6 19 "iYxg6+ gives only
the centre. perpetual check.' Our team analysis ended
7... cS! (this was planned by me back in the with this conclusion, and later Timo-
1985 match) shchenko and I looked at these positions
separately. In the last variation Gennady
found a serious improvement for White,
about which no one knew, apart from the
two of us .. .
8... ctJc6 (8 ... b6 is more passive: 9 ctJeS! ~b7
10 "iYf3 Tal-Abramovic, Moscow 1982) 90-0
8 ctJf3!
Essential flexibility: it is advantageous
for White to maintain the tension in the
centre. After 8 dxcS ~xcS he is unable to
prevent the simplifying ... dS-d4, for exam-
ple: 9 ctJf3 ctJc6 10 0-0 d4 11 ctJa4 ~d6! 12
~xd6 'iixd6 13 ctJxd4 (13 exd4 ~g4) As has already been said, the exchanges
13 ... ctJxd4 14 exd4 ~g4! or 13 ctJc3 ~g4 14 9 ... cxd4 10 ctJxd4 ctJxd4 (10 ... ~g4?! is
ctJbS "iYd7 IS ctJbxd4 ctJxd4 16 exd4 Mad8, weaker: 11 "iYa4 ctJxd4?! 12 "iYxd4! 'iid713 h3
regaining the pawn with full equality. ~e6 14 Mfdl Mfc8 IS ~eS Kasparov-Tal,
In our preparations we initially studied Skelleftea 1989) 11 exd4 favour White, for
8 ctJge2 ctJc6 9 0-0. Now after 9... cxd4 10 example: 11...~g4 12 'iVb3 or 11...'iVb6 12
ctJxd4 ctJxd4 11 exd4 White has a small but Mel ~e6 13 ctJa4 "iYaS 14 a3 and b2-b4 with
enduring plus - all his pieces are slightly unpleasant pressure.
better placed, and this symmetry is more 9 ... ~g4?!
promising for him than that in the 23rd This move surprised me: in contrast to
game of the 1985 match (although Karpov my opponent, I knew that it did not prom-
also suffered in it). Therefore we preferred ise Black an easy life. 9 ... c4 is better, stabilis-
9... ~g4 (even after 8 ctJge2) 10 dxcS ~xcS 11 ing the situation in the centre and transfer-
78
The Third Match: 1986
ring the battle to the wings: 10 .i.c2 .i.g4! 11 Schlechter, Ostend 1905). That only leaves
h3 .i.hS 12 g4 .i.g6 13 CLleS .i.b4! with the the move in the game, which gives White
idea of ... .i.xc3, relieving the pressure on the advantage of the two bishops.
the d5-point (Vyzhmanavin-Timoshchenko, 11 ... .i.xf3 12 'iYxf3 d4
Irkutsk 1986), or 10 .i.bl .i.g4 11 h3 .i.h5 12 For the moment all in accordance with
g4 .i.g6 13 a3 .i.xbl 14 l:!.xbl 'ii'd7, also with our analysis. After making this thematic
unclear play. advance, Black can normally regard the
10 dxcS future with equanimity. It would appear
This exchange is forced: it is impossible that this is also the case here, especially
to maintain the d4-point any longer. It may since it is not apparent how White can gain
seem that now Black easily solves all his any perceptible benefit from his two bish-
opening difficulties, but two moves later he ops: 13 exd4?! CLlxd4 14 'iYxb7? CLle6! - the
encounters a new problem. bishops come under attack, and one of
10 ... .i.xcs them is lost. However, by exploiting tactical
Nothing is changed by 10 ... d4 11 CLle4! nuances associated with the presence of his
(11 exd4 CLlxd4 12 .i.xh7+ 'ii;>xh7 13 'ii'xd4 light-square bishop, White avoids simplifi-
.i.xf3 14 'iUxd8 l:!.axd8 15 gxf3 .i.xc5 gives cation and retains the initiative.
White too little) 11...CLlxe4 12 .i.xe4 .i.xc5 13 13 CLle4!
h3!, and since after 13 ... .i.h5 14 l:!.c1 .i.b6 15 After this move Karpov sank into
g4 .i.g6 16 .i.xc6 bxc6 17 CLlxd4 White has an thought for the first time. I was very
obvious advantage, there only remains pleased to have gained such a promising
13 ... .i.xf3 14 'ii'xf3, transposing into the position, and I was hoping to employ our
game. home preparation.
79
Kasparov vs. Karpov: 1986-1987
jLg5 etc.) 18 jLxg6! hxg6 19 "iUxg6+ 'it'h8, for point (the a2-pawn is not hanging), but
which he was unanimously - and unjustly 14 ... ~6 would gain in strength, since now
- criticised by the commentators. after 15 jLd6(g5) Black equalises by
In fact White does not give perpetual 15 ... CLlxe4! 16 jLxe7 (16 ~xe4 g6 17 jLxe7
check, but includes his rook in the attack - ttfe8) 16 ... CLlxf2!, while 15 CLlg3 g6 is unclear
20 ttadl (it was this that Timoshchenko (but not 15 ... ttfe8?! 16 CLlf5! Milov-Toth,
discovered) 20 ... 'iVf6 21 'iVh5+ ~g8 22 ttd5 Liechtenstein 1996).
ttf7 23 jLg5 'iVh8 24 'llUxh8+ Wxh8 25 ttxc5
tte8 26 jLd2 and g2-g4 with good winning
chances, or, even stronger, 20 'iVh5+ ~g8 21
'iVg4+! ~h8 22 ttadl 'iVe7 (22 ... 'iVf6 23 ttd5)
23 ttd5! (but not 23 ttd7? "iUxd7 24 'iVxd7
ttxf4 25 'iVxb7 tte8 26 'iVxc6 ttel with a
draw) 23 ... ttf7 24 ~g5 'iVf8 25 'iVh4+ ~g8 26
jLh6 jLe7 27 'iVg3+ .l:rg7 28 jLxg7 ~xg7 29
~3 or 21...~h7 22 .l:radl jLd4 23 jLe3 'iVf6
24 jLxd4 CLlxd4 25 ttxd4 ttf7 26 ~5+ ~g8
27 ttg4+ ttg7 28 'iVd5+, winning the f2-pawn
and the game.
The move made by Karpov is far more
rational and sound. It was my turn to think: 14 ... 'iVas! (30)
how should White continue? After all, after The queen takes control of the 5th rank
exchanges his initiative may evaporate ... and, in addition, an attack on the queenside
It is wrong in principle to play 14 exd4 pawns forms the basis of Black's counter-
(it is Black who should be aiming to relieve play. Therefore 14 ... ~6 (14 ... CLld5?! 15 jLh2
the tension) 14 ... 'iVxd4 15 ttadl ttad8, and dxe3? 16 ~c4 or 15 ... ~6 16 CLld6 is obvi-
there is nothing real. White has a stable ously worse) came into consideration, for
plus after 14 CLlxf6+ ~xf6 15 e4 CLle5 16 jLxe5 example:
jLxe5 17 'iVg4 (17 'iVe2 ~g5!) 17 ... ~c8!? 1) 15 jLg5. Now 15 ... CLlxe4?! is unfavour-
(17 ... ~e7 18 f4! Yakovich-Timoshchenko, able because of 16 jLxe7! CLlxe7 17 ~xe4
Barnaul 1988) 18 f4 ~xg4 19 hxg4, but this CLlg6 18 ~xd4 or 16 ... ttfe8 17 .i.xe4 ttxe7 18
simplifies the game too much. jLxc6 bxc6 19 l:i.xd4 'iVxb2 20 'iVxc6. White is
I wanted more: to concentrate my forces also better after 15 ... CLld5 16 ~5 g6
in the centre, quickly complete my devel- (16 ... ~xg5? is bad: 17 CLlxg5 h6 18 CLlxf7! CLlf6
opment and switch to an attack on the 19 CLlxh6+ or 18 ... CLlce7 19 jLc4! Mxf7 20
kingside. With this aim a rook move to dl Mxd4, while if 16 ... ttad8 both 17 exd4 and
suggests itself. And again there is the 17 ~c4 are tempting) 17 'iVh6 f6 (or
eternal dilemma - which rook to place on 17... dxe3 18 jLc4!) 18 jLc4! fxg5 19 jLxd5+
this square, the queen's or the king's? ~h8 20 CLlxg5 (not 20 jLxc6? ttf5! with the
Reckoning that after 14 ttadl White would threat of ... ~f8) 20 ... ~xg5 21 ~xg5 ttf5 22
create a powerful piece grouping, I decided ~4 dxe3 (22 ... l:txd5? 23 'iVf6+ and 'iVe6+) 23
to retain for my rook on f1 the prospect of fxe3 'iVxe3+ 24 ~hl ~e7 (there is nothing
coming into play on the kingside. else) 25 'iVa4(c4), condemning Black to a
14 ttad1 (22) depressing defence.
If 14 ttfdl, then 14 ...'iVa5 would lose its It is hardly any better to try and sim-
80
The Third Match: 1986
plify by IS .. .':t:JeS 16 ~fS iZ:lxd3 in view of 17 idea - White removes a piece from the
~xd3! (in the book Ova matcha I considered centre. But - he prepares an attack on the
only 17 .i.xf6 .i.xf6 18 iZ:lxf6+ ~xf6 19 ~xf6 kingside, in which the rook at f1 will corne
gxf6 20 ~xd3 dxe3 21 fxe3 ~fd8!, when with in useful!
accurate defence Black gains a draw) The 'obvious' IS .i.gS is parried by the
17... dxe3 18 ~xe3 ~c6 (18 ... ~xb2 is dubious accurate IS ... iZ:lxe4! (but not IS ... iZ:leS?? 16
on account of 19 iZ:lxf6+ .i.xf6 20 ~b 1) 19 'fifS - from as the queen does not control
~fel with unpleasant pressure. the f6-square) 16 'iVxe4 (16 .i.xe7?! iZ:ld2!)
2) IS .i.d6! iZ:ldS (IS ... iZ:lxe4?! 16 .i.xe7!) 16 16 ... g6 17 .i.xe7 ~fe8 18 b4!? (after 18 "iVh4
iZ:lgS! (16 'fifS .i.xd6 17 iZ:lxd6 iZ:lf6 18 iZ:lc4 ~xe7 19 exd4 ~4! 20 .i.e4 ~ae8 21 .i.xc6
~c7 19 e4 ~fe8 is unclear, but 18 exd4!? bxc6 the activity of Black's pieces denies the
promises White some advantage) 16 ... iZ:lf6 opponent any hopes of success) 18 .. :iVc7! 19
(in Ova matcha I recommended 16 ... .i.xgS 17 bS ~xe7 20 "iVh4 dxe3! 21 bxc6 e2 with
.i.xf8 iZ:lxe3, but this loses to 18 .i.xg7! 'it'xg7 equality. Here the rook at f1 'carne in
19 fxe3 or 18 ... iZ:lxdl 19"iVhS 'it'xg7 20 'fixgS+ useful' to the opponent ...
'it'f8 21 .i.xh7) 17 .i.xe7 iZ:lxe7, and here However, White would have retained
White has two good continuations: 18 exd4 some advantage with IS .i.c4!?, a move
with the possible sequel 18 ... 'fixb2 19 ~bl underestimated by the commentators, for
'fixd4 20 ~xb7 h6 21 ~dl! 'fieS 22 iZ:lxf7! example: IS ... iZ:lxe4 16 ~xe4 .i.f6 17 exd4
'it'xf7 (22 ... ~xf7 23 ~xe7 and ~xa8+) 23 ~fe8 18 'fif3 or IS ... ~ad8 16 exd4 iZ:lxd4 17
.i.c4+ 'it>g6 24 ~bS, winning, or 18 iZ:le4 dxe3 iZ:lxf6+ .i.xf6 18 ~xb7, and Black still has to
19 iZ:lxf6+ 'fixf6 20 'fixf6 gxf6 21 fxe3 'it>g7 22 think about how to regain the pawn .
.i.e4 iZ:lc6 23 ~d7 ~ab8 24 b3 or 18 ... iZ:ledS 19 Summing up the result of the opening
.i.c4 'fixb2 20 ~d2 'fia3 21 ~xd4 iZ:lxe4 22 duel, it can be said that the variation from
~xe4 iZ:lb6 23 .i.b3, and Black comes under myoId analysis that was employed by
a positional bind. Karpov - 9 ... .i.g4?! 10 dxcS .i.xcSll h3 .i.xf3
Thus 14 ... ~6 IS .i.gS or IS ..td6! would - turned out to be bad (perhaps even more
have given White clearly the better chances. dubious than 11.. ..i.e6). To avoid being
Karpov's decision (14 ... ~aS!) seems to me 'burned alive', Black has to display miracles
to be more critical and correct. of resourcefulness.
15 ... dxe3
15 iZ:lg3! (17)
An unusual and, at first sight, strange 16 fxe3!
81
Kasparov vs. Karpov: 1986-1987
82
The Third Match: 1986
position begins to stabilise: Black loses the Karpov was a great master of defending
exchange (the rook on f8 cannot leave its such positions.
post), but the pawn plus two nimble However, here it would have been eas-
knights will give him definite counterplay.' ier for him to take decisions, and so I pre-
(Gufeld) ferred to complicate the game still more, in
The knight move is necessary, since oth- the hope of exploiting my opponent's
erwise the white bishop switches onto the imminent time-trouble.
a2-g8 diagonal with decisive effect: 21 ... ~f6! (17)
20 ... ~f6? 21 ~c4 i2Je5 22 ~xe6! i2Jxg4 23 This cool-headed reply came at a high
i2Je7+! ~xe7 24 ~xf8 i2Jxe3 25 l:txf7! ~xf8 price to Karpov: he now had just 12 min-
(25 ... i2Jxd1 26 l:tg7+) 26 l:tdd7 with inevita- utes left on his clock.
ble mate, or 20 ... i2Jd6?! 21 i2Jxe7+ 'iVxe7 22
'iVf4! l:tfe8 (22 ... l:tfd8 23 ~g5) 23 ~b1! i2Je5
24 ~a2 i2Jf5 25 ~g5 'iVc7(f8) 26 g4 with an
irresistible attack.
Thus, in a desperate situation, when to
many it appeared that Black's defences on
the kingside would be quickly swept aside,
Karpov displayed admirable tenacity in
defence and with a series of brilliant moves
succeeded in parrying the first wave of the
attack. I was faced with a dilemma ...
22 ii.bS?! (10)
After 22 ~xf8 (the computer plays this
automatically) 22 ... 'it'xf8 23 i2Jd4 'iVe7 24
~e4 i2Jd6 25 ~d5 Black has obvious com-
pensation for the exchange - control over a
complex of dark squares, centralised pieces,
and weakness of the white e3- and b2-
pawns. At the board I came to the conclu-
sion that I had incorrectly assessed my
possibilities on the previous move, when I
assumed that the opposite-colour bishops
21 'iVg3?! (11) would help the attack: the black bishop is
White's striving to decide the game at too strong! Even so, White would still have
all costs by a direct attack on the king leads had some advantage - even with the oppo-
him astray. After the logical 21i2Jxe7+ 'iVxe7 site-colour bishops he would have chances
22 ~xf8 'it'xf8 23 iYf4 he would have won of converting the exchange after e3-e4, b2-
the exchange and taken the game into a b3 and i2Jf3, with the exchange of one of the
technical phase (say, 23 ... 'it'g7 24 ~e4 i2Jd6 black knights (against Andersson in Mos-
25 ~d5 l:td8 26 'iVd4 etc.), although the cow 1981 I tried for a long time to win a far
conversion of the advantage would not less promising position).
have been a simple matter, considering that After 22 ~b5 the chances become objec-
83
Kasparov vs. Karpov: 1986-1987
tively equal, but... The battle is approaching ~xhS with an easy draw.
its height and it demands unusual deci- 26liJfs
sions by both players, and for this reason I I think it was here that Karpov decided
chose the bishop move, sharply inflaming to play for a win. His extra pawn and solid
an already tense situation. It was important position may have seemed sufficient
that by this point both players (especially grounds for optimism. But the point is that
Karpov) were seriously short of time. I in his impetuous attack White has not
realised that such tactics were risky, but I overstepped the boundary of acceptable
intuitively sensed that the activity of risk, and the position is still in a state of
White's pieces should guarantee him dynamic equilibrium.
against any troubles. The initiative at any Incidentally, in my clashes with Karpov,
cost - this is the keynote of White's subse- such situations (a complicated, double-
quent play! edged struggle and a time scramble) arose
22 ... liJg7! (OS) quite often, and each time I was left with
To his credit, Black accepts the chal- the feeling that after a successfully con-
lenge, skilfully discharging the tension on ducted defence he was very resolutely
the kingside. inclined, and would aim to seize the initia-
23.i.xg7 tive and tip the scales in his favour. He was
The retreat of the knight - 23 liJd4 especially successful with this in our first
would have signified the collapse of match. But now it was not the autumn of
White's hopes: 23 ... 'iYe7 24 'iYf4 (the reckless 1984, but the summer of 1986 ...
24 lIxf6? 'iYxf6 2S lIn is parried by 2S ... liJhS Over the next two or three moves Kar-
26 lIxf6 liJxg3 27 ..if4 liJe4 28 ..ixeS lIae8!) pov had the opportunity to oblige me to
24 ... liJhS 2S ..ixf8 lIxf8 26 'iYf2 a6 27 ..id3 force a draw, but he deliberately avoided
lIe8 with full compensation for the ex- doing so, reckoning that White's attacking
change. potential was insufficient and that sooner
23 ... ..ixg7 24 lId6 'iYb3 (24 ... gxfS? 2S lIxe6 or later Black's material advantage would
fxe6 26 e4 is bad for Black) 2SliJxg7 'iYxbs decide the outcome. But the resulting
problems were very complicated and with
very little time on the clock he cracked
under the tension.
26 ... lIad8
On this move Karpov used nearly all of
his already little remaining time, leaving
himself with just one minute for 14 (!)
moves. It would seem that he was looking
for winning chances - and could not find
any. A draw would have resulted from
26 ... lIae8 27 'iYgS f6 28 lIxf6 lIxf6 29 liJh6+
'it'g7 30 'iYxf6+ 'it'xh6 31 lIf4 liJf3+! or, more
quickly, 26 .. .f6! 27 liJh6+ (27 liJd4? ~xb2)
The ranks of White's attacking pieces 27 ... 'it'g7 28 liJfS+ 'it'g8! (but not 28 ... 'it'h8?!
have been greatly thinned, and the impres- 29 liJd4, and White has a dangerous initia-
sion is that his activity will soon peter out. tive).
26 liJhS is parried by 26 ... liJf3+! 27 'iYxf3 27 lIf6 (08)
84
The Third Match: 1986
I was also by no means insisting on the (31...'it>h8 32 ct:Jd6!) 32 hxg6 hxg6 33 ct:Jd4!
draw, which was possible after 27 ~xd8 Me8 34 ~6f4 l':td5 35 'iVh3 etc.;
~xd8 28 'ir'g5! ~d7!, when White has two 3) 27 ... ~h8!? (the safest) 28 ct:Jd4 (28 ct:Jd6
variations of perpetual check: 29 ct:Jh6+ ~g7 'ir'd5 29 ~lf4! Mxd6 30 l::.d4 'iVxd4 31 exd4
30 ct:Jf5+ (30 'iVf6+?! ~xh6 31 ,Uf4 ct:Jf3+! 32 .uxf6 32 'iVxe5 ~g7 33 g4 h6 34 h4 also
~xf3 'iVg5 favours Black) 30 ... ~g8 31 ct:Jh6+, maintains equality) 28 .. :~d5 29 ct:Je6! ct:Jd7!
or 29 ct:Je7+ 'it>f8 30 'iVf6 'iVxb2 31 ct:Jxg6+ 30 e4! 'iVa5 31 ct:Jxf8 ct:Jxf6 32 ct:Jxg6+ hxg6 33
hxg6 32 'iVh8+ ~e7 33 'ir'f6+. .l:!.xf6 with a roughly equal heavy piece
At that moment I was convinced that ending.
the conflict was not yet exhausted. The 28 'iVg5?!
rook move, increasing the tension, was The only moment in the game when
psychologically unpleasant for my oppo- White overstepped the mark. The surest
nent, who was in desperate time-trouble. way of maintaining the balance was 28 b4!,
removing the pawn from attack and hin-
dering the activation of the black queen.
Now after 28 ... ~h8?! 29 ct:Jd6 'iVd5 30 'it>h2!
'it'g8 31 ct:Jf5 'i'e4 32 'iVg5 'it>h8 33 ct:Jh6 'iVd5
34 h4! White has a dangerous attack, and
therefore 28 ... ct:Jd7 29 ct:Jh6+ 'it'g7 30 Mxf7+
~xf7 31 ct:Jxf7 'iVxb4 is better, forcing a
draw: 32 ct:Jg5 (32 ct:Je5 ct:Jf6!) 32 ... 'iVd6 33
'iVg4 ct:Jf6 34 ct:Je6+ ~f7 35 ct:Jg5+ or 32 'iVg5
'iVc5 33 'iVh6+ 'it'g8 34 'iVf4l':te2 35 ct:Jh6+ ~g7
36 ct:Jg4 'iVd5 37 'iVh6+ ~h8 38 .l:!.f8+.
27 ... ~d2
Intuitively Karpov makes a speculative
move, which in itself does not spoil any-
thing - Black trains his fire on the 2nd rank,
hoping to create counterplay against the
white king. But, in readjusting to playing
for a win, Black wasted precious seconds ...
Let us analyse the other possibilities:
1) 27 ... 'iVxb2 (the move that suggests it-
self) 28 ~g5 ~h8 (28 ... ~de8 or 28 ... ~d2(d7)
is also possible, but not 28 ... 'iVb5? because
of the deadly 29 ct:Jh6+ ~h8 30 M!) 29 ct:Jd6! 28 .. :iVxb2?!
(29 ct:Jh6 ~d5!) 29 ... ~g8 30 M!? 'ii'c3!, main- Black chooses the wrong time to treat
taining equality: 31 ct:Jf5 l':td7 32 h5 l':te8 33 himself to a second pawn. Apparently the
hxg6 fxg6 34 ~d6 Mxd6 35 ct:Jxd6 ~f8; absence of real threats and the feverish
2) 27 ... ~d7 28 M! 'iVxb2 29 h5 with suffi- haste of time-trouble dulled Karpov's
cient counterplay: 29 ... a5 30 ct:Jh6+ 'it>g7 31 vigilance. By 28 ... 'it'h8! 29 'ii'h6 .l:tg8 he could
'iVg5l':td5 32 ~h1! 'ii'e2 33 ct:Jf5+ ~h8 34 ct:Je7, have gained an advantage, for example:
or 29 ... MC8 30 ct:Jh6+ ~g7 31 ct:Jf5+ ~f8 1) 30 ct:Je7 'ir'xb2 (30 .. J:tg7!?) 31 'iVg5 ~g7!
85
Kasparov vs. Karpov: 1986-1987
32 l:te6!. A spectacular reply, but after 31 .u6f4! (31 l:!.6f2? ~6!) 31...tLlc6 32 tLlf5+
32 .. .'~jc6! 33 .l:txf7 (alas, 33 tLlxc6? fxe6! 34 ~g8 with a draw, or 31...f6!? 32 .l:txf6 tLlf7 33
tLld4 is bad because of 34 ... h6! 35 'iVxh6+ 'iVf4! ~3 34 ~h2! b5! 35 e4 'iVc4 36 tLlg4
~g8 36 'iVg5 .l:txd4 37 exd4 'iVxd4+ 38 ~h1 l:!.e8, forcing White to give perpetual check
'iVd5 39 'iVg4 'iVe5) 33 ... .l:txg2+ 34 ~xg2 - 37 .l:txg6+ hxg6 38 'iVf6+ etc.;
~xg2+ 35 ~xg2 .l:txf7 36 tLlxc6 bxc6 37 .l:txc6 2) 29 ... a5 (even 29 ... b6 will do) 30 tLle7+
~g7 38 e4 White has only one hope - to (30 tLlh6+ ~g7 31 .l:t6f4 f6 32 .l:txf6 l:!.xf6 33
save himself in a rook endgame where he is 'iVxf6+ leads only to equality) 30 ... ~g7, and
a pawn down; after 31 h4 tLld7 32 'iVh6+! ~xh6 33 tLlf5+
2) 30 tLld6 'iVd5! (30 ... 'iVxb2 is weaker: 31 ~h5! 34 g4+ ~xg4 35 .l:tf4+ ~h5 36 tLlg3+
~g5 .l:tf8 32 tLlxf7+) 31 'i/Vg5 .l:tf8! (now ~h6 37 tLlf5+ ~h5 or 31 l:!.6f4 h5! 32 tLlf5+
... tLlf3+ is threatened) 32 h4 .l:td1 33 tLlxf7+ ~g8 33 tLld4 tLld7 34 'iVd5! 'iVa2 35 'iVxd7
tLlxf7 34 'iVxd5 .l:txd5 35 .l:txf7 l:!.xf7 36 l:!.xf7 .l:txg2 36 'iVxb7 .l:th2+ 37 ~gl l:!.xh3 38 l:!.4f2
l:!.b5 37 l:!.f2 ~g7, and again White has an 'iVc4 39 'iVe7 (39 'iVg2 .l:!.xe3) 39 ... l:!.g3+ 40 ~h2
inferior rook endgame. .l:tg4 Black gains a draw.
However, in time-trouble such as Kar- But the prolonged and unpleasant pres-
pov was experiencing, an objective assess- sure finally disconcerted Karpov, and in
ment of the position is out of the question: severe time-trouble he was unable to adjust
at any moment White may surprise his to the rapidly changing situation.
opponent with an unexpected move. 30 tLld4!
Cutting off the black queen from the de-
fence of the knight. Such unexpected 'back-
wards moves' are most easily missed in
time-trouble. Now White's threats are
irresistible.
30 ... .l:txd4
A sad necessity. 30 ... .l:!.e8 would have
led to a catastrophe on the f7-point - 31
.l:!.xf7 tLlxf7 32 'iVf6+, and mate in three
moves.
31 'Yixes!
The final touch.
29~h1!
Essential prophylaxis and a strong psy-
chological resource: White 'stands still' -
perhaps Black can do the same?
29 ...~h8?
An instantaneous reply and, in the
unanimous opinion of the commentators,
the decisive mistake. Indeed, equal chances
would have been retained both by the
recommended 29 ... .l:td7, and by 29 ... a5, for
example:
1) 29 ... l:!.d7 (the simplest) 30 tLlh6+ ~g7
86
The Third Match: 1986
Here my opponent played 31...Itd2, but nevertheless to turn the course of the game
- the flag on his clock fell and he over- to my advantage? Or for the fact that with
stepped the time limit (1-0). However, after 28 ~gS?! I gave my opponent an unex-
32 ~e7! Itdd8 33 Itxf7 Itxf7 34 Itxf7 'it'g8 3S pected opportunity?
'it'h2! (3S e4 also wins) 3S ... aS 36 e4 or The intensity of the 8th game, its at
3S ... Itc8 36 e4 Itd8 (36 ... h6 37 ~e6) 37 Itxh7 times unexpected clashes and upheavals,
Itf8 38 eS all the same White would have the broad range of measures used by both
won. Times: 2.26-2.30. players, the fierce nature of the struggle - is
this not the essence of genuine chess crea-
Overstepping the time limit 10 moves tivity, is this not why we especially love
before the control in a lost position was a and value chess?!
shocking occurrence, a unique one both in It is hard to over-estimate the psycho-
matches for the world championship and in logical significance of the 8th game for the
Karpov's career (in Linares 1993 he broke entire subsequent struggle. It was a defin-
his anti-record - and again in a game with ing moment: in contrast to earlier times
me: playing White, he lost on time 13 (remember the 6th game of the first match),
moves before the control, also in a hopeless Karpov was unable to punish me for my
position). Let us try to work out what audacious, provocatively sharp play.
happened. After such a knock-down my opponent
A complicated opening, based on nu- was forced to take a time-out, but even so
ances. Karpov employed his 'drawing' he did not manage to fully recover in time
variation, which proved to have a 'hole' for the next game. Before the start of it I felt
(another amazing coincidence of analytical extremely anxious: I remembered my loss
preparations by the two contestants!), and in the Sth game after my win in the 4th,
as a result, after avoiding 13 .. .tLlxe4 14 Jtxe4 especially since I was again intending to
dxe3?, Black ended up in a difficult posi- play the Griinfeld Defence. However, on
tion. A well-conducted transition into the this occasion I was able to equalise, and
middlegame (IS iLlg3!) gave White real Karpov immediately forced a draw by
chances of a combinative attack on the king repetition.
(17 iLlfS, 18 Jth6!, 19 ~S!). Excellent de-
fence by Black (14 ... ~aS!, 19 ... g6!, 20 ... iLleS!,
21...iLf6!, 22 ... iLlg7!) enabled him to parry Game 9
the attack and gain good counter-chances, A.Karpov-G.Kasparov
and then, between the 26th and 28th World Championship Match,
moves, he made an attempt to play for a 9th Game, London 20.08.1986
win. In his desperate search for an advan- Grunfeld Defence 093
tage Karpov got into terrible time-trouble,
but he failed to find the key to the position,
chased after a mirage (28 ... ~xb2?!) and 1 d4 iLlf6 2 e4 g6 3 iLle3 d5 4 iLf4 iLg7 5 e3
went down in flames (29 ... 'it'h8?). 'Leave well alone'. Karpov was obvi-
And for what can I be reproached? For ously satisfied with the variation played in
the fact that with 21 ~g3?! and 22 iLbS?! I the Sth game, but this time I was fully
intensified an already complicated strug- ready for it.
gle? For the fact that with the far from 5... e5 6 dxe5 ~a5 7 Itel dxe4
obvious 27 Itf6 I avoided a draw, hoping Instead of 7 ... iLle4 (Game No.5) Black
87
Kasparov vs. Karpov: 1986-1987
chooses a variation which used to be con- the exchange counts for something! We
sidered unfavourable for him. After study- were able to improve the variation by
ing it, we came to a different opinion. 15 ... ttJc6!? 16 ttJc7 i.xg2 17 .ugl iLh3 with a
8 i.xC4 complicated game. And later I also found
If 8 'iVa4+ 'iVxa4 9 ttJxa4 i.d7 10 ~xc4 I 14 ... i.d3! 15 ttJxa8 ttJd5, when it is now
was planning 10 ... ttJd5! 11 i.g3 ttJc6, al- White who has to fight accurately for
though 11 ... i.b5 12 l:tM i..xfl 13 'it>xfl b5 14 equality (Inkiov-Lputian, Saint John 1988).
cxb6 axb6 is also interesting - in both cases Thus the myth about the strength of 10
White is unable to hold on to his extra ttJb5 was dispelled (which, in all probabil-
pawn. ity, was also not a secret to Karpov), and 10
8... 0-0 9 ttJf3 'iVxC5 i.b3 again began to come to the fore. At
one time, before the early 1980s, this con-
tinuation used to occur quite often, and
Black usually extinguished his opponent's
slight initiative.
10 ... ttJc6 (in 1988 Gavrikov introduced
10 ...'iVa511 0-0 ttJa6!?) 11 0-0
10 i.b3
Played without hesitation, although at
that time the main reply was considered to
be 10 ttJb5. Thus one of Karpov's helpers
quickly gained an obvious advantage after
10 ... 'iVb4+ 11 ttJd2 ttJa6 (11...ttJe4!?) 12 a3
'iVa5 13 b4 ~6 14 ttJc7! (Salov-Ebeling, 11 ... 'iVa5
Groningen 1983). 13 ... 'iVd8 14 i.e5! also The best square, in my opinion. Earlier
favours White (Huzman-Tseshkovsky, (and sometimes also later) 11...'iVh5 was
Tashkent 1987). played, after which, apart from 12 h3
Therefore during the two weeks which (Larsen-Tal, 5th match game, Bled 1965),
had passed since the 5th game, after 10 ttJb5 F.Portisch's move 12 ttJg5!? is also interest-
we had prepared an exchange sacrifice - ing.
10 ... i.e6! 11 ttJc7?! (11 i.xe6 'iVxb5 with 12 h3
equality, Dlugy-Kasparov, Saint John (blitz) White has to concern himself with his
1988) 11...iLxc4 12 b3 'iVa5+ 13 'iVd2 'iVxd2+ bishop on f4: in the event of 12 'iVe2 ttJh5 13
14 ttJxd2 i.d5 15 ttJxa8. The source game iLg5 i.g4 (13 ... h6?! is weaker: 14 i.M g5 15
Ftacnik-Uhlmann (Bucharest 1978) contin- l:tfdl! Yusupov-Anand, 4th match game,
ued 15 ... i.xg2 16 l:tgl iLc6 17 i.xb8 l:txb8 18 Wijk aan Zee 1994) 14 iLM Black has a
ttJc7 ~d8 19 a4, and although Black has choice between the sharp 14 ... g5 15 i.g3
bishop and pawn for a rook, nevertheless ttJxg3 and the quiet 14 ... 'iVb4 with equality
88
The Third Match: 1986
(Petursson-Smejkal, Thesalloniki Olympiad the central squares and increases the power
1988). of the bishop on b3: 16 ... e4 17 i.f4 l:!fd8 18
12 ... iHS l:!fdl l:!ac8 19 l:!xd8+ l:!xd8 20 l:!dl.
Also a logical move: while developing, 14 ... LtJh5?! is also inadequate: 15 i.h2 LtJxd4
Black prevents e3-e4 and prepares ... LtJe4. (15 ... i.xd4 16 exd4 LtJxd4 17 'iVxe7 LtJxb3 18
axb3 i.c6 19 b4! or 17... i.c6 18 i.c4 with
advantage) 16 exd4 i.xd4 17l:!fdl! (17 'iVxe7
i.c6) 17... e5 18 i.xe5 l:!ae8 19 l:!xd4 l:!xe5 20
'iVd2 i.c6, since here 21 .ic4! 'iVb6 22 b4 is
very strong (for example, 22 ... a6 23 a4 LtJf6
24 b5 axb5 25 axb5 LtJe4 26 LtJxe4 l:!xe4 27
l:!dl! l:!xd4 28 'iVxd4 'iVxd4 29 l:!xd4 i.e8 30
l:!d6 and wins).
15 exd4 e6
Now the light-square bishop, after re-
treating from f5, is ready to occupy a no
less active position at c6. Compared with
13 ... LtJxd4 Black has succeeded both in
13 LtJd4 taking control of the d5-square, and in
A rare and essentially harmless move, successfully arranging his pieces. Since the
with the idea after 13... LtJxd4 14 exd4 of activity of the bishop on b3 and knight on
creating pressure on the e-file by d4-d5 and c3 is minimal, White is unable to exploit the
l:!fel (14 ... e6 leaves the bishop at f5 in a weakening of the dark squares; moreover,
dangerous position). if Black should safely develop his rooks,
In the 11 th game Karpov followed the White will stand worse (there will be no
main line - 13 'lWe2 and after 13 ... LtJe4 14 compensation for the weakness of the d4-
LtJd5! e5 he released the novelty 15 l:!xc6!, pawn).
which had been prepared, according to At this point Karpov takes a very com-
him, before the 9th game. Why then did he petent, correct decision (correct both from
not play it that day? Perhaps he needed an the position, and from the match situation)
additional respite after the catastrophic 8th - he aims to 'curtail' the game.
game? Or he was not convinced that the
prospects were good for White? At any
event, against the usual 15 .1i.h2 my team
had found 'cast-iron' equality, and Karpov,
one way or another, had to seek new
paths ...
13 ... i.d7!
Of course: now White will not play 14
l:!el - the e-file is still closed!
14 'iVe2 LtJxd4! (24)
Undoubtedly the best move, solving all
Black's problems. If 14 ... l:!ac8, then 15 LtJf3
followed by e3-e4 is unpleasant. With
14 ... e5 15 LtJxc6 i.xc6 16 i.g5 Black weakens 16 i.d2!
89
Kasparov vs. Karpov: 1986-1987
16 i.eS, maintaining the tension, appears 25 Mcd1 i.f4 26 Me2 Mac8 27 Mee1 with
more natural, but Karpov does not like play- equality.
ing with an 'isolani' without clear prospects But I no longer saw any reason for play-
- after 16 ... ..tc6 Black has an excellent posi- ing on: the day's objective had been at-
tion, as confirmed by the games Lukacs- tained.
Schneider (Budapest 1977) and Petursson- 20 ..te3 "iVas
Ivanchuk (Reggio Emilia 1989).
16•.."iVb6 (not 16 ... i.c6? 17 dS) 17 Mfdl i.c6
After 17... "iVxd4? 18 i.gS Black would
have lost a piece.
90
The Third Match: 1986
14 i.b3!? (15)
A cunning - and also new! - move,
clearing the c-file and preparing lLld5. If 14
lLld5, apart from 14 ... cxd4 or 14 ... lLlb6 Black
can play 14 ... jLxd4 15 lLlxd4 cxd4 16 'ijVxd4
lLlb6 or 16 lLlc7 l:tb8 17 'ijVxd4 lLlb6 with
simplification and inevitable equality. 14 d5
At the time it was hard to imagine that lLlb6 is also unpromising - 'In the Queen's
this would become one of the key lines in Gambit Black does not often succeed in
the match: it was to occur in three games, obtaining a King's Indian bishop: (Tal)
and the last of these, the 22nd, was to bring The Tukmakov-Abramovic game went
me a decisive win. The intriguing point of 14 lLle4 cxd4 15 lLlxf6+ (15 l:te1 lLlb6 16 jLb3
the 10th and 12th games was that in both jLd7 with equality, Dokhoian-Pigusov,
Karpov chose rare variations with ... c6-c5, Irkutsk 1986; Timman-Korchnoi, Amster-
which I myself had prepared for Black, dam 1987) 15 ... lLlxf6 16 'iWb3! ~6 17 l:tfd1
since I was always keeping the Queen's jLd7 18 l:txd4 'ijVxb3 19 jLxb3 l:tac8 with a
Gambit in reserve. quick draw, but after 17 'ijVxb6! axb6 18
11 ... e5 (11 ... c5!? - Game No. 12) 12 h3 lLlxd4 White's chances are somewhat better.
An idea which became popular after the It would be interesting to know, was Kar-
23rd game of the previous match. Then pov intending to go in for this endgame?
Karpov suffered for a long time, without Or perhaps he was planning 16 ... lLle4 and
coping with the problem of equalising, and ... 'ijVf6 (Ionov-Goldin, Minsk 1986)? How-
later games too showed that 12 h3 did not ever, in either case it is hard for White to
only possess novelty value. count on anything serious here.
12 ... exd4 13 exd4 C5?! 14... cxd4
Instead of the usual 13 ... lLlb6 (Game Black would have liked to refrain from
No.22), this is a radical attempt to solve this exchange and play 14 ...b6, but then
Black's problems, already known from the White creates some pressure by 15 jLd5
source game Tukmakov-Abramovic (Bor l:tb8 16 lLle4 jLa6 17 l:te1 cxd4 18 b4 l:tc8 19
1983). In our analyses we did not especially l:txc8 jLxc8 20 lLlxd4lLle5 21 ~3 or 15 ... jLa6
favour this move, but even so we did not see 16 l:Ie1 l:Ic8 17 lLle4 cxd4 18 l:txc8 jLxc8 19
a clear plan for improving White's position. lLlxd4 lLle5 20 lLlb5 etc.
91
Kasparov vs. Karpov: 1986-1987
92
The Third Match: 1986
'iUxg5 hxg5 25 ~xc8 ~xc8 26 ct:Je2. 'iUd4! (if 28 'iUe5? Black saves himself by
Need I say how disturbed and agitated I 28 ... ~xh3! 29 ~xh3 ~f4! 30 'iUe7 ~f3+)
was when a position that we had analysed 28 ... ~b1 29 a4 'iVe2 30 ct:Jc3! 'iUxb2 (30 ... 'iUe1
for Black was reached on the board?! I had is hopeless: 31 ct:Jxb1 'iUxb1 32 b4) 31 'iUd5!
to force myself to calm down and think ~g6 32 'iUa8 ~f5 33 ~h8+ ~g6 34 'iVc6+ ~g5
seriously. 35 'iUf3 'iUd2 36 'iUg3+ ~f6 (36 ... ~h5? 37
ct:Je4!) 37 ~e8! etc.
And if 20 ... ~e8 White has a choice be-
tween 21 ct:Jf4 and 21 ct:Je3. Therefore it is
better to give up the bishop immediately
for this nimble knight.
20 ... ~xd5!
Black avoids the obvious dangers, but
now White's advantage of bishop against
knight, although small, becomes a stable
one. As a result, the scenario of the 2nd
game is repeated: in a slightly inferior
position Karpov is quite unable to escape
from his minor difficulties, whereas I
Our analysis had ended with the con- persistently try to set him new problems,
clusion that after 20 b4 ct:Je6 21 'iUe5 'iUh4 and this almost brings me success.
(21...~xd5!?) Black is alright (22 ct:Je3 'iUf4!). 21 ~xd5 'iVe7
As Tal put it, he 'simply acquires an addi- 21...'iUf6!? carne into consideration. After
tional resource - in a number of variations this 22 ~c3 seemed strong, but by activating
it is significant that the b4-pawn is unde- his rook Black would succeed in repelling
fended.' the attack: 22 ... ~ce8 23 ~f5 ~e1 + 24 ~h2
20 ~c3 was also suggested, but it too is 'iUe7 25 ~g3 ct:Je6.
unconvincing: 20 ... ~e8 21 ~g3 ct:Je6 22 'iUg4
'iUg5 or 22 'iUd3 ~xd5 23 ~xd5 'iUf6 with
equality.
But isn't there anything stronger? After
all, White is excellently developed and
under the cover of the knight at d5 he can
commence specific action. And I was able
to find a more accurate move.
20 'iUg4!
Primarily preventing the activation of
the black queen. Threats associated with
the opposition of queen and rook also arise:
20 ... ~h8? 21 b4 ct:Jd7(e6) 22 ct:Jf4, or 20 ...b5?
21 ct:Jc3 'iUb6 22 ct:Jxb5 ct:Je4 23 .ixf7+! with a 22 ~cd1
terribly strong attack - 23 ... ~h8 24 ~xc8 22 ~f5!? was more energetic, preventing
'iUxf2+ 25 ~h2 ~xc8 26 ~g6 ct:Jf6 27 'iUg3! simplification: Black has to defend against
'iUe2 28 ct:Jc3 or 23 ... ~xf7 24 ~xc8+ ~h7 25 the threat of ~xf7, and the inclusion of the
~dd8 'iUxf2+ 26 ~h2 ~xc8 27 'iUxe4+ ~f5 28 other rook is also prepared (~c3). For
93
Kasparov vs. Karpov: 1986-1987
example: 22 .. J~cd8 23 ~c3 ~e4 24 ~f4!? (24 since his control of the d-file does not give
~xf7 ~xf7 25 ~f3 ~xg4 26 ~xf7+ 'it>h7 27 anything (the d7 invasion square is securely
hxg4 ~d4 gives too little) 24 ... 'iibl + 25 'it>h2 covered), and the rook at c8 is too active.
with definite pressure - 25 .. .'iVxb2 26 ~d3 25 ... ttJxd5 26 ~a6 ttJf6
or 25 ... 'it>h8 26 ~e2 ttJe6 27 ~xe6 fxe6 28 Previously I thought that this was the
~g4 Itd7 29 b4 etc. most solid, but now I consider that the best
22 ~f5!? ~fd8 23 ~cdl ttJe6 24 b3 is also defence was the active 26 ... ttJb4!? 27 ~c4
interesting - at any event, this is not yet a ~c8 28 b3 ~c7 29 a3 ttJc6 or 28 ~b3 ~c7
draw. (and if 29 a3, then 29 ... ttJa6 and ... ttJc5) with
However, even without the queens a more durable position, very close to a
White will have his chances ... draw.
22 ••• ~e4! The position now reached still cannot be
If 22 ... ~cd8, then 23 ~f5. Exchanges sig- called drawn (although this result seems
nificantly ease Black's task but the position the most probable). By activating his king
does not become drawn - the bishop is still and advancing his kingside pawns, White
stronger than the knight. Remember with can strengthen his position.
what skill Fischer used to create problems
for his opponents in such instances!
23 ~xe4 ttJxe4 24 ..\ta6
Perhaps White should have kept all four
rooks on the board with 24 ~5d4!? followed
by b2-b3 and the advance of the kingside
pawns - if 24 ... ttJf6 the knight would have
been passively placed, while after 24 ... ttJc5
it could have come under attack by b2(b3)-
b4.
24 ...ttJf6!
94
The Third Match: 1986
White also exchanges the second pair of 36 ~dS! 'itf6 37 ~f3 or 36 ... etJxdS 37 'itxdS
rooks, which enables him to centralise his 'it'd7 38 g4 'itc7 39 h4 'itd7 40 hS 'itc7 41 gS
king, but eases Black's defence. I went in 'itd7 42 a4 'itc7 43 fS and wins) 3S 'itxc4
for this exchange under the influence of 'itd6 (3S ... a6 36 'itdS 'itd7 37 g4 as 38 a4 'itc7
Fischer's games, but although the resulting 39 M 'itd7 40 hS 'it>c7 41 gS 'itd7 42 gxh6
position seemed to me to be dangerous for gxh6 43 'iteS 'ite7 44 'itfS and wins) 36 'itbS
Black (it appears difficult for him, without 'itc7 37 'ita6 'it'b8 38 fS! with a decisive
creating additional weaknesses, to prevent advantage:
the entry of the white king into his terri- 1) 38 ... 'ita8 39 g4 'itb8 40 M 'ita8 (if
tory), it is objectively drawn. Therefore 40 ... g6, then 41 fxg6 fxg6 42 gS hS 43 a4) 41
White should not himself have exchanged hS! (threatening 42 gS and 43 f6) 41...f6 42
rooks - 30 ~c4!? was stronger, followed by a3 'itb8 43 a4 'ita8 44 as bxaS 4S 'itxaS 'it'b8
g2-g4 and possibly h3-M. 46 'itbS 'itb7 47 'it'cS 'itc7 48 'itdS 'itd7 49 b3!
30 ... l::teS 31 MxeS+ etJxeS a6 SO b4 etc.;
Black switches his knight to the queen- 2) 38 ... hS 39 h4! 'ita8 40 a3 'it>b8 41 a4
side, and in this respect 31...'itxe8 is simply 'ita8 42 as bxaS 43 'itxaS 'itb7 44 'itbS 'itc7 4S
a loss of time: after 32 'ite3 'ite7 33 'it'd4 all 'itcs 'itd7 46 'itdS 'itc7 47 f6! gxf6 48 'ite4
the same it is unfavourable to play 33 ... etJhS 'itd6 49 'itfS 'ite7 SO b4 a6 Sl g3 or 47 ... g6 48
34 'it>eS f6+ in view of 3S 'itfS!, when 3S ... g6+ 'itcS 'it'd7 49 'itbS 'it'c7 SO 'ita6 'itb8 Sl b4
36 'it>xg6 etJxf4+ 37 'it>xh6 etJxg2 38 ~b7 leads 'ita8 S2 bS 'itb8 S3 b6 axb6 S4 'itxb6 etc. I
to the desired position for White with a won a similar ending against Sveshnikov
passed pawn. (47th USSR Championship, Minsk 1979).2
32 'ite4 'it'e7 33 ~C4! 34 'ites f6+
A trap: White invites his opponent to go Black sets up a fortress on the dark
into a pawn ending, which I sensed (calcu- squares. The only way of taking it will be
lation at the board was difficult) should be by creating a zugzwang position.
won.
3S 'it'fSetJeS 36 'ite4
33 ... etJC7 It would have been better to play the
Karpov quickly - after three minutes'
thought - declined the invitation. And he 2 Cf. Endgame Strategy by M.I.Shereshevsky
was right - 33 ... etJd6+? 34 'itdS etJxc4 p.27 (Pergamon, 1985) - Translator's note.
(34 ... etJfS is too late in view of 3S 'it'c6! etJe3
95
Kasparov vs. Karpov: 1986-1987
immediate 36 'it'g6! 'it'f8 37 M ttJd6 38 ~d3, positions with the white king on e5 against
then advance g2-g4-g5, retain the remote h- king on e7 and knight on c7, where ~h3 is
pawn and try to play for zugzwang (as in decisive - d. the note to White's 43rd
the note to White's 40th move). But before move), for example: 50 ~d5 (50 h5 ttJe4)
the time control I began moving my king 50 ... ttJc8 51 'it'f5 ttJe7+ 52 'it'e6 ttJg6 or 50 ~e6
backwards and forwards, in order to await ttJb5 51 'it'f5 ttJd4+ 52 'it'e5 ttJf3+ and ... ttJxM.
the adjournment and only then find a clear 40 ... 'it'e7 41 'it'e4
plan. Alas, things did not come to this ... Again not hurrying with 41 'it'g6 or 41
36 ...ttJC7 37 h4 g4 ttJe8 (41...b5!?) 42 'it'g6 (but not 42 'it'e4?
f5+!).
37 ... 'it'd6
37... ttJe6! was more accurate, with the Here the game could have been ad-
key idea ... ttJf8 - this is an impregnable journed, but (to ease my home analysis) I
fortress! However, just in case Karpov wanted greater clarity, since now Black has
again decided to avoid going into a pawn to choose one of two methods of defence -
endgame - 38 .ixe6 'it'xe6, fearing tricks of with his knight on c7 or on f8. Karpov
the type 39 h5! f5+? 40 'it'd4 'it'd6 41 'it'c4 could also have adjourned the game, but he
'it'c6 42 b4 a6 43 a4 'it'd6 44 'it'd4 'it'e6 45 a5 too wanted clarity - he thought that the
bxa5 46 bxa5 'it'd6 47 g3 and wins, although position could be spoiled only by White.
after 39 ... 'it'd6! 40 'it'f5 'it'e7 41 'it'g6 'it'f8 the And he was proved right!
position is a draw. 41 ...'it'd6 42 g4 'it'e7?!
38 'it'fs 'it'e7 39 'it'g6 'it'f8 40 'it'fs A superfluous move, but, fortunately
And again a step back, whereas 40 a3!? for Black, not a losing one. Again 42 ... ttJe6!
ttJe8 41 g4 ttJd6 42 ~d5 would have re- with the idea of ... ttJf8 would have solved
tained winning chances. Thus after 42 ... ttJc8 all his problems, since 43 ~xe6 'it'xe6 44 h5
43 'it'f5 it is bad to play 43 ... ttJe7+? 44 'it'e6 gives only a draw: 44 ... a6 45 b4 b5 46 'it'd4
ttJg6 45 f5 ttJxM 46 ~e4! 'it'g8 47 'it'd6 g5 48 'it'd6 47 a3 'it>c6 (but not 47 ... 'it'e6? 48 'it>c5 f5
fxg6 'it'g7 49 'it'c7 ttJxg6 50 'it'b7 ttJe5 51 'it'xa7 49 g5 hxg5 50 fxg5 f4 51 'it'd4 'it'f5 52 h6,
b5 52 'it'b6 ttJc4+ 53 'it'c5 ttJxb2 54 'it'xb5 ttJd 1 creating an outside passed pawn) 48 'it'e4
55 .if3 ttJb2 56 ~e2 and wins, but with 'it'd6 49 'it'f5 'it'e7 50 'it'g6 'it'f8.
43 ... 'it'e7 44 ~a2 ttJd6+ 45 'it'g6 'it'f8 46 g5 43 b4?
fxg5 47 fxg5 hxg5 48 'it'xg5 'it'e7 49 'it'g6 'it'f8 A rapid and unfortunate reply. White's
Black holds on (he only needs to avoid plan is to create a zugzwang position after
96
The Third Match: 1986
g4-g5 and exchanges on g5, and then break This move was at last sealed, but I de-
through with his king on the queenside. cided not to resume the game and the
But now Black has the possibility of ex- following day I offered a draw (Yz-Yz). In
changing another pair of pawns by ... a7-a5, analysis we were unable to find a winning
reducing White's chances to the minimum. plan after 44 ... ~e7 (the drawbacks of 43 b4
Instead of this it was possible to play 43 also told):
~f5, ruling out ... ct:Je6-fS and trying to 1) 45 h5?! ct:JeS 46 ~g6 ~fS 47 a3 ct:Jd6 4S
revert to the main plan (true, here White ~f1 ct:JcS 49 ~f5 ~e7 50 ~d3 ~f7 51 ~c4+
has to reckon with 43 ... b5!? 44 i.gS ct:Ja6), or ~e7 52 ~a6 ct:Jd6+ 53 ~g6 WfS, and White
calmly play 43 a3, since Karpov, fearing the cannot break through;
pawn endgame, would most probably not 2) 45 g5 fxg5 46 fxg5 hxg5 47 ~xg5 (47
have replied 43 ... ct:Je6!, and after the me- hxg5 ct:JeS 4S ~e5 ct:Jd6 49 ~d5 a5!) 47 ... a5 4S
chanical 43 ... ~d6?! White could have ~g6 ~f8 49 b5 ct:Je8 50 ~d3 ct:Jd6 51 a4 ct:Jb7
repaired his omission - 44 Wf5 ~e7 45 ~g6 52 ~f5 ct:Jc5 53 ~c2 'it>e7 54 'it'e5 'it'd7 55 'it'd5
~fS 46 g5 fxg5 47 fxg5 hxg5 4S ~xg5. It ~c7 with a draw;
would have been very difficult for Black to 3) 45 ~e4 ~d6 46 a3 ~e7 (46 ... ct:Je6 will
defend this position: 4S ... ~e7?! 49 ~g6 ~fS also do) 47 a4 ~d6 48 ~f5 'it'e7 49 g5 fxg5 50
50 ~f5 ~e7 51 ~e5 Wd7 52 i.f1 ~e7 53 fxg5 hxg5 51 hxg5 ct:JeS 52 ~e5 ct:Jd6 53 ~d5
i.h3, and here neither the knight nor the ct:Jf7+! 54 ~xf7 Wxf7 55 Wf5 a5! 56 b5 g6+ 57
king can move without allowing the white ~e5 ~e7 or 56 bxa5 bxa5 57 g6+ ~e7 58 ~e5
king across to the queenside (say, 53 ... a5 54 ~d7 59 ~d5 ~e7 with a draw.
i.f1 or 53 ... ct:JeS 54 ~d5 ct:Jd6 55 b3! and Even in a zugzwang position with ad-
Wc6). True, after 4S ... ct:JeS! I have not found vanced white a- and b-pawns:
a win.
43 ... Wd6
Avoiding 43 ... ct:Je6 44 i.xe6 ~xe6 45 b5!,
in order not to have to calculate the lengthy
variation 45 ... h5! 46 f5+ ~d6 47 gxh5 Wc5 4S
~f4 ~d4 49 ~g4 ~e4 50 a4 We5 51 h6 gxh6
52 Wh5 ~xf5 53 Wxh6 ~g4 54 ~g6 f5 55 h5
f4 with a draw.
44WfS
Analysis diagram
97
Kasparov vs. Karpov: 1986-1987
98
The Third Match: 1986
Analysis diagram
99
Kasparov vs. Karpov: 1986-1987
White's idea becomes clear: he has a 23 ... 'iVd4!), but it is more accurate to play
powerful knight on dS, his rook has in- 22 ...'iVa2 23 4JeS (23 'u'dl 'iVe6) 23 ... ,U,fe8, or
vaded the 7th rank, and two pawns are 22 ...'iVa4 23 4JeS (23 'iVh4 .u.d7; 23 4JgS 'iVb3)
attacked. Black has to act resolutely to 23 ... aS! 24 4JdS 'iVb3 2S 4Jf6 'iVe6 26 4Jed7
avoid ending up in a difficult position. .i.xf6 27 'iVxf6+ 'iVxf6 28 4Jxf6 axb4 29 'u'dl
16 ... .i.e6 (37) 'u'c8.
After a long think I decided to begin an The move in the game is more danger-
operation to drive out the rook from c7. If ous for Black, since it forces him to balance
16 ... fxe3? 17 'iVxe3 .i.e6 18 'iVel! White has on the edge of the precipice.
an obvious advantage: 18 ... 4Jc3 19 4Je7+ 17 ... 'iVbS (OS)
Wh8 20 'u'xc3 or 18 ... 'iVxel 19 .u.xel .i.xdS 20 The only move. The endgame after
.i.xdS 4Jd6 21 .i.xb7. 17 ... 'iVxel? 18 'u'xel is obviously unfavour-
able for Black: 18 ... fxe3 19 'u'xe3 Ji..xdS 20
.i.xdS 4Jd6 21 .i.xb7 (Grigore-Perez, Sitges
2007), 18 ... .i.xb2 19 4Je7+ Wh8 20 .i.xe6 fxe6
21 exf4 'u'xf4 22 'u'xb7, or 18 ... b6 19 4Je7+
Wh8 20 .i.xe6 fxe6 21 exf4 .u.xf4 22 b3 etc.
Here, for the first time, Karpov thought
for a long time (nearly half an hour). Which
is quite understandable: the variation with
the exchange sacrifice arose in the course of
the match and there simply wasn't time to
analyse in detail the subsequent complica-
tions. White has a wide choice of tempting
continuations, but only at first sight.
17 'iVel!
To judge by the speed with which Kar-
pov made this move, this was all prepared
in his home laboratory and the unexpected
queen manoeuvre was the point of White's
plan. In the book Ova matcha I recom-
mended 17 4Je7+ (but not 17 .u.xb7? 4Jd6!)
17 ... Wh818 'u'fc1 .i.xb319 axb3 fxe3 20 'iVxe3
4Jd6 21 'iVf4 'with a strong attack', but after
21...,U,ad8 Black, who has no weaknesses,
quickly achieves equality:
1) 22 4JgS h6 (22 ... 'iVb6!?), when 23 4Je4?
(Levitt-Mittelman, London 1990) is weak in
view of 23 ... 'iVb4!, while 23 4Je6!? gS! 24 18 4Je7+ (27)
'iVe3 fxe6 2S 4Jg6+ Wg8 26 'iVxe6+ 'u'f7 27 Now Black can breath more freely - the
.u.xf7 4Jxf7 28 4Je7+ Wf8 29 4Jg6+ leads only opponent is preparing the knight sacrifice
to perpetual check; on g6, and his replies are fairly obvious.
2) 22 b4 'iYb6?! 23 4JeS! with some ad- 'The interesting move 18 .i.c4 provoked
vantage for White (Vasyukov's recommen- contradictory assessments by grandmas-
dation 23 4JgS is not good on account of ters' (Vasyukov), but it does not give any-
100
The Third Match: 1986
thing in view of 18 .. :iVxb2 19 exf4 ctJg3! 20 White's deeply entangled knight escapes,
fxg3 (20 ctJgS? ctJxfl) 20 ... i1.xdS, and White and he retains a clear advantage: 21...'iVxa2
is forced to worry about his rook on c7: 21 22 'iVb4 or 21...'iVb6 22 ~c1.
'iVf2 (21 i1.xdS? 'iVb6+) 21...'iVc3 22 'iVcs i1.xc4 20 'iVb1! (OS)
23 'iVxc4 'iVxc4 24 ~xc4 bS with equality. An interesting plan of attack - the queen
But 18 ctJd4!? i1.xd4 19 ctJe7+ came into stands in ambush, aiming from afar at the
consideration, for example: g6-point. Now White's advanced force (~c7
1) 19 ... ~h8 20 exd4 'iVb6 21 'iVc1! (but not and ctJe7) gains effective support and it
21 ~c1 fS!) 21...i1.xb3 22 axb3 'iVd6 (22 ... f3 23 unexpectedly proves to be in the right place
'iVf4!) 23 ~e1 fS 24 ~xb7 ~ab8, and the (this key idea was most probably studied in
utmost accuracy is demanded of Black: 2S preparations for the game). Black must
~xa7 'iVxd4 26 ~xe4 fxe4 27 ctJxg6+ hxg6 28 immediately undertake something.
'iVc7 'iVxa7 29 'iVxa7 ~fe8! 30 'iVf7 e3 31 'iVxf4
e2 with equality, or 2S ~xb8 l:i.xb8 26 f3
'iVxd4+ 27 ~h2 (27 ~h1 ~e8) 27 ... ctJd2! 28
ctJc6 'iVf2! 29 'iVc3+ ~g8 30 ~d1 ctJxf3+ 31
'iVxf3 'iVxf3 32 gxf3 ~xb3 33 ~d2 ~xf3 34
ctJxa7 gS, and the advancing pawns guaran-
tee Black a draw (3S ~g2 h6 36 ctJc6 ~e3 37
ctJd4 g4! 38 hxg4 f3);
2) 19 ... ~g7!? (with the king on g7 Black
has his problems, but on the other hand the
white queen cannot reach h6) 20 exd4 'iVb6
21 'iVc1 i1.xb3 22 axb3 f3! with counterplay.
18 ... ~h8
20 •.• ctJg5! (12)
This is the safest, since it forces drawing
simplification. In the event of 20 ... 'iVb6 21
~fc1 White completes the coordination of
his pieces and is ready to attack.
It looks dangerous to play 20 ... ctJd6 21
ctJxg6+ (after 21 exf4 ctJe8! 22 ~c8 ~xc8 23
ctJxc8 ctJf6! Black has no reason for com-
plaint) 21...hxg6 22 'iVxg6 ctJe8 23 ctJgS,
when 23 ... 'iVxgS?! 24 ~xgS ctJxc7 2S exf4 etc.
favours White. But 23 ... 'iVfS! 24 'iVhS+ ~g8
2S e4! ctJf6! 26 exfS ctJxhS 27 fxe6 b6 28 b3
~fS would nevertheless retain the balance.
19 i1.xe6 21 ctJh4! (28)
If 19 exf4?, then 19 ... ctJg3! is strong (20 To judge by the time spent by Karpov,
fxg3 'iVb6+). 19 ctJd4 'iVb6! 20 ctJxe6 fxe6 21 my 20th move came as an unpleasant
~c4 ctJd6 22 ~xf4 ~ae8 23 ~xf8+ i1.xf8 is also surprise to him, but he nevertheless found
dangerous for White. the best practical chance. After 21 ctJxgS
19 ...fxe6 (07) 'iVxgS 22 exf4 (22 'iVe4 fxe3 is equal)
After 19 ... 'iVb6? 20 ctJdS 'iVxe6 21 ctJxf4 22 ... l1xf4! 23 l1xb7 l1af8 or 23 ... i1.d4 (but not
101
Kasparov vs. Karpov: 1986-1987
Dlugy's move 23 ... l::!.e8?! because of 24 'iYcl!) lLle7+ with perpetual check, but not 24
Black has excellent compensation for the 'iYxg6? 'iYe5!) 23 ... fxe3 24l::!.xf8+ l::!.xf8 25 'it>h3
pawn. (again avoiding the draw after 25 lLlhxg6+)
With the move in the game White guar- 25 ... e2 (25 ... 'iYh5 26 lLlexg6+ hxg6 27 'iYxg6
antees himself a draw by perpetual check - 'iYh6 28 .l:i.e7 'it>g8 29 l::!.xe6 'iYxg6 30 lLlxg6 l::!.f2
for the moment he cannot hope for more. is also sufficient, or 25 ... l::!.fl 26 l::!.c8+ .if8 27
21 ...lLlxh3+?! (12) l::!.xf8+ l::!.xf8 28 lLlhxg6+ 'it'g7! 29 lLlxf8 'iYh5+),
I could not resist the temptation to make nevertheless forcing White to sue for peace
this showy (but, alas, second-rate) move, - 26 lLlexg6+ (26 'ii'e4? 'ii'h5 with the threat
although by 21...fxe3! I could have forced a of ... g6-g5) 26 ... hxg6 27 lLlxg6+ 'it'g8 28
draw: since Black is clearly better after 22 lLle7+;
fxe3? lLlxh3+! 23 gxh3 'iYg5+ 24 lLlg2 .ie5! 25 2) 22 ... fxe3! (an invitation to 23 lLlhxg6+)
l::!.d7 (other moves are even worse) 25 ... 'iYg3 23 'it'xh3! in the hope of 23 ... e2? 24 l::!.h1! (24
26l::!.f4 .ixf4 27 exf4l::!.xf4 28 'iYe1 (or 28 'iYc2 'iYe4?, with the mating idea 24 ... exfl'ii'? 25
~af8) 28 ... 'iYxe1+ 29 lLlxe1 .l:i.af8, all that lLlhxg6+, loses to the very pretty stroke
remains for White is to give perpetual 24 ... l::!.f3+!! - a unique diverting sacrifice!)
check - 22 lLlhxg6+ hxg6 23 lLlxg6+ 'it>g8 24 24 .. .'~'h5 25 lLlexg6+ hxg6 26 'iYxg6 'ii'e5 27
lLle7+ (24 lLlxf8? e2 25l::!.e1l::!.xf8) etc.. l::!.xb7 l::!.xf2 28 l::!.xg7 'ii'xg7 29 'it>g3 with
However, I was seized by a desire to winning chances. But the accurate 23 ... l::!.xf2!
play brilliantly. I assumed that the draw again forces 24 lLlhxg6+ hxg6 25 lLlxg6+ 'it>g8
would not run away, but as a result I 26 lLle7+ 'it>h8! 27 lLlg6+ with a draw.
devised the only way of creating real 23 lLlexg6+ (04)
problems for myself. It would have been naIve to expect 23
22 'it'h2! 'it>xh3? g5!. Which knight White captures
The knight is taboo: 22 gxh3? 'iYg5+ 23 with on g6 is of no particular importance
lLlg2 f3 and wins. (slightly later he can reach the same posi-
tion), but even so 23 lLlhxg6+ hxg6 24 'ii'xg6!
was more correct, maintaining the attack-
ing construction with the knight on e7.
Now 24 ... 'iYa5? 25 lLld5 is bad for Black,
while if 24 ... 'iYe5? the brilliant manoeuvre
25 'it'xh3! l:tf6 26 'it>g4!! decides matters.
Polugayevsky recommended 24 ... 'iYh7, but
after 25 gxh3! l::!.f6 (25 ... .ie5 26 l::!.c5 or
25 .. .fxe3 26 'iYxe6 is even worse) 26 'ii'xh7+
(26 'iYg4 fxe3 27 fxe3! is also strong)
26 ... 'it>xh7 27 l:tg1 White has a clear advan-
tage: 27 ... l:tf7 28 lLld5! l:taf8 29 l:txf7 l:txf7 30
lLlxf4 .ixb2 31 l:tg5 e5 32 lLld3 .ic3 33 'it>g3
22 ... 'iVhS? with an extra pawn and winning prospects.
A small mistake gives rise to a bigger The only defence is 24 ... l:tf5! 25 'ii'xh5+
one ... Here, to maintain equality, Black by l:txh5 26 gxh3 fxe3 27 lLlg6+ 'it'g8 28 fxe3
now had to display some resourcefulness: .ixb2 29 l:txb7 .ie5+ with good chances of a
1) 22 ... lLlxf2!? (a provocative sacrifice) 23 draw, despite the pawn deficit: 30 lLlxe5 (30
l::!.xf2! (23 lLlexg6+ hxg6 24 lLlxg6+ 'it>g8 25 lLlf4 l:th7!) 30 ... l:txe5 31 l:tff7 l:txe3 32 l:tg7+
102
The Third Match: 1986
~f8 33 Mbf7+ ~e8 34 Mxa7 Mxa7 35 Mxa7 e5 1) 25 Mfc1 (not 25 Mxg7? fxe3+ and
or 31 I:tg1 + ~f8 32 l:i.g3 a5 with the idea of ... 'iVxg7) 25 .. .fxe3+ 26 ~xh3 exf2 27 'iVg4! (27
.".l:!.d8(c8). The weakness of White's sepa- 'iVd3? 'it'g8! and wins) 27 ... 'it'g8 28 tDf3 (28
rated pawns significantly hinders the tDg6? loses to 28 ... 'iVe3+ 29 g3 fl'iV+! 30 Mxfl
conversion of his advantage. But Black Mxfl) 28 ... 'iVf6 29 ~g3 Mad8 with equality,
would still have had to survive as far as or 27 .. .'iiVe3+!? 28 g3 'iVh6 with the promis-
this endgame ... ing idea of a counterattack by ... .uf6 and
23 ... hxg6 ... Mg8;
Another critical moment. 2) 25 'ii'c2 fxe3+ 26 ~xh3 ~g8 27 f4 e2 28
Mel Mxf4 29 Mxe2 'iVh5 30 g3 'ii'g4+ 31 ~h2
Maf8, and Black's chances are not worse: 32
Mxg7+ (32 .l:!.xb7? .ufl!) 32 ... ~xg7 33 gxf4
'iVxh4+ 34 'it'gl Mxf4 or 34 ... Md8 35 Mg2+ ~f7
36 'iVc7+ 'it'f8!;
3) 25 Mxb7 tDxf2! 26 tDf3 'Y!lVf6 27 'iVh5+
'iVh6 28 'Y!lVxh6+ iLxh6 29 Mxf2 fxe3 or, more
precise, 27 ... ~g8 28 Mxf2 fxe3 29 Me2 Mab8
with equality.
After White's incorrect move the initia-
tive passes to Black, and it is now he who is
thinking of winning. Why did Karpov
spend so much time and still commit an
24 'iVxg6? (03) oversight? Perhaps he was reproaching
The chain reaction of mistakes contin- himself for his mistake on the previous
ues. This hasty move, made almost without move ...
thinking, is an obvious oversight: Karpov
overlooked my reply. 24 tDxg6+ ~g8 25
tDe7+ ~h8 (25 ... ~f7? is bad because of 26
tDf5+ ~f6 27 tDxg7 'iVe5 28 'iVh7! or 27 ... 'iVh4
28 g3! 'iVh6 29 g4! and tDh5+) 26 'iVg6! was
correct, transposing into the advantageous
position which could also have arisen after
23 tDhxg6+.
24 ...'iVe5! (03)
The only reply, but an adequate one (in
contrast to the timid 24 ... 'inl7? 25 ~xh3!
fxe3 26 f3 etc.). Black eliminates all danger,
and it is now White who has to play accu-
rately. 25 ... MXf7! (13)
25 l:i.f7?! (41) Simple and strong. 25 ... tDg5? was bad in
This move came as a complete surprise view of 26 'iVh5+ ~g8 27 tDg6! with the
to me. My calculations involved White's threat of tDe7 mate or tDxe5, as was
numerous other possibilities, and in the 25 ... ~g8? 26 tDf3! 'iVxb2 27 Mbl. 25 ... fxe3+?!
first instance I was seeking equality, for was also insufficient: 26 ~xh3 Mxf7
example: (26 ... exf2? 27 tDf3!, while if 26 ... e2 27 Mel
103
Kasparov vs. Karpov: 1986-1987
.uxf7 28 'iVxf7 'iVxb2, then 29 ~h1!!) 27 ~xf7 31 ttJxe6 with equality. Now, however,
'it>h7 (27 ... "iVf6 28 fxe3!) 28 ~e1, winning a White has problems.
pawn and retaining the initiative. 30 ttJd6 fxe3 (the bishop, which for a long
26 ~xf7 time has been asleep, finally wakes up!) 31
ttJC4!
Already the only move: White must get
rid of the insidious pawn on e3. After 31 f4?
~d8 or 31 ttJxb7? ~f8 32 ttJc5 (32 Wg3 ~e5+)
32 ... ~xb2 33 ttJe4 .i.d4 it would have be-
come a powerful passed pawn.
31 ... exf2 (06)
Now the e6-pawn becomes passed. Less
was promised by 31...b5 32 ttJxe3 ~xb2 33
~b1 ~e5+ 34 g3 a6 35 ~c1 with a probable
draw. However, with accurate defence by
White the game continuation should have
led to the same result.
26 ...ttJgS! 32 ~xf2 bS H ttJe3 as
Perhaps Karpov had been hoping for
26 ... iVb5 27 ~c1 (in the variation 27 ttJg6+
Wh7 28 ttJe7 ~f8 29 ~g6+ Wh8 the limit of
White's dreams is a draw) 27 ... ttJg5 28
~5+ 'it>g8 29 exf4 ~xb2 30 ~c7 'iVf6 31 ttJg6
ttJf7 32 ttJe7+ with perpetual check?
27 ttJg6+ (not the impulsive 27 "iVxb7? on
account of 27 ... fxe3+ 28 f4 'iVe4!) 27 ... Wh7 28
ttJxes ttJxf7 29 ttJxf7
104
The Third Match: 1986
34 ~g3 (08) 34 ... a4 35 Me2 Mf8 36 ~g4 (05) bxc3 Mc4 46 'It>e3 ~g4 47 ~d3 ~xg3 48 !1e2.
36 ... .id4 (07) Of course, I should have displayed
36 ... -Ith6 37 Mc6! (37 Me2 Mf4+ and ... Mb4) greater persistence - in matches I was
37... Mf4+ 38 ~g3 Me4 was slightly stronger, usually rewarded for this. But with
although here too after 39 ttJd5 Me5 40 ttJc7 37 ... -Itxe3 I hurried to secure victory in the
'It>f5 41 ~f3 Me3+ 42 ~f2 the position is a London half of the match (after all, in the
draw. 12th game I had the white pieces). An
37 Me2 important landmark, nevertheless ...
Not 37 Mc6? in view of 37... Mf2 38 ttJc2 38 Mxe3 Mf2 39 b3
-Itxb2 39 Mxe6+ ~f7 40 Mc6 Mxg2+ 41 ~f3 The most precise, although White also
Mh2, and the endgame a pawn down is lost holds on in the event of 39 g3 Mxb2 40
because of the poor position of the knight. Mxe6+ or 39 Mxe6+ ~f7 40 Wg3 (40 !1b6
Mxg2+ 41 ~f5 Mxb2 42 Mb7+ will also do)
40 .. Jhb2 41 Mb6 b4 42 'It>f4!, and the threats
of ... a4-a3 and ...b4-b3 are not dangerous.
39 ...MXg2+ 40 ~f3 Mxa2 41 bxa4
If White greatly desired, it was still pos-
sible to lose: 41 Mxe6+? ~f7 42 Mb6? (42 Me3
is essential, agreeing to 42 ... a3 43 b4 Mal)
42 ... Ma3!.
105
Kasparov vs. Karpov: 1986-1987
to both players for the 11 th game, the I was in no doubt that Karpov would
English displayed their customary tact and choose the Queen's Gambit, and I decided
political correctness, once again confirming to again play for a small advantage, by
their neutrality. And as a result Karpov and repeating the variation from the 10th game
I both received 64 gold sovereigns from the and the 23rd game of the previous match.
time of Queen Victoria! The outcome of the opening in them could
At the time the Observer wrote: 'Kasparov not have satisfied my opponent, and I was
and Karpov displayed the greatest mastery curious to see what improvement he would
accessible to human reason; in fact it lies beyond prepare.
human reason, because it is not possible to learn 4 ... liJf6 5 iLg5 h6 6 iLxf6 iLxf6 7 e3 0-0 8
to play like this. They play at such a level, at such 1:1c1 c6 9 iLd3 liJd7 100-0 dxc4 11 iLxC4
a degree of complexity, that it is accessible only to
them. Even when some of their games are close to
conclusion, world famous grandmasters are
unable to decide which of them will win.'
However, a careful analysis of the 11 th
game shows that such a rapturous assess-
ment of our play was an exaggeration. Of
course, this game was visually very interest-
ing - outwardly it was highly spectacular,
but the quality of the play by both sides in
the second half of it leaves much to be
desired ... Incidentally, Karpov also rated it
very highly, and even compared it with the
statue of Venus. A curious comparison, but 11 ... c5!?
if one supposes that for each mistake an arm A very interesting moment! This ex-
was cut off a wonderful statue, it would be tremely rare move (instead of the usual
more appropriate to put in its place the 1l ... e5 - Game Nos.lO, 22) shocked me: I had
Indian god Shiva, with its many arms ... analysed it in detail for Black and prepared
After the turmoil that preceded it, the it as my main counter to the variation with
12th game seemed rather insipid and it 6 iLxf6. And here Karpov is the first to
became the quietest of those played in employ 'my' variation ...
London: neither of us wanted to take a risk, I liked the move 11.. .c5 for the reason
and we were already thinking about the that it changes the character of the play and
second, Leningrad half of the match. leads to more 'Queen's Gambit-like' posi-
tions than 11...e5; for the moment the
bishop at c8 is shut in, but on the other
Game 12 hand the diagonal of the bishop at c4 is not
G.Kasparov-A.Karpov lengthened and the f7-point is not weak-
World Championship Match, ened. Black exchanges c-pawn for d-pawn,
12th Game, London 27.08.1986 and if the opponent does not do something
Queen's Gambit D55 extraordinary, he will quickly solve the
problem of developing his c8-bishop and
obtain a comfortable position without any
1 d4 d5 2 c4 e6 3liJC3 iLe7 4liJf3 weaknesses.
106
The Third Match: 1986
But suddenly I had to uphold the varia- position, but at that time I evaluated it in
tion for White and make the moves which I favour of White. As did Karpov! He de-
considered the most unpleasant for Black. cided not to give his opponent an isolated
12 'iUe2!? pawn (which he always used to do with
Before this there were only two known relish), but followed our analysis, although
practical examples - 12 ct:Je4 cxd4 13 ct:Jxf6+ the subsequent complicated play harbours
(13 ct:Jxd4 ct:Jb6 is equal) 13 ... 'iUxf6 l/2-lIz a definite strategic risk for Black. This
(Ivkov-Geller, Sochi 1983), and 13 ... ct:Jxf6 14 decision also surprised me, since it was
'iUxd4 'iUxd4 15 ct:Jxd4 ~d7 with equality atypical of Karpov.
(Fuster-Foltys, Munich 1941), while here 13.l:i.fd1
Psakhis and Dorfman (Minsk 1986) agreed
a draw. 12 ~b3 cxd4 13 exd4 b6 (Huzman-
Timoshchenko, Novosibirsk 1986) and 12
dxc5 ct:Jxc5 13 b4 ~xc3 14 .l:i.xc3 ct:Je4
(A.Petrosian-M.Gurevich, Baku 1986) are
also not dangerous.
13 ... cxd4
In the event of 13 ... b5?! 14 dxc5! bxc4 15
c6 Black loses a pawn for unclear compen-
sation - possible, and not bad, but such
play is not at all in Karpov's style!
14 ct:Jxd4
12 ... a6 Here the capture 14 exd4 is weaker, giv-
With the intention of ... cxd4, ... b7-b5 and ing White only an outwardly threatening
... ~b7 - again Karpov acts exactly as I was position: in fact, after 14 ... b5 15 ~b3 (15
planning to play! I was hoping for 12 ... cxd4 'iUe4 .l:i.b8 16 ~d3 g6 is no better) 15 ... 'iVb6!
13 exd4 ct:Jb6 (after 13 ... a6 14 ~d3 Black Black deploys his pieces comfortably, and
does not have 14 ... b5? because of 15 'iUe4) 14 the 16 d5 breakthrough is neutralised by
~d3 with some pressure for White: 14 ... g6 16 ... ct:Jc5 17 'iUe3 .l:i.b8! 18 dxe6 (18 ~c2 .l:i.d8
(14 ... ~d7?! 15 'iUe4) 15 ~e4 .l:i.b8 16 .l:i.fdl is equal) 18 ... ct:Jxb3! 19 exf7+ .l:i.xf7 20 'iUxb6
ct:Jd5 17 ct:Je5 ct:Je7 18 ~f3 ~d7 19 d5 (our .l:i.xb6 21 axb3 ~b7. In the 3rd game of the
analysis of that time) or 16 ... ~g7 (Shabalov- 1985 match Karpov carried out a similar
Pigusov, Tashkent 1987) 17 ct:Je5 ~d7 18 plan with Black, also not preventing d4-d5,
'iUf3 etc. and in my analysis I was guided by this
I now think that after 12 ... cxd4 (as well idea.
as after 12 ... 'iUe7 13 .l:i.fdl .l:i.d8 with the idea Now, however, White avoids an 'iso-
of ... cxd4 and ... ct:Jb6, Ribli-Spassky, Reggio Iani' and tries to exploit his lead in devel-
Emilia 1986) Black has a solid enough opment.
107
Kasparov vs. Karpov: 1986-1987
16 tLlf3
When I was proposing to play this
variation with Black, my trainers and I
mainly studied this continuation (the game
showed that it is not very dangerous) and
16 f4, after which 16 ... ~b8 followed by
... tLlf6 is good.
In the 14th and 16th games of the match
I opened with 1 e4, and in the meantime a
new idea of mine was analysed by our team
- 16 ~5!? (and if 16 ... 'it'h7, then 17 ~b3 g6
18 ~3 tLlb6 19 f4! ~g7 20 tLlf3 with quite 17 ~d2!
good attacking prospects). It was prepared Combining pressure on the d-file with
108
The Third Match: 1986
the plan of switching the bishop to the long :s.d8 with approximate equality.
diagonal. Logically I should have played 19 ctJc5,
17 ... bS 18 .iLe2 ctJf6 trying to exploit Black's weaknesses, but I
This solid simplifying move was the one had no wish to check how far Karpov had
we were counting on in our home analysis. gone in his 'parallel' analysis. And, since 19
lS ... ctJe5?! is obviously worse: 19 ctJd6 .iLxd6 ctJd6 i.xd6 20 ~xd6 ~xd6 21 :s.xd6 .iLb7 is
20 ~xd6 ~xd6 21 :s.xd6 ctJxf3+ 22 .iLxf3 :s.a7 absolutely harmless (22 :s.c7? ctJe8), I chose
23 b4!, and the difference in the activity of a sideline, guaranteeing White a durable
the two sides' pieces is enormous. position, although without any particular
But it was possible to play 18 ... .iLb7!? 19 pretensions.
~xd7 ~xd7 20 :s.xd7 i.xe4 21 ctJd2 .iLg6 19 ... ~xf6 20 ~d4
(21.. ..iLd5?! 22 e4 .iLxa2 23 :s.a1 .iLc4 24 .iLxc4
bxc4 25 ctJxc4 favours White) 22 .iLf3 :s.a7 23
:s.xa7 .iLxa7 24 :s.c6 :s.d8, and Black main-
tains the balance, even after losing a pawn:
25 ctJb3 a5! 26 :s.a6 (26 ctJxa5 :s.d2) 26 ... a4 27
:s.xa7 axb3 28 axb3 :s.d3 or 25 ctJf1!? b4 26
:s.xa6 .iLc5 27 :s.c6 i.e7.
109
Kasparov vs. Karpov: 1986-1987
110
The Third Match: 1986
to be wrong: it was in the second half of the matches, there was no reason to revise this
match that the main events were to de- figure for a return match. As the match had
velop ... been divided into two halves, the same
However, despite all these problems, I must be done with the prize money: 36,000
was pleased with the overall nature of the roubles in Leningrad and 36,000 (this time
contest: I couldn't see how Karpov would in foreign currency roubles) in London.
be able to win by a two-point margin in From this it followed that Kasparov and
Leningrad. What's more, during the break I Karpov could dispose only of this sum. In
calmed down somewhat, since I thought other words, out of the 691,000 Swiss
that I was playing better than Karpov and I francs, which at the official rate of exchange
was intending to increase my lead. was then 290,000 foreign currency roubles
The first part of the match was splen- (paid to us by the London organisers in the
didly organised, largely thanks to the ratio of five eighths to the winner and three
generous financial assistance bequeathed eighths to the loser), the USSR Sports
by the Greater London Council before it Committee paid into the Chernobyl fund
was 'buried' by the Thatcher government. only' our' 36,000 foreign currency roubles!
At the Closing Ceremony the former Prime There is an obvious difference between
Minister Lord Callaghan made an ironic 290,000 and 36,000, but still more obvious
comment about this: 'Shakespeare is not in this case was the difference between
always right', he said, quoting from Julius human morality and bureaucratic morality.
Caesar: 'The evil that men do lives after them. On 29 August, a month after the start of
The good is oft interred with their bones.' the match, we left London. As the plane
The prize fund for the London half of gained height and set course for Leningrad,
the match could indeed have helped many I glanced down at the broad London river
people. The point was that back in the below and wondered: 'All's quiet on the
spring of 1986, when I heard about the Thames, but what's in store for me on the
Chernobyl tragedy, I announced my will- banks of the Neva?'
ingness to donate the foreign currency part Karpov was in a different mood: 'The
of my future prize to the purchasing of London half of the return match left an unpleas-
medical supplies for victims of the nuclear ant taste in the mouth. The organisers devoted
reactor explosion. Soon afterwards, Karpov most of their attention to Kasparov, and for
did the same. But our perfectly natural them I was merely an annoying necessity for the
human response did not meet with any performance. And so I really wanted to annoy
understanding at either FIDE or the USSR them with my play! - but it didn't happen. I
Sports Committee. had ideas, and I played no worse than Kasparov,
Initially Campomanes categorically de- but in the fifth hour of play I would become
manded that, as before, one per cent of the dejected, get into time-trouble, and begin
prize fund for each draw should be paid making incredible oversights ... A point behind, I
into FIDE's coffers. Then it was the turn of flew to Leningrad, and on the way I caught a
the Sports Committee bosses to demon- cold; I was still not in the mood. And how I
strate their financial wizardry. We were lacked that - being in the right mood!'
presented with the following chain of Karpov and I flew in the same plane,
logical deduction. Since the USSR Council and the two of us played cards with my
of Ministers had decreed that there should trainers, to pass the time and to try - alas,
be a prize of 72,000 roubles for the Moscow unsuccessfully - to put chess out of our
111
Kasparov vs. Karpov: 1986-1987
minds for at least the few hours of the Karpov, with the support that he enjoyed from
flight. When we landed in Leningrad, it the city authorities, would not have had any
was noticed by the journalists that the difficulty in finding a quiet place further away
champion left the airport in an ordinary from us. But when we arrived in Leningrad we
Volga with a Baku number plate, whereas learned our opponents had set up residence in
Karpov was driven off in an official Chaika the most inconvenient place on Kamenny Island
limousine escorted by a military traffic - but just a few hundred metres from us!
police vehicle. Straight away the local Alongside their house were a tumour clinic and
authorities made it clear which of us was a factory, teeming with rats. In addition, there
the outsider. Subsequently too the organis- were other people living in this house. This was
ers made sure I didn't forget that I wasn't a strange choice, most unnatural for Karpov,
playing on my home ground. Whereas in who was not accustomed to being inferior to his
Moscow the situation was now balanced opponent in any way. '
and I had plenty of supporters, in Lenin- The match itself was held in the Concert
grad I found myself in an inconceivably Hall of the Leningrad Hotel. Sevastyanov,
hostile environment. who here for the last time (as I imagined)
So, from the banks of the Thames we was representing the USSR Chess Federa-
had moved to the banks of the Neva. A tion in his capacity of Chairman, said to the
new city, with new impressions. Our team press: 'We hope that the second half of the
was put up in a three-story mansion on match staged in Leningrad will be just as
Kamenny Island (again, as in the previous creative, even more interesting in the competi-
match, I was helped by the Trades Union tive sense and equally well organised as the
leader Stepan Shalaev). Karpov settled in London part of the event.' But not all the press
nearby, on the same island. It was an representatives felt comfortable here, and
excellent place for walks, and both contest- some 'sinners' never received their accredi-
ants and their trainers frequently availed tation for the match.
themselves of this opportunity (but Karpov The journalist Yuri Rost wrote in Lit-
and I bumped into each other only once - eraturnaya Gazeta: 'Technically, one can have
before the 24th game). For the first time in no complaints about the organisation of the
three matches, the headquarters of the match. The hall is marvellous, there is plenty of
opposing sides were in close proximity, in light, the chairs are comfortable, the press centre
direct line of vision! That we unexpectedly is superbly equipped and communications have
found ourselves to be neighbours was been established. So what is the problem? Only
Karpov's wish: he had declined the out-of- when your pass has been checked a hundred
town residence that had been made ready times, only when you are made to leave your
for him. notebook, folder, newspaper or God knows what
Nikitin: 'This decision of Karpov seems else at the desk and you enter the hall do you
suspicious in the extreme. Both in London and realise what the problem is: you are under
in Seville the residences of the two 'Ka's were suspicion. Most of the audience are not those
separated as widely as possible, which is under- who really wanted to see the match, but those
standable: with such enormous and prolonged who succeeded in getting through. Those who
nervous tension, even a chance encounter when get in have been carefully selected by special
out on a walk provokes a mass of negative personnel, so numerous that you gain the
emotions. Our place of stay in Leningrad was impression that what is taking place is not a
decided on before the start of the match, and gathering of admirers of two outstanding chess
112
The Third Match: 1986
players or lovers of the wisest of sports, but a match were so marked, that the 13th game
convention of underworld leaders.' provoked strong associations with the first.
I wasn't aware of all this, since I was not At 17.00, using long poles, the demon-
in the auditorium, but on the stage, at the strators reproduced White's first move on
chess board. But I was aware that Lenin- the large demonstration boards. After the
grad was Karpov's kingdom and that I 'computerisation' of London, this proce-
needed to keep my wits about me, to pre- dure looked extremely archaic. But the
vent him from gaining any home advan- Soviet Chess Federation and the Sports
tage. Committee proudly declined the offer by
I later found out how difficult it had the English to provide (for free!) all the
been for the head of my team to arrange the necessary equipment for the Leningrad half
catering and see to other everyday matters. of the match. They supported Campomanes
He even had to write an official letter to the in every way, whereas the prestigious
city authorities. Of course, many things British Chess Federation was adopting a
were kept from me, but now I know just neutral position and calling for a change of
what my circle of people had to endure FIDE President at the forthcoming Con-
both then, and in the previous two matches. gress in Dubai. The result was that techni-
As the years pass, more and more is gradu- cal progress did not reach Leningrad ...
ally coming to light about the various
methods that were used in a cold war, one
which could not have been waged without Game 13
official support... It never ceases to astonish A.Karpov-G.Kasparov
me how we managed to endure it all and World Championship Match,
not falter. 13th Game, Leningrad 05.09.1986
Back at the negotiation stage, through GriJnjeld Defence 079
my youth I had insisted on the rapid re-
sumption of the match - a week's break
seemed an eternity (now I think that this 1 d4 ttJf6 2 c4 g6 3 ttJf3
was necessary, and I would also have In the interval we intensively analysed
benefited from the acclimatisation). The various lines with 3 ttJc3 dS: in particular,
thorny question of the maximum duration of course, 4 .i.f4 (Game Nos.S, 9, 11), and
of the break was solved by Karpov in a then also 4 ttJf3 ~g7 S 'iYb3 (Game Nos.lS,
very simple way: he prolonged it at the cost 17, 19, 42, 48) - Mikhail Gurevich, who
of a time-out, when on 3 September (the came to help me, demonstrated here some
day that the 13th game should have taken new ideas for White. But the time for the
place) he presented a doctor's certificate main battles in the Griinfeld Defence had
confirming that he had a cold. However, it not yet arrived ...
seemed to me that the organisers were also 3....i.g7 4 g3
counting on a 'cold', since on 2 September, Following the 3rd game - for the mo-
at the traditional examination of the venue ment Karpov is willing to be satisfied with
before the start, I found it unprepared. a small advantage. Before the game we
But by S September all was ready, and hardly looked at this variation, considering
the match began. Or more precisely - its it to be second-rate.
second half began, but the changes in the 4 ... c6 (I again chose the symmetric varia-
situation and the very atmosphere of the tion, which at the time was reputed to be
113
Kasparov vs. Karpov: 1986-1987
114
The Third Match: 1986
of 12 ... f5 13 ct:Je5 ct:Jb6 (13 ... ct:Jxe5?! 14 dxe5 is with excellent play) 15 ... ct:Jc4 16 exd5 exd5
unfavourable for Black) there is the re- 17 ct:Jxd5, when White supposedly has the
source 14 b3 i.d7 15 "ii'd3 ct:Jc8 16 i.c1!, advantage: 17 .. :~xd5? 18 ct:Je5 "ii'b5 19 a4!
which gives White the initiative (Kasparov- ct:Jb4 20 "ii'b3 "ii'a6 21 ct:Jxc4 i.e6 22 "ii'xb4 etc.
Topalov, advanced chess, Leon 1998). However, later Black gained a draw
13 i.f2 with the ugly 17 ... ct:J6a5 or 17... ct:Jd6 (Khari-
Consistent play. Karpov made this tonov-Nunn, Leuwarden 1995), although
move quite quickly, although here too 13 he still had some problems. In Informator
b3!? i.d7 14 i.c1 is interesting, restricting No.64 Nunn also considers 17 ... i.e6(!) 18
the knight on b6 and activating the bishop ct:Jc3 ct:Je7(?!), while the computer suggests
(Chiburdanidze-Xie Jun, Groningen Candi- an important improvement - 18 ... l:!.c8! 19
dates 1997). l:!.fel i.f7, and Black is perfectly alright!
I did not see this equalising possibility
either at the board or in later years. And,
judging the variation with 10 .. .£6 to be
favourable for White, before the match in
Seville we worked out another system of
defence - with 1O ... ct:Jc6 (although most
probably it was no better than 10 ... f6).
13 ... fS
Played after some thought. At that time
many decisions in the opening were taken
largely by intuition - thus here I did not
want to allow e2-e4. The difference in the
placing of the black pieces in the variations
13 ... i.d7 14 e4! dxe4 15 ct:Jxe4 (Karpov-
Gelfand, 4th match game, Sanghi Nagar 14 ct:Jes i.d7
1995; Game No.101 in Volume V of My Great The manoeuvring battle that now com-
Predecessors) and 12 e4 is, of course, in mences promises to be protracted. White
White's favour. has a slight initiative (mainly due to the
13 ... ct:Jc4!? looked tempting, with the difference in the strengths of the knights on
idea of 14 "ii'c2 ct:Jd6 (Sakaev-AI Modiahki, e5 and b6), but the character of the position
Singapore 1990) or 14 ct:Jd2 ct:Jxd2 15 "ii'xd2 does not promise him much, so long as the
f5, when the exchange of the b6-knight for opponent is patient and accurate. One
the powerful f3-knight is an enormous would think that, with his style, Karpov
achievement by Black. But I avoided this could not have wished for anything more,
because of the 'dangerous' 14 e4 ct:Jxb2 15 but in this match he did not play the set-up
"ii'c2 (if 15 "ii'b3 there is 15 ... ct:Jd3 16 exd5 with g2-g3 again.
ct:Ja5 17"ii'b5 ct:Jxf2 18l:!.xf2 i.d7 19 "ii'd3l:!.c8 15 "ii'd2
115
Kasparov vs. Karpov: 1986-1987
16 .. .'it'hS
This move prepares 17 ... lLld6 (17 lLlxc6
~xc6! 18 ~xe6?? l:leS 19 ~f7 lLld6) and is
useful for carrying out the plan with ... g6-
gS. Black is simply forced to seek counter-
chances on the kingside (I think that Kar-
pov did not take these chances seriously).
ls ... lLlcS!? After all, after playing his knight to e4
IS .. :~e7 followed by ... l:lfcS and ... ~eS is Black 'equalises' the strength of the minor
seemingly more natural, but after b2-b3 all pieces, but in the meantime White will
the same the problem of the passive knight seize control of the c-file. The ... g6-gS
at b6 would have to be solved. And 1 de- advance has the aim of clearing the g-file
vised a plan which may change the content for the operation of the black rooks.
and direction of the play after the knight 17 l:lfd1
manoeuvre to e4 and the future possible
advance ... g6-gS (when White is diverted
by actions on the queenside). The same aim
is pursued by IS ... <t>hS 16 b3 lLlc8 17 l:lael
~e8 18 l:lfdl lLld6 19 ~el ~e7 (Sturua-
Nataf, BielI999).
16~e3
By trying to hinder Black's regrouping
(16 ... lLld6? 17 lLlxc6 bxc6 18 lLla4), White
wants to continue the game at a slow
tempo. The alternative was the preparation
of g3-g4, but in itself this plan is effective
only when Black is engaged in operations
on the opposite wing, as otherwise it leads White prepares ~el and is not in a
merely to simplification: 16 h3 lLld6 17 g4 hurry to occupy the c-file, assuming that
lLle4 18lLlxe4 fxe4 19 l:lael l:lc8 20 l:lc3lLlxeS one or two tempi in such a situation will
21 l:lxc8 ~xc8 22 dxeS b6 23 )::tel ~7 24 not decide anything. However, if 17 l:lfel
~d4 l:lc8 2S e3 ~bS with equality. there was the satisfactory reply 17 ... lLld6 IS
116
The Third Match: 1986
b3 Mc8 19 .ie1 CLJe4. tage was the clever 25 .ib4!? In the event of
17 ... CLJd6 18 b3 Mc8 25 ... gxf4 26 CLJxf4 Me8 27 .ixe4 fxe4 28 Mfl
Not 18 .. .'iVe7? 19 CLJxc6, and Black is .ig5 29 'iYf2 White's chances are slightly
obliged to recapture with the pawn better even after the exchange of all the
(19 ... .ixc6 20 .ixd5). rooks on the c-file: 29 ... a5 30 .id2 'lWf6 31
19 Macl .ie8 20 .iel Mc2 .id7 32 Mfd Mxc2 33 Mxc2 Mc8 34
Intending the following plan: CLJa4-b2- Mxc8+ .ixc8 35 h4 .ih6 36 .ie3, and his
d3, CLJxc6, CLJe5 and ~b4, increasing the control of the dark squares causes the
pressure on the queenside. It is time to act! opponent a certain discomfort. But after
20 ... .if6! (an important move) 21 CLJa4 b6 25 ... Mg8 26 CLJe5 .ie8 Black has a solid
22 CLJb2 CLJe4 enough position: 27 Mxc8 'lWxc8 28 .ixe4
dxe4 29 CLJc4 b5 30 CLJd6 'iYc2 or 29 fxg5
~xg5 30 'iYc3 'iYxc3 31 .ixc3 'it'g7.
25 ... gxf4 26 gxf4 .ie8
There was no sense in hurrying:
26 ... .ixe5 27 fxe5 Mg8 28 .ig3, and White's
position is the more pleasant. But now he
has to play accurately.
Over the last ten moves Black has
achieved more: the minor pieces are
equivalent, and although the c-file has been
lost, the g-file has been opened and is ready
for use. After ... .ixe5 and ... Mg8 it will be
possible to create serious threats along it -
23 CLJbd3 after all, this is a file opposite the white
A crucial moment. It is clear that Black king! Why did this happen? The point is
has made preparations for 23 ... g5, so now that I have not wasted a single tempo (both
Karpov hurries with his knight manoeuvre, the e8-bishop and the king are in the right
not wishing to weaken his king's defences place), whereas Karpov's play has been
by 23 h4, which would forestall the devel- slow and not altogether concrete.
opment of the opponent's counterplay.
Then 23 ... h6 and ... g6-g5 would expose the
black king too much, but by 23 ... Mg8 and
... Mg7-c7 Black would have held the c-file,
for example: 24 CLJbd3 Mg7! 25 g4 (25 a4
Mgc7) 25 ... .ixh4 26 CLJxc6 .ixc6 27 g5 ~xe1
28 Mxe1 Mgc7 with equality.
23 ... g5!
From now on the rook on f8 no longer
has to perform nonsensical manoeuvres:
after ... gxf4 it will find work on the open g-
file (while if White recaptures on f4 with a
piece, this will weaken his control of e5).
24 CLJxc6 ~xc6 25 CLJe5 27 'iVh3
The last attempt to fight for an advan- White's main trump now is his more ac-
117
Kasparov vs. Karpov: 1986-1987
tive queen. In the event of 27 J::txc8 'iVxc8 28 31...J::tc7 32 J::txc7 'iVxc7 33 e3 ..txe5 34 fxe5
.l::tc1 i¥b7 29 'iVh3 (with the idea of ..tM; it is 'iVc2 35 'iVh4 'iVd3+ 36 ..te2 'iVxe3 37 'iVd8
inappropriate to play 29 J::tc2? J::tg8 30 'iVc1 'iVh3+ 38 <;t>gl with a draw.
"iWg7 31 e3 ..th5, when White's position is
already difficult) 29 ... J::tg8 (it is also possible
to prevent the exchange of bishops -
29 ... 'iVe7, and if 30 J::tc8, then 30 ... ~g8 31 <;t>f1
i¥b7, evicting the rook) 30 ..tM ..txM 31
'iVxM 'iVg7 32 'iVh3 ..th5 33 J::tc2 ..tg4 34 'iVh4
..th5 White must accept the peace offer (35
'iVh3), since after 35 e3? ..td1! 36 J::tb2 'iVc7
Black's activity becomes dangerous (37
'iVh6? ..tf3! and wins).
27 ... J::tg8 28 <;t>f1
And again a non-concrete, 'generally
useful' move. 28 'iVh6! was more energetic,
restricting Black somewhat and gaining a 30 ... as?!
draw without difficulty after 28 ... J::tc7 29 Better chances were offered by the im-
J::txc7 "iWxc7 30 <;t>f1 "iWc8 31 J::td3 (31 ..txe4 mediate 30 ... .txe5!? 31 dxe5 J::tg4, since 32
..tg7!) 31.. ...tg7 32 'iVh4 'iVc1 (32 ... ..tf6 33 e3? loses ignominiously to the unexpected
'iVh6) 33 ..txe4 dxe4 34 J::tc3 or 28 ... iixe5 29 32 ... ..tb5+! 33 <;t>gl d4!, for example: 34 ..td6
fxe5 J::tg6 30 'iVf8+ (30 J::txc8?! 'iVxc8 31 J::tc1 (34 'iVxh5+ <;t>g7 35 'iVh3 ..te2) 34 ... d3 35
'iVa6 is unfavourable for White) 30 ... J::tg8 31 'iVxh5+ <;t>g7 36 J::tc7+ 'iVxc7 37 .i.xc7 d2, or 34
'iVh6 J::tg6 32 'iVf8+. "iWf3 as! 35 ..td6 (35 h3 axb4!) 35 ... d3 36 h3
Karpov probably thought that the 'iVh6 d2 37 J::td1 'iVh4.
manoeuvre would not run away from And if 32 ..tf3 it was possible to sacrifice
him .. . the exchange by 32 ... a5 33 ..tel J::txf4 34 'iVg2
28 ... J::tXC1 29 ~xc1 hS! d4! 35 ..tg3 J::tg4 36 ..txg4 hxg4 with a rather
Paradoxical and very strong - from the unpleasant initiative, although after 37 J::td1
purely chess, and the psychological point of ..tb5 38 h3 White nevertheless holds on:
view (an audacious pawn move away from 38 ... tt:Jc3 39 J::td2 'iVg5 40 'iVa8+ <;t>g7 41 i¥b7+
the king!). Black tries to restrict the white <;t>g6 42 'iVc8! or 38 ... 'iVg5 39 J::txd4 ..tc6 40
queen (it is essentially out of play) and the J::txe4 ..txe4 41 'iVh2 <;t>g8 42 hxg4 'iVxg4 43
roles of the two sides are reversed - it is 'iVh4 with a draw.
now White who has to think of how to 31..ta3?
equalise (in impending time-trouble!). A serious mistake. The loss of two tempi
30 iib4?! would be unpleasant, but after 31 ..tel
An 'old-schedule' move, which contains Black did not have anything real (say,
considerable dangers for White. It would 31...J::tg7 32 ..txe4 dxe4 33 'iVc3 with a prob-
appear that Karpov felt unsure and had not able draw). The withdrawal of the bishop
noticed the change of scene. 30 ..tf3! was to the wing strongly affects the entire
more cautious, when it is hard for Black to situation around the white king - Black
create direct threats: 30 ...h4 31 .l:'tc6 "iWe7 32 begins a swift attack.
J::tc8 tt:Jd6 33 J::ta8, or 30 ... J::tg7 31 a4 (31 ..txe4 31 ... ..txes! 32 dxes (32 fxe5 'iVg5 was even
dxe4 32 'iVc3 <;t>h7 33 ..tf2 is also possible) worse) 32 ... J::tg4 33 ..txe4
118
The Third Match: 1986
33 ... dxe4?
It was my turn to go wrong in time-
trouble. The threat of including the queen
in the attack is parried by White by giving
up a pawn. A far more powerful attacking
potential with opposite-colour bishops
would have been retained by 33 .. .fxe4!, 34... ~xf4+
effectively giving Black an extra piece and How can one avoid capturing a pawn
the constant threat of ... d5-d4. For example: with check, especially in time-trouble! But
1) 34 e3 a4! (a brilliant reply, preparing nothing better is apparent: 34 ... e3 is parried
the invasion of the queen) 35 ~c6 axb3 36 by 35 'iY'f3! with a draw after 35 ... 'iVh4
axb3 'iVa8! 37 ~xe6 'iVxa3 38 ~xe8+ 'it'g7, and (35 ... i.d7 36 h3) 36 ~c8 ~xf4 (36 ... ~g8 37
the white king is defenceless; h3) 37 ~xe8+.
119
Kasparov vs. Karpov: 1986-1987
Game 14
G.Kasparov-A.Karpov
World Championship Match,
39 ... h4!? 14th Game, Leningrad 08.09.1986
For some reason I judged this time- Ruy Lopez C92
trouble move to be dubious, whereas in fact
it is the last - and an excellent! - chance of
success. By 39 ... a4 40 h4 .i.d7 Black would 1 e4 e5 2 ct:Jf3 ct:Jc6
simply have forced a draw (41ttc7 .i.b5). I was intending once again to test
40 h3! Black's defences in the Petroff Defence -
Karpov instinctively made the only against 2 ... ct:Jf6 (Game No.6) a new concep-
move! If 40 .i.c7?!, then 40 ... h3! 41 .i.xb6 tion with 3 d4 had now been prepared. But
'it>h7 is very strong, and White is again on Karpov pleasantly surprised me ...
the verge of defeat, despite the opposite- 3 .i.b5 a6 4 .i.a4 ct:Jf6 5 0-0 .i.e7 6 tte1 b5 7
colour bishops. 42 .i.xa5? ttg1 + 43 'it>d2 .i.b3 d6 8 c3 0-09 h3
l:i.g2+ 44 'it>e1 .i.h5 is bad for White, while if The classical 'Spanish', that inexhausti-
42 ttd8 Black also activates his rook - ble mine of original strategic plans, subtle
42 ... ttg1 + 43 'it>f2 ttg2+ 44 'it>e1 .i.b5 45 l:i.d2 positional manoeuvring, swift attacks and
120
The Third Match: 1986
counterattacks, was Karpov's favourite pov chose the solid but passive 13 ... tiJb8 14
opening. Since his youth he felt confident in ~d3 c6, but now he goes in for the most
it, irrespective of what colour he was play- topical and sharp continuation: Black
ing. The range of openings played in our concedes the centre, obtaining in return the
games was quite diverse, but up till then in b4-square for his knight, and the possibility
the 'Spanish' I had not managed to win: in of beginning active play on the queenside
three games I had achieved highly promis- and then attacking the opponent's pawn
ing positions, but they all ended in draws, centre. It should be remembered: at that
and one game I even lost. Therefore this time the theory of this variation was only
next dispute was of a crucial nature, and beginning to take shape, and many of the
we prepared very seriously for it. subsequent moves, which nowadays young
players make automatically, were found
only with great difficulty.
14 cxd4 tiJb4 15 ~bl cS
The main reply, although we also al-
ready had experience of playing the varia-
tion 15 ... bxa4 16 .l:!.xa4 a5 17 .l:!.a3! (Kas-
parov-Beliavsky, Moscow Interzonal 1982;
Balashov-Karpov, 50th USSR Champion-
ship, Moscow 1983). Life has shown that if
17 ... g6 the best is 18 .l:!.ae3 (Sax-Beliavsky,
Moscow Interzonal 1982), if 17 ... 'iVd7 - 18
tiJh4 (Ehlvest-Beliavsky, 51st USSR Cham-
pionship, Lvov 1984) or 18 tiJh2, and if
9 ... ~b7 10 d4 .l:!.e8 11 tiJbd2 ~f8 12 a4 17 ... .l:!.a6 - 18 tiJh2 g6 19 f3! (Kasparov-
The main line. Others are examined in Karpov, 2nd match game, New York 1990).
Revolution in the 70s (pp.243-256). 16 dS
12 ... h6 (12 .. .'iVd7 - Game Nos.50, 57 in To fight for an advantage White has to
Kasparov vs. Karpov 1975-1985) 13 i..c2 close the centre. In the game Kasparov-
One of the numerous Spanish tabiyas. Balashov (Kislovodsk 1982), one of the first
on the given theme, after 16 dxc5 dxc5 17 e5
tiJd7! it transpired that it was difficult for
White to complete his development. He did
no better with 16 b3 cxd4 17 tiJxd4 bxa4 18
.l:!.xa4 a5 19 ~b2 g6 20 ~c3 .l:!.c8 with equal-
ity (Tseshkovsky-Balashov, Minsk 1982).
16 ...tiJd7
Black is hoping to exploit the position of
the knight on b4 to undermine the centre
by ... £7-f5, or, after ... c5-c4, to occupy d3-
point. After 16 ... g6 17 tiJf1 (Sax-Greenfeld,
Luzern Olympiad 1982) or 16 ... bxa4?! 17
.l:!.xa4 (Grunfeld-Frey, Luzern Olympiad
13 ... exd4 1982) it is more difficult for him to create
In the 9th game of the 1985 match Kar- counterplay.
121
Kasparov vs. Karpov: 1986-1987
Analysis diagram
122
The Third Match: 1986
We analysed 19 ...'iVb6 20 tLlfS a great the passed d3-pawn is strong!) 24 tLlf3 ~c4
deal and considered it promising for White 2S ~xd3 tLleS 26 ~e3 tLlxf3+ 27 gxf3 ~xa4 28
(a part of this analysis is still topical to- tLlc3 ~b4 or 27 ... ~xa4 28 tLlc3 ~d7 etc.;
day!), but modern theory and practice do 2) 22 ~e3!' Before the 1987 match we de-
not confirm our optimism. Nevertheless, in cided that this was the main way to fight
his 'parallel' analysis Karpov rejected this for an advantage, for example:
line and prepared a far from obvious pawn a) 22 .... tLleS 23 tLl4f3! fS (23 ... tLlc4?! 24
sacrifice, which I had overlooked. It was ~xd3 Ivanchuk-Kruppa, Frunze 1988) 24
clearly aimed at unsettling me, but here he tLlxeS ~xeS 2S i..b2 ~e7 26 exfS!? (we looked
was unlucky: on encountering a new posi- at 26 ~xd3 fxe4 27 ~g3 with a small plus)
tion and forced to work out its subtleties 26 ... ~xe3 27 fxe3 i..xdS 28 ~g4, retaining
directly at the board, I did not become the initiative (Glek-Kharlamov, correspon-
flustered and, despite a couple of slips, I dence 1988);
played one of my most subtle positional b) 22 ... tLlcS 23 i..b2 ~aS (if 23 ... i..c8, then
games ... 24 tLlc6 ~4 2S i..d4 Anand-Beliavsky,
20 bxa3 tLld3 21 i..xd3 cxd3 Madrid 1998) 24 tLlfS! g6 2S tLlg3 i..g7 (also
after 2S ... i..e7!? because of the weakness of
the long diagonal the black king cannot feel
safe) 26 i..xg7 'it'xg7 27 tLlb3 'iVxa3 28 tLlxcs
~xcS 29 ~xd3 with somewhat the better
chances for White (Kotronias-Gligoric,
Niksic 1997).
However, 22 i..b2 proved to have one
undoubted virtue: Karpov thought for a
long time over his reply, and it became
clear that he had not considered this move
at home (years later Igor Zaitsev confirmed
that the ex-champion had focused all his
attention on the variations with 22 ~e3 and
Here I thought for a long time - it was especially 22 tLlxbS).
essential to find a way to retain the initia-
tive.
22 i..b2 (17)
White sets his sights on the opponent's
kingside. At the time I attached an exclama-
tion mark to this move, but White also has
two other possibilities:
1) 22 tLlxbS. Although I took my oppo-
nent 'at his word', Black does indeed have
good compensation for the pawn: 22 ... ~aS!
(22 ... i..a6?! 23 tLld4) 23 a4! (23 tLld4 ~c3!)
23 ... ~c8 (23 ... tLlb6 24 .ia3!; 23 ... i..a6 is not so
clear: 24 tLlc4 'iVb4 2S tLlcxd6 i..xd6 26 i..a3
~xa3 27 tLlxa3 i..xa3 28 ~d2 i..d6 with three 22 ... ~a5! (4S)
minor pieces for a queen and two pawns - The correct reaction. 22 ... tLlcS suggested
123
Kasparov vs. Karpov: 1986-1987
itself, but then by 23 ~e3! I could have However, 24 ttJe3! is far stronger, for ex-
transposed into a favourable position from ample:
the previous note. 1) 24 ... ..ta6 (24 .. :iVa4? 25 'iVbl! ttJc5 26
23 ttJfS! (30) ttJg4 is altogether bad for Black) 25 ttJb3
A 'squabble' over the winning of the far- 'iVa4 26 'iVxd3 'iVxe4 27 'iVd2 'iVf4 28 ttJc1!,
advanced black pawn is inappropriate here, retaining some advantage;
as the following lines are weak: 23 ~e3 2) 24 ... ttJe5 25 ttJb3 'iVa4 26 ..txe5!, and
ttJc5(e5), as well as 23 ttJ2b3 'iVa4 24 'iVxd3 after 26 ... dxe5 27 'iVxd3 h5 28 ~c1 or
..txd5 or 23 ttJ2f3 ttJc5 24 ttJb3 'iVa4! 25 ttJxc5 26 ... ~xe5 27 ttJg4 ~e7 28 ttJf6+! Wg7 29
'iVxdl 26 ~xdl dxc5 27 ~xd3 ~xe4 28 d6 'iVxd3 b4 (29 ... Wxf6? 30 'iVc3+, while if
~e8, and Black's chances are already better. 29 ... 'iVxa3, then 30 ~bl) 30 'iVe3 'iVxa3 31
Little is also given by 23 ttJ4b3 'iVa4 ~bl White continues playing for a win
(23 ...'iVc7!?) 24 'iVbl ttJb6 25 ~e3 (25 'iVxd3 without any risk;
..txd5) 25 ... f5 26 'iVxd3 ttJc4!, forcing perpet- 3) 24 ... ..tg7 25 ttJb3 'iVa4 26 ..txg7 Wxg7
ual check: 27 ttJxc4 bxc4 28 'iVd4 ~xe4 29 27 'iVxd3 ~xe4 28 'iVc3+, and Black's defence
~xe4 fxe4 30 'iVxe4 'iVxb3 31 'iVe6+. is difficult: 28 ... ttJf6?! 29 ct:Jd2 ~e5 30 ct:Jf3
~e4 31 ~bl ~f4 32 ct:Jg4 'iVc4 33 'iVaI or
28 ... ttJe5 29 ttJa5 b4 30 axb4 'iVxb4 (30 ... ~xb4
31 f4!) 31 'iVaI ~f4!, holding on with literally
the only moves.
It is most likely that Karpov sensed
simply intuitively that after 23 ... g6 Black
would have a dangerous position, and he
chose a more solid move.
24..txes!
I replied very quickly, although in such
positions one's hand naturally reaches out
towards the f-pawn. But after 24 f4 ct:Jc4!
(more active than 24 ... ttJg6?! 25 ~f1) White
23 ... ttJes! (20) could have hoped only for a draw: 25 ttJxc4
On his last two moves White had spent (not 25 ..txg7? 'iVxd2 26 'iVe4 'iVxe 1+ 27 Wh2
nearly an hour, and Black even more - in a ttJe3! 28 ttJxe3 ..tc8! and wins) 25 ... bxc4 26
critical position the choice of a move is ttJxh6+ (alas: 26 ..txg7? ..tc8! 27 ..txf8 Wxf8
always difficult. Karpov blocked the dan- and wins) 26 ... gxh6 27 'iVg4+ Wh7 28 'iYf5+
gerous diagonal with his knight, and this with perpetual check, or 25 ct:Jb3 ct:Jxb2 26
was the correct decision. ct:Jxh6+! (26 ttJxa5 is less good in view of
In the book Ova matcha I recommended 26 ... ttJxdl 27 ~xdl ..ta6 28 ttJg3 h5)
23 ... g6, 'in accordance with the spirit of 26 ... Wh7! 27 ct:Jxa5 ct:Jxdl 28 ~xdl gxh6 29
fighting for the initiative': 24 ttJb3?! 'iVa4 25 ttJxb7 ~xe4 30 ~xd3 ~xf4 31 ~b3 ~a4 with
'iVxd3 ttJe5 (but not 25 ... gxf5? 26 'iVg3+ Wh7 equality.
27 'iVf3 ttJe5 28 'iVxf5+ Wg8 29 ~e3 with a 24 ... dxes (14)
strong attack for White) 26 i.xe5 (26 'iVg3 It is unfavourable for Black to sacrifice
..txd5!) 26 ... ~xe5 27 f4 ~e8 28 ttJg3 ..tg7 29 the exchange - 24 ... ~xe5 25 ct:Jf3 ~xf5 26
~c1 ~c8 or 25 ... b4!? with more than suffi- exf5 ..txd5, since after 27 ~e3 ..tc4 28 'iVc1
cient compensation for the pawn. he does not have sufficient compensation.
124
The Third Match: 1986
125
Kasparov vs. Karpov: 1986-1987
However, in the computer era it is not 'iVb5) 32 ~c3 'iVb8, and Black's fortress is
difficult to establish that the energetic 27 difficult to approach.
d6! was the only way to retain the advan- 28 ct:Je3
tage:
1) 27 ... g6 28 ct:Je7+ 'itg7 29 l:tc1 .l:td8 30
'iVg3 'ith7 31 ct:Jc5 l:txd6 32 ct:Jxb7 ~xb7 33
ct:Jd5 l:tc6 34 UbI!, and with queens and
rooks on the board, the' eternal' knight on
d5 makes it hard for Black to defend;
2) 27 ... Ud8 28 'iVg3 g6 29 J:i.e3! i..xd6 30
ct:Jxh6+ 'itg7 31 Uf3! or 29 ... Ue8 30 M!, and
Black still needs to extinguish the flame of
the opponent's initiative: 30 ... 'ith7 31 'iVf3!
'itg8 32 ~g4 h5 (32 ... 'ith8?! 33 Ug3) 33 'iVg5
'iVd8! 34 ct:Jc5 i..c6, or 30 ... Ue6 31 h5 'ith7 32
hxg6+ fxg6 33 'iVg4 Uf6 34 'iVh4 g5 35 'iVh5
i..xd6 36 Ud3 i..f8 37 'iVe8 'itg8 38 ct:Jd2 Ue6 28 ... i..xa3!
39 'iVh5 Uf6 etc. Again the best move. 28 ... h5 was rec-
And today, taking into account the ob- ommended, but after 29 ct:Jc2 it is now bad
jective evaluation of the position, I am to play 29 ... i..xa3? because of 30 ct:Jxa3 Uxa3
reluctantly forced to assess the move 27 Uc1 31 d6!, and otherwise there follows l:tbl and
as dubious. ct:Jd2-f3, when Black has no compensation
for the pawn.
28 ... Uxa3 also looked insufficient in
view of 29 ct:Jg4 Ua2 (29 ... i..g7 30 d6) 30
'iVbl! with an obvious advantage to White
(Ova matcha), but after 29 ... f6! 30 'iVc2 h5 31
ct:Je3 Ua8 (31...'iVa7!?) 32 'iVc7 'i¥xc7 33 Uxc7
i..c8 the advantage evaporates. More prob-
lems are set by 29 'iVc2, for example:
29 ... h5?! 30 ct:Jc5 i..c8 31 ct:Jd3 'iVd4 32 'iVxc8!
Uxd3 33 'iVe8! 'iVxe4 (33 ... Uxe3? 34 Uc8) 34
Uc7 'iVf4 35 g3 ~f6 36 Uc8 'iVg7 37 'iVxb5,
and Black's position is unenviable. But
29 ... Ua8 30 ct:Jg4 ~d6 31 'iVd2 f5 32 ct:Jxh6+
27 ... g6 i..xh6 33 'iVxh6 fxe4 and ... i..xd5 is more
The normal reply. 27 .. Jha3? (weaken- solid, with good drawing chances.
ing the back rank) would have lost after 28 29 Ual J:i.a4
d6 Ua8 29 ct:Jc5 l:tc8 30 'ith2! or 28 ... g6 29 29 ... i..d6? is weak: 30 Uxa8+ i..xa8 31
lLle7+ ~g7 (29 ... ~h8 30 Uc7!, while if lLlg4 h5? 32 lLlf6+ ~h8 (32 ... ~g7 33 ct:Je8+,
29 ... 'ith7, then 30 'iVf3 f5 31 ct:Jxf5! is also ct:Jxd6 and 'iVxb5) 33 'iVd2 with a decisive
good) 30 'iVg3 'it f6 31 Ud 1. attack. White would have been left with
But 27 ... i..xa3 28 'iVg3 'iVf6 was not as some advantage after 29 ... i..f8 30 Uxa8
bad as I thought, for example: 29 Ual i..f8 i..xa8 31 Wic3 'iVb8 32 ct:Jg4 i..g7 33 Wie3! (33
30 Uxa8 i..xa8 31 'iVd3 ~6 (31...b4?! 32 ct:Jc5 h5 34 ct:Je3 'iVc8 is less good) 33 ... h5 34
126
The Third Match: 1986
tiJh6+! .¥t.xh6 35 'iVxh6 or 33 ... 'it>h7 34 'ii'f3 extra pawn: 36 f3 'it>f8 37 'it>f2 ..\ta6 38 'it>e1
'iUa7 35 tiJf6+ 'i&th8 36 tiJe8. 'it>e7 39 tiJd1 'it>d6 40 'it>d2 f5! 41 'it>c3 ..\te2 42
Therefore the most accurate continua- tiJe3 f4 43 tiJg4 Jl.f1 44 'it>xb3 ..\txg2 45 tiJf2
tion would have been to restrict the knight ..\txf3 with a draw.
by 29 ... h5! 30 'iVc3 f6. Now in the event of 31 However, Karpov did not yet see the
d6 'iVc6! it is unclear how White can need to find a way to save the game - with
strengthen his position, while after 31 g4!? 30 ... ..\tf8 he was undoubtedly pinning his
(breaking up the opponent's defences on hopes on 31 tiJxe5? ~g7 32 tiJd7 Mxa1+ 33
the kingside) 31...hxg4 32 tiJxg4 Ma4! (but tiJxa1 'iUc7, but ...
not the passive 32 ... Jl.e7?! 33 Mxa8+ ..\txa8 34
'iUc8+ 'iUd8 35 'iVa6) Black has an equal
game: 33 Md1 'iUd6 or 33 'iUg3 ..\tf8 34 Mel
~c8, while if 33 Mxa3 there can follow
33 ... Mxa3 34 'iVb4 Mxb3 35 'iUxb3 ..\tc8 or
even 33 ... b4 34 'iVf3 Mxa3 35 tiJxf6+ 'it>g7 36
tiJe8+ 'it>g8, forcing a draw.
After 29 ... Ma4 the position remains
roughly equal, but Black has some prob-
lems to solve.
30 tiJg4
31 ~c1!
Here this move is indeed strong. With
subtle manoeuvres White has impercepti-
bly achieved complete domination - the
rook 'pendulum' has disrupted the har-
mony in the opponent's ranks. Black's
defensive problems were also aggravated
by severe time-trouble.
30 ... ..\tf8?
Simple solutions no longer work and
Black needed to display some ingenuity -
30 ... h5! 31 tiJxe5 ~b2!, when 32 Mxa4 (32
tiJd7?! Mxa1+ 33 tiJxa1 'iUa5 34 tiJc2? 'iUc7 35
'iVb3 ..\tg7 36 'iVxb5 .i.c8 or 34 tiJb3 'iVe1 + 35
'it>h2 ..\tc8! is unfavourable for White)
32 ... bxa4 33 tiJc4 'iVxb3 34 'iUxb3 axb3 35
tiJxb2 M! (fixing the enemy pawns on light
squares) would have led to a study-like
endgame with knight against bishop, 31 .. :iVd6? (08)
where White is unable to convert a sound Worn out by the complicated struggle,
127
Kasparov vs. Karpov: 1986-1987
128
The Third Match: 1986
44 d6!, for example: 44 ... .\ta5 45 CLlc4 .\td8 46 Before the 15th game I took a time-out.
g3+ ~xg3 47 CLlxe5, or 44 ... .\tb4 45 d7 .\te7 46 Not only in order to 'digest' the important
~d3 ~g3 (46 ... ~g5 47 ~c4) 47 CLlc4 etc. win and restore my emotional energy, but
Black resigned without resuming the also for additional preparation. In the
game (1-0). Times: 2.32-2.28. Griinfeld Defence a storm was expected -
the competitive situation obliged Karpov to
In this 14th game there was a genuine try and make maximum use of the white
'Spanish' battle, with a clash of strategic pieces.
ideas. First there was a fierce skirmish on Just before the start of play, something
the queenside, and with an interesting strange occurred. I was asked, as was
pawn sacrifice Karpov managed for a time Karpov, apparently, to agree to a little
to seize the initiative there. Then came a ceremony involving FIDE Vice-President
swift change of scene - a demonstration of Tuleda. Due to the ex-champion'S habit of
activity on the kingside by White led after arriving late, this ceremony began a couple
the unexpected 24 .i.xe5! to a favourable of minutes before the clocks were due to be
stabilisation of the centre and the creation started. Campomanes came onto the stage
of a protected passed pawn. Black slightly and introduced Tuleda to the audience,
overstepped the mark (26 .. J:ta8?!), which said something with the most radiant smile,
White failed to exploit (27 .l:tc1 ?!), after and then Karpov and I were presented with
which Black tried too directly to break badges from the Venezuelan Chess Federa-
through on the queenside and through tion ... This was an absurd and ridiculous
over-ambition he lost his objectivity spectacle, and the game began some 5-6
(29 ... .l:ta4; 30 ... .\tf8?; 31...'iVd6?). Severe minutes later than usual.
punishment followed (31 .l:tc1!; 32 CLlc5!), During the ceremony I noticed how irri-
and a catastrophe became inevitable. tated Karpov was by all this - he was dying
Even today I rate this win among by to get on with the game itself! And, of
best creative achievements, especially since course, not without having something up
it was gained on 'Spanish' territory. Yes, his sleeve.
Karpov made a few errors, and on the 27th
move I also played inaccurately, but in
general to gain decisive positional gains Game 15
original play was demanded of White ... A.Karpov-G.Kasparov
'However, what is a correct game? There are World Championship Match,
hardly any games played ideally by both sides, 15th Game, Leningrad 12.09.1986
since people are not machines. There can only be Griinfeld Defence 098
various degrees of mistakes.' (Karpov)
Thus I increased my lead to two points 1 d4 CLlf6 2 c4 g6 3 CLlC3 ds 4 CLlf3 .\tg7 5 'iVb3
(8-6), although my opponent again easily And so, no continuations such as 5 .\tf4
found his bearings in an opening position (Game No.1) or 4 .i.f4 (Game Nos.5, 9, 11) -
that was new to him and skilfully avoided White confronts the Griinfeld Defence
my home preparation. Therefore the joy of head-on. It was said that at this moment the
victory was tinged with a feeling of alarm. leader of Karpov's delegation confiden-
After all, Timoshchenko had now left, but tially informed the journalists in the press
the miracles were continuing ... centre: 'Today we begin!'
129
Kasparov vs. Karpov: 1986-1981
130
The Third Match: 1986
to play like Fischer, having prepared his initially only 15 jLd4 'iUc6?! 16 jLxg7 (16
own surprise here. jLc5! Mc8 17 jLh3 Me8 18 0-0 b6 19 jLe3 is
9 ... CDc6 better), but Gurevich embarrassed me with
I swallowed the bait, but it would be in- the new move 15 J.e2!? - after 15 ... 'iUxb2 16
teresting to know how Karpov would have 'iYxb2 jLxb2 17 0-0 White's lead in devel-
responded to 9 ... CDb6. Possibly he had in opment and two bishops give him excellent
mind 10 'iUc5!?, but most probably (to judge compensation for the pawn (despite the
by Beliavsky's words) - 10 'iVb3, automati- draw in the game M.Gurevich-Lputian,
cally reaching the main position of the 54th USSR Championship, Minsk 1987). 15
variation, which we, of course, had also f4!? is also tempting: 15 ... 'iUxb2 16 'iVxb2
analysed. jLxb2 17 Mb5 jLc3+ 18 We2 b6 19 e5, then
jLg2 and Mel (Beliavsky-Krasenkow, Barce-
lona 2007). Karpov would probably have
played 15 f4 or 15 jLe2;
3) 10 ...CDc6 11 d5 CDe5 12 jLe2 CDxf3+ 13
gxf3 jLh5 - an extremely sharp position,
known since the middle of the 20th century.
Here White has tried both 14 f4 ~xe2 15
CDxe2 'iUd7 16 h4 c6 (Lilienthal-Bronstein,
Saltsjobaden Interzonal 1948), and 14 h4
~d7 15 a4!? (Smyslov-Botvinnik, 11th
match game, Moscow 1958; Game No.113 in
Volume II of My Great Predecessors), while
after the game Portisch-Gheorghiu (Manila
Analysis diagram 1974) 14 Mg1!? 'iUd7(c8) 15 Mg3 became
fashionable - mention can be made of fierce
Black has to choose from three continua- Sosonko-Timman duels (Amsterdam 1975;
tions: Wijk aan Zee 1981 and 1982; 2nd match
1) 10 ... e6 11 jLe2 CDc6, and although 12 game, Bergen 1984), as well as wins by
e5 has often been played here (the source Ivanchuk against Timman (Linares 1993)
game: Uhlmann-Larsen, 8th match game, and by Beliavsky against Ristic (Novi Sad
Las Palmas 1971), to me 12 CDg1!? seemed 2000) and Korneev (Moscow 2005).
more unpleasant (Polugayevsky-Bilek, When in the late 1990s I was preparing
Ludwigsburg 1969) - a move in the spirit of the 5 'iVb3 variation for White, this position
Karpov, who did not like to give up seemed to me to be mutually dangerous
squares. White retains some pressure, and I and unclear - all the customary positional
rejected this; guidelines have been violated. How was
2) 10 ... jLxf3 11 gxf3 e6. The weakening Karpov planning to handle it, in view of the
of White's pawn structure forces him to fact that it was not at all in his style? How-
play energetically. Nothing is given by 12 ever, if you foresee in advance that the
h4 CDc6 13 e5 h5. We looked at 12 d5! exd5 opponent will reply 9 ... CDc6, then White's
(Black does not equalise with 12 ... 'iUf6 13 f4 task is greatly simplified ...
or 12 ... 'iUe7 13 dxe6 fxe6 14 jLh3 Eingorn- 10 jLe2 CDb6
Beliavsky, 53rd USSR Championship, Kiev The rare 10 ... jLxf3!? 11 gxf3! e5 has not
1986) 13 CDxd5 CDxd5 14 Mxd5 'iUf6 and been refuted - perhaps that is what I
131
Kasparov vs. Karpov: 1986-1987
should have played in the 18th game. I should have tried to play the Grunfeld in
11 'iYcS! (if 11 'iYd3 the standard 11 ....i.xf3 12 a fundamentally different way, not forcing
gxf3 e5 is quite good, but also the creative exchanges and maintaining the tension. But
11...f5, 11 ... 'iYd7 or 11...'iYc8 Kir.Georgiev- in those days I considered this variation to
Kamsky, Khanty-Mansiysk 2007) 11 ...'iYd6 be completely harmless for Black, and all
our main effort was devoted to an analysis
of the expected 9 'iVb3 c5!?
12 eSt?
An original strategic (mixed up with
some tactical subtleties!) idea. 'At first sight
a paradoxical decision: White not only
hopelessly weakens his e5-pawn, but also
allows the exchange of queens. But all is
not so simple.' (Karpov) In the endgame
the knight on b6 unexpectedly experiences
serious discomfort.
Karpov played 12 e5 quite quickly, and
here I had to think: White's threats are very
In the Botvinnik-Fischer encounter, after obvious (12 ... 'iYd7?! 13 h3!), so that the
12 h3 .i.xf3 13 gxf3l:!.fd8 14 d5 (if 14 e5, then queens have to be exchanged, but what
14 ... 'iYxc515 dxc5l:!.xdl+ 16 'it'xdl ctJd717 f4 then to do with the knight?
g5! 18 fxg5 .i.xe5 is equal- Fischer) 14 ... ctJe5 12 ... 'iYxcS 13 dxcS
15 ctJb5 'iYf6! 16 f4 ctJed7 17 e5 'iYxf4! Black
achieved excellent play (Game No.77 in
Volume II of My Great Predecessors). In
addition, both 12 d5 and 12 0-0 (a recom-
mendation of Botvinnik) have occurred -
we also analysed these moves:
1) 12 d5 ctJe5 13 ctJxe5 ~xe2 14 ctJxg6
~xdl (14 ... 'iYxg6?! is unfavourable because
of 15 'it'xe2 f5 16 l:!.hel! - a computer move!
- 16 ...'iVh5+ 17 'it'd3 fxe4+ 18 'it'c2) 15 ctJxe7+
'it'h8 16 'it'xdl l:!.fe8 17 'iYxd6 cxd6 18 tLlf5
.i.xc3 19 bxc3l:!.xe4 20 .i.d4+ 'it'g8 21 f3l:!.ee8
22 ctJxd6 l:!.ed8 23 ctJxb7 l:!.xd5, and White
has not achieved anything; 13 ... ctJc8!?
2) 12 a-a!? This looks somewhat The retreat of the knight to c8 is strate-
stronger. We were not planning 12 ... .i.xf3 gically extremely risky, but Karpov had
13 gxf3 ctJd7 (13 ... e5? 14 ctJb5 Uhlmann- probably not studied it in his preparations
Jimenez, Marianske Lazne 1965) 14 'iYxd6 and was now forced to solve problems at
cxd6 15 f4 with some advantage to White, the board. The knight is aiming for e7, but
but rather 12 ... e6! 13 ctJb5 'iYxc5 14 dxc5 ctJa4 will it succeed in reaching there, has it got
15 b3 ctJc3 16 ctJxc3 .i.xc3, and Black should time? Nothing forcing for White is appar-
hold on. ent, and in addition for the moment the e5-
I now think that in the existing situation pawn is attacked.
132
The Third Match: 1986
13 ... itJd7 suggested itself, emphasising 16 .l:!.d7! (also in the event of 16 ... a6 17itJxc7
the weakness of the eS-pawn. After 14 e6 .l:!.a7 18 itJdS .ixb2 19 itJb6! e6 20 'it'e2 Black
~xe6 ISitJgS ~fd8 16itJxe6 fxe6 170-0 both is pinned to the ropes) favours White, as
17 ... itJf6 (as in the source game Zaltsman- does 14 ... itJxeS IS itJxc7 .ixf3 (IS ... .l:!.b8?! 16
Henley, New York 1983), and 17 ... ~xc3 18 itJxeS .ixe2 17 'it'xe2 .ixeS 18 ~d7) 16 gxf3!
bxc3 itJf6 (Dlugy-Kramnik, Internet (blitz) .l:!.b8 17 b4. In these variations exchanges
1999) are good. But after 14 h3!, according occur, but the knight on c8 remains a
to our analysis of the late 1990s, White still burden to Black.
retains some advantage: Since if 14 ... e6 there can follow IS ltd7
1) 14 ... ~xf3 IS gxf3! ~ad8 (IS ... itJdxeS? (but not IS itJxc7 .l:!.b8, which leads to the
16 f4, winning the knight) 16 f4 gS! 17 fxgS position in the game), Black made a move
itJdxeS 18 0-0 etc.; which restricts the opponent's choice.
2) 14 ... .te6 IS itJgS itJdxeS. An attempt lSitJxC7 e6!
to maintain the balance with the help of Another strong move, aimed at unnerv-
tactical trickery: 16 f4 itJc4 17 ~c1 itJa3! - a ing White: the eS-pawn will not run away.
brilliant counter! 16 itJxe6 fxe6 17 f4 is But now the knight at c7 is in danger of 'not
better: 17 ... itJf7 18 ~c4 (18 ~d7 ~ac8) 18 ... eS running away' (16 ... a6 is threatened) and in
190-0, or 17 ... ~h6 18 0-0 ~ad8 19 a3! (19 g3 addition the knight at c8 has a future.
itJb4! is unclear, Zigler-Pribyl, Gothenburg 16itJbs (forced)
1997) 19 .. J:Ixdl 20 itJxdl itJf7 21 ~c4 with
excellent compensation for the pawn,
although Black has a tolerable position.
14itJbS
The most plausible move, but it loses
the advantage. Better chances are given by
14 h3! with the idea of paralysing the
knight on c8, which the ex-champion
discovered only in painstaking home
analysis and tried in the 17th game.
16 ... itJSe7
Black's position has become quite attrac-
tive, and Karpov thought for a long time:
does White have any way of retaining the
initiative?
17.l:!.d2!?
A deep prophylactic move: the bishop at
e2 and the b2-pawn are defended, and the
possible doubling of rooks on the d-file is
prepared. For the moment White keeps his
14 ... ~b8! king in the centre, in order after exchanges
It would appear that Karpov underes- to exploit this plus. He needs to find a way
timated this reply. 14 ... ~xf3 IS ~xf3 ~xeS of simplifying that will turn the queenside
133
Kasparov vs. Karpov: 1986-1987
134
The Third Match: 1986
In the event of 20 etJfd4 ~xe2 21 'it>xe2 game of the Spassky-Fischer match (Game
etJfxd4+ 22 etJxd4 (22 ... etJxe5 23 b3!) No.102 in Volume IV of My Great Predeces-
22 ... etJxd4 23 ~xd4 ~xe5 24 ~b4 (aiming for sors), the bishop cannot be caught: 29 ..ixf5
the creation of a passed pawn on the gxf5 30 'it>f3 hxg3 31 fxg3 ..ig1 32 b4! ~c8 33
queenside) 24 ... ~fc8 25 ~c1 ~xc1 26 ..ixc1 ~d1 ..ih2 34 ~d2 with a repetition of
b5!? and ... 'it>f8 Black would have solved all moves, or 29 ... exf5!? 30 gxM (now 30 'it>f3
his problems. hxg3 31 fxg3 ..ig1 32 b4 does not work
However, with the unexpected 20 etJh4! because of 32 ... ~e8! 33 ~d1 ~e3+ 34 'it>g2
White would have exchanged a pair of ~e5 35 etJd6 ..ie3) 30 ... 'it>g7 and ... ~h8 with
minor pieces in favourable circumstances good counterplay.
and obtained a slight, but enduring advan- This was the last moment when any
tage: 20 ... ..ixe2 21 'it>xe2 etJxM 22 ..ixM kind of fight might have flared up .
..ixe5 23 ~c1 or 22 ... etJxe5 23 ~c2. 24 ... ~fd8 25 ~fd1 ~xd2 26 ~xd2 ~c8 27 g3
20 ... h6! (clarifying the intentions of the ~c1+ 28 'it>g2 'it>f8 29 ..ie4 'it>e7
bishop on g5) 21 ..if6
After 21 ..if4 g5 22 ..ig3 ~fd8 followed
by exchanges on g3 and f3 the black knight
goes to e5, and the ..ig7+etJe5 combination
is very effective. White wants to 'barter' his
dark-square bishop for the knight on c6,
rather than the one on f5 - but all the same
he cannot avoid the appearance of oppo-
site-colour bishops.
21 ... ..ixf3 22 ..ixf3 etJxe5 23 ..ixe5 (23 ..ixg7
etJxf3+ 24 'it>e2 etJxd2 25 ..ixf8 etJxb3 with
equality) 23 ... ..ixe5
135
Kasparov vs. Karpov: 1986-1987
made, but on one thing everyone agreed - a quiet development system with White, but
fierce battle was to be expected (there was it is not in my nature to avoid an open
also a direct hit: the Ruy Lopez, wild com- battle. Besides, I believed in the lucky star
plications and a severe time scramble!). of the 16th game, and I was dreaming of
However, it was not hard to predict this, creating another masterpiece. So that
bearing in mind how uncompromising the Karpov could have been in no doubt that
Leningrad encounters had been, and - the he would obtain the desired com plica-
number of the forthcoming game! tions ...
Yes, somehow the number 16 had a
magical effect on the two players. And this
was by no means mysticism, but quite Game 16
specific, even 'living' memories. After all, it G.Kasparov-A.Karpov
was in the 16th game of the previous match World Championship Match,
that the conclusive turning-point had 16th Game, Leningrad 15.09.1986
occurred. Remember with what inexorable Ruy Lopez C92
consistency Black tightened the blockade,
and how the white pieces, paralysed by the
wonderful knight on d3, suffocated within 1 e4 e5 2 ct:Jf3 ct:Jc6
their own territory ... There is no doubt that Of course, not 2 ... ct:Jf6 (Game No.6). The
such crushing defeats are remembered for a Ruy Lopez is far more suitable for the
long time not only by the winner, but also settling of accounts.
by the loser, especially since for Karpov the 3 iLb5 a6 4 iLa4 ct:Jf6 5 0-0 iLe7 6 J:tel b5 7
direct consequence of this defeat was the ~b3 d6 8 c3 0-0 9 h3 iLb7 10 d4 J:te8 11
loss of the title 'world champion', to which ct:Jbd2 iLf8 12 a4 h6
he had become so accustomed for ten years. Following the well-trodden path. Kar-
For this game too, despite the disap- pov demonstrates his readiness to continue
pointments of the two previous ones, the theoretical debate begun in the 14th
Karpov arrived in an extremely determined game.
mood, hoping for a full-scale revenge. A 13 iLC2 exd4 14 cxd4 ct:Jb4 15 iLbl C5 16 d5
year earlier I had gained a win in the 16th ct:Jd7 17 J:ta3!
game by using Karpov's favourite weapon
- domination, and now, in Leningrad, he
decided to fight in a radical, unusual way
for him, by heading for the most forceful,
critical continuations, involving great risk,
and for complicated piece play with the
likelihood of combinative explosions: the
idea being that victory on territory tradi-
tionally regarded as Kasparov's would be a
severe psychological blow for the oppo-
nent. And in general, why, given a bit of
luck, should this game not signal a turning-
point in the return match? Of course, I
could guarantee myself beforehand against 17 ... c4
any great problems, by choosing some And this is a familiar motif, evoking
136
The Third Match: 1986
memories of that 16th game: the black This idea came to me during the night
knight intends to establish itself at d3, after the 15th game, and our entire team
causing confusion in the white ranks. We began looking at it the day before the 16th.
made these opening moves quickly, virtu- Our collective analysis showed that White's
ally at blitz speed. attack is very dangerous - his well-
18 tiJd4 coordinated pieces literally fall on the black
Abandoning 18 axb5 (Game No.14) for king:
the moment, I followed the afore- 1) 21...axb5? 22 tiJf5 g6 23 'iVh5! tiJe5 24
mentioned A.Sokolov-Psakhis game, where ~xh6 ~c8 25 ~xf8 .l::i.xf8 26 'iVh6 ~xf5 27 exf5
I was intending to demonstrate an im- and wins, or 23 ... ~g7 24 tiJxh6+ ~xh6 25
provement that had just been found for ~xh6, and after both 25 .. .'iVc7 26 'iVh4 tiJxd5
White. 27.l::i.c1 and 25 ... tiJe5 26 ~e3 'i!Va5 (26 ... 'i!Vd8 27
f4) 27 .l::i.f1 Black is in a sorry state;
2) 21.. ..ic8. The best reply. Initially here
we looked at 22 ~xh6 axb5 23 b3 with
sufficient compensation for the piece, but
not 23 tiJf3? .l::i.al 24 tiJg5 tiJe5! (24 ... .l::i.xbl? is
incorrect: 25 'i!Vxbl gxh6 26 tiJe6+ 'it'h8 27
tiJxf8 .l::i.xf8 28 'i!V c1 ! Sax-Short, Subotica
Interzonal 1987) 25 'iVh5 g6 26 'iVh4 tiJbd3!,
repelling the attack. But in the end we
deemed the strongest to be 22 b3! tiJe5 23
~e3, when Black's defence is difficult:
23 ... 'i!Vd8 24 'i!Vd2 as 25 ~xh6 with three
pawns for a knight and the constant threat
After 18 ... tiJe5 19 axb5 'iVb6 instead of of tiJc6, or 23 ... tiJg6 24 f4! (A.Sokolov-
the modest 20 tiJ2f3 tiJbd3 (20 tiJf5?! tiJbd3 is Portisch, Brussels 1988).
also dubious, Oll-Goldin, Vilnius 1988) I After checking our old analysis, I have
had devised an unexpected knight sacrifice again ascertained that this sacrifice is very
- 20 tiJxc4!! tiJxc4 21 .l::i.g3, cutting the Gor- favourable for White, since it gives him
dian knot. several important pluses. His rook instantly
swings across to g3, while the black knights
are deprived of their strong point at d3, and
the knight on b4 begins to 'hang', coming
under attack with gain of tempo ('i!Vd2). For
the moment the extra piece does not play
any particular role - White continues to
build up the pressure on the kingside, and
in some cases he can even switch to a
purely positional course.
I remember my joyful anticipation:
would I succeed at last in employing in this
match a genuinely destructive novelty?!
Before the game I warned my mother:
Analysis diagram 'Today I'm going to sacrifice a knight, so
137
Kasparov vs. Karpov: 1986-1987
don't faint.' But my opponent proved ready accept pawns sacrificed by me. Now the
for this turn of events and he was the first eternal creative debate - 'material or the
to employ a novelty. initiative?' - is given a new impulse, but
18 .. :~f6! (06) unexpectedly with reversed 'colours'! It is
This apparently pointless attack on the apparent that Karpov was eager at any
knight at d4, dislodging White from his price to provoke a crisis in the match ...
home analysis, left me slightly shocked - The move 19 ... liJc5 was made without
again Karpov avoided a 'mine'! True, he hesitation, which is especially intriguing,
made the queen move without enthusiasm, since at that moment Karpov disregarded
and soon the impression was gained that he another, no less interesting possibility,
had prepared the entire variation in a which - and this is important - was more in
hurry. accordance with his style: 19 ... liJd3!? 20
19liJ2f3 (33) i..xd3 (nothing is given by 20 ~xd3 cxd3 21
White does not want to remove his axb5 axb5 22 liJxb5 ~al or 22 i..xd3 liJc5)
knight from the centre and lift the attack on 20 ... b4!. At the board this variation con-
the b5-pawn, and therefore he is forced to cerned me, and after the game I discussed it
block the rook's access to g3. Now, having with my trainers.
safeguarded his king, Black can make use It is clear that after 21 ~al cxd3 the
of his trumps, associated in particular with pawn sacrifice is a temporary one: 22 ~xd3
his outpost at d3. liJc5 23 ~c4 (23 ~1? i..xd5) 23 ... a5!
(23 ... liJxe4 is also not bad) 24 liJb5 ~ac8 25
liJa7 i..a6 (25 ... ~a8 26 liJb5 with equality) 26
~d4 ~xd4 27 liJxd4 ~a8 28 liJab5 ~xe4, or
22 liJb3 ~g6! 23 ~xd3 liJf6 24 liJh4 (24
liJbd2?! ..ixd5) 24 ... ~7 25 f3 i..xd5 26 i..d2
i..e6.
Therefore, in search of even a hint of an
advantage, White himself would have had
to decide on an exchange sacrifice:
1) 21 i..xc4 bxa3 22 b3 liJc5 23 ~c2 ~g6
24 liJh4 (24 i..xa3 ~xe4 with equality)
24 .. :iVf6 25 liJhf3 ~g6 V2-V2 (Anand-
Kamsky, 5th match game, Las Palmas 1995)
19 ... liJC5 or 22 b4 liJe5 23 i..b3 liJxf3+ 24 liJxf3 '¥IVaI!,
For the second time in succession in a and White has to keep an eye on the a-
similar position, Karpov goes in for a pawn pawn, which is still alive: 25 ~d4 ~1 26
sacrifice, hoping in return for indefinite liJd2 a2 27 ~2 ~xb2 28 i..xb2 i..c8 29 i..xa2
positional compensation! An occurrence i..d7 and ... ~ab8, or 25 liJd4 ~ac8 (25 ... a2 26
which, frankly speaking, was very rare for ~d2) 26 liJc2 ~xc2! 27 ~xc2 ~c8 28 ~dl
him: 'pawn sacrifice for the initiative' was i..e7 29 ..if4 ~c3 30 b5 axb5 31 axb5 'iVb4 32
never one of Karpov's favourite chess ~e3 ~xb5 - at any event, Black has nothing
procedures. And over the long history of to fear;
our encounters he showed himself to be a 2) 21 ~b3!. The best chance - a para-
staunch supporter of a material advantage doxical exchange sacrifice, which I devised
and at any convenient moment he would before the 1987 match.
138
The Third Match: 1986
Analysis diagram
139
Kasparov vs. Karpov: 1986-1987
21 ... ~xa3 22 tDxa3 27 ... tDcS! (but not 27 ... tDxc1? 28 ~a3 'ii'e7 29
After 22 bxa3?! tDbd3 Black quickly re- 'ii'xa8 .i.xc4 30 'ii'c8 tDe2+ 31 ~h1 .i.bS 32
gains the pawn and he has the resource 'iWb8 and wins) 28 b4 .i.xc4 29 bxcS dxcS 30
.. :~i'al. .i.d2, and White has 'merely' an extra pawn
which is difficult to convert (30 ... ~a6 31
'iWb7 ~b6 32 .i.aS ~xb7 33 .i.xd8 fS 34 tDd2
.i.a2 3S exfS .JtxdS is possible, with drawing
chances).
23 ~e3!
Essential prophylaxis. White prepares in
advance for the appearance of a black
knight on d3.
22 ... .i.a6
Before the invasion at d3 Black tries to
strengthen his position to the maximum,
although 22 ... tDbd3 23 .i.xd3 cxd3 would
also have given him good compensation for
the pawn: 24 tDd2 'ii'd4 2S tDb4 'iWb4 26 tDc3
g6, or 24 b4 tDxe4 2S 'ii'xd3 'ii'g6 (threaten-
ing ... tDgS!) 26 tDh4 (26 ~h2 fS) 26 ... tDg3 27
'ii'd1 ~xe1+ 28 'ii'xe1 'ii'e4 29 'ii'xe4 tDxe4 30 23 ..•~b8
tDfS .i.xdS with a probable draw. The final preparation for the invasion.
But here 23 ... tDxd3?! (instead of Karpov played this move (which was not
23 ... cxd3) would have allowed White to possible after 20 ~ee3) without particular
consolidate successfully - 24 ~e3! .i.a6! thought, and later it became the standard
(after 24 ... tDxb2 2S .Jtxb2 'ii'xb2 26 tDxc4 or one.
24 ... tDcS 2S 'ii'c2 Black is simply a pawn Meanwhile, Black had several other
down) 2S 'ii'a4! (2S 'ii'c2 ~c8 is more pas- tempting moves:
sive) 2S ... ~a8 26 'ii'c6 with a dangerous 1) 23 ... g6? (an attempt to bring the
initiative: bishop at f8 into play) 24 .i.d2! 'ii'xb2 2S .i.c3
1) 26 ... ~a7?! 27 .i.d2 'ii'xb2 28 tDxc4 .i.xc4 'ii'xa3 26 'ii'd4 ~eS 27 tDxeS .i.g7 (after
29 'ii'xc4 ~a1+ 30 ~h2 tDxf2 31 ~e2 tDd1 32 27 ... tDb3? 28 'ii'a7! dxeS 29 ~f3 fS 30 exfS a
eS! dxeS 33 tDxeS or 31...~h1+ 32 ~g3 tDd1 sad fate awaits the abandoned black king)
33 .i.b4 'ii'f6 34 eS! 'ii'g6+ 3S 'ii'g4 'ii'd3 36 28 'ii'd2! 'iWb3 (28 ... tDbd3 29 tDxd3 cxd3 30
'ii'e4, and Black suffers from the fact that his eS!) 29 .i.xb4 'ii'xb1+ 30 ~e1 'iVb3 31 .i.xcs
pieces are stuck in the opponent's territory; .i.xeS (31...c3 32 'ii'e3 .i.xeS 33 f4) 32 .i.d4 c3
2) 26 ... 'ii'd8 27 tDxc4 (in the event of 27 33 'ii'e3 .i.xd4 34 'ii'xd4 c2 3S eS!, and White
.i.d2 tDxb2 28 tDc2 tDd3! 29 tDb4 Black should be able to convert his exchange
defends by 29 ... tDxb4 30 .i.xb4 'iWb8 31 tDd4 advantage;
g6 or 29 ... tDcS 30 tDxa6 ~xa6 31 'iWbs ~a4) 2) 23 ... tDcd3? 24 'ii'a4! (breaking up the
140
The Third Match: 1986
group of black pieces) 24 ... Ma8 (24 ... CLlxc1? Black would have been quite satisfied
25 'lWxe8 'lWxb2 26 e5! CLlbd3 27 e6 and wins) with 24 Mc3 CLlbd3 25 ilxd3 cxd3 26 ile3 (26
25 CLlxc4 CLlxc1 (25 ... ilxc4? 26 'lWxa8 CLlxc1 27 CLld2 'lWd4!) 26 ... CLlxe4 27 Mc6 Ma8 (Anand-
'lWa3!) 26 'lWxb4 ilxc4 27 'iVxc4 'lWxb2 28 'lWc2 Timoshchenko, Frunze 1987), when if 25
or 28 Mel with an extra pawn and winning CLlxc4?? he has 25 .. .'iVxc3!. On observing this
chances; simple trap, for a moment I felt sad: the
3) 23 ... CLlbd3. This was wrongly criti- black knight on d3 reminded me too
cised by the commentators, including strongly of the events in the 16th game of
myself in Informator No.42. Yes, in the the previous match. Besides, the time on
variation 24 ilxd3 cxd3 (24 ... CLlxd3?! 25 my opponent's clock clearly demonstrated
'lWa4! - d. the note to Black's 22nd move) 25 that this was not the first time he had seen
b4 (25 'lWeI CLlb3 or 25 CLld2 'lWd4 is unclear) this position.
25 ... CLle4 26 b5 (26 'lWa4? Mc8) 26 ... ilb7 27 White is not promised anything by 24
Mxd3 (if 27 'lWxd3, then 27 ...'lWa1 28 'lWb1 - 28 CLle1 CLlbd3 25 CLlxd3 cxd3, 24 CLld2 CLlbd3 25
Mel? CLlxf2! - 28 ... 'lWxb1 29 CLlxb1 Mc8 with ilxd3 cxd3, or 24 CLld4 CLlbd3 25 ilxd3 CLlxd3
equality) 27 ... CLlc3? 28 ilb2 CLlxd1 29 ilxf6 (Shirov-Anand, Moscow (blitz) 2007), but
CLlxf2 30 'it'xf2 gxf6 31 CLlc4 Black has major 24 CLlh2!? deserved some consideration:
problems, but after 27 ... MC8! White's extra 24 ... CLlcd3 25 CLlg4 'lWd4 26 CLlc2 CLlxc2 27
pawn is not at all perceptible. ilxc2 h5 28 CLlh2 g6 29 CLlf3 (Milos-Mecking,
23 ... CLlbd3 would probably have led to Buenos Aires 2001) 29 ...'lWc5 with a double-
roughly equal play, but this is of purely edged game.
theoretical interest, since Karpov was In my view, 24 e5 was the most correct
aiming for more, as he was well ahead of move, from both the chess (opening lines
me on the clock and he thought that he had for attack) and the psychological point of
caused me definite discomfort. Even so, view - it came as a surprise to Karpov.
Black's pressure is not so strong as to force 24 ... dxe5
White to passively await the development
of events. 'There is no need to panic', I said
to myself. 'I must boldly press forward: in
the resulting complications everything will
be decided by enterprise and ingenuity.'
25 CLlxe5
By exploiting the removal of the black
rook from the e-file, White has radically
changed the character of the play. True, he
has had to part with his strong pawn centre
24 e5! (19) (which had lost its original value, since
141
Kasparov vs. Karpov: 1986-1987
Black has managed to go round it from the traditional uncertainties after the opening -
flank), and for the moment the dS-pawn is which rook? - here the price of an incorrect
more of a weakness than a strength. In choice may prove very high, since practi-
return White has sharply activated his cally all the pieces are participating in the
pieces and given the opportunity he is play and the position abounds in tactical
ready to create threats to the black king. I possibilities. Of course, at first sight
felt that I would always have sufficient 2S ... tLlbd3 looks more logical - all other
counterplay against the opponent's actions things being equal, simultaneously Black
in the centre and on the queenside: nearly also includes his rook in the play.
all my pieces are eyeing the kingside, However, 2S ... tLlcd3! was necessary, re-
whereas most of the opponent's are on the taining control of c2. White does not have a
other side of the battlefield. The position is great choice, and the simplest is to force an
evidently one of dynamic balance. immediate draw by 26 i.xd3 tLlxd3 27 ~xd3
But here, to my surprise, Karpov cxd3 28 tLld7 'iVd6 29 tLlxb8 'iVxb8 30 tLlb1 (30
thought for more than an hour! This is 'iVa4 is also possible) 30 ... 'iVeS 31 tLlc3 i.b4
another point in the game which is hard to 32 i.d2 ~xc3 33 i.xc3 'iVxdS.
explain. Given thorough home preparation But I was intending the fighting 26 tLlg4.
it is hardly possible to overlook the e4-eS Now where should the queen move? We
breakthrough (it is a very obvious idea), will analyse three replies:
but if it is assumed that the entire analysis 1) 26 ... 'iVh4 27 ~g3 ~h8 28 i.d2 i.d6?!
of the plan with 18 .. .'iVf6 was made in great (Nunn-Psakhis, Hastings 1987/88), and here
haste, the picture begins to become clear. White could have gained some advantage
At any event, the fruit of this long think by 29 i.xb4! tLlxb4 (but not 29 ... ..ixb4? 30
(which, incidentally, equalised the reading tLlxc4 or 29 ... ~xb4 30 ~e3! and tLlxc4) 30 ~c3
on our clocks) was a move which immedi- and the unavoidable tLlxc4. But with
ately placed Black in a critical position ... 28 ... ..icS! Black can prettily force a draw: 29
'iVf3 fS! 30 tLle3 f4 31 tLlfS 'iVxg3 32 tLlxg3
fxg3 33 'YWe4 i.xf2+ 34 ~h1 ..ics 3S i.e3 ~f8!
36 ~gl c3! 37 i.xcS ~f1 +! 38 ~xf1 tLlxcS+ 39
~gl tLlxe4 40 i.xe4 cxb2 41 d6 i.c8 42 tLlc4
and tLlxb2;
2) 26 .. .'iVb6. This looks dangerous: the
queen moves away from the kingside,
where after 27 ~g3 the main events de-
velop. However, after 27 ... tLlxc1! 28 iVxc1
tLld3! (28 ... ~h8? 29 tLleS) 29 tLlxh6+ ~h8
(29 ... ~h7? 30 iVf4) 30 tLlxf7+ ~g8 White has
no more than a draw. The following varia-
tions indicate how important it is to ex-
2S ...tLlbd3? (63) change the bishop on c1: 27 ... ~h8? 28 i.e3!
The start of a comedy of errors, which 'iVd6 (28 ... i.cS 29 iVf3) 29 tLlxh6 iVxdS 30
pursued the two players for three moves in iVg4! iVe6 31 tLlfS with a decisive attack, or
succession. Certainly, the invasion at d3 27 ... i.d6? (putting the dangerous rook
must not be delayed any longer, but with under fire) 28 i.e3! iVc7 (defending the f7-
which knight? Moreover, in contrast to point in the event of 29 tLlxh6+ ~f8 30 'iVhS)
142
The Third Match: 1986
29 ttJxh6+ (nevertheless!) 29 ... 'it'f8 'it'g8 36 l:tf5 'iYg7 37 'iYe3 White's attack is
sufficient for a draw.
In the book Ova matcha I enthusiastically
analysed the sharp variations with 27 ... ~d6
28 ~e3 'iVxb2 29 ttJxh6+ (it appears that the
black king, abandoned to the mercy of fate,
will come to a sad end, but White is unable
to land a decisive blow) 29 ... 'it'f8 30 'iVh5
gxh6 31 l:tf3?! ttJe5! (it is bad to play
31...l:tb7? 32 'iYxh6+ or 31...f6? 32 'iYxh6+ 'it'f7
33 'iVh7+ 'it'e8 34 'iYg8+ 'it'd7 35 'iYf7+ 'it'c8 36
l:txf6 etc.) 32 l:tf6 (32 'iYxh6+ 'it'e8? 33 l:tf5!
favours White, but 32 ... 'it'e7! 33 l:tf6! leads
merely to a transposition of moves)
Analysis diagram 32 ... 'it'e7 33 'iYxh6, and now Black launches
a counter-attack: 33 ... ttJxd5 34 l:txd6 (hop-
30 l:txg7! (the tempo of the offensive ing for 34 ... ttJxe3? 35 'iYf6+)
does not slacken; the rook sacrifice drags
the king out into the fire of the other pieces)
30 ... 'it'xg7 31 'iYg4+ 'it'h7 (31...'it'f8 32 'iYg8+
'it'e7 33 'iYxf7+ 'it'd8 34 'iYg8+ 'it'd7 35 'iYe6+
etc.) 32 'ii'f5+ 'it'g7 33 'iYg5+ 'it'h7 34 ttJf5 ~f8
35 d6 or 30 ... ~h2+ 31 'it'h1 'it'xg7 32 'iVg4+
'it'h8 33 ~d4+ ~e5 34 d6 with crushing
threats.
Black also has a difficult position after
27 ... ~c5?! 28 ~e3! (the sharp 28 ttJxh6+?
'it'f8 29 'iVh5 does not work in view of
29 ... ~xf2+ 30 'it'h1 g6! 31 'iYf3 ttJe5 32 'ii'f4
'iYd4) 28 ... ~xe3 29 ttJxe3 or 28 ... ~f8 29 '!iVf3.
So there is no alternative to the variation Analysis diagram
27 ... ttJxc1! 28 'iYxc1 ttJd3!;
3) 26 ... 'iYd4!. The strongest: in the centre 34 ... ttJf3+! (the white king, which up till
the queen occupies a far more aggressive now has been calmly observing the battle
position and controls the key a1-h8 diago- from its residence, itself comes under
nal. White has two main continuations: murderous fire) 35 gxf3 ttJxe3! (35 ... l:tg8+? is
a) 27 l:tg3 'it'h8! (27 ... ttJxc1? loses to 28 weak: 36 ~g5+! l:txg5 37 'iYxg5+ 'it'xd6 38
ttJxh6+ ~h8 29 ttJxf7+ 'it'g8 30 'iVh5) 28 ~e3 'ii'd8+, and White is out of danger), and
'iVxb2 29 ttJxh6 gxh6 30 'iVh5 '!iVf6, and now now 36 Mxa6? l:tg8+ 37 Mg6 l:txg6+! 38 ~xg6
31 ttJxc4(?) ~xc4 32 'iVg4, which was rec- 'lic1 + 39 ~h2 ttJf1 + leads to the loss of the
ommended in Ova matcha, will not do in queen, but White is saved by the study-like
view of 32 ... ~d6! 33 'iYxc4 ~xg3 34 fxg3 36 l:tg6!! (blocking the terrible g-file)
ttJxd5 35 'iYxd3 l:txb1+ 36 'iYxb1 ttJxe3, 36 ... fxg6 (after 36 ... 'iYc1+ 37 'it'h2 fxg6 38
winning, but after 31 l:tf3! '!!Vg6 32 ~d4+ 'iYg7+ the black king cannot hide from the
~g7 33 ~xg7+ 'iYxg7 34 l:txf7 'iYg5 35 'iYf3 checks) 37 'iYxe3+ 'it'd6 38 ~e4, and al-
143
Kasparov vs. Karpov: 1986-1987
though Black retains the advantage, the ~f3? .id6) 36 .. .fxg6 37 ~g4, picking up the
insecure position of his king sharply re- last pawn.
duces his winning chances, for example: Thus the correct 25 ... iDcd3! would have
38 ... ~b3 39 ~f4+ ~e5 40 ~f8+ ~e7 41 ~f4+ fully justified Black's opening idea and
'it'c5 42 iDc2 etc. given him a good game, whereas after
But under the strict eye of the computer 25 ... iDbd3? the scales could have tipped in
it transpires that White has two clear ways White's favour.
of forcing a draw by perpetual check. The
first is 31 .ixh6+ (instead of 31 ~f3?!)
31...'it'e7 32 .ig5+ f6 33 'iVh7+ (33 ~e3+? .ie5)
33 ... 'it'd8 34 .ie3! (but not 34 .ixf6+? ~xf6
35 ~g8+ .if8, as given in Ova matcha)
34 ... .ixg3 35 ~g8+ etc. The second is 28
..ixd3!? (instead of 28 .ie3) 28 ... iDxd3 29
iDc2 'iVb6 30 iDxh6+ 'it'f8 31 ~f3!, and since
31...~c7? is spectacularly refuted by 32
~xg7! ~xg7 33 ~g4+ 'it'h7 34 ~e4+! 'it'h8 35
'iVh4! ..ih2+ 36 'it'h1 jLe5 37 d6! ~d7 38
iDg4+ 'it'g8 39 ~g5+ 'it'f8 40 iDxe5 ~xd6 41
iDxd3 cxd3 42 .if4 ~g6 43 ~c5+ 'it'g8 44
jLxb8 dxc2 45 'it'h2 .ifl 46 .ig3, there only 26 iDg4? (14)
remains 31...gxh6 32 .ixh6+ 'it'e7 33 .ig5+ Played without particular hesitation, al-
'it'f8 34 .ih6+; though this places White on the verge of
b) 27 iDc2 (a more positional method - a defeat! In such positions, moves made on
bad piece is exchanged, and in addition the general grounds, not supported by accurate
annoying knight at d3 is deprived of sup- calculation, often lead to disaster.
port) 27 ... iDxc2 (the incautious 27 ... ~xd5? The problem was that for me Black's
leads to instant disaster: 28 iDf6+! gxf6 29 25th move came as something of a surprise.
~g3+ .ig7 30 'iYg4) 28 .ixc2. In the book Awaiting Karpov's reply, I sat in my rest
Ova matcha I restricted myself here to the room, mentally calculating the unclear
recommendation 28 ... .ic5(?) with the variations with 25 ... iDcd3 26 iDg4. There-
assessment 'unclear', but this is bad be- fore, when I encountered a surprise, I did
cause of 29 ~f3! ~d8 (29 ... iDxc1 30 iDxh6+! not instantly readjust, but decided to con-
Dvoiris-Timoshchenko, Barnaul 1988) 30 tinue as planned. Of course, the opened b-
~e4 ~xd5 31 iDxh6+! gxh6 32 ~g4+ 'it'h7 33 file and the agile knight on c5 are serious
~f6 .if8 34 ~xa6 and wins. arguments for Black in the coming battle,
But in fact Black has a reasonable choice changing the situation to his advantage, but
between 28 ... .id6!? 29 b3 ~a1 30 bxc4 .ixc4 nevertheless I was hoping that the immi-
31 .ixd3 ~xd3 32 ~e1 ..ig6 with a rapid nent attack on the king would safeguard
draw (Dvoiris-Timoshchenko, Na- White, even if his queenside were to be
berezhnye Chelny 1988) and 28 ... h5! 29 iDe5 completely destroyed.
~xd5 30 b3 iDf4 (30 ... .ib7 31 iDf3 is equal) However, for a top-class grandmaster
31 ~e4 cxb3 32 ~xf4 ~xd1+ 33 ~xd1 b2 34 this is not really a justification: with an
.ixb2 ~xb2 35 ~xh5 g6, and it is now White hour for thought, he should find both the
who has to make a draw - 36 .ixg6 (36 drawbacks to the planned move, and the
144
The Third Match: 1986
best decision, which in the given instance Let us consider the four possible rook
was 26 ·iY'c2!. With this brilliant reply White moves:
could have gained a clear advantage. In this a) 27 ... l:!.b3? 2S l:i.eS c3. This is refuted by
case Black's activity would have corne to the fantastic 29 i.e3! - Black's downfall is
standstill, but above all - the scattered caused by his weakness on the b1-h7 di-
white pieces would have begun coordinat- agonal: 29 ... MXb2 (29 ... g6 30 i.d4) 30 i.xc5
ing harmoniously. Added to this should be l:i.xc2 31l:i.xfS+ 'it>h7 32 i.xc2 and wins;
White's material advantage and the possi- b) 27 .. J:tb7? 2Sl:i.eS g5 (the threat of i.e3!
bility of winning the c4-pawn. forces Black to open wider the window for
However, we should not be in a hurry his king; 2S ... g6 29 i.xh6! l:i.xb2 30 i.xfS
to declare that 26 "iVc2 would have auto- 'it>h7 31 i.e7! "iVf4 32 g3 and wins) 29 ctJa5!
matically won the game, as many commen- (this is even stronger than 29 MaS, or 29 f3,
tators did. Black's resources are far from which I suggested in Ova matcha), and with
exhausted, and on the way to his goal the fall of the c4-pawn Black's position
White would have had to display consider- collapses 29 ... l:i.e7 30 l:i.xe7 ~xe7
able resourcefulness: (30 ... "iVxe7 31 i.e3!) 31 ctJ3xc4 ctJxc1 32 "iYxc1
1) 26 ... ctJb3(?) 27 ctJd7! "iVd6 2S ctJxbS i.b5 33 ctJe3, or 29 ... l:i.d7 30 ctJ5xc4 l:i.xd5 31
~xbS 29 ctJxc4 ctJbxc1. In the book Ova b4! ctJb3 32 i.e3 ctJd4 33 i.xd4 l:i.xd4 34 b5
matcha I called this 'Black's best chance', i.b7 35 ctJe5! ~e4 36 ctJg4 "iVg7 37 "iVcs and
having in mind 30 l:i.xd3 ctJxd3 31 "iVxd3 g6 wins;
32 b3 i.c5, when 'it is not clear whether c) 27 ... l:i.a4 2S l:i.eS ~b7, and after 29
White's material advantage is sufficient for ctJxc4 i.xc6 30 dxc6 "iVxc6 31 l:i.e3 or 29 i.e3
a win', but after 30 d6! g6 31 "iVc3 or i.xc6 30 dxc6 "iVxc6 31l:i.dS "iVc7 (31..."iVb6 32
30 ... ~dS 31 b3 g6 32 "iYdl, winning, the l:i.cS!) 32 l:i.d4 g6 33 l:i.xc4 White remains
situation becomes clear. with an extra passed pawn on the b-file;
2) 26 ... l:i.b4. The only move - the c4- d) 27 ... l:i.b6! 2S l:i.eS! l:i.xc6 (this exchange
pawn has to be defended (27 ctJaxc4 i.xc4 sacrifice is forced: 2S ... ctJd7? is answered
2S ctJxc4 l:i.xc4 with equality), but after 27 well by both 29 ctJe7+ 'it>hS 30 i.e3 l:i.xb2 31
ctJc6 (27 l:i.e2? ctJb3) Black faces difficult "iYa4 or 30 ... l:i.b4 31 "iVc3!, and 29 ctJa5l:i.d6 30
problems. ctJ5xc4 l:i.xd5 31 b4 'iVg6 32 ctJe3) 29 dxc6
"iVxc6 30 l:i.dS!. White's idea is to return the
exchange at a convenient moment and
obtain an extra passed pawn. And he
achieves this:
30 ... "iVf6 31 ~xfS+! 'it>xfS 32 i.e3 with the
threats of b2-b4 and ctJxc4;
30 ... "iVb6 31 l:i.aS i.b7 32 ctJxc4 'iYc6 33
l:i.xfS+! 'it'xfS 34 f3 or 34 ctJe3;
30 ... 'i'e4 31 ~d2 i.b7 32 f3 'iYh4 33 i.a5!
ctJe6 (33 ... ~a6?! is worse: 34 l:i.aS i.b7 35
l:i.eS! i.d5 36 ~e1! 'iYf4 37 i.f2 ctJe6 3S b3) 34
ctJxc4 ctJxdS 35 'iYxd3 ~c5+ 36 'it>h2 ctJe6 37
.id2;
Analysis diagram 30 ... ~b7 31 f3 "iVb6 32 I:!.xfS+! 'it>xfS 33
ctJxc4 "iYe6 34 ~e3 i.a6 35 i.a2 or 31...'iVc7
145
Kasparov vs. Karpov: 1986-1987
32 ~e8 (it is too early for 32 ~xf8+ 'it'xf8 33 Kasparov 'took him at his word' and
'Yi'xc4 because of 33 ... jLa6 34 'iVc3 ctJxc1 35 missed a winning possibility.
'ir'xc1 'ilVd6 with counterplay) 32 ... ~d7 33 I have serious doubts about the sincerity
~xf8+! 'itxf8 34 jLe3 'iVe6 35 ctJxc4 jLa6 36 of these words. In a bad position it cannot
jLa2 'itg8 37 jLxc5 ctJxc5. be denied that bluff is a normal thing, but is
Black's position before his 25th move really
so hopeless? I cannot believe that during
the game Karpov did not consider invading
at d3 with the knight from c5. And yet after
25 ... ctJcd3, as we were able to establish
above, White has no advantage.
So why did Karpov nevertheless play
25 ... ctJbd3? The answer suggests itself: he
overlooked 26 'iVc2! and found the move
later, after the conclusion of the game (just
as I did, and also all the commentators).
After other moves Black is quite alright.
Thus 26 ctJaxc4?! 'iVxf2+ 27 'ithl ctJf4! 28 'Yi'f3
Analysis diagram 'iVxf3 29 ~xf3 ctJxd5 causes White consider-
able problems in view of his uncoordinated
38 'ith2 (38 'iVc3 ctJa4! 39 'iVb4 jLxc4 40 pieces, although after 30 jLd2 he should be
jLxc4 'i!i'e3+ 41 'it'hl 'iYc1+ 42 'it'h2 'iYf4+ 43 g3 able to maintain the balance. 26 jLxd3 ctJxd3
'iVxf3 44 'iVxa4 'Yi'f2+ 45 'ithl 'iVf3+ is unclear) 27 ~xd3 cxd3 28 ctJd7 leads to the same
38 ... 'ilVd5 39 'iVe2 ctJd3 40 b3. draw as after 25 ... ctJcd3, but now Black has
The other question is how great are 26 ... cxd3 ... Stop! Isn't it here that the riddle
White's chances of converting his material is concealed? Perhaps Karpov wanted to
advantage. The play is very complicated, avoid a forced draw?
and Black may be able to save himself In the book Ova matcha I even attached
(although at the board, with the clock an exclamation mark to 26 ... cxd3, thinking
ticking away, this is highly problematic). that the advantage was now with Black: 27
But in any case after 26 'iVc2 he would have ctJc6 ~xb2 28 jLxb2 'iVxb2 29 ~e8(??) d2! 30
faced a difficult struggle for a draw ... ctJe7+ (30 ctJc2 'i!i'c1 31 ctJe3 jLe2!) 30 ... 'it'h7 31
Thus, after spending more than an hour, 'iVh5 'iVal+ 32 'ith2 dl'iV 33 'iVf5+ g6 34
Karpov made a serious mistake, but I failed 'iVxf7+ 'Yi'g7 etc. But after 29 ctJbl! Black still
to exploit it. Of course, the situation should has to demonstrate that he has sufficient
not be over-dramatised - such mutual compensation for the exchange (29 .. .f5!?),
oversights occur even in matches for the and therefore 27 ... ~a8 28 b4 'iVaI is safer,
world championship, but nevertheless it is with equality. Or 27 ctJbl jLd6 28 ctJc6 ~a8
worth dwelling in slightly more detail on (28 ... ~xb2?! 29 jLxb2 'iVxb2 30 ~el!) with
this interesting moment. good counterplay: 29 b4 (29 ctJc3 jLc4)
According to Karpov, he suddenly dis- 29 ... ctJd7 30 'iVb3 'iVh4! 31 jLb2 (31 ctJd2?! is
covered that the prepared invasion at d3 weaker: 31...ctJf6 32 b5 ctJg4! 33 ~f3 ~e8! 34
was merely bluff because of 26 'iVc2!, but, ctJf1 ~el 35 jLf4 jLc5! 36 jLg3 'iVxg3! 37 hxg4
on failing to find anything better, he never- 'ilVxf3 38 gxf3 d2 39 bxa6 dl'iV 40 'Yi'xdl ~xdl
theless followed this path. But, fortunately, and ... ~al) 31...'iVc4 32 'iVxc4 jLxc4 and
146
The Third Match: 1986
26 ...'ii'b6?
After this fairly quick and 'obvious' re-
ply everything fell into place: Karpov
began preparing to win material, while
Kasparov set about assembling a striking
force for storming the king's fortress; i.e. Analysis diagram
each was engaged in his favourite business!
The withdrawal to b6 seemed to every- 1) 28 .. :~e4!? 29 ~e3 'ii'd4 30 tZJeS 'iVh4; or
one, including myself, to be 'an excellent 2) 28 ... ~e8!? 29 tZJe3 'iVd8(hS) with a dan-
defensive manoeuvre: the queen appears to gerous initiative;
move a long way from the scene of the 3) 28 ... .ii.d6!? 29 tZJe3 (29 .ii.xh6 l:.xb2)
forthcoming battle, but in reality only from 29 ... 'ii'eS 30 g3 .ii.b7 31 ~xf7 tZJxf2! 32 ~xf2
this square can it retain control of the 'ii'xg3+ 33 ~f1 'ii'xh3+ 34 ~e2 .ii.g3; or
important b2- and f2-points, as well as the 4) 28 .. J:td8! 29 tZJf6+ (29 .ii.xh6 tZJe4 or 29
147
Kasparov vs. Karpov: 1986-1987
ttJe3 'iYa8 is no better) 29 ... gxf6 30 .i:i.g3+ itg7 within the bounds of dynamic equilibrium.
31 .i.xh6 Wh7 32 .i.xg7 ttJe4 33 iLxf6 ttJdxf2! The time reserves are roughly equal, al-
with good winning chances. though not very great (what is an hour in
I should mention that instead of 28 .i.a2 such a tense situation?!). There is no ques-
immediate attacks against the king's for- tion of relying here on a full variational
tress do not work - neither 28 ttJf6+?! gxf6 analysis - the situations which can arise at
29 .i:i.g3+ .i.g7 30 .i.xh6 ttJe6 31 ttJxc4 .i.xc4 32 literally every move are too unusual and
.i.xd3 .i.xd3 33 .i.xg7 ttJxg7 34 l:txd3 'iYeS or complex. Positional guidelines are totally
31 .i:i.xg7+ ttJxg7 32 'iYg4 Wf8 33 .i.xg7+ We7, eroded, and the customary scale of values
nor especially 28 .i.xh6? ttJe4! (but not is displaced, since the two sides are attack-
28 ... .i:i.xb2? 29 ttJf6+! gxf6 30 .i:i.g3+ or ing targets of different importance. There-
28 ... 'iYe6? 29 ttJf6+! Wh8 30 .i.f4 'iYxf6 31 fore the qualities that come to the fore are
.i.xb8 'iYxb2 32 .i:i.xf7 'iYxb8 33 'i!fuS+ Wg8 34 those such as intuition and enterprise,
.i:i.f3 'iVb7 3S £i.a2) 29 ttJe3 'iYe6 30 ttJaxc4 which means that to complain of bad luck
ttJdxf2 31 'iYd4 .i.xc4 32 ttJxc4 .i:i.c8 and Black is absurd, to say the least - everything is
wins. within one's own grasp.
And finally, Black has clearly the better 27 .i:i.g3 (14)
game after the quiet 27 .i:i.g3 (first driving An unequivocal hint at a direct attack. It
the king into the corner) 27 ... Wh8 28 .i:i.f3 is interesting that at this very important
'iYxdS 29 ttJe3 'iYd7! 30 ttJaxc4 .i.xc4 31 ttJxc4 moment Karpov replied largely intuitively,
ttJeS! or 30 .i.d2 .i.b7 31 .i.c3 .i.xf3 32 'iYxf3 almost without thinking.
ttJxb2 33 .i.fS 'iVb7 34 'iYf4 Wg8 etc.
Thus with 26 ... 'iYfS! Karpov could have
set me very serious problems. But this did
not happen, and we effectively began a
new game: left behind were all the opening
tricks, the clash of plans and the' courteous'
exchange of mistakes.
27 ... g6 (04)
All the same Black could not avoid a
weakening of his king's pawn screen:
27 ... Wh8 28 ttJxh6!. Now after the simple-
minded 28 ... gxh6? the outcome is quickly
decided by a temporary sacrifice of the
forgotten knight on a3, destroying Black's
At this last moment of comparative calm position in the centre: 29 ttJxc4! .i.xc4 30
before the brief concluding battle, let us try 'iYg4 'iYg6 31 'iYxc4 ttJeS (31.. ..i:i.b4 32 'iYc3+
to assess the situation. Both sides have .i.g7 33 'iYxb4! 'iYxg3 34 'iYxcS!) 32 'iYf4! 'iYxb1
important trumps and the position is (32 ... 'iYd6 33 'iYfS) 33 'iYxeS+ Wh7 34 'iVxb8
148
The Third Match: 1986
'iVxc1 + 35 'it>h2 ~g7 36 'iVc7 with an easy than one way - 28 ... liJexf2 29liJxf2 ~xa3 30
win. liJxd3 cxd3, or 28 ... ~xa3 29 ~xd3 (29 i:.xe4
The only thing that can save Black is liJxc1 30 'iVxc1 'iVxb2 is worse for White)
resolute action - 28 ... liJe4! (forcing White to 29 ... cxd3 30 ~xe4 ~xb2, or 2S .. .£5! - in Ova
think also about defence) 29 liJxf7+ 'it>g8. A matcha I condemned this move because of
very interesting position: 29 liJxc4 (29 'iVf3?! is weaker: 29 ... ~xa3 30
'iVxf5 liJexf2 31 liJxf2 ~c5 32 'iVe4 ~b5! with
the threat of ... ~e8), but after 29 ... ~xc4 30
~xd3 ~xd3 31 'iVxd3 ltcS! Black has the
advantage.
It would seem advantageous for White
to sacrifice the exchange, eliminating the
opponent's outpost and at the same time
getting rid of the inactive knight on a3: 28
liJxc4?! ~xc4 29 ltxd3 (or 29 ~xd3 liJxg3 30
~xc4liJe4 with equality) 29 ... ~xd3 30 'iVxd3
(30 ~xd3 lteS), but after 30 ... lteS this ex-
changing operation does not bring him any
dividends - the d5- and b2-pawns are weak
Analysis diagram and require defending, and with such a
powerful knight Black can combine pres-
If 30 ~e3, then 30 ... liJexf2! 31 'iVh5 ~c5! sure on f2 with threats along the 1st rank.
32 liJg5 ~xe3 33 'iVh7+ 'it>f8 34 liJe6+ 'it>e7 35 Moreover, White's threats along the bl-h7
'iVxg7+ 'it>d6 36 'iVg3+ 'it>xd5 37 liJc7+ (37 diagonal are easily parried: 31 ~e3? 'iVxb2
~xe3 'iVxe6 38 ~xf2 is equal) 37 ... 'it>c6 38 32 f3 ltb8! 33 ~c2 'iVaI + 34 'it>h2 ~d6+ 35 f4
~xe3 (38liJxa6? liJxh3++!) 38 ... 'iVxc7 39 'iVf3+ ltb4.
'it>d6 40 ~xd3 liJxd3 41 'iVf6+ 'it>d7 42 'iVf5+ White is forced to accept the 'invitation'
'it>c6 43 'iVf3+ with perpetual check, or 34 - 28 liJxh6+! 'it>h7.
'iVh8+ 'it>e7 35 'iVxg7+ 'it>d6 36 'iVf6+ 'it>d7!
(36 ... 'it>xd5? 37 'iVf3+! ~d6 38 ~xe3 'iVxb2 39
~c2 with good chances of success) 37 'iVf5+
'it>d8!, and White has to reconcile himself to
a draw, since after 38 'iVf8+ 'it>d7 39 'iVf7+
'it>c8 40 'iVe8+ 'it>b7 41 'iVe7+ 'iVc7! 42 'iVxe3
liJdl Black is not in any danger. No more is
given by 30 ~e3 'iVxb2 31 'iVh5liJxg3 32 fxg3
(as in the variations with 27 ... liJe4, analysed
below).
An obvious question arises: would it not
be better for Black to play 27 ... liJe4!? imme-
diately, in order after 28 liJxh6+ to play his
king to h7? White would have had difficult Analysis diagram
problems to solve. Thus the cowardly 28
~e3?! (hoping for 28 ... liJec5 29 ~g3 with a Here 29lte3? liJexf2 30 'iVf3 looks tempt-
repetition of moves) can be parried in more ing, with the idea after 30 ... 'iVxh6? of creat-
149
Kasparov vs. Karpov: 1986-1987
ing a study-like rook 'staircase' - 31 'uxd3! 36 ~f5, and although Black is a rook up, his
~xc1 + 32 'udl +!. Black can defend by king has been abandoned by its forces to a
30 ... ~xa3 31 tDxf7 WgS! 32 'ue6 ~a7 33 certain death;
tDh6+ WhS 34 ~e3 ~c5 35 tDf7+ WgS 36 3) 31...iYf6?! (the return of the queen to
~xc5 'ifxc5 37 tDh6+ WhS with equality, but the defence does not promise Black an easy
the unexpected 30 ... iYf6! completely refutes life) 32 'ifu5+ Wg8 33 tDg5 ~d6 (33 ... iVf5? is
White's idea: 31 'iVxf6 gxf6 32 tDg4 tDxg4 33 hopeless: 34 ~xd3 'iVxd3 35 iYf7+ WhS 36
hxg4 ~c5, or 31 tDg4 "iixf3 32 'uxf3 tDxg4 33 ~h2! ~b7 37 ~d4 or 36 ... ~d6 37 ~f4! ~xf4
tDxc4 (33 hxg4 ~xa3) 33 ... ~c5+! (a cunning 38 gxf4) 34 'ifu7+ 'it>f8 35 iYe4 iYe7 (it is bad
intermediate check!) 34 Whl (34 Wf1 tDh2+) to play 35 ... iYe5? 36 iYf3+ WgS 37 ~xd3
34 ... ~xc4 35 hxg4. cxd3 38 iYf7+ ~h8 39 ~f4 or 35 ... ~gS? 36
~xd3 cxd3 37 Wh2! g6 3S tDc4 ~xc4 39
'iVxc4 iYf5 40 tDe6 ~f8 41 ~d2! with the idea
of iYd4 and ~c3, winning) 36 'iff5+ ~eS 37
iYg6+, and the black king is at the cross-
roads:
Analysis diagram
150
The Third Match: 1986
draw by 38 ctJf7+ Wc7 39 ~xd6+ ~xd6 40 .lii.e4 is dangerous for Black) 39 ~xe5 ctJxe5
ctJxd6 Wxd6 41 ctJxc4+! .lii.xc4 42 .lii.xd3 or 40 ctJd2 .lii.xd7, or 35 .. .'iVxa3 36 ~8+ 'it>xf7
continue the battle - 38 ctJe4!? .lii.c5! 39 .lii.xc5 37 ~xb8 ~e7 38 d8ctJ+! (but not 38 ~xb7?
ctJxc5 40 Wh2! ctJxe4 41 .lii.xe4 ~f6 ~xe3+ 39 Wh2 .lii.d8!) 38 ... Wg6 39 ~xb7 (39
(41...~xa3? 42 d6!) 42 ~5 ~b3 43 ctJc2, ~f4 .lii.d5 40 ~g4+ .lii.g5 with equality)
without too much danger of losing, in view 39 ... ~xe3+ 40 Wh2 .lii.xd8 41 ~c6+, reaching
of the exposed position of the black king; the haven of a draw after both 41...Wh5 42
4) 31.. ..lii.e7?! (trying to restrict the .lii.dl+ Wg5 43 ~d5+ Wg6 44 ~xd8 ctJe5, and
knight) 32 ~5+ Wg8 33 d6! .lii.f6 (33 ... ~xa3? 41.. ..lii.f6 42 ~xc4 .lii.d4! (not 42 ... .lii.e5? 43
34 ctJh6+! is bad, as is 33 ... ~f6? 34 ctJh6+! ~e6+! and .lii.xd3+) 43 .lii.xd3+ Wf6 44 h4 .lii.e5
gxh6 35 dxe7 ~xe7 36 ~g6+ or 35 ... ~e6 36 45 ~c6+ We7 46 ~g6 .lii.d4 47 .lii.c2 ~gl+ 48
.lii.xh6 ~e4 37 ~g5+ Wf7 38 ~g7+ We6 39 Wh3 ~e3 49 ~g4 ~e6 50 .lii.f5 ~e3;
~f8) 34 d7!, transforming the d-pawn into a 5) 31...Wg8 32 ~5 .lii.xa3! (there is noth-
powerful force. ing else: 32 ... .lii.e7? 33 d6 .lii.xd6 34 ctJh6+!
gxh6 35 ~d5+ and ~xd6, or 32 ... ~al? 33
'it>h2! ~f6 34 ctJg5 ~f5 35 ctJc2! with a strong
attack), or, on the contrary, 31.. ..lii.xa3! 32
~5+ Wg8 33 .lii.xd3+ cxd3 (33 ... ~al+? 34
Wh2 cxd3 35 ctJg5 and wins) 34 ctJg5 d2,
immediately forcing perpetual check (but
not the suicidal 34 ... ~7? 35 ~7+ Wf8 36
~8+ We7 37 ~xg7+ etc.).
After looking at such variations, one in-
voluntarily begins to fear the spectre of a
'drawing death' on the chess board. How-
ever, one is consoled by the thought that
only in analysis is it possible to penetrate so
Analysis diagram deeply.
Black eliminated the threat of the sacri-
White has excellent compensation for fice on h6 in the most radical way - he
the material deficit: removed the pawn from g7, rightly assum-
a) 34 ... ~xa3?! 35 ~8+ Wxf7 36 ~xb8 ing that it would not be easy for White to
'iVe7 37 ~a7!, and Black faces difficult redirect the fire of his attack against the g6-
problems: 37 ... Wg6 (37 ... We6? 38 .lii.xd3 and point. With his cool-headed decision Kar-
'iWxa6+) 38 ~xa6! (but not 38 d8~? ~xd8 39 pov clearly let it be known that he consid-
~xa6 ~d5) 38 ... ~xe3+ 39 Wh2 ~d4 40 d8~! ered White's attack to be short-lived, and
~xd8 41 ~xc4 or 37 ... Wg8 38 .lii.d4 .lii.b5 39 he was ready (guided by Capablanca's
~a8+ Wh7 40 .lii.xf6 ~xd7 41 ~e4+ Wg8 42 well-known principle - the minimum
.lii.c3 with good winning chances; number of pieces in defence) to assail the
b) 34 ... .lii.b7 35 .lii.c2!. The undefended opponent's defenceless queenside with all
bishop at bl and knight at a3 are success- his might. And he rejected 27 ... ctJe4, appar-
fully fulfilling important defensive duties. ently in order to avoid the simplification in
Black can fight for a draw by 35 ... ~al+ 36 the event of 28 ctJxc4 or the unnecessary (in
ctJbl! (36 Wh2? ~f1) 36 ... .lii.c6 37 ~8+ Wxf7 his opinion) complications after 28 ctJxh6+.
38 ~xb8 ~e5! (38 ... .lii.xd7?! 39 ~c7 ctJe5 40 In my view, this episode is interesting from
151
Kasparov vs. Karpov: 1986-1987
the standpoint of understanding the subse- side. The black king begins to feel uncom-
quent events in the game. fortable - the defensive lines around it look
unconvincing compared with the oppo-
nent's powerful piece grouping.
29 ...tLld7 (07)
By covering the f6-square, Karpov fol-
lows his conception (defence with minimal
means), but clearly underestimates the
opponent's attacking potential. Strangely
enough, this natural move, which in itself is
not bad, hinders Black's subsequent play.
The commentators, who unanimously
attached an exclamation mark to it, did not
notice its significant drawback - the weak-
ening of control over the important strate-
28 .i.xh6! (14) gic d3-point. Now at an appropriate mo-
In this specific instance the knight is a ment White can get rid of the knight on d3,
far more valuable attacking piece than the which is restricting his actions. However, at
dark-square bishop. After 28 tLlxh6+? .i.xh6 this point the entire press centre, hypno-
29 .i.xh6 "iVxb2 30 .i.xd3 cxd3 nothing tised by the inevitable loss of the knight at
comes of the attack. 28 tLlxc4?! .i.xc4 29 a3, thought that a win for Black was not far
.i.xd3 tLlxd3 30 ~xd3 .i.xd3 31 'iVxd3 h5! is off, and therefore Karpov's unwillingness
also unfavourable for White. to force a draw was taken as being self-
28 ...'i¥xb2 (07) evident.
At first sight Black has clearly been After 29 ... 'iVxa3 30 tLlf6+ 'it'h8 the attack
more successful: he has already broken and the material advantage would have
through on the queenside and is ready to balanced each other.
win the knight on a3, whereas White has
not yet created any real threats to the king.
Analysis diagram
152
The Third Match: 1986
Wg7! 34 lL:lh5+ Wg8 35 lL:lf6+ (35 'iVf6? 'iVh6) 35 'iVe3! .l:te6 36 ~d4! "iVb2
35 ... ~g7 with a repetition of moves, and the
spectacular 31 'iVh5!?, leading again to
perpetual check: 31.. ..l:txb1 + (31.. ..id6? 32
i..f8+! or 31...i..g7? 32 'iVh4! leads to mate)
32 i..c1 + (simpler than 32 ~h2 .l:th1 + 33
~xh1 lL:lxf2+ 34 ~gl! 'iVxg3 35 i..g7+! ~xg7
36 lL:le8+) 32 ... ~g7 33 lL:le8+ Wg8 34 lL:lf6+.
It is not possible to gain an advantage
by 29 ... .l:tb6 on account of 30 ~h2!. Now
30 ... 'iVxa3? loses to 31 .ixf8 ~xf8 32 lL:le5!,
30 ... f5?! 31 d6! is dubious, and if 30 ... lL:ld7,
then 31 lL:lxc4 or 31 i..e3 is not bad. 30 ... .id6
is more critical, but after 31 il.e3!? with the
threat of lL:lh6+ (the modest 31 i..xd3 is also Analysis diagram
possible, with equality) Black must in the
first instance think about how not to lose: 37 dxe6!! ~xd4 38 e7 - Black is now a
31.. ..l:tb7? 32 .ixd3 lL:lxd3 33 lL:lf6+ ~h8 34 queen up, but he is absolutely helpless:
lL:lc2! or 31...lL:le5?! 32 lL:lh6+! ~f8 33 ~f4 38 .. :iVd7(e4) 39 lL:lf6+, while the defence
lL:lcd3 (33 ... 'iVxa3 34 .ixg6) 34 ii.xd3 cxd3 35 38 ... .ib5 is killed by the 'dead' knight on
.id4 "iVb4 36 lL:lb1! with a dangerous attack. a3! Therefore 32 .. .f5!? is correct: 33 lL:lxc4
31...~f8? 32 lL:lxc4! is also weak, but .ixc4 34 i..xc4 lL:le1! 35 ~d1 lL:lc2 36 i..xc5
31...~g7 is more solid. i..xc5 37 i..d3 il.d6! 38 .ixc2 fxg4 with
equality.
The immediate 29 ... .id6 is also interest-
ing, attacking the dangerous rook, although
the dark-square bishop, which remains
alive, again successfully replaces it in the
attack - 30 i..e3! i..xg3 31 lL:lf6+ Wg7 32
'iVxg3, reaching a very sharp position,
which is difficult to assess:
1) 32 ... ~xf6 33 .ixd3 cxd3 (33 ... lL:lxd3?! is
worse: 34 iYh4+ ~e5 35 'iVe7+ Wxd5 36
~d7+, when 36 ... ~e5? 37 lL:lc2! "iVb1+ 38 Wh2
lL:lxf2 39 'iVe7+ Wd5 40 'i!Vxf7+ We5 41 ~e7+!
Wd5 42 'i!Vc5+ We6 43 lL:ld4+ leads to mate,
Analysis diagram while after 36 ... ~e4 37 'i!Vc6+ ~f5 38 g4+
~e5 39 'i!Vxa6 "iVb4! 40 lL:lxc4+ ~d5 41 lL:lb6+
Here, however, apart from the prosaic .l:txb6 42 iYxd3+ ~e6 43 i..xb6 'ti'xb6 Black is
32 lL:lxc4 i..xc4 33 ii.xd3 i..xd3 34 .ixc5 i..xc5 obliged to go into a queen endgame a pawn
35 ~xd3 with equality there is the incredi- down) 34 ~f4+ Wg7 35 .id4+ ~g8 36 ~e5
ble resource 32 .ia2!! with a spectacular 'i!Vxd4 37 'iVxd4 lL:lb3, forcing a draw: 38 ~e5
knock-out in the event of 32 ... .ixg3+ 33 fxg3 .l:tc8 39 lL:lb1 .l:tc1+ 40 Wh2 .l:txb1 41 'ih>8+
'iVxa2? (after 33 .. .f5 34 lL:lxc4 or 33 ... lL:le1 34 ~g7 42 'i'e5+ Wh7 43 'i!Ve7;
'iVf2 White is simply better) 34 .ih6+! ~g8 2) 32 .. .'iWe5 33 lL:lh5+! (not 33 i..h6+? Wxf6
153
Kasparov vs. Karpov: 1986-1987
or 33 'iVh4? l1hS) 33 ... 'it'gS! (the dark-square .iL.xd2 'iVxd2 36 .iL.xd3 cxd3 37 'iVe5+ 'it'gS 3S
'draught' is Black's undoing after both tLJf6+ 'it'fS 39 d6 'iVc1 + 40 'it'h2 d2, and again
33 ... 'it'h7? 34 'fWxe5 tLJxe5 35 .iL.xc5, and it is a draw!
33 ... 'it'hS? 34 'iVxe5+ tLJxe5 35 .iL.d4 tLJd7 36 In fact, such a result would have been
tLJf6! tLJxf6 37 .iL.xe5 l1b6 3S d6 or 35 ... l1xb 1+ the logical outcome of the purposeful
36 tLJxbl gxh5 37 .iL.xc5) 34 tLJf6+ 'it'g7 35 actions of the two players, who have en-
tLJh5+ with perpetual check; deavoured to decide the game on different
3) 32 ... l1hS 33 tLJg4 with full compensa- parts of the board. But such a development
tion for the exchange: of events did not come into the plans of
a) 33 ... .iL.b7? 34 .iL.xd3 cxd3 35 'fWf4! tLJb3 Karpov, who still did not believe in the
36 tLJc4 'iVal+ 37 'it'h2 .iL.xd5 3S tLJd2 l1dS 39 seriousness of White's attack. After all,
~6+ 'it'gS 40 ~4 and wins; Black needs only a one-move respite, in
b) 33 ... .iL.cS? 34 .iL.xd3 cxd3 (34 ... tLJxd3 35 order to calmly pick up the stranded knight
tLJxc4) 35 tLJc4 'iVaI + 36 'it'h2 tLJb3 37 tLJge5 at a3 and then convert his material advan-
with a very strong attack - 37 ... tLJd4 3S 'iVf4 tage. But he will not gain this respite right
tLJf5 39 tLJd6 d2 40 tLJxf5+ ..Ili.xf5 41 .iL.xd2 etc.; to the end of the game ...
c) 33 ... l1h5?! 34 'iVd6 .i.cS 35 .iL.xd3 tLJxd3 30.iL.xf8
36 tLJb5 with the threat of .iL.d4+:
30 ... 'it'xf8
Analysis diagram By moving his king off the g-file, Black
continues playing for a win. In the event of
36 ... tLJe5! (there is nothing better) 37 30 .. JhfS 31 tLJh6+ he could have hoped
tLJxe5 ~1+ 3S 'it'h2 .iL.xh3! (this counter- only for a draw - 31...'it'g7 (but not
sacrifice saves Black; 3S ... 'iVxb5? 39 .iL.d4) 39 31.. .. 'it'h7? 32 tLJxf7 'iVxa3 33 'iVg4! or
gxh3 'iVf5 40 'fWd7 'iVxe5+ 41 'it'gl 'iVxd5 42 32 ... tLJ7e5 33 tLJg5+ 'it'gS 34 'iVe3! etc.) 32
tLJd6 l1g5+! 43 .iL.xg5 'iVxg5+ 44 'it'f1 'iVc1 +, tLJf5+, and again a choice has to be made:
with difficulty gaining a draw; 1) 32 ... 'it'h7 33 'iVe3! gxf5 (eliminating the
d) 33 ... l1eS 34 i..h6+ (34 .iL.xc5 tLJxc5 35 powerful knight; 33 ...'iVe5 34 'iVh6+ 'it'gS 35
.iL.xg6! is also not bad) 34 ... 'it'gS 35 'iVd6 tLJe4! tLJe3 tLJf6! 36 'it'h2 tLJxd5 is also acceptable)
(but not 35 ... 'iVxa3? 36 tLJf6+ or 35 ... l1bS? 36 34 .iL.xd3 cxd3 35 'iVg5 'iVf6 36 ~5+ ~6 37
.iL.e3) 36 'iVd7 l:i.bS 37 'iVe7! with a guaran- 'iVxf5+ 'it'hS 38 'iVxd7 'iVc1 +! 39 'it'h2 l1gS!,
teed draw; and the threat of ... d3-d2-dl'iV produces a
e) 33 ...tLJe4! (the safest) 34 'iVc7 tLJd2 35 draw, or 34 'iVe7!? 'iVxf2+ 35 'it'h2 'iVf4! 36
154
The Third Match: 1986
155
Kasparov vs. Karpov: 1986-1987
2) 31..:~c1?!. Now quite incredible decisive turning point has already oc-
variations arise - 32 i..xd3 (not fearing the curred. But apart from the continuations
creation of a powerful passed pawn for the examined, Black has an energetic manoeu-
opponent, White eliminates the knight on vre, which appears to put everything in its
d3, in order to effectively include his queen place.
in the attack) 32 ... cxd3 33 CDf6!. 31 ... ~b3!
After making this move, Karpov calmly
stood up and began leisurely strolling up
and down the stage, transferring his trium-
phant glance from the board to the audito-
rium and back again. Indeed, the rook
comes into play with great effect, creating
an unpleasant 'X_ray' along the 3rd rank,
the aim of which is to reach the rook stand-
ing in ambush at g3 (with its exchange the
potential of White's attack will be ex-
hausted). And after safeguarding himself
against the combined threats, Black will at
last be able to pick up the knight on a3.
Analysis diagram The grandmasters watching the game
were sure that I was losing. Now White
The knight on d7 has three possibilities essentially has no choice, but his only
(not counting 33 ... CDb6? 34 d6 d2 351!Vh5!): sensible reply radically changes the situa-
a) 33 ... CDxf6 34 ~xf6 Wg8 (34 ... d2 35 tion ...
.l:!.xg6 d1'iY 36 'iVd6+ and mate) 35 .l:!.g4! 'iVc7+
(35 ... 'iYb2 36 "iUxa6 d2 37 CDc4!) 36 d6 'ii'd8 37
"iUe5! .l:!.b6 38 l:ld4, and things are hopeless
for Black;
b) 33 ... CDe5 34 'iYe4! d2 (34 ... "iUc7 35 ~f4!,
and if 35 ... Wg7, then 36 CDh5+!) 35 "iUxe5
d1'iV 36 CDd7+! Wg8 37 "iUxb8+ Wh7 38 ~e3 or
37... i..c8 38 .l:!.e3! "iUd4 39 CDb5, and Black
loses, despite having two queens. And after
the more tenacious 34 ... 'iYb2! (34 ... "iUc3?! 35
'iVa4) 35 CDb1! .l:!.b6 (35 ... Wg7 36 "iUf4) 36
"iUxe5 "iUxe5 37 CDd7+ c,t>e7 38 CDxe5 .l:!.xb1 39
CDxd3 Wd6 40 CDf4 White has a technically
won endgame; 32 i..xd3!
c) 33 ... CDc5 34 d6! CDe6 (if 34 .. :iha3? or By this point I had less than 10 minutes
34 ... .l:!.d8?, then 35 iVh5! is decisive, while if left on my clock, whereas Karpov had
34 ... 'iYb2 35 "iUf4!) 35 CDd7+ Wg8 36 "iUd5! slightly more than half an hour, but after 32
"iUe1! (36 ... d2? 37 .l:!.xg6+) 37 .l:i.xd3 i..xd3 38 i..xd3 I became absolutely calm, since I felt
CDxb8, and again Black is balancing on the intuitively that White had nothing to fear.
edge of a precipice. Of course, Black can win a piece, but now
The impression may be gained that the the white queen gains the chance to steal
156
The Third Match: 1986
Analysis diagram
157
Kasparov vs. Karpov: 1986-1987
obtained because of mate in two moves, d6! or 39 .. .£6 40 ~7+ ~h6 41 d6 does not
while 38 ... ~f6 leads to the loss of the old help) 40 ~8+ ~g7 41 'i¥e5+ ~g8 42 'i¥e8+
one after 39 'iVxf7+ 'it>xe5 40 'iVg7+) 39 iVc3! 'it>g7 43 d6, or instead 39 ... ~h7 40 'ii'f6 'i¥f5
(after 39 ~xc2? Black emerges unscathed - 41 'i¥xf5 gxf5 42 d6 c3 43 d7 c2 44 d8'i¥ c1 'i¥
39 .. :ii'xd5 40 ~2 'iVe6!), and since 45 'i¥f6.
39 .. :~xd5? 40 etJd7+ or 39 .. :~b1 40 etJd7+ It is very noticeable how important it is
'it>g8 41 ~e5! is bad for Black, as is 39 ... f6 40 in such endings to have a secure shelter for
etJg4 'iVf7 41 d6 ~d3 42 'iVxd3 c1"if 43 d7, he the king. The draw is achieved by a para-
is forced at any price to divert the white doxical diverting manoeuvre: 37... ~a8! 38
queen from the a1-hB diagonal: 39 ... c1~ 40 'i¥d8 'i¥d4! (38 ... ~b7? loses to 39 h5! gxh540
~xc1 ~xd5 41 ~a1! (a victorious return to ~g5+ ~f8 41 etJe5) 39 'i¥xa8 'i¥xg4. Little
the chosen diagonal) 41...f6 (41...1We6 42 seems to have changed compared with the
etJd7+ f6 43 etJc5 - this is why the queen is previous position, but the lack of a check at
needed at a1!) 42 etJxg6, and the win for b2 prevents White from successfully com-
White is merely a question of time; bining the advance of his d-pawn with
b) 36 .. :iVd4? 37 1Wc6! c3 38 ~xa6 c2 39 checks to the king: 40 ~a1 + ~f8! (after
'iVa3! 'iVd2 40 'i¥b2+ ~f8 41 'iYh8+ ~e7 42 other moves the difference in the placing of
~f6+ ~e8 43 d6 or 40 ... ~g8 41 ~8+ 'it>h7 the white queen is not a factor) 41 d6 ~e8
(41...~g7 42 'iYe5+) 42 ~f8 ~d4 43 ~xf7+ 42 'iVa4+ ~d8 43 'i¥a5+ (43 'i¥c6 'iVc8)
'iVg7 44 'YWf4, and now the curtain comes 43 ... ~e8 44 'i¥a8+ (44 ~5+ 'i¥d7) 44 ... ~d7
down; 45 ~8 'i¥e6 46 'i¥c7+ ~e8, and White is not
c) 36 ... ~b7!. The correct defence. Now able to strengthen his position - 47 g4 c3 48
after 37 etJe3 'iVd4 38 'iVxb7 'iVxe3 39 d6 'iVd2 'i¥xc3 'iVxd6+ 49 ~h3 leads to a slightly
40 'iVe7 c3 41 d7 c2 42 d8'iY 'iYxd8 43 'ii'xd8 better, but drawn endgame.
c1 'iV Black saves himself in the queen 2) 34 ... ~e8 35 'i¥xa6 l:!.xg3, and after 36
ending, but after 37 M! he still has to play d6 ~f8 37 ~xg3 c3 38 etJe3 c2 39 'i¥c8+ r3!ig7
with extreme accuracy. 40 etJxc2 'i¥e5+ 41 f4 'i¥xd6 or 37 fxg3 c3 38
etJe3 'i¥d2 39 'iVc8+ ~g7 40 'iVxd7 'i¥xe3 41
'i¥c8 'i¥d2 42 d7 c2 43 d8'iV 'i¥xd8 44 'iVxd8
c1 'i¥ Black exchanges his c-pawn for the
dangerous passed d-pawn, and with pawns
on one wing he reaches the haven of a
draw. White can keep his d-pawn for the
moment - 36 ~xg3 c3 37 etJe3, but after the
accurate continuation 37 ... ~4 it becomes
hard for him to breach the opponent's
defences.
Karpov rejected 32 ... l:!.xd3 and 32 ... l:!.xa3,
as they would have forced Black to fight
only for a draw a pawn down, and he chose
Analysis diagram the sharp capture with the pawn, hoping to
queen it in the time scramble. This was a
For example, after 37 ... 'i¥d4? 38 'iVxb7 speculative move, but not bad luck. ..
'iVxg4 39 ~2+ the queen ending surpris- 33 'i¥f4
ingly proves to be lost: 39 ... 'it>g8 (39 ... ~f8 40 Of course!
158
The Third Match: 1986
Analysis diagram
159
Kasparov vs. Karpov: 1986-1987
his pawn and spend a tempo on removing l:hb3). For example, 35 .. .'it>g7? leads to an
his king from the danger zone - 37 ... 'it>e8?, immediate rout: 36 ct:Jf5+ 'it>h7 37 'li'f8 etc.,
but then the idle knight on a3 gains the and so it has to come out into the open field
chance to exact terrible revenge for the - 35 ... 'it>e8. However, White lacks the force
debasement to which it has been subjected. to land a decisive blow: 36 'li'c6+ 'it>f8! 37
'li'd6+ 'it>e8, or 36 l::tg5 d1'1i' 37 l::te5+ 'iVxe5+
38 'li'xe5+ 'it>f8! 39 'li'xf6 'li'xd5 40 'li'xa6 'li'e5+
41 g3 'it>g7! 42 ct:Jc4 'li'e1 43 ct:Jg4 l::tb1 44 'li'f6+
'it>g8, in each case with a draw.
The main resource is 36 'li'xa6 d1'1i' (this
cannot be delayed: 36 ... l::txg3? 37 'li'c8+ 'it>e7
38 d6+ 'it>xd6 39 ct:Jc4+) 37 ~c8+ 'it>e7 38 ~c5+
(against two queens one cannot take liber-
ties: 38 ct:Jc4? 'li'xf2! 39 d6+ 'li'xd6) 38 ... 'it>d7!
(otherwise White would decisively include
his a3-knight in the attack - 38 ... 'it>e8? 39
ct:Jc4 etc.). Now after 39 ct:Jc4 'iVba1! the
threat of mate obliges White to force a
Analysis diagram draw: 40 'li'c6+ 'it>d8! 41 'li'd6+.
But he also has another way of attacking
38 ct:Jb1! d1'1i' 39 ct:Jc3!. A fork of two - 39 'li'c6+ 'it>e7 40 d6+ 'it>f8 (it is dangerous
queens! When has a knight been so lucky! to play 40 ... 'it>e6 41 'li'c8+ ct:Jd7 42 'li'e8+ 'it>d5
Black cannot avoid a prosaically lost end- 43 'li'xf7+ 'it>c5 44 l::tg5+!, while 40 ... 'Ii'xd6 41
game, since if 39 ... ~5d4 White has the 'li'xd6+ 'it>xd6 42 ct:Jc4+ 'it>e7 43 ct:Jxb2 l::txb2 44
decisive 40 'li'e6+. l::tf3 l::tb7 45 g4 'it>e6 leads to an uninteresting
38 ct:Jd6+ 'it>g8 39 'li'xg6+ 'it>f8 40 'li'f6+ endgame, where White has few chances of
'it>g8. Draw? Not immediately, since White converting his pawn advantage, but with
has a strong measure in reserve - 41 ct:Jf5!. an extra queen it is not easy to agree to go
The black king is in a mating net, and it into such an ending) 41 'li'c8+ 'it>g7 42 ct:Jf5+
appears that even the two queens are 'it>h7 43 'li'f8 ct:Jg4+ 44 l::txg4.
unable to prevent its execution. But, re-
phrasing a well-known chess saying about
the advantage of the two bishops, it can be
said: 'The advantage of two queens (!) is
that one of them can always be advanta-
geously given up!'. Black replies 41..:~xf5!
(the over-showy 41 .'Ii'xg2+ does not
achieve its aim: 42 'it>xg2 ..ib7+ 43 'it>g3 'li'f3+
44 'it>h4 'li'xf2+ 45 'it'h5! 'li'f3+ 46 'it>g6 'li'g2+
47 'li'g5) 42 'li'xf5 'li'd6+ 43 f4 'li'xa3, and here
the three pawns for the piece give White no
more than a moral advantage.
But all these fears seem child's play
compared with the dangers awaiting the Analysis diagram
black king after 35 'li'd6+ (instead of 35
160
The Third Match: 1986
44 ... 'iVe5+? 45 tiJg3 (but not 45 f4? 'iVxf5 also wins the opponent's! As if at the
46 .l::i.h4+ 'iVdh5) 45 ... g5 (if 45 .. .'iVf6, then 46 waving of a magic wand, the scattered
tiJe4 'iVe5+ 47 g3 with irresistible threats) 46 white pieces achieve amazing harmony.
'iVxf7+! 'iVg7 47 'iVe6, and White should win.
Now let us display a little imagination
on the part of Black ... And instead of the
plausible 44 ...'iVe5+ we discover an amazing
possibility - 44 ... .l::i.xh3+!!. Grandiose! A
forcing variation leads us to a draw: 45
'it'xh3 'iVh1 + (not 45 ... 'iVc3+? because of 46
'it'h2! 'iVxg4 47 'iVh6+, and mate with the
knight on the next move) 46 'it'g3 'iVe5+
(46 ... 'iVxa3+? 47 f3 'iVe1+ 48 'it'h2 'iVe5+ 49 f4
is bad for Black) 47 .l::i.f4 g5! 48 'iVxf7+ 'it'h8 49
'iVf8+ 'it'h7. A desperate attempt to play for
a win - 50 'iVe7+ 'iVxe7 51 dxe7 gxf4+ 52
'it'xf4! (after all, it is not easy for the black When I landed this deadly blow with an
queen to battle on its own: 52 ... 'iVc1+? 53 undefended pawn, a storm of applause
'it'g4 'iVc8 54 tiJb5 'iVc4+ 55 tiJbd4) is most broke out in the auditorium. The chief
simply parried by exploiting a study-like arbiter Lothar Schmid vigorously waved
motif: 52 ... 'iVxg2! 53 e8'iV 'iVg4+ 54 'it'e5 'iVe4+ his arms, calling on the spectators to be
55 'it'xe4 - stalemate! quiet, and for a time he succeeded.
Compared with the' canvasses' we have 37 ... 'it'e6 (time-trouble agony) 38 .l::i.e8+ 'it>d5
seen (normally, alas, remaining off-stage), 39 .l::i.xe5+ tiJxe5 40 d7 .l::i.b8 41 tiJxf7 1-0
the combination carried out by White in the
game looks very amateurish!
34 tiJh6 'iVe7 35 .l::i.xg6 'iVe5
For an instant Black has everything in
order - White's strongest piece is crippled ...
161
Kasparov vs. Karpov: 1986-1987
I cannot describe that intoxicating wave among those lost through 'bad luck'. It
of happiness which engulfed me at the end turns out that in this simple way every-
of this game. The emotional rejoicing was thing can be explained! Nothing particular
perhaps comparable only with my reaction happened, just one blunder caused by bad
to the final, 24th game of the previous luck. .. But is it really only bad luck that is to
match. Of course, a brilliant and memorable blame? Where is the clash of aggressive
win over a formidable opponent is always a plans, the sacrifices and counter-sacrifices,
red-letter day for any genuine chess player, the labyrinths of mind-boggling variations
but that day I intuitively felt that something - all that which embodies a struggle of the
more significant had occurred. The time had highest intensity and demands the total
not yet come for a detailed analysis of all the commitment of the two players?
upheavals of this most fascinating encoun- No, the role of the 16th game cannot be
ter, and as yet it was difficult to give an belittled by a reference to bad luck, under-
accurate assessment of the two sides' plans mining the very essence of chess justice! In
or delve into all the nuances of the swiftly my view, for all the mistakes by both sides,
changing situation, but I was in no doubt it is unique as regards the intensity and
that none of our other games compared with scale of the problems which the players
this one for richness of content. faced in a new, very complicated position,
Later, when I was working on the book which later became one of the fashionable
Dva Matcha, the meticulous analysis of this tabiyas of the Ruy Lopez (incidentally,
game took me two weeks of painstaking Karpov's desire to get even in this opening
work and convinced me of the correctness of was to cost him our 1990 match, where he
this conclusion. In my opinion, the enor- suffered two crushing defeats in the Zaitsev
mous amount of commentary is fully justi- Variation).
fied: the result is a fascinating journey into The outcome of this grandiose battle es-
the colourful world of the most varied ideas, sentially symbolised the collapse of Kar-
which comprise the basis of modem chess. pov's hopes of gaining revenge. After
As they quickly spread, many of the side seizing the opening initiative and ceasing to
variations begin to acquire independent avoid an open battle, he was undoubtedly
significance, and the 16th game essentially aiming to demonstrate his superiority in all
unites a great number of fine and frenzied the crucial debates. Break the opponent's
battles, the analysis of which gives genuine resistance using a broad arsenal of means,
pleasure to chess connoisseurs. and in particular the opponent's own chess
Unfortunately, in general the chess methods - this was the strategic aim which
commentators were not especially enthusi- Karpov probably set himself. But on the
astic about this game, but regarded it as an whole his blows missed their target; more-
irregular accumulation of accidents. The over, the knock-out in the 16th game essen-
sharp disruption of the balance on different tially decided the match - my enormous
parts of the board, the lack of correspon- lead (9V2-6V2) and obvious playing advan-
dence with generally accepted dogmas, the tage, shown by my confident actions in the
raid of cavalier recklessness - all this, in 13th, 14th and 15th games, did not leave
their opinion, deprived the game of the this in any doubt.
necessary aesthetic consistency. However, But it is one thing to discuss this calmly
the commentators were successfully per- when it was all long ago, and quite another
suaded by Karpov, who listed this game to come to such a conclusion during the
162
The Third Match: 1986
course of the match. It is a very dangerous After this game, and especially the 19th,
feeling - that's it, the match is over, it's only many asked the question: should the Griin-
the formalities that remain ... By not ridding feld Defence have been employed in such a
myself of this feeling, I committed a psy- favourable match situation? After all, I
chological blunder and was unable to realised perfectly well that the tone of
maintain that maximum level of concentra- Karpov's play could be raised only by
tion which had served as a reliable shield successful home preparation. And it proba-
during all my trials. However, both earlier bly made sense to change the pattern of the
and later, a distinct slump occurred in my play, for which a time-out would have
play after great great emotional stress. come in useful.
4 tt::lf3 ~g7 5 'irYb3 dxc4 6 'irYxC4 0-0 7 e4 ~g4
Three Zeroes
Of course, the extreme match tension had
exhausted both players, but whereas for
Karpov the next game was his only chance
of lifting the burden of failure, for me, on
the contrary, a slight break was required in
order to gather my thoughts and genuinely
motivate myself for the final games of the
match. In a similar situation after my win in
the 14th game I took a time-out and man-
aged to prepare successfully (mainly emo-
tionally) for the next two encounters. But at
this moment, not wishing immediately to
make use of my last break, I refrained from Again the Smyslov Variation. I think that
such a sensible step. this was a psychological mistake. Before the
After missing this chance and going game I had doubts about whether to repeat
along to the 17th game in a disorientated 7 ... ~g4 or play 7... tt::la6 - we had already
state, without myself realising it I brought intensively prepared this variation, al-
back the disappearing intrigue to the though, as it transpired in the 19th game,
match. Effectively this game opened a new here too there was a novelty awaiting me.
stage of the match - a completely different Nevertheless I should have 'changed
contest began! It should not have been tune'. If not 7... tt::la6, then check what had
forgotten that Karpov always fights to the been prepared against 7 ... a6 - we had also
last second ... looked at this move, but I was afraid of the
reply 8 eS, and, perhaps, not without rea-
son: many years later in this way Karpov
Game 17 overcame Kamsky (9th match game, Elista
A.Karpov-G.Kasparov 1996), Leko (Tilburg 1996) and Svidler (Dos
World Championship Match, Hermanas 1999), although on the whole
17th Game, Leningrad 17.09.1986 Black has sufficient counterplay, and the
Grunfeld Defence 098 sharpness of the position is not altogether
in accordance with Karpov's style. More-
over, in this game the ex-champion would
1 d4 tt::lf6 2 c4 g6 3 tt::lC3 ds have been encountering 7... a6 for the first
163
Kasparov vs. Karpov: 1986-1987
164
The Third Match: 1986
ISth game, when at the board he was Black regains the pawn with sufficient
unable to work out what to do after the counterplay: 22 CtJa4 CtJdS 23 ~c1 (if 23 'it>f2
unexpected move 13 ... CtJc8. But here I or 23 ~d2, then 23 ... ~xb2) 23 ... CtJb4 or 22
wrongly went in for this position a second ~c1 ~d4 23 CtJe4 ~xb2 24 'it'e2 CtJdS 2S 'it'f3
time, since it was a position that clearly lent l:tb4 etc.
itself to home analysis, and as a result I too In addition, 16 ... a6!? was also not bad:
was unable to work out what to do at the 17 ~xc6 bxc6 18 f4 ~f6 19 l:txc7 CtJa7 20 0-0
board - such tedious endings were not to (20 a4 l:tfb8!) 20 ... CtJbS 21 l:txc6 ~xc3 22 bxc3
my taste ... CtJxc3 with equality (Kaufman-Kudrin,
14 ... ~xf3 (14 ... ~e6? IS CtJgS) 15 ~xf3 Chicago 1997).
But Karpov had other ideas.
16 ... bxc6 17 ~d4!
The most accurate solution from the po-
sitional point of view. White intends an
exchange of minor pieces, leaving the
opponent with a broken pawn structure on
the queenside and a restricted knight on c8.
'Without this subtle manoeuvre, Karpov's
opening idea would have lost its point.
Now White's initiative more than compen-
sates for his minimal material deficit.'
(Taimanov)
Zaitsev assessed this position more
15 ... ~xe5 modestly in Infarmatar NoA2 - simply
Most of the commentators recom- 'compensation for the pawn'. And he was
mended IS ... CtJxeS(?) as the lesser evil. I proved right! In fact the knight can never-
cannot agree with this. After 16 ~xb7 l:tb8 theless escape from c8 - via either a7 or e7.
17 c6! CtJc4 18 CtJdS! (18 l:td7 is inaccurate: But first Black must decide whether or not
18 ... CtJxe3 19 fxe3 ~eS 20 CtJdS e6 21 ~xc8 to exchange bishops. I replied quite
~g3+! 22 'it'd2 Itxb2+ 23 'it'c1 l:txa2 or 22 quickly.
'it'dl exdS 23 Si..b7 l:tbe8 24 'it'd2 fS and .. .fS-
f4 with equality) 18 ... CtJxe3 (18 ... CtJxb2?! 19
l:tbl) 19 fxe3 CtJd6 20 CtJxe7+ 'it'h8 White
gains a great advantage by the exchange
sacrifice 21 l:txd6! cxd6 22 'it'e2!, for exam-
ple: 22 ... ~xb2 23 l:tbl i1i.f6 24 CtJdS ~d8 2S c7
~xc7 26 CtJxc7 fS 27 l:tbS l:tf7 28 i1i.dS MxbS
29 CtJxbS l:td7 30 'it'd3 with a won endgame.
16 ~xc6!
A surprise. I was mainly expecting 16
l:td7 (we had looked at this move), planning
16 ... e6 17 ~xc6 (if 17 ~h6 l:te8 18 CtJbS CtJ8e7
19 CtJxc7 there is 19 ... l:tad8!) 17 ... bxc6 18 f4
~g7! (18 ... ~xc3+ 19 bxc3 is too depressing) 17 ... ~f4
19 l:txc7 l:te8 20 l:txc6 CtJe7 21 l:ta6 l:teb8, and Black intends to free himself by ... e7-eS
165
Kasparov vs. Karpov: 1986-1987
and ... etJc8-e7 (for example, 18 g3 e5 or 18 able to come into play via e7).
etJe2 e5). 17... iH6?! was not good in view of I also analysed the direct 19 l:td7(!) etJa7
18 i..xf6 exf6 19 l:td7 l:tb8 20 b3 l:te8+ 21 20 a4! (of course, not 20 l:txe7 etJb5 21 etJxb5
'it'd2. l:tfe8! or 20 l:txc7 etJb5 21 etJxb5 axb5 22 a3
In the heat of the moment I also as- b4, and Black becomes active) 20 ... Ufb8 21
sessed the position after 17 ... i..xd4!? 18 l:txe7 l:txb2 22 0-0 Uc2 23 etJe4 l:tc4 24 l:td1
l:txd4 as 'very difficult' for Black. And only l:txa4 25 l:tdd7 l:tf8 26 etJg5 etJc8 27 l:txf7 l:txf7
years later did I discover that this was the 28 etJxf7 'it'f8, doubting 'whether White's
simplest way of equalising. I must correct advantage is sufficient for a win', but after
the pessimistic variations given in the book 28 l:txf7! it is quite sufficient. Moreover,
Dva matcha: nothing better for Black is apparent: the
1) 18 ... Ub8 19 b3 a5, and if 20 l:.d7 Black familiar idea 19 ... l:tb8 20 b3 etJa7 (20 ... l:tb7 21
is saved by 20 ... etJa7! 21 l:txc7 etJb5 22 etJxb5 'it'd2 etJa7 22 l:te1 is also insufficient) 21
l:txb5 23 l:txc6 a4 or 21 0-0 etJb5 22 etJxb5 Uxc7 etJb5 is not supported here by the ... a5-
l:txb5 23 l:td l:tb7 24 l:txe7 l:td8. Also noth- a4 resource, and 22 etJxb5 axb5 23 l:txc6 l:tfc8
ing is given by 20 'it'e2 etJa7 21 l:td7 l:tfc8 22 24 l:txc8+ ltxc8 25 b4 is clearly advanta-
l:txe7 etJb5 23 etJxb5 Ue8!. After 20 l:ta4, on geous to White.
the other hand, it would appear that 'Black The conclusion: after 17 ... i..xd4 18 l:txd4
is unable to escape from the vice'. If both 18 ... ltb8 and 18 ... a5 would have led to
20 ... l:ta8 the commentators suggested 21 a draw (but not 18 ... a6?! because of 19
'it'e2 (21 0-0 etJa7 and ... etJb5 is equal) l:td7!). With the bishops still on the board,
21...l:td8(?) 22 l:td1 l:txd1 23 etJxd1 followed the situation remains more tense.
byetJe3-c4.
However, after 21...etJa7 (22 l:txa5? etJb5)
or even 21...e5 22 Ud1 etJa7 23 Ud7 etJb5 24
etJxb5 cxb5 25 l:te4 f6 26 l:te3 (26 l:txc7 l:tfc8)
26 ... l:tf7 Black easily gains a draw;
2) 18 ... a5. Now after 19 Ud7, apart from
19 ... l:tb8 20 b3 etJa7, either 19 ... a4 20 l:.xc7
l:ta5 21 l:txc6 etJa7 or 19 ... etJa7 and ... l:tfb8 is
not bad. Therefore I recommended 19 'it'e2
etJa7 20 a4!, and 'Black is still unable to
solve satisfactorily the problem of his
knight' (although, I should add, after
20 ... Ufb8 21l:.b1 ~f8 his defences are solid).
But after the standard 19 ... l:tb8! White has 180-0
not even a hint of an advantage: 20 l:tb1 'The opening battle has been won by
l:tb4 21 l:td7 l:tc4 or 20 b3 etJa7 (d. variation White'. (Taimanov) As we will now see, all
1); is not so gloomy, and castling, although it
3) 18 ... a6?!. After this I recommended brought Karpov a quick win, does not give
the 'suffocating' 19 a4(?!) etJa7(?!) 20 'it'd2 any advantage. Nor does 18 etJe2 e5 19 i..c3
Ufd8 (20 ... l:tfb8 21 'it'd) 21 l:.xd8+ l:txd8+ 22 (19 etJxf4 exf4!) 19 ... etJe7 20 g3 i..h6 21 i..xe5
'it'c2, aiming for the exchange of rooks and etJd5 with equality.
a favourable knight endgame. But after Therefore 18 g3 e5 (White is slightly bet-
19 ... Ub8! Black is alright (his knight will be ter after 18 ... i..h6 19 i..e5 l:tb8 20 b3) 19 gxf4
166
The Third Match: 1986
exd4 20 .l:i.xd4 deserves some attention, Thus 18 ... eS! would have allowed Black
although after 20 ... .l:i.b8 21 b3 .l:i.e8+ and to hope for a draw (in my games with
... ttJe7 Black again does not experience any Karpov I have saved far more difficult
particular difficulties. positions). But the most interesting thing is
18 ... aS? that even after the 'dreadful move' 18 ... aS,
This incorrect plan is another confirma- to which I attached two question marks in
tion of the fact that I should have taken a Ova matcha, Black is still by no means lost!
time-out. One involuntarily recalls 20 ... aS 19.1:i.fel!
from the Sth game, but there the pawn was 'That's it - the knight on c8 cannot man-
at least aiming for the queening square ... age to get to the centre, and the game is
18 ... f6 would have enabled Black to re- essentially decided' (Ova matcha). Well, let's
sist, for example: 19 ite3 .i.xe3 20 fxe3 .l:i.b8 see ...
21 b3 'it'f7 22 .l:i.d7 .l:i.b7 with the intention of
... a7-aS(a6) and ... ttJa7 - White's advantage
is not apparent.
But the best was 18 ... eS! 19 .i.e3 .i.xe3 20
fxe3 ttJe7 (the e-pawn is alive and the
knight is in play). Karpov was clearly ready
for this turn of events, but in a game with
Timman, played 20 days after the match
(Tilburg 1986), he was unable to show an
advantage for White, although he undoubt-
edly demonstrated his analysis: 21l:i.d7 ttJfS
22 lhc7 (if 22 'it'f2, then 22 ... .l:i.ad8 23 .l:i.fdl
.l:i.xd7 24 .l:i.xd7 .l:i.c8 2S ttJe4 ~f8 26 g4 ttJg7!
and ... ttJe6 is good) 22 ... .l:i.fc8! 23 .l:i.d7 .l:i.d8 24 19 ... a4?
.l:i.fdl .l:i.xd7 2S .l:i.xd7 ttJxe3 26 .l:i.c7 (26 ttJe4 Only this is indeed the decisive mistake.
ttJdS 27 ttJgS f6 with equality - Karpov) 'A continuation of the faulty plan. Of
26 ... .l:i.b8! (the key move, apparently not course, 19 ... fS was not an easy move to
taken into account by the ex-champion's decide on, but Black's control of e4 would
team) 27 b3 .l:i.d8! 28 ttJe4 (28 .l:i.xc6 .l:i.d2!) have given him some chances.' (Karpov) In
28 ... .l:i.d4 with equality. To the credit of the my view, after 20 g3 .i.h6 21 ~e6 or 20 ttJe2
Dutch grandmaster, he was not afraid to .i.gS 21 .i.eS ttJa7 22 ttJd4 these chances are
check the depth of White's plan, and he very slight.
upheld the Smyslov Variation. 19 .. .f6 was far stronger (a move aimed at
True, according to Karpov 'the impres- solving the nagging problem of the knight
sion remained that somewhere White could on c8; 19 ... e6? is bad in view of 20 .i.f6! with
have played more strongly'. But where? complete paralysis), after which the com-
Say, 23 .l:i.xc8+ (instead of 23 .l:i.d7) 23 ... .l:i.xc8 mentators, including myself, unanimously
24 'it'f2!? (24 e4 ttJd4 is equal- Karpov) with recommended 20 .l:i.e6 .l:i.a6(?) 21 ttJdS. But
the hope of gaining a microscopic plus after after 21.. ..i.d6! 22 ttJxf6+ exf6 23 cxd6 cxd6
24 ... .l:i.d8 2S tLle4 .l:i.d3 26 .l:i.el. However, the 24 .i.xf6 cS Black is not yet losing (although
accurate manoeuvre 24 ... ttJg7!? and ... ttJe6, it is doubtful whether anyone would want
disclosing the weakness of the cS-pawn, to play such a position). However, 21 ttJe2!
maintains the balance. .i.h6(gS) 22 .i.c3 'it'f7 23 ttJd4 ttJa7 24 .l:i.del
167
Kasparov vs. Karpov: 1986-1987
gives White an overwhelming advantage. Taimanov and Zaitsev gave 24 ... l:IfdS, but
But after 20 ... l:IdS! 21 l:Ixc6 l:Ia7 22 g3 after 25 l:!.xdS+ l:!.xdS 26 ~xg7 'it>xg7 27 l:!.xe7
~h6 Black can still put up a fight - thus, in l:!.d3 2Sl:!.xa7 (2Sl:!.e3?! l:!.d2! 29l:Ie2 does not
the variation 23 ~e3 l:Ixd1+ 24 ttJxd1 ~xe3 work because of 29 ... l:Ixe2! 30 ttJxe2 'it>f6,
25 ttJxe3 'it>f7 26 ttJd5 'it>eS! he confidently equalising) 2S ... l:!.xc3 29 l:!.xa3 l:!.xc5 30 l:!.a6
gains a draw (27 l:Ixc7 l:Ixc7 2S ttJxc7+ 'iitd7 Black has to save a rook ending a pawn
29 ttJd5 'it>c6 etc.). down.
White can consider 20 g3 ~h6 21 f4!? Therefore 24 .. .£6! is better, with highly
(instead of 21 l:Ie6 l:IdS! 22 l:Ixc6 l:Ia7 - d. unclear play: 25 ii.d4 'it>f7 or 25 ~f4 f5! 26
above) 21...l:IdS 22 ~f2 l:Ixd1 23 l:Ixd1 'it>f7 l:!.e3 l:IfdS! 27 l:!.xe7 l:!.xd1 + 2S ttJxd1 ~fS 29
24 l:Id7 l:Ia7 25 l:IdS! l:IaS 26 ttJe4 e5 (only l:!.d7 ~xc5, then ... ttJb5, and White's advan-
not 26 ... ttJb6? 27 l:IxaS ttJxaS, and Black is tage evaporates.
effectively a knight down) 27 fxe5 fxe5 23 ...ii.C1 24 l:!.xC7 ii.b2 (the only move) 25
when he has excellent compensation for the ttJa4!
pawn, but whether he can win is still an
open question.
The remainder is not so interesting:
White methodically converts his advantage
into a win.
2ol:!.e4
'Initially I wanted to stop the pawn with
20 a3, in order to then deal with it. But then
I decided not to waste time on it: after the
capture on a4 Black could play his knight to
b5, with hopes of equalising.' (Karpov)
2o ... ~h6 21 ~e5 a3 22 b3 ttJa7
At last the luckless beast comes into
play, but its imaginary freedom is acquired Again the simplest and most convincing
at too high a price. solution. 'After the incautious 25l:!.xe7 there
could follow 25 ... i.xc3 26 ~xc3 ttJb5 27 ~f6
l:IadS, aiming at the a2-pawn' (Karpov). But
after 2S l:!.e2 nothing would come of this.
25 ... l:!.feS! (Taimanov, Zaitsev) is more
tenacious, for example: 26 ~d4 l:!.xe7 (but
not 26 ... 'it>fS? 27 l:!.xeS+ l:!.xeS 2S l:!.xeS+ 'it>xeS
29 'it>f1 etc.) 27 l:Ixe7 l:!.dS 2S i.e5 ttJb5 29
ttJxb5 i.xe5 30 l:!.xe5 l:!.d1 + 31 'it'h2 cxb5 32
b4 'it>fS, and Black can still struggle on.
25 ... ttJb5 (25 ... ~xe5 26 l:!.xe5 e6 27 ttJb6 or
26 ... ttJb5 27 l:!.xc6 e6 2S l:!.b6 was even more
hopeless) 26 l:!.xc6 l:!.fd8
If 26 .. .f6, then 27 i.xb2 axb2 2S l:!.e1!, for
23l:!.d7! example: 2S ... ttJd4 29 l:!.b6 ttJe2+ 30 'it>f1 ttJf4
Consistent. 23 ~xc7? ~g7 24 ~e5 (24 31l:!.b1 ttJd3 32 c6l:!.fcS 33 ttJxb2 and wins.
ttJb1 ttJb5) was far less accurate. After this 27l:Ib6
168
The Third Match: 1986
'The main idea of White's strategy. He is game I was in the mood to engage my
not diverted by the capture of the bishop opponent in a decisive battle, assuming
on b2, but strengthens his domination of that Karpov would not avoid one.
the board.' (Karpov) And a battle took place! In the 18th
27 ... MdS game, in the words of the observers, I
An attractive trap before the curtain played 'brilliant and original chess', with
comes down, but there are to be no mira- two of my moves proving especially strik-
cles in this game. ing. One commentator wrote: Typical
Kasparov! The entire board was in flames'. The
game nearly became my best creative
achievement in the matches with Karpov.
Alas, only nearly ...
Taimanov: 'For dramatic complexity this
game was unequalled in the return match. It
was not only the longest, but also the most tense
and exciting. Never before had the chances of
the two contestants, playing at the limit of their
strength and emotion, swung so violently -
from victory to defeat. There were more mis-
takes than usual. Yet as regards originality of
plans, richness of ideas and diversity of events,
28 ii.g3 the 18th game can be called the culmination of a
Of course, not 28 tLlxb2? Mxe5! 29 J::txe5? fascinating duel.'
(29 tLlc4 Mxe4 30 Mxb5 is essential) 29 ... axb2 The miracles in this unforgettable game
30 Mel Mxa2 31 ~h2 Mal, and White even began right from the opening. There is
loses. every reason to assume that both my first
28 ... tLlC3 29 tLlxC3 i.xC3 30 c6 ii.d4 (30 ... MC5 move and the position that was soon
31 c7) 31 Mb7 1-0 reached did not come as a surprise to
Times: 1.38-2.12. Karpov.
169
Kasparov vs. Karpov: 1986-1987
6 e3
Today 6 ct:Jd2!? is more often played,
avoiding the variation with ... g7-g5 and
... ct:Je4, and after 6... ..txc3 and ... d7-d6
immediately setting up a 'big centre' by f2-
f3 and e2-e4. I was also one of the first to
begin employing the knight move - against
Psakhis (2nd match game, Murcia 1990)
and Onischuk (Tilburg 1997). True, with
Psakhis after 6 ... ..txc3 (nowadays even
'After four moves an argument broke 6... h6 7 ..tM ct:Jc6!? is being tried) 7 bxc3 h6
out in the press centre: what should this 8 ..tM g5 9 ..tg3 d6 10 e3 a position from the
opening be called - a Nimzo-Indian De- present game was reached, but 10 f3!? was
170
The Third Match: 1986
possible with the idea of e2-e4. stricted, and the importance of White's
6 ... h6 7 .Jth4 .JtxC3+ centre and his spatial advantage are in-
But not immediately 7... d6? because of 8 creased.
'liUa4+ ct:Jc6 9 dS. Earlier here Black used to Incidentally, this plan is quite old and is
play the more' sweeping' 7... gS 8 .Jtg3 ct:Je4, known from the game Botvinnik-Keres
but in the 1980s he often experienced (12th USSR Championship, Moscow 1940):
difficulties both in the gambit line 9 ct:Jd2 9 ... eS 10 f3 'liUe7 11 e4 ct:Jbd7 12 .Jtd3 gS
ct:Jxc3 10 bxc3 .Jtxc3 11 .l:!.c1 .Jtb4 12 h4 gxM (12 ... ct:Jf8 13 cS?! Tal-Hecht, Varna Olym-
13 .l:!.xM (2nd game of my match with piad 1962 - Game No.132 in Volume II of My
Timman), and after the classical approach 9 Great Predecessors; 12 ... cS 13 ct:Jf1 Azmai-
'liUc2 .Jtxc3+ 10 bxc3 d6 (instead of 10 ... ct:Jxg3, parashvili-Khalifman, Kuybyshev 1986) 13
Langeweg-Karpov, Amsterdam 1981) 11 .Jtf2 ct:JhS 14 g3 etc. Later the strategy of
.Jtd3 fS 12 dS ct:JcS 13 M! (4th and 6th quickly seizing the weak fS-square became
games). topical - 14 ct:Jf1! (Tal-Mnatsakanian, 30th
8 bxc3 d6 USSR Championship, Yerevan 1962).
It is clear what attracts those playing 9 ... gS
Black to such set-ups: defects in the oppo- At that time this half-forgotten move
nent's pawn chain, a lack of scope for the had only just once again appeared on the
white bishops, and the flexible placing of his scene. The bishop is driven away from h4
own pieces. The virtues of the position from before White prepares a retreat for it at f2,
White's standpoint are also apparent: pow- and now carrying out the 'Botvinnik plan'
erful pawn centre, potentially dangerous will involve the loss of a tempo. But White
bishops ... The answer to the question 'who can change his plan: the early ... g7-gS
will win?' is even clearer - it is the one who allows him to seize the h-file by h2-M.
can demonstrate the correctness of his own Therefore it is better not to hurry with
view on the position: what in fact are more either 9 ... eS or 9 ... gS - later games showed
important - the defects in the pawn chain or that 9 ... ct:Jbd710 f3 'liUe7 is more flexible, and
the strength of the centre, the 'dangerous' then 11 'liUa4 eS 12 e4 0-0 or 11 e4 gS 12 .Jtf2
white bishops or their 'lack of scope' ... ct:JhS (13 g3 fS) with roughly equal play.
9 ct:Jd2 10.Jtg3
In my opinion, the most logical plan: af- The first critical moment. 'Whereas the
ter f2-£3 and e3-e4 the bishop at b7 is re- two contestants played this part of the
171
Kasparov vs. Karpov: 1986-1987
game very quickly, almost without pausing Also a rapid reply, but an unexpected
for breath or hesitating over their choice of one for me - and yet another novelty by
moves, the spectators had much to con- Karpov in this match! By retaining the
sider. Initially they were all surprised that option of ... ctJb8-c6, he diverts me from my
Kasparov, after winning twice in succession intended new plan of attack. It can be
in the Ruy Lopez, reverted to 1 d4. Then imagined how shocked I was: 'Again the
there was Karpov's decision to revert to the opponent anticipated an improvement on
Nimzo-Indian Defence, and finally, his my part! From where does he get such
choice of a variation where his opponent insight?!'
had had some happy experiences, whereas Light on the origin of this phenomenon
he himself, as far as I recall, had no such was shed by an article entitled' Autumn on
experience.' (Taimanov) Kamenny Island' by a member of Karpov's
team, the psychologist and journalist Igor
Akimov, published soon after the match in
the magazine Studenchesky Meridian. Here
there is some amazing evidence - it turns
out that on the eve of the game Karpov had
spent a sleepless night analysing the posi-
tion after 10 ... lIVe7, since he was completely
sure that it would occur.
It is already well past midnight, and Karpov
is still looking at the position ... I am sitting
opposite him. Between us is the board. To me
White seems to stand well, but I don't like
Karpov's position at all. I ask:
After the usual10 .. .'~Jbd7 I was planning 'Where does this position come from?'
11 M! ~g8 12 hxg5 hxg5, and now not 13 'It'll be reached tomorrow ... '
lIVc2 lIVe7 14 e4 0-0-0 15 ~e2 e5 16 f3 ~h8 (It's already been today for a long time: it's
(Miles-A.Sokolov, Bugojno 1986) or 13 f3 after two in the morning.)
lIVe7 14 lIVa4 a5 15 0-0-0 0-0-0 (Legky- 'It'll be reached tomorrow ... ' This is said as
Shneider, Lvov 1986) with good play for though there is no question about it - simply,
Black, but 13 iVb3!?, preventing queenside without any emphasis ... I realise that I've been
castling (with the idea of 13 ... lIVe7 14 c5) or given an honest reply, but as a dilettante I can't
13 lIVa4!? (the source game: Bareev- accept it.
Gavrikov, 54th USSR Championship, 'But listen Tolya, on the way Kasparov can
Minsk 1987), judging the position to be in deviate several times ... '
favour of White. 'Why should Kasparov deviate if he has al-
Similar motifs are in evidence after ready gone in for this line several times? And
10 ... .'~Jc6 (a novelty of the 21st century) 11 always with success.'
M! ~g8 12 hxg5 hxg5 13 ~h6 lIVe7 14 lIVf3! 'It must be a difficult line?'
~g6 15 ~xg6 fxg6 (Bacrot-Adams, Sarajevo 'Not particularly. You see this pawn group?
2000) 16 ctJe4 (16 c5!?) 16 ... ctJxe4 17 lIVxe4 This is from Nimzowitsch. And this set-up is
lIVf6 18 c5! or 14 ... ctJd7 15 c5! (Bareev- from the Queen's Indian.'
Karpov, Ajaccio (blitz) 2007). Karpov makes a flurry of moves, his fingers
10•. :~Ve7!? flying over the board. He has assessed the
172
The Third Match: 1986
resulting position. He doesn't like it ... was he so sure that Kasparov would stumble
'I don't have a feeling for it,' he says. 'I into it and that the entire subsequent course of
move the pieces about, and everything is alien. the game would depend on whether or not he
There is no cohesion. It feels unnatural. The figured out the subtleties? A mystery ...
construction is alien, and so are the dynamics. Akimov immediately hastens to assure
A typical Kasparov set-up.' the reader that 'such foresight is in the
'Why then do you want to go in for it?' nature of things'! But this raises uncomfort-
'You see, here things are not straightfor- able questions for Karpov. Why after two
ward. Kasparov has honed this position at home, crushing wins in the Ruy Lopez should I
it is familiar to him. He knows its possibilities change tack, and begin the game not with
and its pitfalls. He is not afraid of it and he will the king's pawn, but the queen's pawn? And
go in for it with an easy heart. But, on the other how after such a night was he going to reply
hand ... do you see where my queen is? I move it to 1 e4, in view of his crisis in the Ruy Lopez
here on the 10th move - a small, quiet move, but (in world championship matches a player
it changes the assessment of the entire situation usually gives up 'losing' set-ups) and the
and sets the tone for the subsequent play. It's excessively peaceful character of the Petroff
never been played before. If at the board Kas- Defence, against which, incidentally, I had
parov fails to understand the nuances, I obtain prepared a new line? Why after 1 d4 tZJf6 2
excellent counterplay; but if he does, then I end c4 e6 was I bound to play 3 tZJf3 (instead of
up in this box, from which for the moment I the successful 3 tZJc3) and after 3 ...b6 - 4 tZJc3,
don't see a way out ... For him this move is like a rather than 4 g3 or 4 a3, inviting the 'mod-
red light! You'll see tomorrow, how after my em' variations with 4 ... ~a6 or 4 ... c5, which
10th move he'll seize his head in his hands and were then coming into fashion? Let us
will sit and think, until he realises what the suppose that Karpov made a brilliant guess.
point of it is ... ' But, even knowing the opening, how could
'Very well, let's suppose that you don't find he picture in such detail the structure of the
a way out of this quagmire; what will you do? opponent's preparations? How could he
Go in for some other, approved variation?' possibly foresee exactly what would be 'on
'Tomorrow in any case I will play this posi- the board tomorrow', if I was intending to
tion. Here sharp, concrete play is inevitable. employ a continuation that I had never
And if I'm lucky ... ' played before?! And at the same time to be
I was some twenty minutes late arriving for sure that I had 'honed this position at home',
the game, - Akimov continues. Great was my and that it was familiar to me (indeed, we
astonishment when I saw on the demonstration had several pages of analysis), and that
board the same position as the night before. Black's 10th move would come as a surprise
Kasparov was alone on the stage. He was to me! It has to be agreed that there is only
squeezing his temples, rubbing his chin, adjust- one sensible explanation ...
ing his jacket - to all appearances, he felt On the day of the game the incredible
uncomfortable. happened: from Karpov's chess table at
'How did Karpov take him aback?' I asked a home the position migrated on to the stage
master player sitting alongside. of the concert hall of the Leningrad Hotel.
'The queen move to e7.' Instead of the natural 10 ... tZJbd7, Black
Everything happened exactly as Karpov pre- preferred the clever 10 .. :ilVe7 - what Karpov
dicted. A simple move, in no way remarkable ... was avoiding, no one realised, apart from
But what did it signify for Karpov? And why he and I.
173
Kasparov vs. Karpov: 1986-1987
Nikitin: 'The position did not appeal to the tions, the most topical of which is 13 .l:th6
former champion - it seemed "alien" to him, a ttJbd7 14 'iVa4 (Beliavsky-Gulko, Reggio
"Kasparov" position - and yet he went in for it. Emilia 1991; Kasparov-Zarnicki, Buenos
f think that this deliberate psychological trick Aires (simul) 1992; Bacrot-Karpov, Cannes
had the aim of again demonstrating that he 2002; Topalov-Karpov, Vitoria Gasteiz
"knew everything", and it would thereby upset 2007), but in general he has not achieved
his opponent's mental state. Although Kasparov any tangible results.
easily solved the chess problem he had been set The unclear gambit 13 cS!? dxcS 14 .tbS+
and gained an enormous advantage, in time- has also occurred (V.Mikhalevski-Ligterink,
trouble he became over-anxious ... ' Leeuwarden 1995) or immediately 11 cS!?
dxcS 12 .tbS+ (Zviagintsev-Solozhenkin,
Elista 1995). Generally speaking, pawn
sacrifices and other 'drastic' measures are
typical of White's play in this opening
scheme. And this is understandable: the
static elements of the position (such as, for
example, the integrity of the pawn struc-
ture) 'vote' for Black, and White has to play
energetically!
11 ... a5
lla4
After some thought I made a useful
move, directed against Black's queenside
castling, but with an accurate reply he can
solve his opening problems. Therefore later
the main line became 11 M .l:tg8 12 hxgS
hxgS. Here 13 a4 is now weaker in view of
13 ... ttJc6! (with the idea of ... 0-0-0 and
... .l:th8) 14 ttJb3 ttJe4 - the g3-bishop has
been prematurely deprived of support. In
the game Bareev-Dolmatov (S4th USSR 'This contradicts the idea of 10 ... 'iVe7.
Championship, Minsk 1987) White success- Karpov played 11...aS very quickly, having
fully sacrificed a pawn - 13 .te2 .txg2?! 14 probably decided that the inclusion of 11 a4
.l:th6! (threatening .tf3) 14 ... g4 IS .tM ttJbd7 as does not greatly change the situation on
16 'iYa4 Wf8 17 O-O-O! Wg7 18 .l:thS etc. How- the board. However, this is not so. 11 ... ttJc6!
ever, 13 ... ttJc6! is sounder: 14 'iYa4 0-0-0 was more logical.' (Makarychev). Appar-
(Savchenko-Lerner, Simferopol 1988) or 14 ently Karpov was concerned about 12 ttJb3
'iYc2 0-0-0 IS 0-0-0 .l:th8 with equality, as with the threat of a4-aS, but after 12 ... hS! 13
Psakhis played against me (d. the note to f3 (13 M ttJe4 with equality) 13 ... h4 14 .tf2
White's 6th move). as and ... 0-0-0 Black has a good game
Apart from 13 a4 and 13 .te2 White has (Dokhoian-Lerner, Lvov 1990).
tried a further nine (!) different continua- It would appear that during his urgent
174
The Third Match: 1986
night-time analysis my opponent simply Mdl. After 14 ... CLlc6 15 c5! (if 15 Mbl, 15 e4
did not have time to delve into certain or 15 ~e2 Black successfully plays
nuances of an unfamiliar position. 15 ... 0-0-0) 15 ... dxc5 16 dxc5 'iYxc5 17 ~xc7
12 h4 CLld5 18 CLle4 (Makarychev) 18 ... CLld4!
'Play over the entire front, emphasising (18 ... 'iYe7? 19 ~d6) 19 cxd4 'iVxc7 20 ~d2
the insecure position of the black king. and Mad the initiative is again with White.
White's desire to open the position is However, 14 ... ~c6!? would have re-
associated with the presence of his two tained the possibility of the normal devel-
bishops: (Gurevich) opment of the knight at d7. After this the
12 ... MgS immediate 15 c5 is not very dangerous for
Over a period of 22 years no one has Black in view of 15 ...bxc5 (15 ... dxc5?! 16
tried 12 ... g4 (avoiding the opening of the h- ~b5!) 16 dxc5 d5 (Makarychev), but 15 Mbl
file) 13 h5 CLlbd7 (if 14 ~h4 Gurevich rec- CLlbd7 16 c5 bxc5 17 ~b5! is more interest-
ommended 14 ... 'iVf8). ing, with sharp play (Rashkovsky-
13 hxgs hxgs Dolmatov, 54th USSR Championship,
Minsk 1987).
Of course, at a6 the knight is not very
well placed. However, Karpov is hoping to
castle and then seize the h-file.
15 Mbl!
'This rook will exert considerable (al-
though indirect) pressure on the black king,
bishop and knight at a6, and will make the
c4-c5! breakthrough possible. Kasparov's
play in a very complicated position just
after the opening creates a strong harmoni-
ous impression - a combination of a deep
strategic plan with controlled concrete
14 'iVb3! decisions. However, the world champion
A very important moment - White tries soon began to run seriously short of time:
to prevent the opponent from completing (Makarychev)
his development. 'The main attacking
resource is the c4-c5 breakthrough, and
operations on the h-file are kept in reserve.'
(Taimanov). The modest 14 Mh2 CLlbd7 15
~e2 (Kozlov-Tukmakov, Novosibirsk 1986)
would have allowed 15 ... 0-0-0 and ... MhS.
14 ... ttJa6
'A courageous decision, signifying an
almost complete rejection of active coun-
terplay for the sake of erecting a powerful
defensive wall.' (Makarychev). 14 .. .ct:Jbd7?
does not work because of 15 c5! dxc5 16
~xc7, and 14 ... CLle4?! is also insufficient on
account of 15 CLlxe4 ~xe4 16 c5! dxc5 17 ls ...'it>fS!
175
Kasparov vs. Karpov: 1986-1987
176
The Third Match: 1986
to h8 ... 'In the event of 17 ... ctJb8?! 18 jLd3 Now 19 dxe5?! dxe5 is weak, while if 19
ctJbd7 19 e4 Black's strategic problems l::i.b5!? l:!.e8!. After 19 f3 again 19 .. J:te8 is
would merely have been aggravated: good, with the threats of 20 ... exd4 and
19 ... Wg7? is still dangerous on account of 20 20 ... jLxa4. Therefore White must try and
e5 dxe5 21 dxe5, while if 19 ... e5, then 20 f3! break through:
followed by ctJfl-e3-f5.' (Makarychev) 1) 19 c5 exd4 20 cxd4 (20 exd4 bxc5)
But perhaps Black should have decided 20 ... ctJb4 21 l::i.xb4 (there is nothing better)
on the seemingly risky 17 ... 'iVd7!? After this 21...axb4 22 cxd6 cxd6 23 ctJc4 'iVe6 24 ctJxb6
in the book Dva matcha I suggested 18 jLd3 l::i.d8, and White has no more than sufficient
We7 (18 ... jLxa4? 19 'iVf3) 19 'iVe2 'with the compensation for the exchange;
threat of c4-c5'. However, after 19 .. J:th8 2) 19 Wfl Wg7 20 c5!? exd4 21 cxd6 dxc3
(but not 19 ... ~xa4? 20 f4! gxf4 21 jLh4) the 22 jLe5 l::i.h8! 23 l::i.h5 l:!.ag8!, and after both
line 20 c5?! l::i.xh2 21 jLxh2 l::i.h8 is unclear, 24 .i.xa6 cxd2 25 dxc7 jLxa4 and 24 jLxc3
and it is better to be patient with 20 Wfl ctJc5 25 .i.xf6+ Wxf6 26 jLh7 'iVxd6 White's
l::i.xh2 (20 ... ~xa4?! 21 c5!) 21 ~xh2 l::i.d8 22 attack is sufficient only for a draw.
llb2! (avoiding a trap: 22 c5? ctJxc5! 23 dxc5 What then should be done? After
dxc5) 22 ... ctJb8 23 e4, and the threat of c4-c5 17 ... 'iVd7 Gurevich recommended an inter-
is now real. esting exchange sacrifice - 18 l::i.b5!? We7 19
The immediate 19 c5!? (instead of 19 .i.d3. I agree that after the possible 19 ... 'iVd8
'iVe2) 19 ... bxc5 20 'iVe2 cxd4 21 exd4 leads to 20 e4 jLxb5 21 cxb5 ctJb8 22 'iVf3 ctJbd7 23
very complicated play. And, apparently, ctJc4 White has an excellent position. But
the clearest way to gain an advantage is 19 here too 18 ... e5! is stronger, exploiting the
Wfl!? l::i.h8 (19 ... jLxa4?! 20 'iVe2l::i.ad8 21 ctJf3! fact that after 19 dxe5 dxe5 the e5-pawn is
with an attack) 20 Wg1! l::i.xh2 21 jLxh2 jLxa4 taboo because of 20 ... ctJg4!, while if 20 ..te2
(21...l::i.h8?! 22 c5!) 22 'iVf3 l::i.d8 23 e4 or both 20 ... e4 and 20 ... ctJc5 are possible.
21...l::i.d8 22 ~c2, combining the threats of At any event, 17... 'iVd7 followed by ... e6-
e3-e4 and c4-c5. e5 would have given Black good counter-
But the whole point is that instead of the chances.
human 18 ... We7 Black has the unexpected
computer resource 18 ... e5!.
18 cs!
'Again, for the umpteenth time, we see
Analysis diagram Kasparov's favourite idea - a pawn sacri-
fice for the initiative.' (Gurevich). This
177
Kasparov vs. Karpov: 1986-1987
sacrifice is very timely: Black, who was all 22 dxc5 e5!) 22 ... cxd4 23 cxd4 e5! 24 CLlf3
ready to neutralise the consequences of his CLlbd7 25 Wf1!? with the idea of 25 ... ..Itxa4?
dangerously-played opening, now faces 26 ~a2! ..Itc6 27 ~c2 or 25 ... exd4?! 26 CLlxd4
new problems. ..Itxe4 27 f3! ..Itxd3 2S ~xd3, and the opening
18 ... bxcS of the game is to White's advantage
It is easier for White to attack in the (2S ... CLlc5 29 CLlf5+ WfS 30 CLlxe7 CLlxd3 31
event of lS ... dxc5?! 19 ..Ite5! (threatening CLlxgS WxgS 32 .l:tb5 CLlc5 33 ..Ite5 etc.), but
'iVh5) 19 ... WfS (19 ... .l:thS? 20 ~g4) 20 ..Itb5, after the cautious 25 ... .l:tabS there is still all
for example: 20 .. .':t:JbS? 21 dxc5 bxc5 to play for;
(21.. ...Itxb5 22 .l:txb5!) 22 .l:th6 CLleS (22 ... CLld5? 2) 19 ... ..Itxb5!? (instead of '?') 20 axb5 (20
23 c4 ..Itxb5 24 cxd5) 23 ~5 f6 24 .l:th7! (24 .l:txb5 .l:thS!) 20 ... CLlbS 21 dxc5 CLlbd7! 22 cxd6
..Itxc6 CLlxc6 25 ~f3 is also not bad) 24 ... CLlg7 cxd6 23 CLlc4 d5 or 23 b6 e5, repelling
(24 ... .l:tg7 25 ~6) 25 ~3! and wins, or the White's attack.
more tenacious 20 ... ..Itb7 21 .l:th6 .l:tg6 20 dxcS
(21...CLld5? 22 e4 CLlxc3 23 'iVh5 CLlxb1 24 Now Black really does have a difficult
..Itg7+! with mate) 22 .l:txg6 fxg6 23 ~e2 etc. position and it is hard to offer him any
19 ..Itbs (the Ruy Lopez!) good advice.
178
The Third Match: 1986
'iYhs f6 24 l:!.h7! (as in the note to Black's f4) 28 ~a3+ 4Jd6 29 ~xd6+ cxd6 30 l:!.h7 g3
18th move). 31 f4 ~g4 32 ~d3 ~e7 33 ~d4 ~g6 34 l:!.hl
The move made by Karpov shows con- and wins.
cern for his pawn chain, but at the same From the practical point of view I
time it so weakens the dark squares should perhaps have played this, since the
(around his king!) that it induces White to variations after 23 'iYhs are more compli-
look for a winning continuation. In fact, cated, and I already had little time left...
there is no direct win. Was this a lure on 23 ...f624l:!.h7
Karpov's part? If so, it hit the target: here I
spent much time seeking a clear way to
win, bu t, alas, I failed to find one - in the
position it turned out that there were too
many complicated tactical nuances, too
many tempting possibilities ...
21 ~e5 'it'f8 22l:!.h6 4Je8
The manoeuvre 22 ... 4Jfd7? 23 ~xc7
~xcS (23 ... 4JxcS 24 ~xb8) has a tactical
refutation - 24 4Je4!.
24 ...4Jg7
The only move for the second successive
time. For some reason I thought that
24 ... l:!.g7 forced a draw, but after 2S 'iVh6
there is no defence: 2S ... .\txbS 26 l:!.h8+! Wf7
27 'iYhS+ l:!.g6 28 l:!.h6 .\td3 29 e4, or 2S ... fxeS
26 l:!.h8+ Wf7 27 4Jf3 ~f6 28 'iVhs+ l:!.g6 29
l:!.h6 Wg7 (29 ... e4 30 l:!.xg6 ~xc3+ 31 4Jd2
4Jg7 32 l:!.xg7+ Wxg7 33 'iYxgS+ Wf7 34 'iYf4+
and 'iYxc7+) 30 l:!.h7+ Wf8 31 4JxeS! etc.
23 ~h5 25 'iYf3!
Such a move (and with a trap - Unfortunately, here I spent 9 out of my
23 ... ~xbS? 24 l:!.h7! and 'iYh6+) is hard to IS minutes remaining to the time control on
avoid making, and I quite unjustly criti- calculating a bishop sacrifice - 2S 'iYh6? fxeS
cised it in Dva matcha. It is most probably 26 4Jf3. How I could have done with those
the strongest, although after 23 c4!? ~xbS minutes later' on! But after 26 ... ~xbS the
24 cxbS 4Jd7 2S ~b2 White, having main- sacrifice would merely have led to great
tained his dark-square bishop on the long complications and a draw:
diagonal, would also have had an over- 1) 27 axbS (27 4JxeS? ~e8 and wins)
whelming position. The difference in the 27 ... 4Jd7 28 c6 g4! 29 4JgS 4Jf6 30 b6! 4Jxh7
strength of the pieces is such that the stra- 31 4Jxh7+ We8 32 bxc7 l:!.c8! or 29 4Jh4 ~f6!
tegic battle could be considered decided, 30 cxd7 ~xh6 31l:!.xh6 ~e7;
for example: 2S ... 4JxcS 26 ~c2 4Jd7 27 l:!.al 2) 27 l:!.xbS 4Jc6 28 4JxgS We8 29 c4! ~d7
or 2S ... g4 26 c6 4Jb6 27 ~c2! ~gS (27 ... g3 28 30 cxdS exdS 31 ~h3+ We8 32 'iVh6 or
179
Kasparov vs. Karpov: 1986-1987
immediately 28 c4! (winning the e4-square 28 ... ct:Jd7 29 i.xc6 ct:Jxe5 30 "iVh5+ 'i¥tf8 31
for the knight) 28 ... dxc4 29 ct:Jxg5 'i¥te8 30 i.xa8 ct:Jxh5 32 l:tb8+, while if 28 ... i.xb5 or
ct:Je4 'i¥tf8. 28 ... g4 29 "iYf4 i.xb5 any of the three cap-
White's total domination would have tures is decisive.
enabled him to hold on even a piece down, 28l:th6 reduces the tempo of White's of-
but what was the point of spending so long fensive. Alas, in this game the highly
studying such continuations in a winning promising career of my king's rook in the
position?! end turned out badly ...
2S ... 'i¥tf7 26 "iYhS+ 'i¥tfB 27 "iYf3! 'i¥tf7 2B .•. ct:JeB
28 ... ct:Jd7? will not do on account of 29
i.xf6! ct:Jxf6 30 i.xc6 ct:Jf5 31 l:txf6+ 'iYxf6 32
i.xa8 .l:.xa8 33 l:tb7 l:tc8 34 e4 and wins.
180
The Third Match: 1986
30 cxd5!? i.d3(?) 31 d6!, but 30 ... exd5 31 White's chances are slightly better, but the
l:th7+ ttJg7 is better, with great complica- position is now close to dynamic equilib-
tions: 32 i.xc7! 'it>g6 33 i.d6 'iVe6 34 l':!.xg7+ rium: Black has managed to develop his
l:txg735 axb5l':!.h7 36 ttJf1 a4 37 'iVd1, never- pieces ...
theless retaining the initiative.
The cool-headed 29 i.g3 also came into
consideration. In any case this is not at all
what White was dreaming of just one move
ago, but after 29 e4 the situation becomes
altogether more complicated.
29 ... g4 30 'i!Vf4 i.xb5 31l:txb5
After the hasty 31 i.xf6? ttJxf6 32 e5
Black had prepared 32 ... i.d3! (covering the
h7-square) 33l:txf6+ 'it>e8.
31 ... ttJd7
31...ttJc6? would have allowed a destruc-
tive piece sacrifice: 32 l:th7+ l:tg7 33 l:txg7+
ttJxg7 34 i.xc7 e5 35 'iVxg4 'iVxc7 36 exd5 33 ... ttJxe4?
and wins. An incorrect, essentially losing move.
33 ... ttJxc7 34 l:txc5 was advantageous to
White, although in that case the tension
would have been retained.
But the fearless 33 ... ttJxa4! 34 i.xa5 'iVd7
would have led to wild complications - 35
exd5! (35 'i¥d3 l1c8 and ... ttJc5) 35 ... 'iVxb5
(35 ... exd5!? 36 'i¥d3 ttJb2) 36 'iVxe6+ ~g7 37
l:th4! <;!>f8 38 l:th7! ttJg7, holding out against
the attack:
1) 39 i.b4+ ttJc5 40 i.xc5+ 'iVxc5 41 'iVxf6+
We8 42 .uxg7 l':!.xg7 43 'iVxg7 ~xd5 44 'iYxg4
l:ta1+ 45 'it>e2l':i.a2;
2) 39 'iVxf6+ 'it>e8 40 d6 'iY'd7 41 ttJc4! .uxa5!
32 i.XC7!? 42 ttJxa5 ttJc5! 43 'iVg6+ 'iVf7 44 'iYxg4 'iYd5 45
Desperately fighting for the initiative. 32 'iVg6+ 'it>d8 46 'iVf6+ 'it>c8 47 'iVd4 'iVxd4 48
i.xf6 also looked tempting, but after cxd4 ttJce6, in each case with equality.
32 ... ttJexf6 33 c6! (not 33 e5 ttJxe5 34 'Yixe5 34 ttJxe4 dxe4 35 i.xa5 f5 36 i.b4
l:th8! with counterplay) 33 ... ttJf8! 34 e5 ttJg6 Or 36 ~d4! immediately. Material is still
35 l:th7+ l:tg7 36 l:txg7+ 'it>xg7 37 exf6+ 'it>f7 equal, but White's pieces (including his
or 32 ... ttJdxf6! 33 e5 l:tg6 34 l:th8 l:tg8 35 bishop!) are taking part in the attack, and it
l:txg8 'it>xg8 36 exf6 ttJxf6 Black would have cannot be parried. But... 'The group of
retained possibilities of resisting. pawns in the centre and the opponent's
32 ... ttJXC5 (if 32 ... e5? the simplest is 33l:th7+ severe time-trouble give Black definite
l:tg7 34 'iVh6! ttJxc5 35 'iVh5+ 'it>f8 36 .l:Ih8+ counter-chances.' (Gurevich)
l:tg8 37 i.xa5 ttJd3+ 38 'it>f1) 33 'iY'e3! 36 ...~d7 37 ~d4! (centralisation - and a
Both forced, and strong! In principle trap: 37 .. :~xd4? 38l1b7+!) 37 ...l:ta7
181
Kasparov vs. Karpov: 1986-1987
The only defence against a rapid rout: position, and no time-trouble can justify
37 .. J~d8 38 .l:i.b7! ~xb7 39 ~xd8 g3 40 'iVh4! them.
gxf2+ 41 ~xf2 ~a7+ (41.. ..l:i.g6 42 .l:i.h8) 42 38 ...'Dg7 (I overlooked this defence) 39 as??
~f1 'Dg7 43 .l:i.f6+! ~e8 44 ~7. During the last few seconds before the
In this absolutely won position White fall of my flag I completely lost my head. 39
had three moves to make before the time .i.c5! (Gurevich) would still have won after
control - any three normal moves, and that 39 ... ~xd4 (there is nothing better) 40 .i.xd4
would have been the end both of the game, .l:i.d7 41 .l:i.h6 and a4-a5 or 40 ... ~g6 41 .i.xa7
and of any struggle in the match! The most ~xh7 42 .l:i.b7(b8). Instead of this White
practical was 38 .i.c5 ~xd4 39 .i.xd4 ':d7 even manages to lose!
(39 ... .l:i.c7 40 a5) 40 .l:i.b8 (precisely three
moves) or 38 ... .l:i.b7 39 ~h7+ l:tg7 40 .l:i.xg7+
'Dxg7 41 ~xd7+ .l:i.xd7 42 a5 with an elemen-
tary win. The most surprising - 38 .i.d6!?
.l:i.g6 39 .l:i.h8 .l:i.g8 40 .l:i.b8!, and the most
'attacking' - 38 ~e5!? .l:i.g6 (38 ... .l:i.xa4? 39
.l:i.b7!) 39 .l:i.h7+ .l:i.g7 40 .l:i.xg7+ 'Dxg7 41 ~b8!
'De8 42 ~8! and wins.
I can't find any explanation for what
happened next... For 37 moves in this game
Black has not once 'raised his head'. And if
in general one can talk about 'gifts' as
applied to the result of a chess game, then -
your attention please! - you are about to 39 ... ~g6?!
witness the handing over of a substantial Karpov had about two minutes left and
gift! in the time scramble he hurried to exchange
the queens, although he could have gained
a decisive attack by 39 .. .'iVxb5! 40 'iWxa7+
~g6 41 .l:i.h4 l:td8! (Makarychev; 41...'it'd3?!
42 ~d4!) 42 ~e3 'Dh5 etc. After 39 ... ~g6 the
win for Black is by no means so obvious.
40~xd7 .l:i.xd7
Here the game was adjourned. For some
ten minutes I stared at the board, trying to
understand what had happened in the time
scramble and what had become of White's
advantage ...
Initially I wrote on my scoresheet the
panicky 41 .l:i.xg7+?, in order to rid myself of
38.l:i.h7+? the rook which was now out of play (the
Even now, nearly a quarter of a century conversion of the exchange advantage
later, I cannot look at this and my next would not have presented a difficult prob-
move without a shudder - a 'combination' lem for Black). But then I managed to take
of the two weakest moves in my career! myself in hand and I realised that I could
They disrupt the entire harmony of White's still fight for a draw: in the end, material
182
The Third Match: 1986
was still equal, and White also had his is more tenacious, but here too after
trumps - the passed pawns on the queen- 46 ... Md2+ 47 'it'f1 Ma2! things are difficult for
side. White: 48 Me2 Mxe2 49 'it'xe2 Me8+ 50 'it'f2 f4
41Mh4 51 gxf4 'it'h4 52 f5 g3+ 53 'it'f3 'it'h3 54 ii.c5
The sealed move. 'From h4 the rook is Me5 or 48 Mf6 'it'g5 49 Me6 Mh8 50 'it'gl Mhh2
able to prevent the activation of Black's 51 ii.d6 f4! 52 ii.xf4+ 'it'f5 53 Me5+ 'it'f6 54
pawn phalanx.' (Makarychev) Mel Mhc2 and wins. Even if the defence can
Our team analysis went on until six in somehow be improved, this is unlikely to
the morning. It was established that, of the change the evaluation of the position.
main possibilities available to Black, two 42 (4
were the most unpleasant - 41...iiJh5 and Preventing the invasion of both black
41...Mgd8. rooks. Many of the experts thought that 42
On the resumption, miracles with ii.c5 was better, but then White is not able
'parallel' analysis again began to occur - to set up anything even resembling a
identical key ideas and even oversights by fortress: 42 ... Md 1+ 43 'it'e2 M8d2+ 44 'it'e3
the two players. Ct:Jh5 45 Mxh5 'it'xh5 46 'it'f4 g3! 47 'it'xg3 e5
48 i.e3 Md5 49 Mxd5 Mxd5 50 ii.b6 Md3+.
42 ... Mdl+ 43 'it'e2
41 ... Mgd8?!
The first amazing coincidence: we came
to the conclusion that 41...iiJh5(!) was The key position of this variation: which
somewhat weaker - and Karpov did too! of the white pawns should Black first take
Against it we had prepared 42 g3 (forced) under control - 'a' or 'c'?
42 ... e3! 43 Mb6! exf2+ 44 'it'xf2 'it'g5 45 ill..d6?! The a-pawn appears to be more danger-
(not 45 Mxe6? Ct:Jxg3) 45 ... Mgd8 46 Mxh5+! ous (it is closer to the queening square),
(the only chance) 46 ... 'it'xh5 47 ill..f4, reckon- and this is indeed so. Therefore in our
ing that the threats of Mxe6 and a5-a6 night-time analysis we focused on 43 ... Ma1!.
would enable White to save the game. Extreme accuracy is required of White: 44
However, after 47 ... Me8! 48 a6 Ma7 49 c4 ii.c3 (everything else is bad: 44 ii.c5? Ma2+
Mea8! 50 Mxe6 Mxa6 51 Mxa6 Mxa6 52 c5 Me6 45 'it'e1 Ct:Jh5 46 g3 Md3 47 Mb6 f4!, or 44 g3?
53 ii.e3 'it'g6 54 'it'e2 'it' f6 55 'it'd3 'it'e7 56 ill..f4 Ma2+ 45 'it'e1 e3 46 fxe3 Md3 47 i.c5 Ct:Jh5
'it'd7 followed by ... Me4 and ... 'it'c6-d5 Black and wins) 44 ... Mc1! 45 ii.e5! (not 45 i.xg7?
would have won. because of 45 ... Mc2+ 46 'it'e3 'it'xg7 47 'it'f4
The immediate 45 Mxh5+ 'it'xh5 46 Mxe6 'it'g6 48 Mb7 Mxf2+ 49 'it'e5 Ma2 50 Ma7 e3 or
183
Kasparov vs. Karpov: 1986-1987
46 ..t>el ..t>xg7 47 a6 .udd2! 48 .ub7+ ..t>g6 49 e5 52 .te3! (with the saving idea of sacrific-
.uh8 .ue2+ 50 'it'dl .l::tcd2+ 51 'it'c1 .l::ta2 52 'it'bl ing the bishop) 52 ... f4 53 l::txg4+ Wf5 54 .l::tg8.
Iha6 and wins), pinning his hopes on the White is on the verge of disaster, but I have
bishop. not in fact found anything decisive:
1) 54 ... fxe3 55 g4+, and there is no win
after either 55 ... Wf6 56 fxe3 l::tal + 57 Wd2
tLJg3 58 l::tf8+ 'it'e6 59 l::tff7, or 55 ... ..t>e6 56
fxe3 tLJf6 (or 56 ... tLJg3 57 .l::te8+ Wd6 58 a8"iV
.uxa8 59 .uxa8 l::txa8 60 Wf2, and the knight
is trapped) 57 l::tgg7 l::ta6 58 a8'iV l::txa8 59
l::tb6+ Wd5 60 l::txf6 l::th8 61 l::tgf7, gaining a
draw;
2) 54 ... l::tal + 55 Wd2 fxe3+ 56 fxe3 tLJf6 57
g4+ tLJxg4 58 l::tf7+ We6 59 ltff8 J:rla2+ 60
Wc3 .uxa7 61 l::txg4 lt7a3+ 62 Wb4 l::txe3 63
Mg6+ We7 64l::tf5, again with a draw.
Thus, after 43 ... l::tal! White would have
Analysis diagram saved himself literally by a miracle. Our
hasty analysis suggested that 43 ... .l::tc1
After 45 ... l::tc2+ 46 'it'el l::tdd2 White is would have given White a rather simple
saved by 47 l::th6+! 'it'g5 48 l::th7!, for exam- draw, but... a 'hole' crept into our calcula-
ple: 48 ... l::te2+ 49 Wdl tLJe8 50 .uh8, or tions.
48 ... tLJh5 49 ~b8 ltxf2 50 g3. And after Karpov was about 10 minutes late for
46 ... .l::txc4 at the time we considered 47 the start of the adjournment session. Ac-
l::tb6(?) to be sufficient for a draw, on the cording to his second, 5ergey Makarychev,
basis of the variation 47 ... e3 48 fxe3 l::te4 49 'which of the two continuations (43 ... l::tal or
l::td6. However, 47 ... l::tc1 + 48 We2 l::tc2+ 49 43 ... l::tc1) to choose was decided by the ex-
WeI l::ta2! is stronger, and therefore 47 a6! world champion only five minutes before
should be played (see below). the start of play: analysis showed that Black
As it seemed to us, a draw would also did not have a win.' 50 what should be
have resulted from 45 ... l::txc4 46 l::tb6 l::tc2+ preferred - 43 ... l::tal or the obviously
(after 46 ... l::td5 47 .tf4 tLJh5 48l::txh5 Wxh5 49 weaker move?
l::txe6 White squeezes out a draw, even
though the exchange down: 49 ... g3 50 f3
exf3+ 51 Wxf3 l::td3+ 52 l::te3 l::txe3+ 53 .txe3
l::ta4 54 Wxg3 l::txa5 55 Wf4 Wg6 56 g3 l::ta4+
57 Wf3 ..t>f6 58 .tb6) 47 WeI, but we over-
looked the powerful resource 47 ... .l::ta2! 48
.txg7 (48 .tf4? tLJh5 and wins) 48 ... g3!!
(creating a passed pawn) 49 fxg3 Wxg7 50
l::txe6 l::tc8 51 'it'dl l::tc3!, and there is no
defence.
What, then, should be done? Probably
the only chance is 46 a6! l::tc2+ 47 'it'el l::ta2
48 l:tb7 tLJh5 49 a7 l:td5 50 .tb2l::tda5 51 .td4
184
The Third Match: 1986
185
Kasparov vs. Karpov: 1986-1987
l:i.c1+ 50 l:i.d1 is unclear) 49 ~f1! l:i.c3 50 ~gl 55 ... l:i.cc1? 56 l:i.b1!) 55 ... ct:Jf5 56 l:i.b1 l:i.aa2
e3 51 l:i.b1, and the battle still continues: there is no way of saving the game.
51...l:i.xa6 52 l:i.e1! or 51...ct:Jh5 52 g4 etc. 55 ...~f5
True, after 49 ... l:i.gxg2! 50 a7 l:i.gc2 51 l:i.b1 55 ... ct:Jh5 was simpler, including the
l:i.xc4 52 ~e3 ct:Jh5 53 l:i.h2! l:i.a3! 54 l:i.g2+ ~f6 knight: 56 l:i.xe4 ct:Jg3+ 57 l:i.xg3 fxg3 58 l:i.g4+
55 ~b6 ct:Jf4 56 l:i.g8 l:tca4 it is doubtful ~f5 59 l:i.xg3 l:txc6, and White can resign.
whether White can save the game. 56 l:i.b5+ e5 (not 56 ... ~g4? 57 l:i.h4+ with
47 ...l:i.al+ 48 ~e2l:i.a2+ perpetual check) 57l:i.a5l:i.dl
'With the aim of keeping a large reserve The next time control had been reached,
of time for the last moves before the time and Black could have played more accu-
control, which had extended beyond our rately: 57 ... l:i.ac1! 58 c7 e3. Now, however,
home analysis.' (Makarychev). One senses White gains a chance.
that Karpov too had not analysed 43 ... l:i.c1
very thoroughly.
49~el g3
Unfortunately for White, this is also
good enough, but 49 ... ~g5! would have
won much more simply, including all the
forces in the attack: 50 c6 (50 l:i.h7 ct:Jh5)
50 ... ~xh4 51 c7 l:i.a1+ 52 ~e2 f4 with un-
avoidable mate, or 50 g3l:i.a1 + 51 ~e2 f4.
50 fxg3l:i.xg3
58 a7?
The chance consisted of 58 c7!. In the
event of 58 ... l:i.xc7 59 a7 l:i.cc1 60 a8~ l:i.xe1+
61 ~f2 l:i.f1 + 62 ~e2 f3+ 63 l:i.xf3+ exf3+ 64
~e3 l:i.ce 1+ 65 ~d3 ct:Je6 Black would still
have had to work hard to convert his
advantage.
After the immediate 58 ... l:i.cc1 there fol-
lows 59 c8~+ l:i.xc8 60 a7 l:i.a8 61 l:i.b3 with
the threats of l:i.b8 and l:i.bb5. And if 58 ... e3!
51~fl there is the clever trick 59 l:i.h2!, when
, An interesting idea, which almost 59 .. ..l:hh2? 60 c8~+ ct:Je6 is dangerous
proved successful. If 51 l:i.h2, then ... f5-f4-f3 because of 61 l:i.xe5+! (but not 61 ~c3 l:i.f2+!
would have been decisive.' (Makarychev) 62 ~gl ct:Jd4 63 ~c8+ ~e4, and the black
51 ...l:i.gxg2 52 ~el (at least someone has to king hides from the checks) 61... ~xe5 62
try and help the king!) 52 ... l:i.gc2 ~c3+ ~f5 63 a7 l:i.f2+ 64 ~gl l:i.e2 65 a8'iY
Black combines attack with prophylaxis. l:tdxe1+ 66 'iYxe1 l:i.xe1+ 67 ~g2 - again a
53 c6 l:tal 54 l:i.h3 f4 (of course, 54 .... ct:Jh5 position is reached with rook, knight and
would also have won) 55l:i.b4 two pawns for a queen, where White can
Also after 55 l:i.hb3 (in the hope of resist.
186
The Third Match: 1986
Even so, with subtle play - 59 .. J:kc1! 60 myself was bothered by all this. I asked
c8'iV+ Mxc8 61 a7 f3! (with the threat of myself: 'Have I really deserved this?! Is this
... Mxel +) or 60 Me2iDe6! 61 a7iDxc7 62 a8"iY all a punishment for insufficient effort in
iDxa8 63 Mxa8 ~e4 64 Ma4+ Md4 65 Ma3 Md3 the past?'
66 Ma4+ ~f5 67 Mh2 Md2 68 Me2 Mcc2 - Playing after two successive defeats is
Black would have been able to place his very difficult, in fact it's pretty terrible -
opponent in a hopeless position. you imagine cracks appearing in the most
S8 ... e3! 0-1 reliable set-ups. It was true that I had never
Mate is unavoidable, aM. White re- lost three games in a row, but there is a first
'...
signed. Times: 3.39-3.42. time for everything... It was decided to
continue the debate in the Griinfeld De-
All I could do was bitterly re'gret that I fence, although it would have been more
had ruined a masterpiece (there also sensible, of course, to change openings. But
flashed through my memory the tragic 6th I did this only in the 21st game, whereas -
game of the first match, where I also unex- there is every reason to assume! - the
pectedly returned my queen to dl, then employment of a new opening in the 19th
failed to find the murderous 25 'iVh5! and would not have affected the opponent's
also contrived to lose). The situation in the state of preparedness ...
match again changed - my lead was re- Zaitsev: 'There was a good mood in the
duced to the minimum and the psychologi- Karpov training team as we prepared for this
cal initiative had obviously passed to my game. Indeed, we were heartened by the results
opponent. of the two preceding games: in the 17th we were
Nikitin: 'The team meeting late that eve- able to employ a strong opening novelty, and in
ning after such a disastrous adjournment the 18th fortune smiled on us. For perhaps the
session was an uneasy one. When we began first time in the match after the difficult start
reflecting on possible channels for the leakage of (and it could have been even worse, but for the
information - and virtually no one any longer miraculous escape in the 2nd game) and the
had doubts about this - one of the trainers depressing results of the 7th and 8th games, we
behaved strangely: he parried questions in an felt that there was finally a ray of hope.'
inappropriate and, above all, very nervy way.
But there was no direct evidence against him,
and all our logical conclusions could have Game 19
contained a mistake. Finally it was decided to A.Karpov-G.Kasparov
increase our vigilance and to restrict ourselves World Championship Match,
to the motto: "Lads, let's live together harmoni- 19th Game, Leningrad 24.09.1986
ously! '" Grunfeld Defence 097
I took a time-out, in order to corne to af-
ter such a hail of blows. The commentators
were also bewildered, and one of them 1 d4iDf6 2 C4 g6 3iDc3 ds
wrote: 'The opinion is that after the brilliant 'So, the ninth occurrence of the Griinfeld
chess pyrotechnics in the 16th game, the young in the match. Just in case we looked at one
champion has become the victim of a double of the variations of the Queen's Gambit, but
whammy - over-confidence plus the mistaken we were secretly hoping that the opponent
certainty that the outcome of the match is would continue the opening dispute in the
already decided. This may cost him dearly.' I Griinfeld Defence.' (Zaitsev)
187
Kasparov vs. Karpov: 1986-1987
4 ttJf3 i.g7 5 "iVb3 dxc4 6 "iVxC4 0-0 7 e4 ttJa6 ttJd7 22 e5 l:i.b6 with reasonable compensa-
Instead of Smyslov's 7... i.g4 (Game tion for the pawn (Piket-Kasparov, Wijk
Nos. IS, 17), of which, naturally, there aan Zee 2000).
remained negative impressions, after three As will be seen, in fighting for an open-
days of intensive work with the entire ing advantage White must evacuate his
training group I chose for the first time one king from the centre as quickly as possible.
of the 'modern' ways of fighting against the S i.e2 c5 9 d5
centre. But here too an opening surprise Otherwise White cannot hope for any-
awaited me. thing, as was shown by the very first
games: 9 dxc5 i.e6! 10 ~5 l:i.c8! (Kmoch-
Prins, Amsterdam 1940), or 9 0-0 cxd4 10
l:i.dl e5! 11 ttJxe5 ttJd7 12 ttJxd7 i.xd7 13
ttJd5? l:i.c8 14 "iVb3 ttJc5 15 "iVa3 l:i.e8 16 f3 f5!
17 "iVxa7 fxe4 18 fxe4 ttJxe4 19 i.f3 i.c6 20
ttJb4 ttJf2!! and wins (Kramer-Najdorf, New
York 1948).
9 ... e610 0-0
I would have been happy with the sharp
10 i.g5 exd5 11 ttJxd5 i.e6 12 0-0-0 i.xd5 13
l:i.xd5 ~6 14 i.xf6 'iVxf6 15 e5 'iVf5 16 i.d3
"iVc8 (16 ... "iVe6!?) 17 l:i.dl?! b5! 18 "iVh4 ttJb4,
and Black seized the initiative (Beliavsky-
Black's idea is simple - to strike at the Kasparov, Belfort 1988).
centre with 8 ... c5 and initiate piece play, 10... exd5 11 exd5 i.f5
exploiting the advanced position of the
white queen, for example:
1) 8 i.g5 c5 9 d5 h6 10 i.h4 b5! 11 ttJxb5
"iVa5+ 12 ttJd2l:i.b8! 13l:i.dl? ttJb4 14 a3l:i.xb5!
with crushing threats (Panchenko-Sideif-
Zade, Tashkent 1980);
2) 8 "iVb3 c5 9 d5 e6 10 i.xa6 bxa6 11 0-0
exd5 12 exd5 ~6 (12 ... i.f5 or 12 ... l:i.e8 is
also played) 13 l:i.dl i.b7 14 i.f4 l:i.fe8
(Kozul-Dorfman, Sarajevo 1988) or
14 ... l:i.fd8! (Kozul-Tukmakov, Solin 1999),
and the bishops compensate for the defects
in the pawn chain;
3) 8 i.f4 c5 9 dxc5 i.e6 10 "iVb5 i.d7! 11 Of course, this variation also has its
'iVxb7 ttJxc5 12 ~4 ttJe6 13 i.e5 a5 14 'ii'a3 drawbacks: there is a knight on the edge of
'iib6 15 i.c4l:i.fc8 16 i.xf6 i.xf6 17 ttJd5 "iVb8 the board, and from being a target the
with equality (Topalov-Kasparov, Sarajevo isolated d5-pawn may be transformed into
1999) or 9 d5 e6 10 d6!? e5!? 11 i.xe5 ttJb4 12 a dangerous passed pawn. These two
l:!.dl i.e6!? 13 'iVxc5 ttJc2+ 14 'itd2 ttJd7 15 factors concerned me, and it was with
"iVc7 ttJxe5 16 "iVxd8 l:i.axd8 17 'itxc2 ttJg4!? difficulty that I forced myself to play
18 l:i.d2 i.h6 19 l:i.e2 l:i.xd6 20 h3 ttJf6 21 g4 7 ... ttJa6. This position of the knight always
188
The Third Match: 1986
provoked a certain superstitious fear in me, backs of 12 .if4, is more logical than
and an instinctive desire to get rid of this 13 ... ttJd7 or 13 ... ~6 (after which 14 ~5! is
'bad' piece as soon as possible - hence the now good, Dlugy-Gavrikov, Tunis Inter-
psychological mistake which I made on the zonal 1985).
15th move. In addition, during the few
days of analysis we were unable to delve
into all the nuances of the position.
Although I imagined tricks everywhere,
my trainers nevertheless convinced me that
here there were sufficient counter-chances,
and that the complicated, dynamic play
was fully in keeping with my style. And
today, with the benefit of the accumulated
experience (7 ... ttJa6 was in my arsenal to
the end of my career), I can state that for
Black this is not the most dangerous varia-
tion of the Grunfeld Defence.
12.if4 Black appears to stand well: active
At the time this confidently-made move pieces, open files and diagonals, and the
became the main line, but already in Seville threat of ... ttJxc3 followed by ... l:te4. After
it was supplanted by 12 l:tdl (Game Nos.42, the standard 14 .id3 I was not planning the
48), and then also 12 .ie3. Thus life itself retreat 14 ... ttJd6, which had occurred in
made an assessment of Karpov's novelty in 1986, but I wanted with the help of attrac-
this game ... tive tactics to demonstrate how insecurely
Incidentally, against 12 .ie3 we had the white pieces are placed: 14 ... .ixc3! 15
prepared an interesting temporary pawn bxc3 b5 16 'li'xb5 ttJxc3, when the exchange
sacrifice, suggested by Magerramov: sacrifice 17 'li'c4 ttJxdl 18 l:txdl .ixd3 19
12 ... ~6! 13 b3 (after 13 ttJh4 .id7 14 l:tabl l:txd3 (M.Gurevich-Kotronias, Reykjavik
l:tfe8 15 b4 l:txe3! 16 fxe3 cxb4 17 'li'f4 'li'c5 1988) is dubious because of 19 ... 'Ii'a5!, while
Black has an excellent game, Timman- 17 'li'xa6 .ixd3 leads to equality - 18 'li'xd3
Kasparov, Sarajevo 1999) 13 .. JUe8 14 l:tadl ttJe2+ 19 'it'hl ttJxf4 20 'li'c4 'li'd6
Mad8!? with the idea of 15 ttJa4 'li'd6 16 (M.Gurevich-Kasparov, 55th USSR Cham-
ttJxc5 ttJxc5 17 .ixc5 'li'd7. Or 15 h3 'li'a5 16 pionship, Moscow 1988) or 18 l:txd3! ttJe2+
l:tfe 1 ttJd7 17 ttJa4 ttJb4! with chances for 19 'it'hl ttJxf4 20 l:td2 'li'd7! etc.
both sides (Khalifman-Kasparov, Linares Because of these tactics 12 .if4 seemed
2000). to us to be a second-rate move (and the
12 ... l:te8 main attention was focused on 12 .ie3). But
Natural development and preparation my opponent again hit upon a weak spot...
for ... ttJe4. Grandmaster Gavrikov, an Zaitsev: '7 ... ttJa6 did not catch us un-
expert on this variation, used to prefer awares. Back at the start of the Leningrad
12 ... ~6, but against this 13 .ie5! proved half of the match in a joint analysis with
unpleasant (the source game: Ivanchuk- Mark Tseitlin (one of the leading experts on
Lputian, Irkutsk 1986). this variation for Black) I discovered a
13 ~ad1 ttJe4! novelty, which later, by the efforts of the
This advance, emphasising the draw- entire group, was brought into a state of
189
Kasparov vs. Karpov: 1986-1987
190
The Third Match: 1986
17 ... ~d7 18 .l::i.xe4 .l::i.xe4 19 'iUxe4 ~fS 20 'iUe3 ~xbS 19 'iVxbS, forcing Black to give up
~xd3 runs into the pretty 21 ~eS! 'iUfS 22 something: 19 ... 'iUxe2 20 ~xe2 tDc3 21 'iUxb7
tDd6 'iVd7 23 ~xg7 ~xg7 24 tDxf7! with a tDxe2+ 22 ~hl tDxf4 23 'iVxa6 tDxdS with
dangerous attack. Apparently it is better to chances of a draw;
play 17 ... gS 18 d6! ~d7! (18 ... gxf4 19 ~xe4 4) IS ... ~d7!.
tDb4 20 tDc7 .l::i.xe4 21 .l::i.xe4 ~xe4 22 'iUxe4
.l::i.d8 23 d7!) 19 .l::i.xe4 gxf4 20 l:tdl l:tad8,
continuing the fight for equality;
2) IS ... ~g4. The pin suggests itself, and
in the event of 16 ~eS! .l::i.xeS 17 tDxeS ~xdl
18 ~xe4 'iUxeS 19 .l::i.xdl .l::i.e8 20 ~f3 (or 20
tDc3 'il'd6 21 ~f3 ~xc3 22 bxc3 tDc7)
20 ... 'iUel + 21 'iUfl 'iUxfl + 22 ~xfl ~xb2 Black
gains a draw without any particular prob-
lems, but 17 ~xe4!? .l::i.ee8! 18 .l::i.fel is better,
retaining a small plus (18 ... 'iVxb2 19 .l::i.bl
'iVf6 20 d6 etc.).
3) IS ... .l::i.ad8!? (also a beginner's move!)
16 .l::i.del (after 16 .l::i.fel 'iVxb2 the f2-pawn is Analysis diagram
hanging, but 16 a3!? is interesting with the
idea of 16 ... ~d7 17 ~c1 or 16 ... ~g4!? 17 The most accurate implementation of
~eS) 16 ... 'iUxb2. the 14 ... 'iVf6 idea, effectively putting an end
to the line with 12 i.f4. It was annoying
that I saw this move at the board, but in my
nervous and depressed state I was unable
to calculate it to the end. For the 1987 match
my trainers and I analysed it, but, alas, this
novelty saw the light of day only later and
in an unimportant game involving one of
my helpers (you can imagine my disap-
pointment): 16 ~eS ~xbS! 17 'iVxbS .l::i.xeS! 18
tDxeS (18 'iUxb7 is no better: 18 ... .l::i.ee8 19
'iUxa6 'iUxa6 20 ~xa6 ~xb2 21 d6 .l::i.ad8 22 d7
.l::i.e7 23 ~bS ~f6) 18 ... tDd6 19 tDg4 (or 19
tDd7?! 'iUd4 20 'iVaS c4) 19 ... 'iVf4 20 'iVd7 c4
Analysis diagram 21 g3 tDcS! 22 'iVc7 'iUxg4 23 'iVxcs 'iVd7! 24
~e2 ~xb2 with full compensation for the
Now 17 tDc7? is weak: 17 ... tDxc7 18 ~xc7 exchange (Ivanchuk-Dorfman, Lvov 1988).
tDd2! 19 .l::i.xe8+ .l::i.xe8 20 tDxd2 'iUxd2 21 ~xfS I felt that I had to initiate active piece
gxfS 22 g3 ~d4!... 0-1 (Dzhandzhava- play, but the method chosen (IS ... tDb4) was
Kasparov, Baku (simul) 1987). Instead 17 inappropriate. What told was that I was
.l::!.e2 'iVb4 18 a3 'iVxc4 19 ~xc4 ~d7 20 l:tbl playing the variation with 7 ... tDa6 for the
tDf6 is unclear, but 17 a3!? is tempting: first time and I underestimated Black's
17 ... ~d7 (the exchanges after 17 ... tDc3? 18 possibilities (since in similar positions I had
.l::i.xe8+ merely activate the d-pawn) 18 .l::i.e2 several times advanced my d-pawn to the
191
Kasparov vs. Karpov: 1986-1987
16 ctJc7! (21)
The first long think by the ex-champion.
'This move reflects Karpov's character -
sober assessment, deep and accurate calcula-
tion,' Gufeld enthused. 'In order to accept an
exchange sacrifice in such a sharp situation,
one must not only be a realist, but also have
strong nerves and a firm character.'
'There is no time for 16 i..bl because of
16 ... i..dT, I asserted in Dva matcha. But here
too after 17 i..e3 'iVa6 18 a4 White has some 19 l:tde1? (18)
advantage: 18 ... ctJf6 19 d6!? b6 20 i..g5 or 'Such a move would never have oc-
18 ... 'iVxa4 19 ctJc7 ctJa6 20 'iVxa4 i..xa4 21 curred to me: the rook was already stand-
ctJxe8 l:txe8 22 d6 i..xdl 23 l:txdl ctJf6 24 d7 ing behind the passed pawn', wrote Gufeld.
l:td8 25 ctJe5 etc. In event of 16 ... i..g4 17 i..e5 'Who would have rejected the obvious 19
l:txe5 18 i..xe4 l:tee8 (18 ... l:th5 19 a3!) 19 h3! d6! ctJc3 20 d7 ? What would have hap-
'iVf4 (19 ... i..d7 20 ctJc7) 20 hxg4 l:txe4 21 pened in this case? Again complications,
'iVxc5 the d-pawn also continues to cause again lengthy calculations. But Karpov
Black problems. simply restricts the opponent's counterplay,
16... ctJxd3 (16) 17 ctJxe8 l:txe8 18 'iVxd3 after which much becomes clear.'
Not 18 l:txd3? ctJd6. But now the black 'A technically perfect and simple deci-
knight has no favourable discovery in view sion in Karpov's style,' Zaitsev echoed him.
of 19 'iVb5. 'There is no need to demonstrate tactical
192
The Third Match: 1986
193
Kasparov vs. Karpov: 1986-1987
to give up a piece for the d-pawn. by either 23 h3 iDxf2 24 l:!.xf2 1!Vxf4 2S 'it'g2
Also, nothing real is given by 20 iDh4 'iVd6, or 23 ·~c7 l:!.f8! 24 ..teS i.xg4 2S ..txg7
..td4 21 ..te3 ..td7 22 ..txd4 1!VxdS 23 f3 iDd6 'it'xg7 26 1!VeS+ iDf6 27 l::ta1 1!Vb3 or 24 h3
24 l:!.xe8+ iDxe8. And if 20 g4?! there is ..tbS! 2S d6 h6 26 iDd2 iDf6 and ... i.x£1;
20 ... c4 21 1IVd1 (21 'iVd2 'iiVa4!) 21.. ...txg4 22 3) 21 ..tc7. According to Zaitsev, this is
d6 ..tf8!, again with very unclear play (say, 'unpleasant for Black', but here too after the
23 d7 l::td8 24 l::txe4 l::txd7 2S 'Yi'e1 ..txf3 26 accurate 21...a6! White has nothing:
l:!.e8 'iVa3, and Black is alright: 27 ..th6 l:!.d4! a) 22 'iiVxb7 1!VxdS 23 1!VxdS (23 1!Vxa6
or 27 'iVeSl:!.dS etc.). l:!.a8) 23 ... l:!.xdS, not fearing 24 iDh4 ..tf6 2S
But, of course, Karpov had in mind 20 iDxfS gxfS;
1!Vbs - it was because of this move that I b) 22 1!Vb6 l:!.xdS
rejected 19 .. :iWxa2, erroneously assuming
that my position would collapse. But in fact
Black has adequate counter-chances here in
view of the superior mobilisation of his
pieces (the rook on £1!). There is even some
point in trying to confuse matters -
20 ... l::tf8!? 21 1!Vxb7 c4, but 20 .... l:!.d8! is more
precise, for example:
Analysis diagram
194
The Third Match: 1986
195
Kasparov vs. Karpov: 1986-1987
Suppressing Black's last activity. In a 29 ... ctJf5 30 iL.xc5 ~xe1+ 31 Wxe1 b6 32 d6! -
good position a good move can always be Black 'battles on', merely in order to ad-
found, if it is properly sought. In such cases journ the game ...
Petrosian used to ask himself: 'What have I 30 iL.xcs ~d8 31 ~e5
done wrong, so that I should be punished?'
25 ... iL.bS 26 f3
It transpires that after 26 ... iL.xfl 27 Wxfl
ctJf6 28 ~xe8+ ctJxe8 29 iL.e5! a rare position
arises: 'the d-pawn finally shows what it is
capable of' (Karpov). For it the knight has
to be given up - 29 ... f6 30 d6! or 29 ... ctJg7 30
d6 ctJe6 31 d7 Wf8 32 iL.f6 and wins. This
was what I feared! Black's downfall is
caused by the presence of the a2-pawn. If
he had eliminated it in good time, he could
now have fought on even a piece down.
26 ... gS
With the hope of saving a rather diffi- 'In such positions the goal is achieved
cult endgame with opposite-colour bishops. on autopilot.' (Zaitsev)
But it too is dashed ... 31 ... f6 32 ~f5 b6 33 iL.d4 ctJe7 34 iL.xf6 Mxd5
27 iL.xgS! iL.xfl (after 27 ... ctJxg5 28 ~xe8+ 35 ~g5+ ~xg5 36 ~xg5 ctJc6 37 We2 Wf7 38
iL.xe8 White picks up the knight - 29 h4) 28 Wd3 We6 (after 38 ... ctJb4+ 39 Wc4 ctJxa2? 40
Wxfl ctJd6 Wb3 the knight would have been trapped)
39 WC4 ctJe5+ 40 Wd4 ctJc6+
Here the game was adjourned.
29 iL.e7
The final touch. Black would still have
retained some chances after 29 ~xe8+ ctJxe8 41Wc4
30 iL.e7 f5 31 iL.xc5, but now the endgame a The sealed move. The continuations 41
pawn down is completely hopeless. We4 and 41 Wc4 ctJe5+ 42 Wb5 are equally
29 ... ctJc8 good, and so the following day I resigned
Unable to find an acceptable square for without resuming (1-0). Times: 2.43-2.17.
his knight - 29 ... ctJc4 30 d6 ctJb6 31 ~b1!
(threatening ~xb6) 31...ctJd7 32 l::txb7 or So, the third zero in a row! With the ac-
196
The Third Match: 1986
tive support of his opponent, Karpov had struck with great precision directly at the
managed to perform a minor miracle. My key points of our analysis. He successfully
recent enormous advantage had evapo- avoided all the traps and effectively antici-
rated like the morning mist. The score was pated all the strikes we had prepared ...
now equal: 91/2-9 1h and I had plenty to think In the 5th game I chose a rare and risky
about: how to explain my frankly weak plan, and quite quickly Karpov found the
play in the 17th and 19th games, and the refutation of the entire idea. This made me
breakdown in the 18th ... What had become suspicious: Karpov could hardly have
of that enormous playing advantage that I prepared so thoroughly before the match
had enjoyed before the 17th game? All for this new opening. And if the problem
kinds of thoughts came to mind ... was new for him, he should have taken
Karpov: 'When the score became critical - more time thinking about it at the board.
'minus three' - nearly everyone had written me So, was home preparation something with
off. But at that moment a renaissance began. a London hallmark? But at the start Karpov
There was nothing to lose, and I began playing had encountered other urgent problems, in
extremely sharply. Kasparov did not expect this the Nimzo-Indian Defence, for instance,
and he cracked under the sudden tension.' where in the 2nd game he failed to gain full
Zaitsev: 'I will not hide the fact that for us equality. It was hard to understand how, in
these were the most cheerful moments of the less than a week, his team had managed to
match. After three successive wins it unexpect- prepare a complicated variation with an
edly seemed that the wildest dreams of Karpov's important improvement, and, moreover,
supporters were about to be realised. Only one not in the main direction, but in a sideline.
more effort was required.' This showed an amazing gift of foresight!
Even so, I was still 'leading' - 12-12 Nikitin: 'From the 5th game onwards Kar-
would suit me, and so Karpov still had to pov began accurately hitting the weak points of
win another game. With the three wins he our opening repertoire, and Garry, mystified by
had, so to speak, saved his honour, but he the reasons for such foresight, became upset and
had not yet won the match. I had to cast nervous. In the second match we had enjoyed a
gloomy thoughts aside and prepare for the clear advantage in the opening, but it had
next game. suddenly disappeared, although on this occasion
we had prepared even more novelties. This was
Stab in the Back an extremely unpleasant blow. Garry was very
It was then that the unforeseen happened: distressed and became lost in conjectures, until
after the 19th game, Vladimirov, who for he gradually came to the terrible conclusion.'
five years had been my closest assistant, left But in the 4th game Karpov effectively
the team. It was his own decision. This closed an entire variation of the Nimzo-
sudden move shed a little light on the Indian Defence, by employing a continua-
mysterious events that had been occurring tion which was one of our analytical se-
during the match. crets. As will be seen from the notes, it was
I will summarise the unusual facts which not only the analyses of the two players
have been mentioned earlier in the game that coincided, but also the 'holes' in their
commentaries. It will be remembered that analyses! Our short-sightedness is easy to
back in London, when Karpov encountered explain: we were looking for an advantage
my new opening - the Griinfeld Defence - for White and we overlooked the plus
not only was he not taken aback, but he points of 12 .. :~c7!. But surely Karpov, in
197
Kasparov vs. Karpov: 1986-1987
198
The Third Match: 1986
compound and the trainers were to work mirov also handed his telephone in. But
separately. Vladimirov flatly refused to when at 6 o'clock on the morning of 25
comply (Nikitin: 'In such a situation this was a September my mother instinctively dialled
foolish, revealing step, to say the least'). I tried his number, someone at the other end lifted
to persuade him that it shouldn't be hard to the receiver without answering. And sure
do this - the match was nearly over. And enough, that day a second telephone which
after all, during the match Karpov's trainers he'd been concealing was found in Vladi-
hadn't even attended any of the games, let mirov's room. We then decided to discon-
alone taken long walks in the town. He just nect all the phones in the house except the
had to show a bit of restraint ... But before main one.
the 19th game Vladimirov went off anyway, When Vladimirov discovered that his
quietly, no one knew where. This by itself secret telephone had been disconnected, he
was reason enough to get rid of him. was enraged. He accused us of spy mania
Nikitin: 'In the 19th game the former and announced that he was leaving the
champion displayed remarkable erudition in a team. Then the head of our delegation
complicated variation which we had prepared demanded that he return his chess notes.
long before, but kept deep in reserve. "Karpov Vladimirov replied that there weren't any
already knew how to breach the Griinfeld such notes. But Litvinov quickly refreshed
Defence, and it again collapsed", Akimov wrote his memory and he was forced to produce
later. A seemingly harmless sentence, but in them. And here we saw what an extensive
fact excessive frankness by a friend of Karpov 'correspondence game' had been going on
who was unfamiliar with chess nuances. And, behind my back. For some reason, Vladimi-
indeed, such aggressive and rapid play in the rov's range of interests did not coincide
opening is possible only when you know what with the openings that he usually used, but
that day is going to happen on the board.' included only those that had been my main
After the 19th game Litvinov and weapons against Karpov.
Nikitin finally informed me that for the last We were disheartened and didn't try to
six months Vladimirov had been making persuade Vladimirov to stay. All the night
copies of our joint opening investigations. of 26 September (the day of the 20th game)
They had thought that he was doing this he waited by the main telephone for some
with my knowledge... We immediately call, and after it he left the team - early that
remembered Vladimirov's frequent disap- morning our chauffeur took him to the air-
pearances and his complete freedom in port. Before this we asked him for a written
London: before every game he would go explanation, and this is what he wrote:
'for a run'. We remembered that an hour
before the 18th game was due to be re- From February 1986 onwards I made writ-
sumed he had also gone off into the town, ten notes of the analyses made at Kasparov's
and Karpov had been late for the adjourn- training sessions in which I took part. Along
ment session ... Nikitin: 'Our next meeting with the fair copies made for me by Shakarov, I
was stormy and extremely open. The trainers made abridged copies for myself for later use in
again confirmed their readiness to stick to a tournaments. I told no one in the team about the
regime which would exclude any suspicion, and existence of these notes. I made notes on the
only one man flatly refused to do this.' following variations:
To be on the safe side, we decided to - Grunfeld Defence (main variation and
remove all the trainers' telephones. Vladi- variation with 5 .i.f4),
199
Kasparov vs. Karpov: 1986-1987
200
The Third Match: 1986
on, apart from me, a novice in these matters. draw, to regain my mental balance? Or, as
'How's things, Igor Arkadich? What about usual, aim for a fight? 'Why not?', I said. 'I
the variations?' have just lost three games in a row; can I
'Not very reassuring', said Zaitsev. 'The really lose a fourth?' But Misha Gurevich,
lads aren't ready to discuss them with you. No who had great experience not only in chess,
one ... Decide for yourself We won't have time but also of gambling matters, replied: 'The
to answer all your questions, that's for sure. Of theory of probability doesn't work like that.
course, there is a risk ... But you know better.' When you play roulette, you can bet on the
'Why hesitate?', asked the Central Army black each time and still lose several times in a
Sports Club representative in surprise. 'We row!' This is sad, but true: there is no point
must keep up the pressure - strike while the in believing that, after several misfortunes,
iron's hot.' luck is bound to smile on you. There are no
'Don't put pressure on Anatoly Ev- cosmic weights, balancing everything up ...
genevich', another one interjected. 'You've In the 20th game we decided in the end
heard it: the position is complicated, and how to 'not to play' (i.e. to aim for a short draw), in
play is unclear ... ' the 21st to parry the opponent's onslaught,
And five minutes later he went to phone the and to make the 22nd the 'retribution
arbiters that Karpov was taking a time-out. game'. With this in mind we planned the
Later Karpov himself explained that he opening repertoire for the coming games,
had developed problems in the opening. and in particular a new 'Black' opening was
Evidently, something had happened to his hastily prepared - in such a situation the
'foresight', and for the first time in the Gri.i.nfeld Defence was out of the question
match he didn't know what position would (mainly on psychological grounds).
be reached on the board tomorrow. Just as To carry out the first part of the plan, I
inexplicable was the way his indomitable chose the solid Catalan Opening. At that
fighting spirit evaporated at this crucial time it occurred quite often in my games,
stage of the match. but I had not played it against Karpov since
The match was entering the finishing the unlimited match (1984/85).
straight. The two players had used all their
time-outs and now the day for each subse-
quent game was definitely known. By Game 20
taking a break, Karpov had lost the psycho- G.Kasparov-A.Karpov
logical initiative and given me time to lick World Championship Match,
my wounds. I was exhausted: the tension of 20th Game, Leningrad 29.09.1986
the two previous years was taking its toll, Catalan Opening E05
added to which was the burden of the
return match. But I now had three days'
rest, and during this time I was able to 1 d4 tZ.lf6 2 c4 e6 3 g3
come to my senses and plan for the last On this occasion not the move 3 tZ.lc3
stage of the battle. (Game Nos.2, 4) and also not 3 tZ.lf3 (Game
Nos.lS, 21, 24).
Retribution 3... dS 4 .i.g2 .i.e7 5 tZ.lf3 0-060-0 dXc4
In this extreme situation I discussed with Karpov regularly played this variation,
my trainers how I should play with White where Black's chances of seizing the initia-
in the 20th game. Should I quickly force a tive are minimal. Given an elementary
201
Kasparov vs. Karpov: 1986-1987
degree of caution, White does not risk diverts one of the black pieces from the
anything, and this meant that I could preparation of ... c7-cS. If llct:Jbd2 there can
guarantee myself the necessary respite. follow 11...cS! 12 ~xf6 gxf6!?, not fearing
the weakening of the kingside: 13 ct:JgS?!
fxgS 14 ~xb7 .i:!.b8 IS ~e4 fS 16 ~g2 cxd4 is
unfavourable for White, and also after 13
dxcS ct:JxcS 14 ct:Jb3 .i:!.c8 IS .i:!.adl 'iVb6
(Khalifman-Aseev, Leningrad 1989) or 13
.i:!.adl 'iVb6 14 dxcS ct:JxcS IS ct:Jd4 J:l.ac8
(Seirawan-Ivanchuk, Monte Carlo (rapid)
1994) Black has a good game.
11 ... ct:Jxf6
Of course, not 1l ... ~xf6 or 11...gxf6 in
view of 12ct:JgS, when ... c7-cS can no longer
be played: 12 ... ~xgS (12 .. .fxgS) 13 ~xb7
.i:!.a7 14 ~c6ct:Jb8 IS e3.
7 'iVc2 a6 8 'iVxC4 12ct:Jbd2 .i:!.c8 13 ct:Jb3
8 a4 is also possible, as in the 22nd game White has done his utmost to hinder the
of the 1984/8S match. White obtained freeing advance ... c7-cS and he is hoping to
somewhat the better game, but effective exploit the weakness of the cS- and c6-
ways of equalising were soon found for squares, which Andersson sometimes
Black. succeeded in doing. But with accurate play
8 ... b5 9 'iVc2 ~b7 10 ~g5 Black should be able to solve his opening
The soundest. In the 8th game of the problems.
1984/8S match I played 10 ~d2, but there I
was fighting for an advantage, whereas
here it was more important to hold on
tightly to the 'bird in the hand'.
10. ..lZJbd7
202
The Third Match: 1986
203
Kasparov vs. Karpov: 1986-1987
'iVxc5 19 'iYxc5 ~xc5 20 ~ac1 ~xc1 (the c-file ~xd8 19 ~xa6 ~dl+ 20 .i.fl g5! the activity
has to be conceded, since 20 ... ~fc8? loses a of Black's pieces compensates for his mini-
pawn: 21l:f.xc5 l:f.xc5 22 LtJxb5) 21 ~xc1 ~d8! mal material deficit: 21 LtJxg5 .i.xf2+ 22
22 ~c6 LtJe8 23 LtJb3 ~d6 24 ~c8 'it>f8 25 ~a8 'itxf2 LtJg4+ 23 Wg2 LtJe3+ 24 'it>f3 LtJxfl 25 h4
~c6 26 LtJd4 ~d6 27 e3 e5 28 LtJf5 .l:!.d2, LtJd2+ 26 We3 h6 27 LtJf3 LtJxf3 28 'itxf3 ~d4
gaining a draw with accurate play; with equality.
c) 16 ... .i.xc5 17 LtJxc5 (17 .i.xd5 .i.xd4!) 16 ... 'iUC7
17 .. :iWxc5 18 'iUxc5 ~xc5 19 .i.xd5!? (19 l:tac1
l:!.fc8 20 .l:!.xc5 ~xc5 21 .i.xd5 ~xd5 or 21 LtJb3
l:tc8 with equality) 19 ... LtJxd5 20 l:rac1 ~xc1
(20 ... ~fc8? 21 ~xc5 ~xc5 22 e4 and LtJxb5) 21
~xc1, and again White has slightly the
better endgame, but Black has good
chances of a draw;
2) 16 ~ac1. Natural development.
17 a4
'17 ~ac1 .Jtxc5 also promises nothing
real - by accurate play Black should be able
to solve his problems (White cannot reach
the c6-square), (Ova matcha). Even so, it
made sense to play this - after 18 e3 the
position would have remained more tense
and Black would have had to play care-
Analysis diagram fully. In view of Karpov's imminent time-
trouble, this could have proved psychologi-
16 ... .i.xb3 17 axb3!? (as yet only 17 'lWxb3 cally unpleasant for him.
.i.xc5 18 e3 'i!Vb6 with equality has been Thus after 18 ... 'iUb8 19 a4! the weakness
played, but not 18 LtJel? .i.xf2+! 19 Wxf2 of Black's queenside begins to cause alarm,
LtJg4+) 17... .i.xc5 18 LtJd4! 'iUb6 (Black also while after 18 ... h6 19 LtJd4!? 'iUb6 20 LtJc6 to
fails to equalise immediately with 18 ... .i.b6 rid himself of the obstinate knight he has to
19 'iUxc7 ~xc7 20 ~xc7 .i.xc7 21 LtJc6) 19 e3 calculate the dangerous variation 20 ... ~xc6
.l::f.c7 20 b4 .i.d6 21 'i¥b3 ~d8 22l:!.xc7 .i.xc7 23 21 .i.xc6 'iUxc6 22 'iUc2 LtJe4 23 b4 LtJg5 24
~c1 or 17... 'ihc5 18 'lWxc5 ~xc5 19 ~xc5 'iUxc5 etc. White is also slightly better in the
.i.xc5 20 LtJe5 with a slight advantage to event of 18 .. JHd8 19 LtJd4 'iUb6 20 LtJc6
White. .l:!.xdl+ 21 'iUxdl. 18 ... 'iUb6!? 19 LtJe5 .l:!.c7 is
16 'iUxb3 more solid, but here too after 20 ~c3 .l:!.fc8
In the event of 16 axb3 ~xc5 17 'lWxc5 21 'lWc2 the conflict is by no means ex-
.i.xc5! (the' greed' test is not a difficult one: hausted.
17.. :iWxdl+?! 18 .l:!.xdl .i.xc5 19 LtJe5, and However, to be frank, I was not yet
White has unpleasant pressure) 18 l:f.xd8 ready for a large-scale battle ...
204
The Third Match: 1986
19lt:Jd4
I considered the last try to be 19 ~a5 (19 Game 21
It:Jel 'iVb6 or 19".'iVc4!?), but it could have A.Karpov-G.Kasparov
succeeded only in the event of a mistake by World Championship Match,
Black. 19".~fd8 with counterplay is accept- 21st Game, Leningrad
able: 20 ~xd8+ ~xd8 21 'iVxb5 'iVc1 + 22 ~f1 1/2.10.1986
~f8! 23 It:Je5 (23 e3 ~d5 with equality) Queen's Indian Defence E15
23".~dl 24 e3 It:Je4! 25 ~a8 'iVc2! 26 It:Jd3
'iVe2!, forcing perpetual check - 27 ~xf8+
Wxf8 28 'iVb8+. Good enough too is 1 d4lt:Jf6 2 c4 e6
19".lt:Je4!? 20 e3 It:Jd6 (avoiding the calcula- 'An historic moment' (Keene): a match
tion of complicated variations) 21 It:Jd4 ~f6 innovation! I was not intending to give up
and Black's position is safe. the Griinfeld Defence, but the accumulated
19 ... b4 negative emotions (three painful 'punc-
The pawn has moved under the defence tures') demanded a respite. I hastily had to
205
Kasparov vs. Karpov: 1986-1987
master a 'new' opening. Previously I had minimum' - this commentary from Ova
employed the Nimzo-Indian Defence only matcha reflects the state of theory at that
twice - against Psakhis (49th USSR Cham- time. Later Karpov was to achieve serious
pionship, Frunze 1981) and Beliavsky (4th success in this quiet strategic variation,
match game, Moscow 1982). The change of which is fully in accordance with his style
'register' was necessary - what was impor- (d. Volume V of My Great Predecessors).
tant was not only the sound reputation of 4 ... ~a6 5 b3 j,b4+ 6 .i.d2 .i.e7 7.i.g2
this opening, but also the element of sur- This position occurred several times in
prise and novelty introduced into the the first match, but as a rule Karpov was
extremely tense match atmosphere. playing Black and he successfully solved
3 et::Jf3 his opening problems. The spirit of the
I was rather expecting 3 et::Jc3 - although 'unlimited' marathon was involuntarily
Karpov rarely allowed the Nimzo-Indian evoked ...
Defence, from the 13th to the 19th games of 7 ..• 0-0
this match his main motto had been activ- After the 16th and 18th games of the
ity. Besides, he must have known that after first match 7 ... c6 came to be considered the
3 ... ~b4 I had lost both of the afore- most accurate, but castling is also quite
mentioned games (true, in the variation acceptable, and in addition it may lead to
with 4 e3 c5 5 ~d3 0-0 6 et::Jf3 d5 7 0-0 dxc4, more complicated play - to which I was
whereas on this occasion we had hastily provoking my opponent in the opening of
prepared the 'Portisch-Hiibner wall': this game.
5 ... et::Jc6 6 et::Jf3 ~xc3+! 7 bxc3 d6; we had also S 0-0 dS
looked at 5 et::Je2!?, as Korchnoi played in Somewhat provocative: Black begins
Baguio). play in the centre without ... c7-c6, agreeing
2 ... e6 can hardly have caught my oppo- in the future (after ... c7-c5 etc.) to an 'iso-
nent unawares - the probability of this was lani' on d5 or hanging pawns.
not so difficult to work out. Back in the 11th
and 13th games of the first match I had
expressed my readiness to play the Queen's
Indian Defence - 1 et::Jf3 et::Jf6 2 c4 b6, and in
the 15th Karpov finally accepted my invita-
tion with 3 d4 e6 etc. And now he simply
decided to return to the opening debate of
two years earlier.
3 •.. b6
I did not want to play 3 ... d5 - the varia-
tion of the Queen's Gambit which I had
employed in the match with White (8th,
10th, 12th and 22nd games) seemed to me
at that moment to be dangerous for Black. 9 et::Jes
4g3 A psychological phenomenon: Karpov
'Again a move which rather surprised refuses the challenge, although he handled
me (4 et::Jc3 - Game No.IS). After 4 g3 White well the variations with 9 cxd5!. It is suffi-
has few chances of an advantage, but cient to remember the 15th game of our
Black's counterplay is also reduced to the first match and the games from the Karpov-
206
The Third Match: 1986
Sokolov super-final (Linares 1987), which dS or 14 ... b5 IS ~aS, although here too it is
went 9 cxdS ctJxdS 10 ctJc3 ctJd7 11 ctJxdS probable that Black has an acceptable game.
exdS 12 .l::kl I1e8 and 13 ... cS. True, I was But Karpov is aiming for a familiar set-up
intending to play 9 ... exdS 10 ctJc3 ..IiLb7, then and he does not succumb to the provoca-
... ctJa6 and ... c7-cS (like Karpov in the 4th tion.
game of the 1984/8S match), but this com- 14... b5 15 I1el
plicated position is also objectively advan- Hoping at the cost of a pawn to exploit
tageous to White. White's spatial advantage and strong pawn
Nevertheless the ex-champion does not centre. Again White makes a well-known
deviate from his plan, but 'follows Kas- move, ignoring an additional possibility -
parov's own treatment when he had White 15 "iVc2!? bxc4 16 I1fdl, and if 16 ... ctJf6, then
in game 6 of their first match.' (Keene) 17 I1abl I1b8 18 I1xb8 "iVxb8 19 ~al! iYbs 20
9 ... c6 10 ..IiLC3 ctJfd7 (promptly exchanging I1bl iYhs 21 ctJxc4 with the initiative (Bukic-
the powerful knight on eS) 11 ctJxd7 ctJxd7 Tal, Bugojno 1978).
12 ctJd2
For more details of this popular tabiya -
d. the previous volume Kasparov vs. Karpov
1975-1985.
12 ... I1c8 13 e4
15 ... bxc4
In the 6th game of the 1984/85 match
Karpov, playing Black, did not risk accept-
ing the pawn sacrifice - IS ... ctJb6?, and by
16 c5! I could have gained an obvious
13 ... dxC4 positional advantage. Analysing this game
Another 'provocation'. Nowadays they at that time, my trainers and I came to the
play both 13 ... cS and 13 ... dxe4 14 ctJxe4 bS conclusion that after 15 ... bxc4 Black has a
IS I1el bxc4 16 ~f1!? (Kramnik-Aronian, reasonable game. But a considerable time
Mexico 2007; Bacrot-Karpov, Ajaccio (blitz) had passed since then, and at the board it
2007; Kasimdzhanov-Karpov, Bastia (rapid) was with difficulty that I tried to remember
2007). And with the plan chosen by me, it is something about it (I think it was the same
considered more accurate to play 13 ... bS 14 for Karpov).
I1el dxc4 IS bxc4 bxc4, transposing into a 16 "iVa
position which arose in the game. Later, after the game Karpov-
14 bXc4 M.Gurevich (SSth USSR Championship,
Neglecting the 'peripheral' opportunity Moscow 1988), 16 "iVa4! ~bS 17 "iVc2 came
to sharpen the situation by 14 ctJxc4 cS IS into fashion - by drawing out the bishop to
207
Kasparov vs. Karpov: 1986-1987
b5, White intends to gain a tempo for the game White transfers his knight to e3,
a2-a4 advance (d. Game No.10 in the previ- overlooking the opponent's counter-stroke.
ous volume Kasparov vs. Karpov 1975-1985, 17 f4 was more accurate, continuing to play
note to Black's 15th move). for restriction (and if 17 ... ~fd8, then now 18
ttJf1 ~f6 19 ttJe3), but such a committal
advance obviously did not come into
Karpov's plans.
However, the quiet 17 ~f1! was best of
all, aiming to regain the pawn. Now
17 ... e5?! leads to an advantage for White
after 18 ttJxc4 ~xc4 19 ~xc4 exd4 20 ~xd4
ttJe5 21 i..e2! ~fd8 22 ~ed1 c5 23 ~ac1 ttJc6
24 ~e3 (Chernin-Psakhis, Beer Sheva 1993).
It is also not so easy to readjust for the ... c6-
c5 advance: 17... ~b7?! 18 ttJxc4 ttJb6 19
~ab1! (19 ~a5 c5) 19 ... c5 20 ttJxb6 axb6 21
"iVb2 or 19 ... ~b8 20 ~a5 ~a8 21 ~g2 ~fe8 22
16 .. :~C7?! ~ed1! with a clear advantage. Black would
Not the most accurate reply. Black's ex- probably have had to cling on to the mate-
tra pawn and solid position do not guaran- rial by 17 ... ttJb6, but here too after 18 a4 he
tee him a quiet life; his main problem is the would still have had problems: 18 ... c5 19 as
activation of his pieces. I rather vaguely ttJd7 20 d5 exd5 21 exd5 ttJf6 22 ~e5 'iVd7 23
remembered the main conclusion of our old ttJxc4 ttJxd5 24 ttJb6! axb6 25 ~xa6 or
analysis - to aim for ... e6-e5, but during the 18 ... ~fd8 19 as ttJd7 20 ttJxc4! (20 .i.xc4 ~b5)
game I decided also to retain the possibility 20 ... ~b8 21 ~d2, and he still has to find a
of ... c6-c5. way to equalise.
The immediate 16 ... e5 is dangerous in 17 ... eS! (17 ... c5 18 d5 was less good for
view of 17 ttJf3 exd4 18 ttJxd4, as is 16 ... c5 Black) 18 ttJe3 exd4 19 ~xd4
17 d5 (Karpov-Adianto, Bali 2000). The
game Chernin-Hertneck (Luzern 1985)
went 16 ... ttJb6 17 ttJf1 c5 18 d5 exd5 19 exd5
~f6 20 ttJe3 ~b8 21 ~e4, and here 21...'it>h8!
22 ~ad1 ~e8 would have given clear equal-
ity, but a typical reaction to ... ttJb6 is
stronger -17 a4!. Therefore later they began
playing 16 ... ~e8, and after 17 ~ad1 - either
17... ttJb6 18 a4! c5 19 d5 exd5 20 exd5 "iVd7
21 as ttJa4 22 ~a1 c3 23 ttJe4 with the idea
of ttJxc3 or d5-d6 (Dautov-Palac, Pula 1997),
or 17... .i.f8 - this is rather more solid (a
fresh example: Onischuk-Yakovenko,
Poikovsky 2007), but in either case White's 19 .•. ~CS
chances are somewhat better. The commentators unanimously ap-
17 ttJfl?! proved this move, but in fact it is expedient
By analogy with the Chernin-Hertneck only from the viewpoint of Black's competi-
208
The Third Match: 1986
tive aim (to make a draw). 19 .. .ctJeS! was pluses are evident - an advantage in the
more logical, maintaining the tension and centre and play against the weak c6-pawn.
not fearing either 20 i..xeS 'iVxeS 21 i..fl It only remains for him to coordinate his
i..b4!? 22 l:ted1 l:tce8, or 20 'iVc3 f6 21 ct:Jxc4 heavy pieces. To aim for exchanges in such
(if 21 i..xeS, then both 21...fxeS and a position is clearly nonsensical.
21..:~xeS 22 'iVxeS fxeS 23 ..ih3 l:td8 24 i..e6+ Was Karpov really so frightened by a
'it'h8 2S ..ixc4 i..c8! are good) 21...i..xc4! possible invasion at d4 or d3? Then why
(21...ct:Jxc4 22 i..fl!) 22 i..xeS fxeS 23 'iVxc4+ did he not choose the 'purely Karpovian' 22
'it'h8 with a double-edged position with i..fl! with the possible continuation
opposite-colour bishops. This would have 22 ... ct:Je6 (22 ... l:td4 23 ct:Je3!) 23 eS ct:Jd4 24
enabled Black to carry out his main defen- 'iVe4 - White is assured of a strategic initia-
sive idea (the undermining ... e6-eS, and the tive. After 24 ... cS 2S ct:Jd6 i..xfl 26 'it'xfl .l:tb8
creation of the threat ... ct:Jd3) and would 27 l:tac1 or 24 ... i..xc4 2S i..xc4 cS 26 l:tab1
have set White more serious problems. 'iVc6 27 'iVxc6 ':xc6 28 'it'g2 ':c7 29 ':b2 and
For all its apparent consistency and f2-f4 Black would have faced a rather
soundness, the exchange of the dark-square unpleasant defence. Now, however, White
bishops leads merely to the stabilisation of loses any real hope of exploiting the mobil-
the position and frees White's hands for ity of his e-pawn.
play against the weakened c-pawns. So what happened? After all, in the
20 i..XCSct:JXCS 21ct:JXC4 l:tfd8?! 22nd game of the 1985 match, when he also
An error! '21...ct:Je6 also fails to solve needed to win at all costs, Karpov played
Black's problems because of 22 eS - the very actively, seeking the slightest chance
values of the d6-square (for White) and d4 to continue the struggle. But here ... What
(for Black) are clearly not equivalent', I had become of the indomitable spirit of this
wrote in Ova matcha. But that is what I uncompromising chess fighter? My formi-
should have played - after 22 ... ct:Jd4 Black is dable opponent underwent an amazing
very close to a draw: 23 'iVe4 i..xc4 24 'iVxd4 metamorphosis at the finish of the Lenin-
i..e6 2S 'iVcs l:tfd8 or 23 'iVc3 cS 24ct:Jd6ct:Je2+ grad half of the match ...
2S l:txe2 i..xe2 26 ct:Jxc8 l:txc8 etc.
22 ... ':xd1
22 l:tad1? Played with great pleasure: without the
During the last few moves the position rooks it is far more easy for Black to defend.
has considerably simplified, but White's But 22 ... ct:Je6!? would also have done: 23
209
Kasparov vs. Karpov: 1986-1987
ttJe3 'iYa5 24 ttJf5 ttJd4! 25 ttJxd4 l::!.xd4 26 e5 pawns in order - 30 ... g5!. Analysis shows
J::tcd8 27 J::tc1 i..c4 with equality. that this was the most accurate way to
23 .l::i.xd1 J::td8 24 J::txd8+ 'iVxd8 25 h4! (oth- draw:
erwise White might even stand worse) 1) 31 hxg5 fxg5 32 ttJc4 'lite6 (32 ... g4? 33
25 .. :i!Vd4 ttJe5+) 33 g4 (otherwise ... g5-g4, breaking
Centralisation in the endgame is always up White's pawn chain) 33 ... ttJd3 34 a3 (34
attractive, but in the given specific instance 'lite2? ttJc1+ or 34 'litg2 c5) 34 ... ttJe5! (an
it is not essential. Black did not have to unconventional decision) 35 ttJxe5 (it is
provoke the exchange of queens, but could pointless to play 35 ttJd2 c5 36 '.t>e2 c4,
have played either 25 ... g6 26 ttJe5 (26 ttJe3 when the black king goes to c5) 35 ... Wxe5
WVd4) 26 ... ~d6 27 ~c3 h5 or 27 f4 i..b5 with 36 'it>e2 'it>d4 37 'litd2 c5 38 'it>c2 c4 39 '.t>d2
quite good prospects, or 25 ... i..b5!? (the c3+ 40 'litc2 'litc4 with a draw;
most solid) 26 ttJe3 ttJe6 27 a4 ttJd4! 28 'iWc5 2) 31 ttJc4 gxM (or even 31...g4!? 32 'lite2
i..xa4 29 ~xa7 i..b5 30 e5 h6 with equality. gxf3+ 33 'litxf3 ttJd7 34 g4 'lite6 35 'litf4 c5) 32
gxM ttJd3! (32 ... 'it>e6? is inaccurate: 33 ttJa5
'litd6 34 'lite2 ttJe6 35 'lite3) 33 ttJa5 c5 34 ttJb7
'it>g6 35 'it>e2 c4 36 ttJd6 ttJc1 + 37 'it>e3 ttJxa2
38 ttJxc4 'lith5 39 'litf4 'litxM 40 'litf5 ttJc3 41
'.t>xf6 'litg3 42 e5 '.t>xf3 43 e6 '.t>f4 with a
draw.
I have to admit that, anticipating a quick
draw, I did not play this ending in the best
way. Now Black's position is somewhat
worse, and he has to be accurate.
30 '.t>f2 'lite6 31 i..c4+ 'litd6 32 'it>e3 ttJd7
210
The Third Match: 1986
44e5+
A speculative check: of course, Karpov
knew about its consequences, but he de-
cided to see whether or not I would make a
mistake. Other moves also do not give
White's only chance is to activate his White anything real- Black is saved by the
knight, but for the moment it is tied to the great mobility of his bishop:
defence of the e4-pawn. 1) 44 We3 WcS 4S ctJd1 .id7 46 a3 as 47
39 .ib3 ctJa8 40 Wd3 ctJb6 41 i.c2 .ixa4 .ixa4 48 ctJc3 .id7 49 fS .ie8;
Here the game was adjourned. A similar 2) 44 .ixa4 .ixa4 4S Wc3 WcS 46 ctJg4
position arose in the 20th game of the .ic6 47 eS fxeS 48 ctJxeS .ibS 49 fS WdS SO
previous match, but here there are more ctJg4 .ia6 Sl f6 gxf6 S2 ctJxh6 .ic8;
grounds for hostilities. Black can passively 3) 44 ctJd1 WcS 4S .ixa4 .ixa4 46 ctJc3
wait, but that is not the most pleasant of .id7 47 We3 .ig4 48 a3 as 49 fS hS, and
occupations. again the draw is obvious.
41 ....ib5+! 44 ...fxe5 45 ctJe4+
The sealed move. I correctly grasped Hoping for 4S ... Wc6? 46 fxeS with strong
that it would be desirable to exchange my activity. In the event of 4S ... We6 the com-
knight for the opponent's bishop. puter assessment makes a sharp jump in
42 Wc3 ctJa4+ 43 Wd2 White's favour, but it is not difficult to
211
Kasparov vs. Karpov: 1986-1987
ascertain that after 46 .ixa4 (46 f5+ We7!) with unknown opening surprises by his
46 ... .ixa4 47 et:Jc5+ Wf5 48 et:Jxa4 exf4 49 gxf4 opponent, Karpov failed to cope with the
'iitxf4 Black cannot lose (50 et:Jb2 Wg4 51 problems and lost his fighting spirit.
et:Jxc4 WxM etc.). Did my opponent seri- This is also indicated by his pessimistic
ously think that after adjournment analysis assessment of his chances after the 21st
I would not risk giving up the piece? game. In our other matches there was a
4s ...We6! fierce battle to the very end. In the second
Draw agreed on White's proposal match (1985) Karpov won the 22nd game
(Yz-Yz). Times: 2.37-2.38. and fought desperately for a win in the
Although I had undertaken another 24th; in the fourth (1987) he won the 23rd
half-step towards my goal (the score had game, forcing me to try and level the scores
become 10 1/2-10V2), this draw meant more to in the 24th; and in the fifth (1990), when
me than just half a point. In interviews after losing by 'minus two', he again found the
the match, Karpov called this the decisive strength to win the 23rd game. It has to be
game of the final phase: after it he realised agreed that, in comparison, the finish of
that the best he could hope for was to this indomitable fighter in the third match
maintain an equal score and he would be makes a rather strange impression.
unable to change it in his favour. Did this At any event, the opponent's irresolute
mean that this was his 'last and decisive play could not fail to lift my spirits - every-
battle'?! Can such passive play (the rejec- thing was going to plan. It was time to go
tion of 9 cxd5 and 15 ~c2, and the moves 17 on to the offensive: 'fate leads the daring,
et:Jf1 and 22 ~adl) really be called playing but cowards are dragged along'.
for a win? The day before the 22nd game it rained
I think that this game had a serious psy- incessantly, but as we set off for the start it
chological implication. Today it is even suddenly stopped. And when I got out of
more obvious that Karpov's play at the the car, over the Leningrad Hotel there was
finish differed strikingly from that which the most brilliant rainbow I had ever seen
he displayed between the 14th and 19th in my life. It was wonderful! For my ro-
games. Not without reason is it said that mantic and rather superstitious nature, this
the last battle is the most difficult one ... was like a sign from on high. And, indeed,
'And the wins had taken so much effort', ex- this game granted me some unforgettable
plained Zaitsev. 'In those concluding days of moments ...
the match, Anatoly reminded me of a weight-
lifter, who with an incredible effort of will and
talent had raised an enormous weight, but at the Game 22
last moment was unable to maintain it'. G.Kasparov-A.Karpov
But was it not an inflatable weight? Af- World Championship Match,
ter all, it is one thing to 'raise an enormous 22nd Game, Leningrad
weight' in training, and quite another to do 3/4.10.1986
it on the stage, in a fierce competition one Queen's Gambit 055
against one. Three wins by Karpov in this
match, in the 5th, 17th and 19th games,
were gained largely thanks to purposeful, 1 d4 et:Jf6 2 c4 e6 3 et:Jf3 dS
'honed' home preparation. But in a real With his 3rd move Karpov lets it be
situation, when he suddenly had to cope known: he does not want to take a risk in
212
The Third Match: 1986
the variations with 3 ... b6 4 ttJc3 ~b7 5 a3 d5 have come under attack) and wants to
6 cxd5 ttJxd5 7 "iVc2 (as in the 32nd game of provoke the reply a2-a3, removing the
the 1984/85 match) or 4 ... ~b4 5 ~g5 (as in support of the bishop (which would like to
the 18th game) and he is essentially re- remain on the a2-g8 diagonal) and weaken-
nouncing any attempt to win. A debatable ing White's queenside.
decision, although one that could have 16 a3
been predicted, and typical of Karpov: he The only way of disputing the timeli-
did the same in the 23rd game of the previ- ness of 15 ... a5 was 16 ttJe5!? As it was, the
ous match. Incidentally, there the same question of when it is best to advance the a-
opening variation occurred. pawn - on the 15th or 18th move (d. the
4 ttJC3 ~e7 5 ~gs h6 6 ~xf6 ~xf6 7 e3 0-0 8 variation with 15 ... ~e8) - remained debat-
~cl c6 9 ~d3 ttJd7 10 0-0 dXc4 11 ~xc4 eS able. Apparently in either case the position
(11 ... c5!? - Game No.12) 12 h3 is on the verge of equality - the majority of
A year earlier this move had a great White's micro-pluses are illusory.
psychological effect, but I did not manage 16 ...~e8
to win. Now the novelty was lacking ... Soon after the match Karpov himself,
12 ... exd4 13 exd4 ttJb6 (13 ... c5?! - Game playing White, demonstrated to Beliavsky
No.lO) 14 ~b3 ~fS 15 ~el (Moscow 1986) that 16 ... "iVd7, the other
continuation employed here, is insufficient:
17 ttJe5 ~xe5 18 ~xe5! ~fe8 19 "iVe2 ~ad8 20
~e1 ~xe5 21 "iVxe5 a4 (21...ttJd5 22 ttJxd5
and ~c1) 22 "iVc5! axb3 23 ~e7 "iVd6? (after
23 ..."iVc8 24 "iVxb6 ~d7 25 ~xd7 "iVxd7 26
"iVxb3 Black simply loses a pawn) 24 l:te8+,
and White won.
17 ~xe8+ "iVxe8 18 "iVd2
lS ... aS
A slight improvement. 15 ... ~e8!? would
have led to a position from the 23rd game
of the 1985 match, which continued 16
~xe8+ "iVxe8 17 "iVd2 "iVd7 18 ~e1 ~d8 (if
18 ... a5!?, then 19 ttJe4!? is interesting) 19
"iVf4 with some pressure for White.
White's opening plan appears to offer
only modest chances, but after that game An interesting moment. At training ses-
the variation became popular, and for a sions before the match my helpers and I
long time Black was unable to neutralise it. analysed this position, but we did not in
Black's last move was one of the most fact discover Black's best reply, which
recent ideas: he moves his pawn from a7 effectively refutes the main line of the
(where after an exchange of rooks it might variation. It is highly significant that this
213
Kasparov vs. Karpov: 1986-1987
reply also went unnoticed by Karpov. Yet ing the queen from going to f4.
another instance of the amazing coinci- 19 ... ~g6
dence of mistakes in the home analyses of This move is easily explained. Many of
the two contestants ... the commentators suggested 19 ... ~e6 20
18 ...etJd7?! ~xe6 'iWxe6 21 'iWc7! 'i!!Vb3, but I am abso-
One of the axioms of the 12 h3 variation lutely certain that none of them would
was that after the exchange of all the rooks want to defend the position after 22 etJe4!
- IS .. :iHd7! 19 l::tel l::teS (if 19 ... a4 20 ~a2 etJfS 23 etJxf6+ gxf6 24 l::tel! etJg6 (24 .. :~i'xb2?
l:tdS, then both 21 'iWf4 and 21 'iWe3 are 25 etJM! 'iWxa3 26 l::te3 and wins) 25 etJh2! h5
promising) 20 l:txeS+ 'iWxeS there follows 21 26 etJf1 etc. Karpov too did not want to.
'iWf4 when White retains the initiative. But Meanwhile, 19 ... ~g6 is a minor victory
in the game M.Gurevich-Van der Sterren for White: the stock of the b3-bishop, which
(Baku 19S6), played a month later, doubts has escaped from the threat of exchange,
were cast on this: 21...~e6! 22 ~xe6 'iWxe6 has risen and the pressure on the f7-point
(without the moves 15 ... a5 16 a3 the a-pawn has been retained.
would now be lost - 22 'i!!VbS+ and 'iWxa7) 23
'i!!VbS+ 'iWcs 24 'iWa7 etJc4 25 b3 etJxa3 26 'iWxa5
~e7 with a quick draw. Also in the event of
23 ~c7 Black would have solved his strate-
gic problem by tactical means: 23 ... etJc4 24
'iWxb7 ~xd4!, and now if 25 etJxd4 there
follows 25 ... 'iWel + 26 'it>h2 'iWe5+, while if 25
'i!!VbS+ 'it>h7 26 etJxd4 - 26 ... 'iWel + 27 'it>h2
'iWxf2 with a guaranteed (in view of ... etJe3)
perpetual check.
Karpov tries to solve the problem in an-
other way. With IS ... etJd7 Black has taken
control of the e5-square and cleared the
way for a possible queen move to b6; true, 20 h4!
now his supervision of the d5-point has A hard-to-find, multi-purpose move:
been weakened, but d4-d5 leads merely to simultaneously both attacking and, above
the simplification of the position. In what all, restraining and performing a prophy-
direction then should White operate? In the lactic function. Now the g6-bishop feels
most critical! He must demonstrate that uncomfortable! In the event of 20 l::tel 'iWdS!
control of e5 and the threat of ... 'iWdS-b6 are 21 etJa4 (21 etJe4 'i!!Vb6!) Black has the strong
mere details, and not the main thing in this 21...~h5! 22 'iWf5 (after 22 g4?! ~g6 the
position, that the black pieces now suffer move h3-M is no longer possible) 22 ... ~xf3
from a certain lack of harmony, and that 23 'iWxf3 etJb6 24 'iYh5 etJd5 with equality.
White, by playing energetically and accu- 20 ... 'iWd8 21 etJa4!
rately, can nevertheless force the opponent Parrying the threat of 21...'i!!Vb6. Now
on to the defensive. Black faces a difficult choice: White's initia-
19'iWf4! tive, particularly on the kingside, may
This way, rather than 19 l:tel (although become dangerous. I was very pleased to
the rook move rules out 19 ... ~e6), to which have found the idea of h3-h4, and I experi-
Black could have replied 19 ... 'i!!VbS!, prevent- enced an emotional surge.
214
The Third Match: 1986
215
Kasparov vs. Karpov: 1986-1987
opening has turned out favourably for me: squares - in particular, against the h4-
Black no longer has the advantage of the pawn. If 29 .. :iVd6 there would have fol-
two bishops (I have managed to exchange lowed 30 l::!.c1 with the intention of l::!.c6. In
my a4-knight for the bishop on f6!) and the any event, Black has simplified the position
pawn structure is evenly balanced, but and is close to equalising.
White's minor pieces are more mobile. 'Almost the same structure occurred in
27 ...ltJdS! the 23rd game of the previous match. But
Many commentators condemned this there White had a light-square bishop, and
decision by Karpov, which after the ex- here he has a knight, which is clearly to his
change on d5 allows the opponent to ob- advantage. Even so, it is hard to believe
tain, in their opinion, a 'stable' advantage. that Kasparov manages to win within 12
But this is a superficial assessment. The moves.' (Keene)
white knight is indeed stronger than the 30 "iYf3
bishop, but Black has provoked the ex- Unconventional, to say the least. Play on
change of the dangerous bishop on b3, he the dark squares with 30 'iVf4 is more
has got rid of his c6-pawn (which was far natural, but then 30 ... l::!.c8 31 ltJxg6 fxg6 32
weaker than the d4-pawn) and he can l::!.e6 'iVc7! (after 33 Wilxc7 l::!.xc7 the rook
exploit the b- and c-files for counterplay. In endgame is drawn: 34 l::!.xg6 l::!.c1 + 35 'It>h2
my opinion, the exchange on d5 strength- l::!.c4 36 l::!.g5 l::!.xd4 37 l::!.xh5 g6 38 l::!.g5 l::!.xh4+
ens Black's position! Of course, White's 39 'It>g3 l::!.a4 40 l::!.xg6+ 'It>f7) or even 30 ... l::!.a6
chances are somewhat better, but they also 31 l::!.c1 l::!.b6 32 l::!.c5 l::!.b3, and it is unclear
were before the exchange ... how real White's advantage is. Therefore
27 ... ltJg4?! was recommended, but after he also fights for influence on the light
28 'lWc3 'lWd6 29 ltJe5! l::!.e8 30 g3! it would squares!
have left White with a more appreciable
advantage: 30 .. :iVf6 31 l::!.e2 or 30 ... ltJxe5 31
dxe5 Wilc7 32 e6. And after 27 ... a4 28 .i.c4
ltJd5 he is not obliged to play 29 .i.xd5 - 29
Wilc1 is possible, when Black has the same
problems: 29 ...Wild6 30 ltJe5l::!.e8 31 g3 etc.
3 O... l::!.a 6
Black, in turn, disregards the 'banal'
30 ... l::!.c8 (after 31 g3 basically nothing
would have changed) and intends to use
his rook along the 6th rank.
31l::!.Cl
28 .i.xds cxdS 29 ltJes Wild8 The impatient 31 ltJxg6l::!.xg6 (31...fxg6 is
Beginning counterplay on the dark too passive: 32l::!.e5l::!.d6 33 g3) 32l::!.e5 could
216
The Third Match: 1986
have spoiled things - after 32 ... 'iYxh4 33 2) 33 tLlxg6 Mxg6 (safer than 33 ... 'it'xg6)
'iYxdS Mb6! at the cost of a pawn Black 34 llcS 'it'h6 3S g3 'iVg4 36 l:txaS Mf6 (with
would have gained good counterplay: the threat of ... hS-h4) 37 MCS g6 38 ,uc1 ttfS;
1) 34 'iYxaS Mb1+ 3S Mel Mxe1+ 36 'iYxe1 3) 33 tLlf3 'iVg4! 34 tLlgS+ (34 'iVd8 ..te4 is
'iYxd4 37 'iYe8+ 'it'h7 38 'iYxf7 'iYa1 + 39 'ith2 equal) 34 ... 'ith6 3S tLlxf7+ ..txf7 36 ~xf7 Mg6!
'iYxa3 40 'iYxhS+ 'it'g8 with a drawn queen (36 ... Mf6 37 ~c4 Mg6 38 g3! - a variation
endgame; from Ova matcha - 38 .. J:td6!, keeping ... hS-
2) 34 Me8+ 'it'h7 3S 'iYfS+ Mg6 36 'iYxaS h4 in reserve, is also acceptable) 37 'iVdS
'iYg4! 37 'iYdS 'iYd1+ 38 'it'h2 ~g4 39 Me3 (the MgS! 38 llVc6+ g6 39 Mc4 ~e2! or 38 'tWa8
rook ending also does not promise a win - 'iYxd4, and Black succeeds in covering his
39 ~e4 'iYxe4 40 .l':txe4 Md6) 39 ... Mf6 40 Mf3 king - 39 'iVh8+ 'it'g6 40 1:lc6+ 'it'f7 41 'iVc8
Mf4! or 40 Me4 MfS! 41 'i¥b7 Mf4 42 f3 'iVgS 43 ~dS 42 g3 l:teS 43 'iVb7+ <it>g8! (but not
Mxf4 'iYxf4+ 44 g3 ~d2+ 4S 'it'h3 'iYd1, also 43 ... .l::i.e7?? 44 Mf6+!) or 41...'iVa1 + 42 <it>h2
with a draw. ~eS 43 g3l:tfS 44 'iVb7+ 'itg8!.
However, I did not even consider ex- However, as we will see, for the mo-
changing the minor pieces, since I believed ment the position in the game is also on the
in the strength of my knight - and not verge of equality.
without reason! 32 'iVh3!
Preparing an invasion. It is true that
Black also gains such an opportunity, but
with a substantial difference: the white
knight will take part in the coming attack,
whereas the black bishop, alas, will not.
32 ...,Ub6
32 ... !!e6 33 .l::i.cS .l::i.b6 34 .l::i.c8 could have
come to the same thing.
33 1:lc8
31 ... 'it'h7
Karpov is aiming to make' solid' moves,
but in a situation where the opponent has
more strategic pluses, chances such as
31...'iYxh4!? 32 'iYxdS 'it'h7 have to be ex-
ploited. The abstractly unfavourable ex-
change of the central pawn for a wing
pawn would have made the play more
lively and concrete, and would have left
Black with every chance of a draw: H ...'iYd6?!
1) 33 g3 (33 Mc8? Mb6) 33 ... ..te4! (Black An inaccuracy, which was not noticed
shows his teeth!) 34 gxh4 ..txdS 3S MCS Md6 by the commentators. At d6 the queen may
36 .l':txaS f6 37 tLld3 ..tc4 38 MxhS+ 'it'g8 and come under a camouflaged knight fork, and
... Mxd4; therefore 33 .. .'~Ve7! was correct, neutralising
217
Kasparov vs. Karpov: 1986-1987
the threat of 'ii'g3-g5 and intending ... a5-a4 t!,xa5 .l::txd4 etc.
and ... t!,b3 with sufficient counterplay (if 34 35.l::ta8!
t!,a8, then 34 ... ..tc2! 35 t!,xa5?! t!,b3). An unpleasant surprise for Black! The a-
34 'ii'g3 pawn cannot be defended: 35 ... .l::ta6? 36
Patience, above all, patience ... CZixf7! ..txf7 37 'ii'd3+ or 35 ... .l::tb3? 36 .l::th8+!
~xh8 37 CZixf7+. Of course, these tactical
blows are not accidental, but a consequence
of the great activity of the white pieces.
But, by giving up the pawn, Black can
try to exploit the time spent by the oppo-
nent on its capture, in order to activate his
forces.
35 ...'ii'e6
Even now 35 .. .'ii'e7 came into considera-
tion, although after 36 .l::txa4 ..tc2 37 .l::tb4!
White would have retained excellent win-
ning chances - after all, he is now a pawn
up: 37... .l::txb4 (37 ... .l::te6 38 'ii'g5!?) 38 axb4
34 ... a4? 'ii'xb4 (38 ... f6 39 'ii'c3) 39 CZixf7 'ii'xd4 40 'ii'b8
But this is now a serious mistake, lead- 'ii'xh4?! 41 'ii'h8+ ~g6 42 CZie5+ Wf5 43 CZif3!
ing to the loss of a pawn. However, it is 'ii'c4 (43 ... 'ii'a4 44 'ii'xg7!) 44 'ii'xh5+ ~e6 45
already difficult for Black to find acceptable 'YIlYe8+, picking up the g7-pawn, or 40 ... ..tf5
moves. It is dangerous to play 34 ... 'ii'e6?! 35 41 CZig5+ ~h6 42 'ii'h8+ ~g6 43 'ii'h7+ ~f6 44
t!,a8! t!,b1+ (35 ... ..te4? 36 'ii'g5) 36 ~h2 t!,d1 'ii'xh5 etc.
37 t!,xa5 'ii'b6 (37 ... ..te4? 38 'ii'g5; 37 .. .lhd4?
38 CZif3) 38 t!,a8 'ii'b1 (38 'ii'b7 39 t!,d8) 39 'ii'f4
etc. - as in the variation from the note to
White's 37th move.
The counter-attacking 34 ... t!,b1 +?! 35
~h2 'ii'a6 (with the idea of ... 'ii'fl) is parried
by the accurate 36 t!,f8! (36 t!,e8 ~e2!)
36 ... 'ii'fl (36 ... t!,b7 37 t!,d8 'ii'e6 38 t!,a8 is no
better for Black) 37 'ii'f3 'ii'h1 + (37 ... a4 38
CZixf7! ..txf7 39 'YIlYxf7 'ii'gl+ 40 Wg3 t!,b3+ 41
f3 'ii'e1+ 42 ~f4!, avoiding perpetual check)
38 ~g3 'ii'd1 39 'ii'xd5 'ii'b3+ 40 'ii'xb3 t!,xb3+
41 f3 f6 42 CZic4 with a clearly better ending.
However, here too 34 ... 'ii'e7! (controlling 36.l::txa4
the g5-square) was correct, and if 35 t!,a8, First 36 'ii'g5! was more accurate, deny-
then 35 ... ..te4! with the idea of 36 l:txa5?! f6 ing the opponent any illusion of counter-
37 CZif3 ..txf3, equalising. Black should also play: 36 ... 'ii'd6 37 ~h2 'ii'f6 38 'ilVxf6 l:txf6 39
be able to hold on in the ending after 36 f3 ~g3, or 36 ... .l::tb1+ 37 ~h2 'ii'f6 (37 ... 'ii'd6 38
..tf5 37 'ii'g5 'ii'xg5 38 hxg5 f6 39 gxf6 gxf6 40 t!,xa4) 38 'iVxf6 gxf6 39 CZid7 ~g7 40 .l::txa4
t!,a7+ ~h6 41 CZif7+ ~g6 42 CZid8 .l::td6 or 36 ..tf5 41 CZic5, and Black has a difficult end-
'ii'g5!? ~xg5 37 hxg5 .l::tb1+ 38 ~h2 .l::td1 39 game.
218
The Third Match: 1986
219
Kasparov vs. Karpov: 1986-1987
It is too early for 39 liJd7 in view of liJd7 with 41...~e7. After 41 J::td7! ~a2 42
39 ... iVb1 40 'iVf4 J::th1+ 41 ..t>g3 ~d3+ 42 ~e3 J::txd5 J::txa3 43 J::td7 J::ta2 44 'iVg5! Black
'ii'd1 43 liJf8+ ..t>g8 44 f3 J::te1 45 'iVc3 J::te8! 46 would hardly have been able to save the
liJxg6 fxg6 47 J::tb7 ..t>h7, when the open game, but at least he would not have lost
position of White's king makes it extremely by force.
difficult for him to convert his extra pawn. Here the time scramble came to an end
39 ... J::tc2?! imd the game was adjourned. Outwardly
Again it was better to play 39 ... J::tfl! 40 Karpov looked very content - indeed, Black
liJf3 (40 f3? 'iVc8!) 40 .. .f6 41 ~c7!. It was would appear to have achieved much. I
against this threat that Karpov had de- sank into thought...
fended, but after 41...~f7! 42 a4 ~g4 Black
could still have resisted. True, here White
has more chances than after 38 ... J::tfl!.
Firstly, his passed pawn can advance to a7
- 43 as J::txf2 44 ~g3 J::tfl 45 a6 ~xg3+ 46
..t>xg3 ~g6 47 a7 J::ta1 48liJd2, although I am
not altogether sure that this will bring a
win. Secondly, White does not have to give
up his f2-pawn, but as before can play 43
~g3!? - now his rook does not stand in
front of his a-pawn, and in the endgame
this gives him additional chances.
220
The Third Match: 1986
to overwrite the sealed move (a photocopy this move a study-like solution.' (Keene)
of the scoresheet was published soon 41 ... l::txd4
afterwards in Literaturnaya gazeta), and only It is not a matter of the pawn, but of the
then did I seal my scoresheet in the enve- exchange of queens, which Black wishes to
lope. offer (from f4). How to defend against this?
'Either I am going mad, or else I have a 42 ttJf8+ 'it'h6 43 l::tb4!!
forced and very pretty win!' I informed my The crux of the combination! White's
trainers, who, incidentally, were looking small and scattered army operates so
rather despondent (they too had been harmoniously that Black is unable to break
analysing only 41 l::tb4). I remember their up the mating net without great loss of
first reaction: 'No, that sort of thing doesn't material.
happen ... ' But it turned out that it does
happen! We analysed the position for a
long time: it was an incomparable pleasure,
and, in addition, we had to check over and
over again that our eyes were not deceiving
us, and that these beautiful variations really
were possible!
At night the Kamenny Island was
plunged into darkness, and only in the two
houses where the match contestants were
living did the lights continue to shine. At
our place the lights shone right until day-
break - this was an amusing idea of mine:
let Karpov think that I had not sealed 41 The following variation is especially
ttJd7 ... pretty: 43 ... l::txb4 44 axb4 d4 45 b5 d3 46 b6
Of course, my opponent arrived for the d2 47 b7 d1'iV 48 b8'iV (threatening check
adjournment session with a heavy heart. It and mate from f4) 48 ... 'iVc1 49 ttJxg6 'iVxg6
stands to reason that he had found the 50 'iVh8+ 'iVh7 51 'iVgxg7 mate!.
combination, and one can imagine with Here Karpov thought for about a min-
what feelings he awaited the revealing of ute .. .
the sealed move, and with what difficulty 43 ... l::tC4
he appeared to remain completely calm. Other replies would also have lost -
When the arbiter was opening the enve- 43 ... l::td1 44 l::tb8 ~h7 45 'iVg5+ 'iVxg5 46
lope, Karpov stared into the audience. Even hxg5+ Wxg5 47 ttJxh7+ or 43 ... l::td3, and
so, he could not fully maintain his imper- White has several ways to his goal:
turbable appearance: he was unable to 1) 44 l::tb8 ~h7 45 'iVg5+ (the spectacular
restrain himself and he cast a glance at 45 ttJe6!? fxe6 46 l::tf8! is also decisive)
Schmid's hand, when the latter was taking 45 ... 'iVxg5 46 hxg5+ 'it'xg5 47 ttJxh7+ Wh6 48
the scoresheet out of the envelope. Even ttJf8 l::txa3 49 l::td8, and White must be able
before the move was reproduced on the to convert his extra piece;
board, Karpov saw it and understood 2) 44 a4!? (the simplest) 44 ... l::te3 (it is
everything ... even worse to play 44 ... d4? 45 l::tb5 or
41 ttJd7!! 44 ... l::ta3 45 l::td4! 'iVf6 46 l::txd5) 45 l::tb8 'iVe5
'The grandmasters unanimously called (45 ... ~h7 46 'iVg5+, retaining the a-pawn) 46
221
Kasparov vs. Karpov: 1986-1987
'ii'xe5 lhe5 47 CDd7 with the threat of 48 ~d4!! (a study-like move!) 47 ... l:ta5 48 "fic7
CDxe5 or 48 l:th8+ ..th7 49 CDf8 and l:txh7 with an irresistible attack.
mate; It is not often that one encounters such a
3) 44 'ii'e1 ?!. A tempting move, which we 'box' as that in which the black king found
also looked at. White wins spectacularly itself at the end of this game! It is greatly
after both 44 ... d4? (45 'ii'c1+ was threatened) hindered by its own pawns on g7 and h5
45 'ii'g3! followed by l:tb5, and 44 ... ..th7 45 and the ill-starred bishop, and the role of
l:tb6+ g6 46 'ii'c1 + ~g7 47 CDe6+!! fxe6 48 the key to this construction is played by the
'ii'c7+ 'ii'f7 49 'lWe5+ ~h6 50 l:txe6 and l:te7. white knight.
But - let's hear what Botvinnik had to 44l:txc4 dxc4 45 'ii'd6! c3 46 'ii'd4
say: 'In his adjournment analysis Kasparov
was not satisfied with the simple win of a
piece, but sought a problem-like continua-
tion of the attack with 44 'ii'el. Capablanca
thought that it was unaesthetic to play for a
brilliancy if there was a simple win, and in
the given instance he was proved right:
candidate master M.Chudakov (and after
him grandmaster J.Nunn) showed that
44 ... 'ii'c8! 45 'ii'e7 'ii'f5 46 f4 'ii'f6 would have
repelled the attack. Yes, a striving for
brilliancy sometimes prevents Kasparov
from reaching the truth ... Here with experi-
ence everything should come right.' The c3-pawn cannot be defended:
At the time I also thought that 'the 46 .. :iVd3 47 'ii'f4 mate! Black could have
paradoxical 44 ... 'ii'c8! 45 'ii'e7 ~f5 would continued the 'struggle' to the end
have spoiled White's entire plan.' However, (46 ... ..th7 47 "fixc3, and now either 47 ... g5 48
the computer age has changed the assess- 'ii'd4 gxM 49 CDxh7 ~xh7 50 'ii'e4, transpos-
ment of the position. Thus instead of 46 f4?! ing into a pawn endgame, or 47 ... ..tg8 48
it is far stronger to play 46 'ii'c7!, forcing 'ii'e3+ g5 49 'ii'xg5+ 'ii'xg5 50 hxg5+ ~xg5 51
46 ... l:te3 47 a4 'ii'e5+ 48 'ii'xe5 l:txe5 49 CDd7! a4 f5 52 a5 ..tc4 53 CDd7 ~f4 54 CDc5 ~e5 55
l:te2 50 CDb6, when Black again has a diffi- a6 ~d6 56 a7 ..td5 57 CDd3 ~c7 58 CDf4,
cult endgame. And the attempt 45 .. :iVc1 eliminating all the black pawns), but Kar-
(instead of 45 ... 'ii'f5) is refuted by the un- pov stopped the clocks, admitting his
commonly spectacular 46 'ii'd6! f6 47 l:tb1! defeat (1-0). Times: 2.42-2.32.
'ii'e3 48 l:te1! 'ii'd2 49 l:te2! 'ii'c1 50 'ii'g3 ..tf5 This game, which was of colossal com-
51 l:te7 g5 52 l:tf7 and l:txf6+. But the main petitive importance and which essentially
thing is that instead of 45 'lWe7 White has a decided matters, was simultaneously the
crushing stroke: 45 'ii'e5!, since the knight best in the match, thanks to its study-like
cannot be taken - 45 .. .'iVxf8? because of 46 finish.
l:tb8 'ii'xb8 47 'ii'xb8 l:txa3 48 'i!Vh8+ ..th7 49 Ironically, many of my successes are as-
'ii'c8! (pretty geometry: the threat is 'lWc1 + sociated with the number '13': I was born
and 'ii'xa3 or 'ii'g5 mate) 49 ...~a1 50 ~c3 and (and this is a success!) on 13 April; I became
wins. There only remains 45 ... 'ii'f5, but then the 13th world champion in '85 (8+5=13),
follows the unexpected 46 'iVg3! l:txa3 47 winning by the score 13-11; and this game,
222
The Third Match: 1986
already my 94th against Karpov (9+4), was 2 ... c5 3 tDc3 d5 (Game No.50) had not yet
played on 3 October (the 3rd day of the been prepared.
10th month), and was my 13th win over 3 g3 (not 3 d4 e6 - Game No.21) 3.•. c5 4..1tg2
Karpov! How can I help being supersti- i.b7 50-0 g6
tious?!
223
Kasparov vs. Karpov: 1986-1987
Christiansen, Moron 1981). During my Monte Carlo (blind) 1994). 13 ... h5 has also
training match with Andersson (1985) occurred (a recent example: Malakhov-
Adorjan strongly urged me to play 8 ... d6 Grischuk, Elista 2007).
and ... ct:Jbd7. 'There is no point in playing From the late 1990s 13 ... l:c7 came to the
8 ... ct:Jc6', Andras asserted. 'White gains no fore, and here White has a wide choice:
benefit from his queen at d4!'. To be honest, 1) 14 ct:Jel ..Iixg2 (14 ... 'iYb8!? Zifroni-
that was the extent of my knowledge of this Psakhis, Israel 2002) 15 ct:Jxg2 'iYb8 16 ct:Jf4
set-up, which today has been extensively e6! 17 ct:Jd3 (17 l:xd6 l:xc4) 17 ... l:fc8 18 ..Iif4
played. ..Iif8 19 a4 (19 ct:Jb4!? l:a7) 19 ... 'iYb7 20 g4
gl:d1 ct:Jbd7 ct:Je4 with equality (Pigusov-Cvitan, Ohrid
2001);
2) 14 ~h6 'iYa8 15 ..Iixg7 Wxg7 with ade-
quate counterplay: 16 'iYd4 b5! (Topalov-
Kramnik, Wijk aan Zee (blitz) 1999) or 16
ct:Jel b5! (Karpov-Gelfand, Cap d' Agde
(rapid) 2002);
3) 14 g4 h6!, and now not 15 h3 g5 16
'iYg3 b5! (Kramnik-Karpov, Dos Hermanas
1999) or 15 ..Iixh6 ..Iixh6 16 'iYxh6 ct:Jxg4 17
'iYh4 ct:Jgf6 18 ct:Jd5 ..Iixd5! 19 cxd5 l:xc1 20
l:xc1 b5 (Bu Xiangzhi-Sasikiran, Doha 2006)
with approximate equality, but 15 g5! hxg5
16 ct:Jxg5 ..Iixg2 17 Wxg2, and Black still has
10 b3 to extinguish the opponent's initiative
With the intention of ..Iib2, but later it (Khenkin-Parligras, Dresden 2007);
transpired that the best square for the 4) 14 ..Iih3 'iYb8 (not 14 ... 'iYa8? because of
bishop is e3 (from here it both eyes the 15 ..Iixd7 l:xd7 16 ct:Ja4 Anand-Gomez
queenside, and can also jump to h6 or g5). Esteban, Santurtzi (rapid) 2003), and again
After 10 'iYM apart from the usual 10 ... l:c8 White is at a crossroads.
there is 10 ... h6!? 11 ..Iie3 l:c8 12 l:ac1 g5 13
'iYd4 0-0 (Kramnik-Nikolic, Monte Carlo
(blindfold) 1999). From the late 1980s the
main line became 10 ..Iie3 0-0 11 'iYh4l:c8 12
l:ac1 (but not 12 ..Iih3? ..Iixf3 13 exf3 ct:Je5!
Topalov-Kasparov, Wijk aan Zee (blitz)
1999) 12 ... a6, and after 13 ct:Jel ..Iixg2 14
ct:Jxg2l:e8 15 b3 'iYc7 16 ..Iig5 'iYb7 17 ct:Je3 b5
Black has an excellent game (Kramnik-
Kasparov, 14th match game, London 2000).
13 b3 is more promising, after which
they initially replied 13 ... l:e8 - for example,
14 ..Iih6?! ..Iixh6 15 'iYxh6 (Paunovic-Karpov,
Bali 2002) 15 ... b5!, but then rejected this Analysis diagram
because of the sharp 14 ..Iih3! l:c7 15 g4!
(but not 15 ..Iih6 l:c5 Kramnik-Karpov, It began with 15 ~h6 - Topalov played
224
The Third Match: 1986
this against Psakhis (Las Vegas 1999), Van ttJf6 18 f3 'iVc7 (threatening 19 ... bS) 19 a4 dS!
Wely (Monte Carlo (blind) 200S) and Gris- with equality (Andersson-Portisch, Reykja-
chuk (Monte Carlo (rapid) 2006). Then IS vik 1991).
g4 came to the fore: IS ... e6 16 gS ttJe8 17 I was concerned about 12 ttJdS, justify-
ttJe4 bS! (Sorokin-Sadvakasov, Moscow ing the development of the bishop at b2 (in
200S) or 17 ~g2 bS 18 ttJe4 bxc4 19 bxc4 the game it will wander between the
dS!? 20 cxdS ~xdS with chances for both squares al, b2 and a3, but will not find a
sides (Aronian-Kramnik, Saint Vincent place for itself). This appears to be 'refuted'
200S), and soon IS ~gS also appeared by 12 ... ~xdS(?) 13 cxdS ~c2, practically
(Kramnik-Aronian, Turin Olympiad 2006). forcing a queen sacrifice - 14 ~d2 ttJhS
On the whole, although White tirelessly (14 ... ~xb2 IS 'iVxb2 ttJe4 16 ~d4 is no better)
makes interesting attempts to develop his IS ~xc2! ~xd4 16 ttJxd4, but this is clearly
initiative, the diagnosis remains un- advantageous to White (he has the c-file, an
changed: Black has a solid enough position. outpost at c6 and two powerful bishops).
10.. .l:tc8 12 ... ttJeS 13 ttJxeS ttJxdS 14 'i¥h4 is also
Black cannot get by without this move, insufficient, as is 12 ... bS 13 'iVxa7 ~xdS 14
and he makes it before castling, thus avoid- cxdS ttJcS IS 'iVa3 ~a8 16 'iVb4 ttJxdS 17 'iWd2
ing 10 ... 0-0 11 'i¥h4! ~c8 (1 1...ttJe4?! is etc.
worse: 12 ttJgS! ttJxgS 13 ~xb7) 12 ~h6 with However, the simple 12 ... a6! and only
the exchange of bishops. True, here too then ... b6-bS gives Black good prospects - I
after 12 ... ~cS (12 ... a6!? Pantsulaia-Gelfand, came to this conclusion at the board (while
Khanty Mansyisk 200S) 13 oltxg7 ~xg7 14 my opponent was considering his 12th
'iVd4 ~a8! IS ttJel (Pigusov-A.Ivanov, move), and later I tested it in practice: 13
Moscow 2001) IS ... ~xg2 16 ttJxg2 ~g8 17 ~ac1 bS! (Kir-Georgiev-Kasparov, Saint
ttJe3 ~eS White has merely an insignificant John (blitz) 1988). Although in the end I
plus. lost, the opening was by no means to
blame.
Karpov has other ideas: he does not
hurry, but manoeuvres, aiming for a pro-
longed battle ...
225
Kasparov vs. Karpov: 1986-1987
pawn, Black thereby frees his queen. But ttJa4?! is weak because of IS ... OUhS!, forcing
the harmlessness of White's plan would White into the unfavourable exchange 16
have been more simply emphasised by the .ixf6 .ixf6, since after 16 'iVd4? .ih6 Black
exchange of his dangerous knight, which is unexpectedly obtains a terribly strong
aiming for dS: 12 ... ttJe4!? 13 ~ac1 ttJxc3 14 attack: the rook cannot be driven away - 17
.ixc3 ttJf6! IS ttJel 'iVd7 16 .id4 oUfd8 with g4 eS! 18 "iVxd6 oUe6 19 'iYb4 ttJxg4 or 17 .l:tbl
equality (Gelfand-Kramnik, Monte Carlo .ie4 18 g4 eS! 19 "iVxd6 oUe6, while if 17 e3
(blind) 2003). there follows 17 ... 'iVa8! 18 ttJh4 .ixg2 19
13 oUacl a6 ttJxg2 ttJeS 20 ttJel "iVc8 or 19 .. .'iVf3 20 ttJel
Strictly speaking, this concludes the 'iVfS with the threats of ... 'iYh3 and ...b6-bS.
'compulsory program', and now the free ls ... 'iYa8 16 ttJe1
one has to be demonstrated! White is anticipating the exchange of
14.ial bishops and the switching of his knight via
The bishop is moved to a square where g2 to dS ...
it is defended - this prophylaxis is under- It is hard to explain my sudden desire to
standable: now the threat of ... b6-bS can be play freshly, originally and boldly. Appar-
parried by a2-a4, since if ... ttJcS there is the ently after the 22nd game I had regained
reply oUbI. But one cannot help thinking my former confidence, and in addition I felt
that White's play is insufficiently energetic that my impudent rook would unnerve the
to cause Black any discomfort. However, opponent, but after making my next move,
nothing better is apparent: 14 ttJd4 .ixg2 IS I myself was frightened by its boldness!
~xg2 .l:tcs 16 'iit'gl 'iVa8 etc.
The modest bishop retreat suggested to
me a rather amusing idea.
16 .. JUS!?
I considered 16 ... .ixg2 17 ~xg2 oUhS!,
which was also tempting: if 18 ttJf4?! there
14 ... OUcS!? is 18 ... OUxh2 19 ttJcdS oUh6, and White's
Here the rook will aid the typical position is not improved by the weaknesses
'hedgehog' attacks ...b6-bS and ... d6-dS; in after 18 h4 .ih6 19 f4 .ig7 or 19 ... ~cS.
addition, Black prepares to switch his However, by 18 'iVf3 'iYxf3 19 exf3 White
queen to a8. could have obtained a quiet 'Karpovian'
15 a4 ending - this seemed unpleasant to me, but
The most solid move: Karpov wants to objectively White has no advantage, for
be able to forget for ever about ... b6-bS. IS example: 19 ... ttJeS 20 ttJf4 oUfS 21 ~g2 .ih6
226
The Third Match: 1986
22 etJce2 etJed7 23 .l:tc2 .l:tc5 or 21 etJcd5 ..th6! strict the opponent's play. If 18 etJd3 the
22 etJxf6+ llxf6 23 ..txe5 dxe5 24 etJd5 ..txc1 best is 18 ... etJc5 19 etJxc5 .l:txc5! (19 ... bxc5 is
25 etJxf6+ exf6 26 .l:txc1 .l:td8 with equality. risky: 20 'iYd3 .l:th5 - 20 ... 'iYxb3? 21 .l:tbl - 21
16 ... .l:tf5 is more interesting, since it cre- f3 .l:tb8 22 .l:tbl, when the rook at h5 is out of
ates a completely unconventional position. play), intending ... .l:tec8 and a possible ... d6-
Something similar occurred in the famous d5 or ... b6-b5.
4th game of the Tarrasch-Lasker match 18 ... hS!
(Game No.54 in Volume I of My Great Prede- The aim of this timely move is both de-
cessors). Nearly 80 years later the rook finds fensive (concern for the rook!) and aggres-
itself in a more dangerous position - there sive (a potential threat of ... ~h7 and ... ..th6).
are still many pieces on the board! If 18 ... etJc5, then 19 .l:tbl, while after 18 ... etJe5
Of course, the threat of ... etJf6-g4 is not the reply 19 g4?! etJfxg4 20 fxg4 etJxg4 21
the main point: by controlling the d5- 'iVe4 does not work because of 21...'iYc8!, but
square and assisting the ... d6-d5 advance, the simple 19 etJd5 is good.
the rook is performing useful work. And it 19 etJg2
is not easy for White to attack it with his White would no doubt have liked to
passive minor pieces. Meanwhile, it is prevent the rook from returning by 19 etJd3,
psychologically difficult for the opponent but Black is all ready for 19 ... d5; however,
to avoid thinking about the rook, about the 19 ... ~h7! 20 etJe4 ..th6 is more appropriate,
desire to trap it, although there are few with complicated play. And if 19 'iYd2!?
resources for doing this. Incidentally, (with the idea of e2-e4; 19 'iYf2? etJg4) Black
Tarrasch became carried away by thoughts had the good reply 19 ... .l:tc5! 20 e4 (Karpov
of trapping the rook, and in the end he lost was obviously trying to manage without
the game ... this weakening move) 20 ... h4, which the
But if White, after playing f2-f3, is able knight on g2 prevents.
to play e2-e4 at a convenient moment,
Black's entire venture may be called into
question. Interesting tempo-related play
now begins.
17 ..txb7 ~xb7
19 ... .l:tcS!?
The logical sequel to the preceding play
was the simplifying 19 ... d5 20 cxd5 etJxd5 21
etJxd5 .l:txd5 22 ..txg7 (22 .l:txd5 'iYxd5 23
..txg7 ~xg7 24 .l:tc7 e5 is no better for White)
18 f3! 22 ... 'it>xg7 23 etJf4 .l:txdl+ 24 .l:txdl etJf6 with a
This necessary move is an attempt to re- probable draw. But it seemed to me that
227
Kasparov vs. Karpov: 1986-1987
here White retains a minute plus, and I urge (16 ... l:!.f5!?) had gone - from the posi-
again displayed my belligerence - I decided tion in the match I needed a draw!
not to 'make life easy' for Karpov, but to 24 ctJf4 bS!
maintain the tension by returning my rook Success! The pawn sacrifice is, of course,
to 'base'. A loss of time? Yes, but the posi- only temporary.
tion allows this. 2S cxbS axbs 26 ctJxbS
20..ltb2 White needs exchanges: in a compli-
A strange bishop move - it would ap- cated battle with the queens on, the weak-
pear that the audacious manoeuvres of the nesses in his position make his chances
black rook have unsettled Karpov. 20 ctJf4?! worse.
did not work because of 20 ... ..Ith6!, but 20 26 ... ctJxbS 27 'iVxbs 'iVxbs 28 axbs l:!.b8 (it
'iVd3!? followed by ctJe3 would have been in would appear that the scales have tipped in
the spirit of the position. Black's favour ... ) 29 ..Itb2!
20 ... l:!.cc8 21 ..Ita3 White still has resources. He is even
The impression may be gained that the hoping for some advantage after 29 .. J:txb5
black 'hedgehog' is still destined merely to 30 ..Itxf6 ..Itxf6 31 ctJd5 l:!.d8 32 ctJxf6+ exf6 33
stay on the defensive: after all, neither ... b6- e4. But it is not essential to capture imme-
b5 nor ... d6-d5 is possible. But in fact Black diately.
has achieved something: the opponent's
pawn chain has been weakened.
21 ... ctJcS 22l:!.b1
29 ... l:!.b7!
An accurate solution: now there is noth-
ing that White can do - Black is ready for
22 ... ctJe6! 30 ..Itxf6 ..Itxf6 31 ctJd5 ..Itg7 32 b6 l:!.eb8 and
A very strong manoeuvre. The knight ... e7-e6. I had the feeling that this move
has occupied and immediately abandoned broke Karpov psychologically: something
the excellent square c5, for the sake of suggested to me that there would not now
carrying out Black's main strategic objec- be an adjournment...
tive - the ... b6-b5 advance. 30 b6
23 'iVd3 ctJC7 The best chance was 30 b4!? l:!.eb8! 31
After ... b6-b5 the position will simplify ..Itxf6 ..Itxf6 32 ctJd5 ..Ith8 33 l:!.bc1 e6 34 ctJc7
to a draw. Interesting possibilities were with a more comfortable draw: 34 ... l:!.c8 35
promised by 23 ... ctJd7!? (for example, 24 ctJa6 l:!.xc1 36 l:!.xc1 ..Itd4+ 37 'it'g2 l:!.xb5 38
ctJd5 ctJec5 25 'iVc2 b5), but my aggressive l:!.c8+ Wg7 39 l:!.b8 l:!.xb8 40 ctJxb8 Wf8 41 ctJc6
228
The Third Match: 1986
1 d4 ct:Jf6 2 c4 e6 3 ct:Jf3 b6 4 g3
In a situation where the opponent needs
to win having the white pieces is a great
help. And I chose the solid main variation: I
did not want to go in for complicated play
(4 ct:Jc3 - Game No.18) - too much mental
energy had been spent on solving the
grandiose competitive objective. Any
'supernatural' measures aimed at sharpen-
ing the struggle could be left to Black.
4 ... .Jtb7
After the 2nd game of our first match
After 32 ... ~xg7 33 ct:Jd3 Mxb6 the b4- Karpov usually preferred 4 ... jLa6, but then
pawn is a weakness, as long as the four White has sufficient opportunities to sim-
rooks are on the board (34 Mdc1 ct:JcS! with plify the play, and on this occasion he is
an immediate denouement), and therefore aiming for a more fresh position, even
Karpov offered a draw, which, quite under- though it may be slightly inferior.
standably, was accepted. (Yz-Yz). Times: S jLg2 .Jtb4+
2.21-2.10. Deviating from the main paths. If
S... .Jte7 I would not have repeated the sharp
The score became 12-11, and with that Polugayevsky Variation 6 0-0 0-0 7 dS exdS
the fight for the world championship title 8 ct:Jh4! (Game No.6 in Kasparov vs. Karpov
concluded. At the time it was often asked 1975-1985), but would have played 6 ct:Jc3
why in such cases the match should con- ct:Je4 7 .Jtd2 with a slight plus (later Karpov
tinue. It was all very simple: in those himself regularly played this - Game
matches it was not just the title that was Nos.97, 99 in Volume V of My Great Predeces-
contested, but also the monetary prize. The sors).
winner's share - five-eighths of the prize 6 .Jtd2 as
fund - went to the player who scored 12V2 A comparatively rare variation, which at
points. Nowadays FIDE shares the prize that time was employed by Korchnoi,
fund equally, apparently to sweeten the Speelman, Short and also my opponent's
loser's bitter pill. helpers, Lerner and Salov. Korchnoi also
Thus there was only one question to be tried the provocative 6 ... cS, but for Karpov
settled in the concluding game: would this would have been going too far ...
Karpov be able to level the scores and 70-00-0
thereby share the prize fund equally? I would have been quite happy with
229
Kasparov vs. Karpov: 1986-1987
7 ... i..xd2 8 'ii'xd2 (Korchnoi-Spassky, Palma 1) 3 ... i..b4+ 4 i..d2 as S g3 0-0 6 i..g2 b6 7
de Mallorca 1968) or 7 ...eba6 8 i..gS (Gheor- 0-0 i..a6 (up until the end of the 1970s this
ghiu-Larsen, London 1980). was constantly played by Romanishin, and
then Makarychev) 8 'ii'c2 ttJc6 9 i..gS i..e7 10
a3 h6 11 i..xf6 i..xf6 12 l:tdl 'ii'e7 l3 e3 with
complicated play (Kasparov-Yusupov, 49th
USSR Championship, Frunze 1981);
2) 3 ... b6 4 g3 i..a6 S b3 i..b7 6 i..g2 i..b4+
7 i..d2 as 8 0-0 0-0 9 i..gS (Karpov later
played only 9 'ii'c2 or 9 ttJc3) 9 ... i..e7 10 ttJc3
ttJe4 11 i..xe7 'ii'xe7 12 'ii'c2 ttJxc3 13 'iYxc3 d6
with equality (Kasparov-Hiibner, 4th match
game, Hamburg 1985).
8 ... i..e7 (not 8 ... ttJa6? or 8 ... dS? because of 9
cS! and a2-a3) 9 'ii'c2
Aiming to gain a tempo compared with
8 i..g5 9 i..xf6 i..xf6 10 ttJc3 (as in fact happened in
8 ttJc3 d6 (but not 8 ... l:te8?! 9 'ii'c2, as in the game) and again avoiding the possible
the ancient source game Fairhurst-Yates, simplification after 9 ttJc3 ttJe4: I considered
London 1927) 9 'ii'c2 ttJbd7 is considered 9 'ii'c2 to be objectively the strongest move,
harmless, as Sakharov and Matanovic and obvious play for a draw often does not
defended back in 1968. succeed.
8 i..f4 with the threat of 9 cS and a2-a3 is
a worthy alternative. After 8 ... i..e7 9 ttJc3
ttJe4 (9 ... dS?! 10 ttJeS Osnos-G.Kuzmin, 33rd
USSR Championship, Tallinn 1965) Black
exchanges knights and retains a flexible
pawn structure, for example: 10 'ii'c2 ttJxc3
11 'ii'xc3 d6 12 'ii'd3 fS l3 'ii'e3 'ii'd7 with an
unclear game (Timman-Korchnoi, Las
Palmas 1981) or 12 l:tfel ttJd7 13 'ii'c2 cS 14
l:tadl (Tukmakov-Smyslov, Las Palm as
Interzonal 1982), and here Black could have
equalised by 14 ... cxd4 IS ttJxd4 i..xg2 16
'it'xg2 l:tc8 17 ttJbS dS 18 b3 i..b4! 19 l:t£1
'ii'e7. 9 ... h6
This line would appear to be more suit- Clarifying the situation. 9 ... cS!? 10 dxcS
able for White if he wants a 'cast-iron' bxcS 11 ttJc3 d6 12 l':i.adl ttJbd7 l3 l:td2 'iVb6
draw, but 8 i..gS appealed to me because of was possible, with chances for both sides
its relative novelty, and in itself it is quite (Christiansen-Korchnoi, Pasadena 1983;
good: White aims for a spatial advantage in Jaracz-Korchnoi, Warsaw (rapid) 200S), but
the centre, agreeing in return to grant his the weakening of the bS-point is not in
opponent the two bishops. In addition, I Karpov's style. It was also hard to decide
already had experience of employing this on the still unexplored 9 ... dS (Winants-
move in 'related' set-ups after 3 ttJf3: Karpov, Brussels 1987).
230
The Third Match: 1986
15 e5!?
12.l:i.adl White's forces are deployed actively
A continuation of the same simple strat- enough for him to cross the demarcation
egy: the pieces are placed as close as possi- line, and in addition it is unpleasant for
ble to the centre, control of which will Black to have his queen vis-a.-vis the rook
guarantee White a quiet life. on dl. It is interesting that soon after the
12 ... d613 h4 match, in a game with Korchnoi (Tilburg
A typical idea, aimed at weakening 1986), Karpov tried to improve White's
Black's kingside. The threat is h4-hS. play by IS .l:i.fel iLg7 16 ctJbS 'li'b8! 17 dS eS
13 ... h5 18 b3 'li'd8 19 a3 iLh6 20 b4 with an endur-
The 'holes' after 13 ... ~g7 14 hS gxhS IS ing strategic initiative, but by resourceful
dS are not to Karpov's taste (IS ctJgS?! is defence Black saved the half point (similar
now inappropriate: IS ... 'Ii'xgS 16 iLxb7 positions arise in the variation S... iLe7 6
~a7). In Ova matcha I gave only IS ... eS(?) 16 ctJc3 ctJe4 7 ~d2 when Black captures on
ctJh4 'with a clear positional advantage for d2).
231
Kasparov vs. Karpov: 1986-1987
16 dS?!
But this is overdoing it - I thought I had
found an ideal way of prettily concluding
the game by perpetual check. The re-
strained 16 l:tfel!, completing my develop-
ment, was more logical. Now after 16 ... dxeS
17 4Jxe5 .Jtxg2 18 'it'xg2 4Jxe5 19 dxeS 'lie7 16 ... 4Jxes!
20 f4 l:!.ad8 21 4Je4 l:td7 22 l:txd7 'l\\Vxd7 23 The only defence, but a perfectly ade-
.l:.dl 1i'c6 24 'it'h2 White's position is slightly quate one. Of course, 16 ... exdS? was bad in
better, although the most probable outcome view of 17 e6! with a strong attack: 17 ... fxe6
is a draw. The alternative is 16 ... iVe7, for 18 4JgS "iiVf6 (18 .. :~e8? 19 l:.fel! 4JcS 20
example: .JtxdS! l:tf6 21 .Jtxb7 and 4JdS) 19 cxdS 4JcS
1) 17 exd6 cxd618 4JdS (after 18 dS eS19 20 dxe6 .Jtxg2 21 ~xg2 'iVfS 22 'iVxfSlIxfS 23
4Jd4 4JcS 20 4Jc6 'YWd7 Black has an excel- 4JdS etc.
lent game) 18 ... iVd8 19 4Jf4, and 19 ... 'lif6? 17 4Jxes .Jtxes 18 dxe6 .Jtxg2 19 exf7+
20 4Jxh5! gxhS 21 4Jg5 is bad for Black, but
the position is unclear after 19 ... l:te8 20 4JgS
.Jtxg2 21 4Jxf7! 1i'f6! 22 4Jxd6 ~f3 23 4Jxe8
lIxe8 24 l:td3 .Jtg4 2S ~a4 iVf7 or 19 ... 4Jf6 20
4Jg5 .Jtxg2 21 'it'xg2 (21 4Jgxe6 fxe6 22 4Jxe6
is questionable in view of 22 ... "iVd7 23 'i£i>xg2
.l:tfc8 24 dS b5!) 21...iVd7 etc.;
2) 17 4JbS!. No one suggested this move,
which advantageously maintains the ten-
sion: after 17 ... dxeS 18 4JxeS .Jtxg2 19 ~xg2
4Jxe5 20 dxeS iVc5 21 'ilVe4 Mad8 22 f4 ~b4
23 'YWc2 White's chances are somewhat
better, while if 17 ... l:!.ac8 there is a choice
between the quiet 18 exd6 cxd6 19 dS eS 20 19 ...'i£i>xf7!
4Ja7 l:tc7 21 4Jc6 iVf6 22 4JgS 4JcS 23 a3 and This reply came as a big surprise to me.
the amusing 18 dS!? 4JxeS 19 4JxeS .JtxeS 20 I was expecting 19 ... l:txf7 20 'l\\Vxg6+ Mg7 21
232
The Third Match: 1986
~e6+ ~h7 22 ~f5+ (or 22 ~xg2 ~xh4 23 became slightly ill at ease: of course, the
~f5+ ~g8 24 ~e6+, but not 24l:th1? l:txg3+!) position is almost equal, but only almost...
22 ... ~g8 (not 22 ... ~h6? 23 ~xg2 ~xh4 24 The g6- and h5-pawns confidently restrain
l:th1 ~g4 25 l:txh5+! ~xh5 26 l:th1) 23 ~e6+ three white pawns on the kingside,
with perpetual check, but I underestimated whereas Black has the possibility of creat-
the capture with the king, having over- ing a passed pawn by ... c7-c6 and ... d6-d5.
looked Black's next move. In the first instance White must decide into
20~xg2 what type of ending to transpose - queen
or rook?
233
Kasparov vs. Karpov: 1986-1987
example: 23 fUel a4! 24 Md4 Mae8 2S Mf4+ If Black is still hoping for something, he
<J;;g7 26 Mxe8 Mxe8 27 <J;;f3 Mel! 28 Me4 Mb1 has to play ... d6-dS.
29 Me2 <J;;f6 30 <J;;e4 Md1 or 28 ... Ma1 29 cS
bxcS 30 Mxa4 l:i.b1 31 b3 Mb2 32 <J;;e3 <J;;f6
with some pressure for Black.
23 ... Mae8 24 Mfe1
28 ... Mf8
Guaranteeing ... that the game will be
adjourned! The thematic 28 ... dS 29 cxdS
would have quickly led to a draw in both
24 ... Mxe1 the rook ending - 29 ... cxdS 30 Mel, and the
The more flexible 24 ... MeS!? would have pawn ending - 29 ... MxdS 30 MxdS+ cxdS 31
led to roughly the same positions: 2S b3 f3 <J;;fS (31...d4+? 32 <J;;d3 <J;;dS 33 g4 <J;;eS 34
Mfe8 26 MxeS MxeS 27 <J;;f3 <J;;fS 28 Md3 Mel gS! <J;;f4 3S <J;;xd4 <J;;g3 36 <J;;eS <J;;xh4 37 f4
29 ~dS+ <J;;f6 30 l:td3. There are more prob- <J;;g4 38 fS and wins) 32 <J;;d4 We6 33 'it'c3
lems after 2S MxeS <J;;xeS 26 f3 (26 MdS+ (only not 33 g4? gS!) or 30 .. .'it'xdS 31 'it>f4! bS
<J;;e4!) 26 ... c6!? 27 'it>f2! (after 27 Me1+? <J;;d4! 32 'it'gS cS! 33 'it'xg6! c4 34 f4 (the pawn race
28 Me6 Md8 29 Mxg6 <J;;xc4 and ... <J;;b3 does not disrupt the equilibrium) 34 ...b4 3S
Black's pawns on the queenside are swifter fS c3 36 bxc3 b3 37 f6 b2 38 f7 b1~+ 39 'it'g7,
than White's on the kingside) 27 ... dS 28 also promoting to a queen.
cxdS cxdS 29 Md3, nevertheless holding on. While my opponent was considering his
25 Mxe1 Md8! move, I thoroughly checked these varia-
Black's idea is to play ... c7-c6 and ... d6- tions and realised that they were not dan-
dS, after which it will be easier to attack the gerous for White. I don't know if Karpov
weak white pawns on the queenside. also calculated them, but he preferred
26 :d1 c6 waiting tactics, although there is no longer
Black cannot get by without this move, anything for Black to wait for.
although the best chance was 26 ... Me8!? 27 29 f3 Mh8 (preventing g3-g4) 30 l:i.e1 Mb8
<J;;f3 MeS 28 b3 (28 Md4 Mel) 28 ... <J;;fS, forcing This is more of a psychological attack
White to allow the rook to go to el. This than a real one, but nevertheless White
would have been unlikely to bring Black must prepare for possible activity by the
success, but at least he would have retained opponent on the b-file - he drives the black
some hope. Now, however, the white king king to the kingside.
comes to the centre and all the danger is 31 'it>d3+ 'it'f6 32 Me4
over. This involves a little trap: 32 ... bS? 33
27 <J;;f3 <J;;e5 28 <J;;e3 cxbS cxbS 34 b4!, and unexpectedly the
234
The Third Match: 1986
235
Kasparov vs. Karpov: 1986-1987
was quite unrecognisable. I think that this usual unwillingness to engage in double-
happened partly because of the 16th game, edged theoretical disputes, cracked when it
which Kasparov conducted extremely recklessly, encountered an unprecedented, fundamen-
in the style of Tolush, but gained a pretty win. tally new approach by my opponent to the
Apparently after this he decided that he could solving of opening problems. In addition,
get away with anything ... This was one of the during the match I did not display suffi-
reasons for his subsequent poor play. There was cient flexibility and insight...
probably also another reason - but in any event, On the other hand, Karpov was no
Kasparov lost his composure. And here the longer superior in the concluding stage of
champion had to pass an exam in maturity - he the game and in purely technical positions,
had to demonstrate that he was a worthy cham- and at times he even encountered problems
pion. The exam was passed.' precisely in these situations. It can be said
Nikitin: 'In my view, this was Kasparov's that the mutual exchange of experience,
best match. He fought against an opponent in inevitable over such a prolonged duel,
very difficult circumstances, but he was so largely neutralised the difference in the
strong that nothing could stop him. Garry's playing methods resulting from the distinc-
play has become even more universal ... In a tion in the two players' styles. But for all
match where, of the first nineteen games at least the apparent parity of strength and possi-
ten can be considered chess masterpieces, it was bilities, Karpov nevertheless fell behind.
distressing that it should have been overshad- Objectively he was inferior to me in 'my'
owed by off-the-board games, initiated by people positions (this was the only match in which
supporting Karpov and trying to secure victory Karpov regularly went in for them with
for him at any price. ' both White, and with Black!), and he could
There is no doubt that in the period be- do nothing with me in 'his'. His intensive
tween the second and third matches Kar- preparations before the return match
pov worked incredibly hard and prepared proved insufficient compared with the
a mobile opening repertoire, aiming for thorough, systematic and wide-ranging
sharp theoretical debates during the course work carried out by my team over a period
of the match. With the intention of uphold- of several years.
ing his rights in all critical disputes, he Alas, in statements after the match Kar-
seriously expanded his range of playing pov again began referring to chance or bad
methods. I am convinced that all this luck, thus distorting the true picture. This
would have been quite sufficient to give was the reason for my open letter to him,
him victory in 1985, but by the return published in the magazine 64 (February
match I in turn had processed a great deal 1987):
of information and largely eliminated those 'I am appealing to you not in your ca-
defects which appeared in my previous pacity as editor-in-chief of 64, but as my
meetings with Karpov. The return match long-standing opponent in a number of
essentially broke the traditional stereotypes gripping encounters at the chess board.
of the two players' strong and weak points. Unfortunately, games from world champi-
After a careful study of the game com- onship matches often remain in history
mentaries it will become clear why the with superficial, instant commentaries,
Griinfeld Defence proved a failure. My new made in the heat of the struggle. This lends
opening, based on the conservative nature added importance to sensible commentar-
of Karpov's opening philosophy and his ies, made after the elapse of some time,
236
The Third Match: 1986
when passions have died down. Only a the play in that 24th game of the 1985 match,
detailed analysis, based on an impartial about which YOli write, was such that a win for
approach to the solving of chess problems, White would hnue been more logical than a win
can give a complete picture of the play. for Black'.
And here the decisive word belongs to the In the end, an analytical discussion did
participants themselves. not occur, but Karpov again made categori-
' .. .I should like to draw your attention to cal assertions about the unfair outcome of
radical disagreements in our chess assess- the 24th game, and hence of the match as a
ments, in particular regarding the 24th, whole (what White's chances were in
deciding game of the 1985 match. You have reality is shown by my detailed analysis of
repeatedly stated that you could easily the game in Kasparov vs. Karpov 1975-1985).
have won this game and thereby changed These words of the former champion
the course of chess history. For my part, I sounded strange after the return match,
have upheld the opposite point of view, which convincingly demonstrated the
one which, moreover, I have supported rightfulness of my victory in 1985. It is well
with analytical evidence. But you have not known that history, even chess history,
once found the time to give a concrete reply does not have a subjunctive mood.
to my objections. I am convinced that a So, I had now won a second match
creative duel in print between us will lead against Karpov, this time by the score 12 1/2-
to a sharp increase in the popularity of 11 V2, and for the first time I went ahead on
chess ... Therefore I am inviting you to an the overall number of wins in our matches
open chess discussion ... ' (10-9). Psychologically this was an important
Karpov's reply, in which he pleaded turning-point: I was no longer in any doubt
that he was exceptionally busy before his that I was a better player than my opponent!
Candidates super-final match with Andrey The margin could ha\'e been bigger, even
Sokolov, essentially rejected my offer of an crushing - up to the 17th game I was obvi-
open chess discussion on the 24th game of ously superior, and in general the character
the 1985 match, and in my words he eyen of the play was catastrophic for Karpov,
found a 'logical discrepancy'; 'YOli yOllr5e~f despite his numerous novelties. Everything
talk about the "picture of events". But the suggested that I was beginning to feel
events were in fact that White missed ,-'ery good comfortable as champion. I experienced joy
practical (but by no means analyticaiJ winning and relief, little suspecting that a year later a
chances ... It must be agreed that the course of genuine return match awaited me.
Chapter Three I
The Year between Matches
239
Kasparov vs. Karpov: 1986-1987
Campomanes exploited his moment of triumph distressed that the constant struggle for the
to the full: for the first time in the history of mythical chess crown had so damaged their
FIDE he became a paid President!' relations. '
This late-night campaigning did little to Having lost the fight in FIDE, I came to
help my chess. On the whole the Soviet the conclusion that what was needed was
team (Kasparov, Karpov, Yusupov and an independent organisation for grandmas-
Sokolov, reserves Vaganian and ters, capable of defending their rights and
Tseshkovsky) played unevenly, and after I steering chess on to the path of reform.
unexpectedly lost in the 8th round to Many grandmasters responded readily to
Seirawan, which led to us losing to the my suggestion that we should get together
USA, one of our main rivals, it looked as there and then in Dubai and discuss our
though we were in dire straits. Apart from urgent problems. And although we had no
the Americans, the steady and confident advance plan of action, we all agreed that
English team were also in serious conten- the deep conflicts that had come to a head
tion for the gold medals. Had we lost the in world chess needed to be resolved as
Olympiad, there would no doubt have been soon as possible. This applied above all to
recriminations against me for engaging in the burgeoning administrative apparatus of
politics rather than upholding the honour FIDE, which had almost completely
of Soviet sport. squeezed grandmasters out of the system
Even so, a powerful finish enabled the which ran chess. The bureaucrats pre-
Soviet team to snatch victory: 1. USSR - 40 vented the organisation of new events,
points out of 56; 2. England - 39V2; 3. USA - which with the wide support of Western
381J2. And by winning my last three games I sponsors and the press could have greatly
scored 8V2 out of 11 (+7-1=3) and simulta- promoted chess.
neously gained three gold medals: for the I remember that on the day of the 11th
overall team victory, for the best result on round I entrusted board 1 to Karpov and
board 1, and for the absolute best result in spoke at a meeting of the FIDE Central
the Olympiad. True, the metal out of which Committee. One of the officials asked
the medals were made bore no resemblance indignantly: 'What's all this about - today the
to anything precious, although Campo- grandmasters set up an association, tomorrow
manes boasted that the local Sheikhs had it'll be international masters, and then some-
allotted 7 million dollars to the staging of thing else for arbiters?!' I replied: 'These are
the Olympiad ... Playing, unusually for him, people who represent chess. But what you
on board 2, Karpov made a more modest represent, I don't know. It's of no impor-
score: 6 out of 9 (+-1-1=-1). tance whether or not you exist, but it's on
~ikitin: 'In Dlibai for the first time I was them that the development of chess in the
able to assist Garry at an Olympiad (the main world depends!' And indeed, at that time
trainer of our team was Geller). The 50e'iet tealll the clash between the FIDE bureaucracy
won the gold medals after a d~[fiClllt bll t e'ery and the grandmasters reached boiling-
dramatic fight. The world champiolJ defeated the point. The high-handed decisions by Cam-
Polish player Schmidt at a time whelJ all the pomanes & Co. had provoked great dissat-
other games in the tournament had already isfaction with many of my colleagues. The
concluded. In the team the two 'Ka's were situation was completely absurd and
quickly transformed from implacable opponents unacceptable.
into comrades-in-arms, and I continually felt After two days of heated discussion,
240
The Year between Matches
held in my hotel room, we drew up the tion's main tournament, the World Cup,
main points of our program (I remember was designed solely for the top players.
that they were written down on table- Therefore we later began trying to correct
napkins). It was decided to begin by organ- the situation as far as possible, and in
ising a World Cup, a competition consisting particular we staged a serious of mass
of six super-tournaments, in which the 24 qualifying tournaments for the second
strongest grandmasters would take part. In World Cup cycle.
the two-year world championship cycle The guarantee that our undertaking
planned by FIDE there was no room for the would succeed came in the form of active
Cup, and so one of our first demands was a assistance from Bessel Kok, the technical
return to the former three-year cycle. In a director of SWIFT, the banking computer
letter to the FIDE General Assembly we network based in Brussels. He was drawn
stated that 'to avoid further into chess by Timman, who was also ac-
misunderstandings between FIDE and the tively opposed to Campomanes. I first saw
leading grandmasters, and also to facilitate Bessel at a banquet to mark the end of the
the process of decision-making', we were London half of the return match. He ap-
setting up a Grandmasters Association, we peared there as a prominent sponsor, who
set out our program, and we invited the began his activities in the field of chess by
FIDE leadership to hold a joint meeting in brilliantly organising the SWIFT Tourna-
Brussels early in 1987. ment in April 1986, an event which he
While still in Dubai we set up a provi- suggested should be held annually as one
sional Board of seven grandmasters to of the stages of the World Cup. But the
carry out the preparatory work. I was main thing was that he gave the new body,
elected President, with Anatoly Karpov set up under Belgian law, a sound financial
and Jan Timman as Vice-Presidents, the and legal basis. One cannot but admire this
other members being Ljubomir Ljubojevic, experienced businessman's feel for a good
Lajos Portisch, Yasser Seirawan and John proposition: he was the first to see the
Nunn. It was these seven who, on 15 Feb- possibilities for the GMA, which initially no
ruary 1987, were to put their signatures to one took seriously, and he sensed that the
the documents signifying the official crea- new organisation would be able to influ-
tion of a new international organisation, the ence events in the chess world.
Grandmasters Association (GMA). It was Our appeal to the General Assembly
no surprise that the Board comprised only was essentially an offer to sign a peace
top-class grandmasters: the world cham- agreement and begin working together. At
pion and representatives of the six leading the time I said to Campomanes: 'You may
chess nations. This was a natural reaction to be able to defeat any opposition in FIDE,
the domination of FIDE by incompetent but you won't be able to defeat me!' And he
delegates from the lesser chess federations. seemed to be interested in our project - at
The uniting of the elite grandmasters any event he agreed to hold a special
was a necessary condition for forming the meeting of the FIDE Executive Council in
Association, but even while we were still in Brussels, to discuss the founding of the
Dubai it could be seen from the response of GMA and meet its leaders.
rank-and-file grandmasters that the 'second I think that Campomanes didn't believe
echelon' might be unhappy with the the grandmasters could create a strong and
Board's activities. After all, the Associa- independent organisation, assuming that
241
Kasparov vs. Karpov: 1986-1987
we wouldn't be able to agree on a united up, but then the war intervened ... And after
policy. So for the moment he didn't object the war FIDE took everything over.
to the GMA, or at least, he didn't take any I concluded my championship year by
action against it. The President was hoping winning the strong double-round OHRA
that it would function strictly within the Tournament in Brussels (December 1986): l.
framework of the International Chess Kasparov - 71/2 out of 10 (+6-1=3); 2.
Federation. After he won the presidential Korchnoi - 51/2; 3--4. Hubner and Nunn - 5;
election, Campomanes thought that the 5. Short - 4; 6. Portisch - 3.
emergence of a new professional body Nikitin: 'This was Garry's first competition
would actually be grist to his mill, as for three years which did not involve Karpov.
through it he would be able to establish Adapting to tournament play, where each day
contacts with sponsors and the press. It you are up against a new opponent, with his
should not be forgotten that the FIDE own style and his own opening repertoire,
leadership had lost all credence in the eyes demanded intensive preparation and a great
of sponsors, 'Nho had no desire to put up deal of hard work. Before the last round, Kas-
any further money. Besides, many sus- parov was leading the second-placed Korchnoi
pected that part of the revenue intended for by 21/2 points! And in the end his rating ap-
FIDE's coffers in fact ended up in the proached a new peak - 2750! This was a won-
pocket of Campomanes himself. derful happy ending to a difficult, nervy year.'
If FIDE had retained its power over My desire to publicise chess by all the
chess and been augmented by an active means at my disposal - games, books, new
body such as the Grandmasters Associa- forms of competition - soon led to me
tion, then without doubt both Campo- contesting an unusual rapid-play match
manes and FIDE stood to gain. But in with Nigel Short (London, 4-5 February
Dubai, when Campomanes tried rapidly 1987). This was the first promotional enter-
(there wasn't much time) to work out the tainment event of its kind at such a high
implications, he underestimated a few level (apart from some innovative matches
things: the ability of the grandmasters to involving David Bronstein in the 1970s).
unite, the changes in the USSR, my persis- We played in white and black dinner suits
tence and, above all, the energy of Bessel on the revolving stage of a rock club in the
Kok. Campomanes was slow to realise how London Hippodrome, and this spectacle
far-reaching were Kok's intentions with was broadcast on commercial TV - hence
regard to chess. the time control of 25 minutes each for the
An interesting detail: the first such entire game: this made a total of 50 min-
meeting of grandmasters took place fifty utes, and in a program lasting an hour this
years before this in the Amstel Hotel in left time for advertisements and commen-
Amsterdam, after the famous A VRO Tour- tary.
nament. The discussions were about creat- I won 4-2 (no draws) and at that time
ing a 'club of the eight strongest', the eight even in a bad dream I could not have
at that time being Alekhine, Capablanca, anticipated that FIDE would try to make
Euwe, Keres, Botvinnik, Fine, Reshevsky rapid chess an instrument in its battle with
and Flohr. Unfortunately, Alekhine and professional grandmasters. But just ten
Capablanca were not on speaking terms, days after the match, from the FIDE Execu-
and they had to take turns to attend the tive Committee in Brussels came a sugges-
meetings. A draft agreement was drawn tion that, along with the existing world
242
The Year between Matches
championship, another, parallel champion- ments, were Timman and Ljubojevic. And
ship should be held with a shortened time it was only when the GMA began function-
control (30 minutes for each player per ing that the new control finally gained
game), and that a special rapid-play rating acceptance, gradually changing the pace at
list should be set up. Fortunately, in 1987 which chess developed.
FIDE did not try to force matters, but All the indicators were that a new era
invited interested federations to first stage was beginning in chess, giving rise to the
experimental tournaments (it can truly be most optimistic hopes. By an amazing
said that 'the threat is stronger than the coincidence, the date that the Grandmasters
execution'). Association was launched in Brussels, 15
After travelling then to Brussels, I February 1987, was two years exactly to the
showed Kok and Timman a video re- day since Campomanes had stopped my
cording of the match with Short. I remem- first world title challenge. Thus the anni-
ber that their initial reaction was extremely versary of an unprecedented act of tyranny
negative, as it was such an unusual specta- in the chess world was marked by the start
cle. They said that this wasn't a serious of an organised resistance to the forces
game and it would degrade chess (this was which had been behind it.
also Botvinnik's conviction, but in the end Botvinnik: 'Back in 1921 Emanuel Lasker
nothing terrible occurred). And when my predicted that chess masters would create an
manager Andrew Page suggested organis- organisation to defend their professio1Jal i1Jter-
ing a series of rapid-play tournaments, ests. That which the second (['orld champion
Kok, Timman and many fellow- dreamed of has been put il1to effect by the
grandmasters rejected the idea - it had not thirteenth champion. KasparoI' will go down in
yet attracted the masses! And it was only history as the first President of the world-wide
several years later that everyone saw the Grmzdmasters Association. It is to be hoped that
wonderful prospects available to rapid tlze Association will not try to imitate the FIDE
chess thanks to the appearance of electronic leadership, but will use its prestige to establish a
demonstration boards ... iust order in the chess world.'
But at that time there were constant dis- Around then a joint meeting, unique in
cussions about shortening the time control. its way, of the leaders of the GMA and
This was indeed an urgent question: obvi- FIDE took place, as was described in the
ous progress in the study of the openings magazine 64 by Nikolai Krogius, the la-
and the analysis of adjourned games de- mentable head of the chess section in the
manded a reduction in thinking time and a USSR Sports Committee and also a FIDE
longer playing session without adjourn- Vice-President. Here is an extract from his
ments. Earlier, in the autumn of 1984, report:
during my first match with Karpov, the
FIDE Congress had approved a gradual An extraordinary meeting of the FIDE Ex-
transition from the classical five-hour time ecutive Committee took place from 13-15
control (40 moves in 2V2 hours with an February in Brussels. I should remind you that
adjournment, and then 16 moves in every during the 52nd FIDE Congress in Dubai
hour) to a six-hour control (40 moves in 2 (1986) a number of grandmasters led by the
hours plus 20 moves in an hour followed world champion Kasparov suggested forming a
by an adjournment). The initiators of this new 'Grandmasters Association' and staging a
sensible idea, which provoked great argu- World Cup competition.
243
Kasparov vs. Karpov: 1986-1987
The FIDE General Assembly authorised the based on a three-year cycle, but will take place
Executive Committee to meet promptly and in only two of these three years.
come to a decision regarding this suggestion. 2. The World Cup will consist of six super-
This was one of the main reasons for the urgent tournaments of 16 players each, chosen from the
meeting of the Executive Committee in Brus- overall number of24 players.
sels. The leaders of the Association (Kasparov, 3. The 24 players will be selected as follows:
Timman, Karpov, Ljubojevic, Seirawan, Nunn a) the world champion (Kasparov); b) the loser
and Portisch) met at the same time in the of the last world championship match (Karpov);
Belgian capital. On the evening of 15 February c) the participants in the last Candidates
the members of the Executive Committee and matches (Sokolov, Yusupov, Vaganian and
the Association held a joint meeting. Timman); d) the seven players with the highest
... Here is the text of suggestions by the average ratings from the four FIDE rating lists
group of grandmasters, distributed in Dubai of 1985-1986 (the last of these will be the
and later included in the official FIDE docu- January 1987 list); e) on condition that they are
ments sent to national federations: not already qualified on the basis of points a, b
and c; fJ players finishing in the first three in
I. Grandmasters Association. each of the three 1987 interzonal tournaments;
The players council [FIDE] should be trans- g) the two players with the highest ratings for
formed into the 'Grandmasters Association'. the 1987 interzonal tournaments (averaged over
1. The members of the Association will be all two years: 1986-1987), on condition that they
grandmasters with an Elo rating of not less have not already qualified from point f
than 2501. No more than 8 Soviet players will compete
2. Only the 24 World Cup participants will in the World Cup (not including the world
have the right to vote in the Association. champion) ...
3. The Association will have a grandmaster Each of the 24 players will take part in four
governing board consisting of seven players. of the six World Cup tournaments. Account
Initially the board will consist of Kasparov will be taken of the best three results. A result is
(President), Karpov and Timman (Vice- determined, firstly, by the place occupied in the
Presidents), Ljubojevic, Portisch, Seirawan and tournament, and secondly, by the percentage of
Nunn. Before the start of the World Cup, points scored. The method used to arrange the
elections for a new board will be held. Commu- 24 contestants in the six tournaments will be
nications with FIDE will be made via a secre- decided by the Association before the start of the
tary with an office in the FIDE headquarters. World Cup.
4. The Association will arrange a special de- III. World championship.
velopment fund, comprising a percentage of the 1. The world championship will be organised
World Cup prize fund. This fund will be used within a three-year cycle, to be combined with
by the Association in accordance with FIDE the World Cup.
policies aimed at helping developing countries. 2. In view of the large number of World Cup
5. All questions concerning the world cham- tournaments, the world championship will
pionship cycle, the World Cup, rules of play, the mainly be staged as matches.
awarding of titles and other matters directly 3. The Candidates matches will consist of 5
affecting the members of the Association must stages, to start in 1988 ...
be approved by the Association.
II. World Cup. (In the end they were reduced to four
1. The World Cup competitions will be stages:
The Year between Matches
1st stage. Matches of six games, 14 con- qualification, rules of play and others become
testants: the loser in the Candidates super- the prerogative of the GMA?
final match Karpov-Sokolov, as well as The staging of the World Cup involves gen-
Yusupov, Vaganian, Timman, 9 winners of erating a very high prize fund for these events.
the interzonal tournaments and a represen- Will this not lead to problems in financing
tative of the organising country; 'other' international tournaments, official FIDE
2nd stage. Matches of six games, 8 con- competitions and other 'normal' chess events?
testants: the 7 winners of the 1st stage and Why does the Association, which calls itself a
the loser in the world championship match; grandmaster association, want to unite by no
3rd stage. Semi-final matches of eight means all the persons who have this title, and
games; afford the right to vote only to twenty-four of
4th stage. Final match of 12 games. - them?
G.K.) I cannot help mentioning one more very
puzzling question: why is it again planned to
Schedule of competitions for 1987-1990: restrict the number of Soviet grandmasters?
1987: zonal and interzonal tournaments. (The answer is obvious: otherwise the
Karpov-Sokolov match. Match for the world World Cup would have been transformed
championship. into an open USSR Championship, which
1988: 1st and 2nd stages of the Candidates would hardly have delighted Western
matches. Three World Cup tournaments. players and sponsors. - G.K.)
1989: 3rd and 4th stages of the Candidates During the meetings in Brussels, the
matches. Remaining three World Cup tourna- grandmasters acquainted the Executive Com-
ments. mittee with the GMA's draft status. It is now
1990: 5th stage of the Candidates matches. envisaged that membership of the GMA will be
Match for the world championship. Zonal and open to all grandmasters who have gained this
interzonal tournaments of the next cycle. title in over-the-board competitions. However,
the matter of the rights and duties of members
... The members of the executive committee remained open (apart from payment of dues and
pointed out that these proposals raise a number the right to participate in elections to the
of serious questions, of fundamental importance Board).
for the development of the world chess move- ... In conclusion FIDE President Campo-
ment. Will not the staging of the Cup damage manes and Vice-President Gobash announced
the prestige of the world championship? (Major that the collaboration of FIDE and the GMA is
tournaments 'look' better than short matches, a matter requiring further study which will be
and victory in a series of tournaments may seem resolved later (at a session of the Executive
more impressive.) Also, the creation of a closed Committee in May 1987 in Bucharest and at a
circle of 24 will restrict the competitive and FIDE General Assembly in Spain in November
creative contact between them and other play- 1987). At the same time they commented that,
ers. How will this affect the development of in view of the proposals made and with the aim
promising young players? Why talk about a of continuing a constructive dialogue with the
World Cup, when players from Asia, Africa, leading grandmasters, in 1987 FIDE will begin
South America and Australia have practically carrying out its arrangements on the basis of a
no chance of taking part in it? What real rights three-year world championship cycle.
will the FIDE General Assembly have, if ques-
tions concerning the world championship, Subsequently, to all appearances, the
245
Kasparov vs. Karpov: 1986-1987
246
The Year between Matches
In the 9th round, in order to take first 2735 against 2710 on 1 January 1987, and
place on my own, I needed without fail to 2740 against 2700 on 1 July. On the number
beat my rival with White. Alas, things of match games between us (96), we had
didn't work out: Ljubojevic unexpectedly already long since overtaken the previous
employed the Slav Defence with ... g7-g6, record holders Bot\'innik and Smyslov (69),
and I responded badly. Black achieved a although as yet our great predecessors
comfortable game and drew easily. were leading on the total number of meet-
My strict teacher Botvinnik wrote about ings (104), since they had played each other
this: 'The world champion still needs to im- far more in tournaments that we had.
prove. This was confirmed at the finish of the Before our matches Karpov and I had
tournament in Brussels (1987), where Kasparov played only three serious games - two of
was competing with Ljubojevic. The culmina- them in the match-tournament of USSR
tion came in the 9th round, when Kasparov had teams and only one in a normal tourna-
White against his main rival. The game ended ment. Thus in Brussels the champion and
in a draw, and as a result Kasparov and Ljubo- the ex-champion were meeting in only their
jevic shared 1st place. This episode indicates second tournament (a unique occurrence!)
that the present champion has not yet achieved and for the first time with their current
the highest mastery in tournament play. status. When he was champion, Karpov
Meanwhile, Lasker (remember, among others, had carefully avoided such meetings, and
his famous game with Capablanca, St. Peters- now I had to try and surpass him - the
burg 1914), Capablanca, Alekhine, and the winner of SWIFT-1986 - not onl\' in a
author of these lines (I can point to my game match, but also in a tournament!
with Keres, Leningrad 1941) all demonstrated Therefore both the game and the entire
their ability to win "to order" at decisive tournament were of considerable signifi-
moments in a tournament. Yes, Kasparov still is cance, especially before the impending
able and needs to advance, but we shollid not be match in Se\·ille.
too severe on him, si11ce there are spots even 011
the sun ... '
But let's see what happened then in Game 25
Brussels. In the next, penultimate, round I A.Karpov-G.Kasparov
had Black against Karpov, who was a point Brussels 22.04.1987,
behind me and Ljubojevic. For the ex- 10th round
champion this was the last chance to inter- English Opening A31
fere in the fight for 1st place: he needed to
beat me (and then also Ljubojevic) in his
individual games. But for me, by contrast, it 1 d4 ct:Jf6 2 c4 g6
was essential not to lose - and not only Although, in the words of Akimov, six
because of the tournament rivalry, but also months earlier 'Karpov already knew how
to avoid spoiling the picture of our long- to breach the Grtinfeld Defence' (3 ct:Jc3 d5 4
standing duel, in which I was then leading ct:Jf3 ~g7 5 'iVb3 - Game Nos.lS, 17, 19, 42,
by 'plus one'. 48), my choice of opening was clear and
Before my hundredth game with Kar- principled.
pov (not counting the very first one, in a 3 g3
simul') my rating was already definitely In view of my opponent's ponderous
higher than that of my historic opponent: play in this tournament, I assumed that he
247
Kasparov vs. Karpov: 1986-1987
248
The Year between Matches
ing practically 'at sight'. Black could easily was no equivalent alternative. Perhaps he
have equalised with a continuation which I was considering 8 ct:Jb5, which after
pointed out in Informator No.43: 6 ...b6! 7 8 ... 'ilVa5+?! 9 ~d2 'ilVd8 10 cxd5! ct:Jxd5 11
~g2 ~b7 transposing into a good variety of ~c3 (11 ct:Jlc3!?) 11 ... ct:Jxc3 12 'ilVxd8+ ~xd8
'hedgehog'. For example: 8 0-0 (8 ~xb7 13 tLllxc3 a6 14 0-0-0+ tLld7 15 ct:Jd4 prom-
'ilVxb7 is no better) 8 ... ~xg2 9 'it'xg2 ~g7 10 ised White somewhat the better chances,
ct:Jc3 0-0 11 ~b2 ~7+ 12 ct:Jf3 (if 12 f3, then but 8 .. :~d7! 9 ~b2 0-0 10 tLlla3 a6 11 tLld4
apart from 12 ... d6, also 12 ... d5 is possible) b5 would have led to excessively compli-
12 ... d6 13 'it'gl ct:Jbd7 14 .i:tbl .i:tfc8 15 ~al cated play.
.i:tc5! 16 'ilVd3 .i:th5! 17 .i:tfdl ct:Jc5 with equal- 8 ... ct:JxdS 90-0!
ity (Neverovsky-Arbakov, Smolensk 1991). Of course, not 9 ~xd5? 'ilVa5+. For the
7 ~g2 dS! (16) moment White plays correctly - he com-
A committal and now forced advance, pletes his development, following the
which demanded accurate calculation and motto 'clarity and simplicity'. Black must
resolution. While the white king is in the think how to justify his risky experiment.
centre Black must try and exploit the
weakness of the long diagonal and the fact
that the knight on d4 is inadequately de-
fended. After the routine 7 ... 0-0 8 O-O! (8
ct:Jc3 d5! - d. the note to Black's 3rd move)
followed by ct:Jc3 and ~b2 the situation
would have stabilised to White's advantage
and the ... 'ilVc7 manoeuvre would have lost
its point.
At the board I was almost carried away
by the 'unclear' check 7 ... 'ilVa5+ with the
idea of 8 ~d2 'ilVc5, but after 8 ~·d2! "ihd2+
9 ~xd2 ct:Jc6 10 ~c3 White has a stable
positional advantage. 9 ...... d7?! (It')
At the time I was proud of this awk-
ward move. Of course, it was not very
aesthetic to block the path of the c8-bishop,
but I thought that without exploiting the
pin on the d4-knight Black could not disen-
tangle himself. It was bad to play 9 ... e6? 10
~a3 or 9 ... 0-0? 10 ~xd5 'ilVd7 11 ct:Jc3 e6 12
~a3 etc., while if 9 .. :iVc5 10 ~b2 ct:Jc6 White
has 11 ct:Je6!, and after 11 ... ~xe6 12 ~xg7
.i:tg8 13 ~b2 0-0-0 14 'ilVc1! the two bishops
leave him with a small plus.
Annotating the game in Informator, I also
judged White to have the advantage after
8 cxdS (24) 9 ... ct:Jc6(!) 10 ~b2 ~6 (10 ... ct:Jxd4?! 11 ~xd4
It was strange that Karpov thought for ct:Jf6 12 'ilVc1!) 11 ~xd5 ct:Jxd4 12 e3 (12 ct:Jc3
so long over an obvious move, when there ~h3) 12 ... e6 (12 ... ct:Jf3+? 13 'ilVxf3 ~xb2 14
249
Kasparov vs. Karpov: 1986-1987
'iVxf7+ ~d8 15 .l:!,dl ~c7 16 ttJc3! and wins) stroke, which gives White a clear advan-
13 ttJc3 0-0 (13 ... ttJf5?! 14 ttJa4) 14 ttJa4 'iVa5 tage: 13 ... 'iVxb5 14 ttJc3! .i.xc3 (14 ... 'iVa5 15
15 .i.g2, but later I found the accurate reply ttJxd5) 15 'iVxc3! ttJxc3 16 'uxdS+ ~g7 17
12 ... .i.e6! 13 .i.xe6 (forced: 13 .i.g2? ttJf3+) .i.xc3+ f6 18 'uxc8 'iVd7 19 ,uc1! or 13 ... .i.xb2
13 ... ttJxe6 14 .i.xg7 ttJxg7 15 ttJc3 0-0 16 ttJd5 14 'iVxb2 .i.d7 (14 ... 'iVxb5 15 ttJc3) 15 a4 a6 16
(if 16 'iVd7, then 16 ... ,Ufc8! 17 ttJd5 'iVd8 with .i.xd5 axb5 17 ttJc3 etc.
equality) 16 ... 'iVd6, and Black, having Therefore Black must urgently remove
simplified the position, is very close to a his knight from d5:
draw.
This was the logical solution, mobilising
the pieces as quickly as possible. By con-
trast, 9 ... 'i'd7 holds up Black's develop-
ment, although it does create the threat of
10 ... 0-0 and ... :dS.
10 li.,b2 0-0
The critical position of the opening, and
a very unusual one. White hJS to solve the
problem of the pin on his d4-knight. A
queen move suggests itself ...
Analysis diagram
250
The Year between Matches
13 ctJxc6 (11)
Without thinking for long, Karpov satis-
fied himself with the simplest decision,
promising him a favourable endgame,
although White also had other tempting
possibili ties:
1) 13 ctJa3 .i.xd4! (an unexpected ex-
change) 14 .i.xd4 ctJdb4! (not 1-l ....~'g-l 13
~b2!) 15 ~c5 (13 e3 '~'g-l) 1:; ....~·xdl+ 16
'iVxdl ~xdl+ 17 :'xdl .i.e6 (it is dangerous to
12 ... ctJc6! (19) play 17 ... ctJxa2 18 .i.xc6 bxc6 19 :'d8+ ~g7 20
The only move (Black must aim for sim- :'e8!) 18 :'d2 - in Illforllll1tor I assessed this
plification) and a bold decision: now I am position as '±', but after 18 ... a3! 19 ctJb5 a4
left with a weak pawn on c6, and Karpo\' Black is sa\'ed bv a clever tactic: 20 bxa4
was a master at exploiting such weak- :'xa-l 21 a3 :'a:;! 22 :b2! .kc4! 23 .kxb4 ctJxb4
nesses. But I was relying on the dynamic 2-l ~d-l ~ZJd3 25 ':'xb7 ~xa3 with equality;
resources of the position. 2) 13 ctJb5!? .kxb2 (13 ...1IVe6? 14 .kxg7
In the magazine 64 Nikitin recom- ~xg7 15 ctJla3! is bad for Black, as is
mended 12 ... 1IVe8. In the event of 13 e-l ~b-l, 13 .. .'~Jf4? 14 ~xd7 ctJxe2+ 15 'it'f1 ctJxc1 16
13 a3 ctJb6 or 13 ctJc3 ctJxc3 14 .i.xc3 :~c6 Ei ~xd8+ ctJxd8 17 .i.xg7 'it'xg7 18 ctJlc3, trap-
ctJb5 ~xdl+ 16 1IVxdl .i.xc3 17 ~xc3 .i.f3 ping the knight: 18 ... ctJd3 19 ~dl .i.f5 20 g4
Black does indeed have equal chances. But ctJb2 21 ~d2 .i.d3+ 22 'it'el a6 23 ctJd4 ~c8 24
13 ctJa3!?, with the idea 13 ... ctJa6 14 ~db3, is ctJde2 and wins) 14 1IVxb2 ~f5!' The only
far more unpleasant for him, and the im- move, to which there are two good replies:
mediate 13 ctJb5! (as in the variation with 11 a) 15 ctJlc3 ctJxc3 16 ,*,xc311Vxb5 17 .i.xc6
1IVd2) is even better: 13 ... i.,xb2 14 ''ilixb2 'i¥b6! (17 ... ~xdl+ 18 ~xdl 'i¥b6 is dangerous
1IVxb5 15 ctJc3 1IVe8 16 ctJxd5 ctJa6 17 'i'e5! in view of 19 .i.d7! .i.xd7 20 ~xd7 'iVe6 21
i.,e6 18 ctJf4 .i.c8 19 h4, or 13 .. .''ilixb5 14 'iVc7 'iVxe2 22 'iVxb7 ~e8 2311Vxa7 or 19 .. .''ilid8
.i.xg7 'it'xg7 15 ctJc3 1IVc5 16 ctJxd5 '~'xc1 17 20 ~d3! .i.xd7 21 'iVd2 and ~hd7) 18 ':'xd8+
~axc1 ctJc6 18 ctJc7 ~xd 1+ 19 ~xd 1 ~b8 20 1IVxd8 19 .i.f3 'i¥b6 and ... i.,e6, with a
.i.xc6 bxc6 21 ~d8 with an obvious advan- slightly inferior, but perfectly defensible
tage to White. position;
251
Kasparov vs. Karpov: 1986-1987
252
The Year between Matches
18 h3?! (04)
Again uncertain play. 18 ttJc3 (18 J::txc6?
ttJb4 19 J::tc1 J::tac8 etc.) was more accurate
and natural: 18 ... ttJxc3 19 J::txc3 J::td2 20 J::te3,
winning a pawn in a more comfortable 22 :ccl (to parry the mate threat, White is
situation - after 20 ... 'it>f6 21 ii.xc6 :c8 22 forced to retreat his on1\' acti \'e piece)
.i.e4 J::tb2 23 'it>el il..e6 Black has sufficient 22 ...:b2 23 :cbl
compensation, but not more. The only sensible defence and at the
same time an obvious one - the exchange of
a pair of rooks.
23 ...J::tdd2
Exploiting the fact that White has
played h2-h3, 23 ... J::tc2!? was possible with
the idea of 24 .i.e4 (24 J::tc1 J::tb2 is equal)
24 ... .i.xh3+ 25 'it>el J::tc5 26 J::tc1 J::tdc8, when
the game would quickly end in a draw:
White would still have had the possibility
of creating a passed pawn on the queen-
side, but Black, having regained his pawn,
could now have created his own passed
pawn on the kingside.
18 ... .i.e6 24 J::txb2 J::txb2 2S 'it>el .i.fs 26 ~dl gS!
Now 18 ... a4?! has no point in view of 19 Black has greatly activated his forces
b4! (but not 19 hxg4? axb3!) 19 ... .i.e6 20 a3 and he invites the opponent to declare his
'uac8 21 ttJc3 ttJb6 22 ttJe4, and White's intentions. 'Despite being a sound pawn to
253
Kasparov vs. Karpov: 1986-1987
the good, Karpov's position began to give The black king is aiming for c3, and
cause for alarm because of the hopelessly White must prevent this.
passive placing of his pieces. For the first 30 a4 ~e5 31 ~c1 Wd4
time the question half-seriously arose: can
White hold on?' (Nikitin)
254
The Year between Matches
.ii.c6 f5 38 e3 ~h4 39 gxf5 .ii.xf5 with a draw allows Black some hopes of success.
is correct. 41 ... 'it'C4 (03) 42 :'b7 (08)
36 ...~h2! (energetically attacking the white The critical position of the concluding
pawns) 37 f3 phase of the game.
255
Kasparov vs. Karpov: 1986-1987
57 ~g8!, eliminating the pawn avalanche. ment was the quiet, somehow inconspicuous
43 i.e8! (11) 43 ...~a1 44 i.xf7 i.xf7 45 ~xf7 play of Anatoly Karpov', Nikitin wrote in the
~xa4 magazine 64. 'He flew to Brussels after his
Times: 2.23-2.28. 'Such rook endings match with Andrey Sokolov, enjoying the status
cannot be won, and the rest does not re- of unequivocal winner of the Candidates super-
quire either commentary or a record of the final. However, this original way of relaxing
times.' (Nikitin) after a tense match once again demonstrated
46 ~g7 ~a5 47 e3 ~a2+ 48 Wei e5 49 ~xg5 that there is a limit to human resources. In
e4 50 ~g8 'it'd3 51 ~b8 ~e2+ 52 'it'fi ~c2 53 Brussels the ex-champion displayed everything
'it'e1 Yz-Yz chess players call poor form: he would already
begin to tire in the fourth hOllr of play, he
I have to admit that Karpov's uncertain avoided the slightest risk, he made oversights
play in this game disorientated me and and he did not playa single game of which he
created a rather bad psychological effect could feel proud. Only his enormous chess
before the match in Seville. I stopped prestige, which caused trepidation among most
treating him with proper respect, I ceased of iJis opponents, enabled him to take clear 3rd
fearing even his favourite positions where place.'
his opponent had chronic weaknesses, and The day after the conclusion of the main
I came to believe that thanks to dynamics I tournament, the organisers staged a 'world
would be able to extricate myself from any blitz championship' - a grandiose double-
scrape. This was a dangerous delusion. round event. Bessel Kok clearly wanted to
When I was making my last few moves, see as many battles as possible between the
I was already thinking about how to catch two 'Ka's: to increase the competitive
up with Ljubojevic, who that day defeated intrigue he incorporated an additional
Larsen and was leading the race - 8 out of match of two games between the players
10! In the last round he drew with Karpov, who finished first and second. The compo-
but I managed to win 'to order' a good sition of the tournament was also strength-
game against Tal and to catch the leader. ened, and it became an exclusively grand-
The results of the tournament were as master event. The Belgian masters Winants
follows: 1-2. Kasparov and Ljubojevic 81/2 and Meulders were replaced by Sosonko
out of 11; 3. Karpov - 7; 4-5. Korchnoi and and the convalescent Hubner. The twelve
Timman - 6V2; 6. Tal- 6; 7. Larsen - 5 1/2 etc. contestants played two successive games
By finishing one and a half points ahead of against one another, with White and with
Karpov, I gained revenge for the 1981 Black. Instead of drawing lots, the players
Moscow 'Tournament of Stars'. were arranged in order of rating, with No.1
As for what Botvinnik said about me be- on the rating list playing first against
ing unready to win decisive games, I No.12, then No.ll, and so on. Thus Karpov
quickly 'corrected' this by winning in the and I were to meet at the very finish.
final round not only against Tal in Brussels This blitz marathon began more than
1987, but also the 24th game against Kar- well for me: in the first six mini-matches I
pov in Seville 1987 and against Spassky and scored 11 points out of 12 and practically
Illescas at the Barcelona stage of the World ensured I would finish in the first two.
Cup in 1989 (where I again shared first After that I relaxed somewhat and reduced
place with Ljubojevic). the tempo. Before the last round the leading
'The main surprise of the SWIFT Tourna- trio was Kasparov - 15V2 out of 20, Timman
256
The Year between Matches
- 13, Karpov - 12. Clearly there was hardly - Game No.45 in Volume III of My Great
any chance of an additional match with Predecessors). The usual move is 9 e4 (Game
Karpov, since it would have taken a miracle NO.I01 in Revolution in the 70s).
for him to overtake Timman. 9 ••• 0-0 (9 ...'iVc8!? 10 i..b2 c5 Kasparov-
Due to our weak motivation and accu- Ehlvest, Belfort 1988; Timman-Karpov,
mulated fatigue, my two games with Kar- Tilburg 1988) 10 i..d3 h6 11 e4 c5 12 0-0
pov were of very poor quality - even by ~a6
blitz standards. We both made absolutely A minor error: nowadays 12 .. :iVc8! 13
horrendous blunders! There is no sense in 'iVe2 ~a6 is considered to be more accurate.
seriously annotating such play, but some After 12 ... ttJc6 13 i..b2 l:tc8 14 'iVe2 ttJa5 15
points are worth mentioning. l:tadl cxd4 16 cxd4, on the other hand,
White has some advantage (Ribli-Hort,
Germany 1991).
Game 26 13 i..f4 cxd4 (if 13 ...'iVc8 White creates a
G.Kasparov-A.Karpov passed pawn by 14 i..xa6 and 15 d5) 14
Blitz Tournament, i..xa6! (Karpov was hoping for 14 cxd4
Brussels 1987 'iVc8) 14... ttJxa6 15 cxd4 'iVc8 16 'iVe2 'iVb7
Queen's Indian Defence E12 17l:tfdl
Here Black is further from equality than
in the game Gelfand-Lautier (Monte Carlo
1 d4 ttJf6 2 c4 e6 3 ttJf3 b6 4 ttJC3 i..b7 (rapid) 1999), where instead of ... h7-h6
(avoiding 4 ... i..b4 - Game No.IS) 5 a3 d5 6 Black had played ... :ac8.
cxd5 ttJxd5 7 'iVc2
One of the critical lines of the Petrosian
Variation, which I employed in the memo-
rable 32nd game of the 1984/85 match.
7... ttJxc3 (7 ... ttJd7?! - Game No.36 in Kas-
parov vs. Karpov 1975-1985) 8 bxc3 J..e7 9 e3
17 ... l:tfe8?!
Too slow. 17 ... :fd8 18 l:td3 ttJc7! 19 ttJe5
ttJe8 was correct, strengthening the king-
side defences. It is interesting that Karpov
prepares to parry d4-d5, but instead of this
I build up an attack on the king.
In order after i..d3 to provoke ... g7-g6 or 18 ttJe5 l:tad8 19 ~d3 i..f6 20 ~adl itJb8?!
... h7-h6 in the hope of exploiting the (this loses, but it is also hard to recommend
weakening of the black king's defences (I 20 ... i..xe5 21 i..xe5 f6 22 ~f4) 21 itJg4! i..g5
carried out a similar plan against Petrosian 22 i..xg5 hxg5 23 e5 'iVe7 24 'iVe4?!
himself - Game No.45 in Volume III of My
257
Kasparov vs. Karpov: 1986-1987
38 ct:Jg5??
We each had about 20 seconds left, and
instead of 38 ct:Jd6 with a favourable end-
game I suddenly placed my knight en prise,
having for some reason decided that Black
would have to reply 38 ... J:!.e8.
38 ... ct:Jxg5 39 J:!.C3 J:!.xd4 40 J:!.c8+ ~g7 41
J:!.C7+ ct:Jf7
258
The Year between Matches
lost on time (1-0). Alas, he did not manage energetic, as later occurred in many games,
to capture the last white pawn, which including my 'advanced chess' match with
would have ensured him at least a draw. Topalov (4th match game, Leon 1998).
6 ... CDc6 (a more venomous move than
The second game, which no longer had 6 ... 'lixd2+ 7 CDbxd2 cxd4 with equality,
any competitive significance, followed a Akopian-Ivanchuk, Linares 1995) 7 d5?!
similar pattern: again I quickly gained a big 7 'lixa5 is safer (Tukmakov-Kupreichik,
advantage - and again I began to 'flounder' Klaipeda 1980; Smyslov-Vaganian, Sochi
in the last minute of play. 1986).
7... CDe4! 8 'lixa5 CDxa5 9 CDa3 CDd6 10 CDd2
l:tb8
Game 27
A.Karpov-G.Kasparov
Blitz Tournament,
Brussels 1987
King's Indian Defence E60
6 'lid2
If 6 i1I..d2, then 6 ...'iib6 with equality (An-
toshin-Adorjan, Budapest 1973), but the
line 6 CDc3CDe4 7 'lid3! cxd4 8CDxd4 is more Allowing White to free himself some-
259
Kasparov vs. Karpov: 1986-1987
40 ... i.e4+?
Extreme tension! There was a simple
win by 40 ... ttJxc3+ 41 bxc3 i.c4+ or 41 'it'el
ttJa4 42 Irxf7+ 'it'h6.
41 'it'e3 .ii.xf1?? (a blunder; 41...fS! would
2S i.e3? still have left Black with every chance of
Missing the chance 2S b3!, for example: winning) 42 'it'xe4, and in view of the threat
2S ... 'it'g7 26 i.xd3 cxd3 27 'it'f2, and accu- of 43 dS+ (a discovered check by the 'stu-
racy is now required of Black: 27 .. JHc8 28 pid' bishop on c3!), it is now time for Black
'it'e3 Itc2 29 Ita 1 Itbc8 30 Wxd3 Itb2 31 Itf2 to resign. But - time, time! Yz-Yz
gS with equality.
2S ... ttJb4 (the knight returns home) 26 a3 Nikitin: 'These two games showed that blitz
ttJe6 (I somehow did not believe in 26 ... ttJa2 is not a serious game, and that events of this
27 i.el) 27 'it>f2 ttJa7 (27 ... f6!7) 28 Itfe1 Itfe8 type should also not be taken seriously'.
29 Ite2 (29 i.b4! ttJbS 30 'it'e3 was more Results of the experimental SWIFT
accurate, with equality) 29 ... ttJbS World Blitz Championship: 1. Kasparov -
And here again there began some horri- 17 out of 22; 2. Timman - IS; 3-4. Karpov
ble blunders by both sides .. . and Ljubojevic - 121f2; S. Hubner - 12; 6-7.
30 'it'e3? (30 Ircc1) 30 ... ttJd6? (30 ... ttJxa3 Korchnoi and Short - 11; 8. Tal - 10lh etc. In
wins) 31 'it>d2 hS 32 Itbe1 ttJfS 33 g4 hxg4 the additional mini-match with Timman a
34 fxg4 ttJd6?! (after 34 ... ttJe7 Black's draw would have satisfied me (thanks to
chances are better) 3S 'it>e3?! my first place in the tournament), but,
3S i.b4 ttJe4+ 36 'it>e3 was correct. summoning my remaining strength, I
3S ... Irb3 (3S ... ttJe4!7) 36 i.d1 'it>g7 37 Itf2 nevertheless won: 11/2- 1/2.
Ith8!? (not a bad idea, giving Black power- Before his next match with me, Karpov
ful play for the exchange) 38 i.xb3 exb3 39 followed his rule of not taking a break from
260
The Year between Matches
play, and he took part in two more tour- win. It was not surprising that Karpov's
naments: in Amsterdam (Euwe Memorial, preference was for Abu Dhabi, the capital
May 1987: 1-2. Karpov and Timman - 4 out of the UAE. The alliance was still alive and
of 6; 3. Korchnoi - 2lh) and in Bilbao (June well.
1987: 1. Karpov - 7 out of 9; 2. Andersson- Since it was well known how I felt about
6112; 3-4. Chiburdanidze and Ljubojevic - Abu Dhabi, there was altogether no point
5lh; 5-6. Su.Polgar and A.Sokolov - 4V2). in stating my choice. In principle, any city
For my part, I limited myself to another would have suited me except Abu Dhabi.
interesting experiment - a televised match- My first choice would have been Seattle.
tournament of three USSR teams (Moscow, Given the clear improvement in relations
8-10 May), with a time control of 45 min- between our countries, it would have been
utes for a game. Leading the 'young gen- highly symbolic for two Soviet chess play-
eration' team, I scored 21/2 out of 4 against ers to contest the world title in America.
Beliavsky and Tal, and my team also won. Such a match would also have done much
During this match-tournament, for the to promote chess in the United States, a
first time in history - a sign of the times! - potentially major market for the game.
nearly fifty Soviet grandmasters assembled But the final choice was Seville, which
for a meeting chaired by Botvinnik. Apart made a strong application in connection
from me, among those who spoke were Tal, with preparations for the World Fair EXPO-
Geller, Polugayevsky, Razuvaev, Psakhis 92. I had been very well received there
and many others. In a unanimously agreed when I went on a familiarisation \'isit in
resolution, those present stated that 'the August 1987. Spain is a chess-playing
structure of the chess movement in the country country with many lovers of the ancient
is out of date and requires radical reform', they game, and it fully desen-ed the right to
approved 'the participation of world champion stage a ,\-orld championship match.
Kasparov and ex-world champioll Karpo(1 ill the Prior to the match in Seville there was
international Grandmasters Associatioll', and an ob\-ious regrouping of forces. In early
they recommended that their colleagues 1987, the Sports Committee and those of its
should join this organisation on an indi\-id- senior officials \\'ho had previously toler-
ual basis. A Council of Soviet grandmasters ated me, sharply changed their attitude.
was created, comprising Kasparoy, Bot Yin- The reason was clear: by speaking out in
nik, Karpov, Alexandria, Beliavsky, Tuk- Dubai against the election of Campomanes,
makov and Yusupov. I had become the first Soviet sportsman to
Meanwhile FIDE was deciding the mat- openly defy a directive from the Sports
ter of where to stage the world champion- Committee. Not content with that, three
ship match that autumn. Again, as in 1983, months later, at the Trades Union Con-
the United Arab Emirates cropped up. The gress, I had spoken from the rostrum of the
Olympiad in Dubai had left no doubt in my Kremlin Palace of Congresses of the need to
mind that counting on the hospitality of the introduce professional sport in the USSR as
organisers there, who still openly sup- the only means of saving the Soviet sports
ported Campomanes, was out of the ques- movement from collapse. I had thereby
tion. New FIDE rules stated that players opposed the official stance of the Sports
could each now choose only one city, and if Committee, which categorically rejected the
they disagreed, then the city offering the possibility of introducing the status of
highest prize fund would automatically professional sportsman in our country. It
261
Kasparov vs. Karpov: 1986-1987
was a question of giving sportsmen, who have a marked advantage in the analysis of
were humiliatingly dependent on the adjourned games.
sports officials, their legitimate rights and, Early in June, four months before the
in a broader sense, of abolishing the State match, I was once again reminded that in
monopoly over the individual. The express- my encounters with Karpov for the world
ing of such' anti-Soviet' ideas immediately championship title it was not only the black
increased the number of my enemies in the and white pieces that took part - the battle-
establishment. field extended far beyond the bounds of the
In FIDE it was also understood that it chess board. My opponent cast doubts on
was no use hoping for peaceful co-existence whether Iosif Dorfman, at that time the
between me and Campomanes. only grandmaster among my helpers, could
Thus the opening of a second front in continue working in my team.
chess and in sport led to the creation of a How this occurred is described in the
powerful coalition, whose members be- service reports of Viktor Litvinov - internal
lieved it essential to put a stop to Kas- KGB documents, which came to light years
parov's 'destructive activities' as soon as later in a sensational publication by Vitaly
possible. Melik-Karamov, 'A trap for the champion'
(in the magazine Ogonyok, 1993, No.28).
'A Trap for the Champion' Report (June 1987): 'On 1 June 1987 a
And so my new match with Karpov was to stranger from Moscow phoned Kasparov at the
start on 10 October 1987 in Seville. At the Zagulba sanatorium and stated that he had
end of May, after returning from Barcelona, some serious information, which Kasparov
where I had been presented with what was should know about in his preparations for the
now my fourth chess 'Oscar', I began match with Karpov. In connection with this he
preparing for the match. The training wanted to travel to Baku and meet him. A few
sessions were held, as usual, close to Baku, days later Kasparov was visited by Alexander
at the Zagulba sanatorium on the banks of Petrovich Feldman (registered at this address:
the Caspian Sea. ... ), who informed him that for a considerable
After the 1986 match a difficult problem time he had been on friendly terms with Karpov
arose with my group of trainers, since, with and had carried out various "delicate" assign-
the departure of Timoshchenko and Vladi- ments for him, including some at the time of the
mirov, only three remained of those with matches for the world chess championship in
whom I had begun the struggle: Nikitin, 1984 and 1985 in Moscow. But subsequmtly
Shakarov and Dorfman. Therefore in March Karpov had shown a lack of gratitude by failing
I entered into discussions with the young to fulfil his promises ... '
Georgian master Zurab Azmaiparashvili What was it that the uninvited visitor
(within a year he became a grandmaster), revealed? It turned out that a leak of infor-
who immediately joined our preparation mation from my camp had been arranged
work and, in particular, helped me to right from the start of the first match! This
master a new match opening - the English news left me in a state of shock!
Opening. Even so, the team still lacked a Report (from the words of Feldman):
strong grandmaster. At the end of August I 'Shortly before the Karpov-Kasparov match in
approached Sergey Dolmatov, and, fortu- 1984, Dorfman (who had only just agreed to
nately for me, he agreed to help. Sergey's work in my team - G.K.) said to his friend
arrival ensured that in Seville we would Pyotr Marusenko, a post-graduate at the
262
The Year between Matches
Philological Faculty of Kiev University, that Game' in Kasparov vs. Karpov 1975-1985).
during the match he wanted to play on the After this all the members of my team left
totalizer, using the information that he had, and for their various homes. The next training
asked him to help. Marusenko, not being an session was planned for April.
expert on these matters, approached the Presi- Report (from the words of Feldman):
dent of the USSR Draughts Federation Vadim 'On 6 March 1985 Feldman travelled to Lvov,
Bairamov, a close friend of Karpov. From to come to an agreement with Dorfman about
Bairamov this information reached the Karpov further collaboration. During the discussions
camp. It was here that Feldman came on to the Dorfman agreed to hand over information and
scene; he had known Dorfman since childhood requested 100,000 roubles for it. It was sug-
and knew of his liking for games of chance. He gested that the delivery of information and the
agreed to implement Dorfman's idea of playing corresponding payment should be divided into
on the totalizer. Their first conversation took three stages: delivery of material relating to the
place in the Hall of Columns during the 4th preparation and course of the first match
game of the 1984185 match. Dorfman himself (1984185); information about the results of
suggested the conditions and the volume of preparation for the second match; delivery of
information: the opening, sealed move, analysis information during the second match with
of adjourned games and information about time- Kasparov.
outs (the price for the information about each 'Between March and 1 June 1985 Feldman
game was to be 150-200 roubles). travelled three times to Lvov for discussions
'In the period between the 4th and the 11 th with Dorfman, accompanied by KGB official
games they met every day at noon at the corner Volod, who gave him instructions, provided
of the Rossiya Hotel (where the Kasparov team insurance, and for each meeting supplied a
was staying). Dorfman passed on the chess miniature tape recorder for recording the
information and received the money. Hardly conversations.
anyone in Karpov's circle had any suspicion of 'On 1 June 1985 at their next meeting in
Feldman's mission ... Lvov, Dorfman handed him 25 sheets of infor-
'When in the 11 th game Karpov made the mation about Kasparov's preparations for the
move 1 0,13, Dorfman realised that the latter first match (ideas, opening developments,
was fully informed (d. Game No.15 in Kas- analysis of adjourned games) and replied to 10
parov vs. Karpov 1975-1985 - G.K.), and he chess and 5 general questions, formulated by
demanded that Feldman should sharply increase one of Karpov's helpers. There were also ques-
the payment for his information. But Karpov, tions about Kasparov's trainers, with particular
who was then leading 4-0, did not consider it interest being shown in Vladimirov's personal-
necessary to meet this demand. Feldman's ity. Feldman paid Dorfman 7,000 roubles.
contact with Dorfman ceased, and the latter left 'At subsequent meetings Dorfman began
the team, going off to the USSR Champio11ship refusing to hand over information, demanding
First League in Tashkent. After Kasparov's win an advance of 100,000 roubles. The discussions
in the 32nd game Dorfman returned to the team went on until 5 August 1985, without the two
and resumed his contacts with Feldman ... ' sides coming to an agreement. However, a week
It will be remembered that on 15 Febru- before the start of the second match, on 27
ary 1985 the FIDE President Campomanes August in Moscow, Dorfman handed Feldman
ordered the termination of the first match the scores of two games from a secret Kasparov-
and announced that a new match would Dorfman training match (apparently, as a sign
start in September (d. the chapter 'The 49th of agreement to collaborate).'
263
Kasparov vs. Karpov: 1986-1987
There was an almost immediate re- be a practical joke, but Feldman said in all
sponse, and events began developing as in seriousness that he was ready to bring 50,000
a detective novel. What happened during roubles immediately, or the equivalent in
those July days of 1987 I learned not only foreign currency, and the remaining 50,000
from Feldman's detailed account, but also after the information was supplied. Dorfman
from Litvinov, who revealed to me that he took fright at this offer, he did not give a definite
knew about Dorfman's machinations reply, and since then he had begun avoiding
during the second match, but did not have meetings with Feldman. Then strangers had
the right to talk about them - such was the begun blackmailing Dorfman by telephone and
decision of his superiors. threatening to kill him.
Report (October 1985): 'On 1 October 'Rzaev did not greatly believe in the truth of
1985 (the day of the 11th game - G.K.) this story, but, with the aim of somehow relieving
Gennady Rzaev, the chairman of the Azerbaijan the situation around Dorfman and safeguarding
State Sports Committee (the leader of Kas- him from the pursuit of unknown crooks
parov's team), telephoned from Moscow to a (Dorfman was completely demoralised by fear
KGB officer in Baku, and reported that during and his help in the match had been reduced to
the match Kasparov's trainer and second, Iosif zero), he suggested that he should meet Feldman
Dorfman, who was in an excitable and fright- and make out that he was agreeable to passing
ened state, had come to him and stated that information to Karpov, but on condition that the
strangers were phoning him in his room, latter should arrange through his contacts to help
threatening to kill him, and demanding that he him obtain a residence permit for Moscow. The
should stop helping Kasparov. In addition, in thinking was that even the all-powerful Karpov
the city and in the Tchaikovsky Hall he was and his highly-placed patrons would be unable to
constantly being shadowed by suspicious fulfil this condition at short notice. Time would
people. pass, the match would end, and thus Dorfman
'During the course of an extensive conversa- would be able to escape from the situation. On 28
tion between Rzaev and Dorfman, about what September 1985, following Rzaev's recommenda-
could have led to this unusual situation, after tion, Dorfman met Feldman and said that he was
some hesitation the latter told how at the start of prepared to pass information to Karpov, if the
the second Karpov-Kasparov match in the latter would help him to move from Lvov to a
Tchaikovsky Hall he had been approached by a constant place of residence in Moscow. This
childhood friend (they had grown up together in request caught Feldman unawares, and he asked
Zhitomir), Alexander Feldman, a teacher of for time to discuss it. But already on 30 Septem-
physics and mathematics who was living in ber he handed Dorfman an application by the
Moscow. Dorfman learned that Feldman was chairman of the Moscow Sports Committee
closely acquainted with the leading players in addressed to the deputy chairman of the Moscow
Moscow and was one of the organisers at the Council A.I.Kostenko, No.01-22/5031 dated 30
match of a secret chess totalizer (as we see, for September 1985, with a request to assist grand-
the moment Dorfman was keeping quiet master Dorfman in moving his residence to
about his contacts with Feldman during the Moscow in order for him to prepare the Moscow
first match and between the matches - chess team. In this application there was a
G.K.). During one of the meetings, on behalf of resolution by Kostenko: "Please bring it before
Karpov, Feldman offered Dorfman 100,000 the commission and assist if possible." At the
roubles for information about Kasparov's chess same time Feldman strictly warned Dorfman
preparation. Dorfman took this conversation to that if he were to tell anyone about this document
264
The Year between Matches
he would be in big trouble. After this Feldman reported to Yusif-zade, the chairman of the
demanded that Dorfman should begin providing Azerbaijan KGB, who promptly informed
chess information immediately, warning him Bagirov, the first secretary of the Azerbaijan
that if he were to behave dishonestly his wife and Communist Party Central Committee, and
daughter would be in deadly danger. This unex- Abramov, the head of the Fifth USSR KGB
pected turn of events completely frightened Directorate, about the situation. That same day,
Dorfman - he realised that he had ended up in on the instructions of Bobkov, the first deputy
the net of a cruel and powerful set of criminals ... ' chairman of the USSR KGB, an operational
Litvinov: 'Of course, Rzaev underestimated group was set up to study the information
the situation. After all, Dorfman appealed to obtained about the Feldman group... In addi-
him on 25 September, and Rzaev phoned me at tion, every measure was devised and adopted to
work only on 1 October ... The first thing he ensure that Dorfman's link with Feldman and
decided, when Dorfman brought him the letter Karpov did not become public knowledge and
from the Moscow Council, was that it was a known especially to Kasparov and the members
fake and that the signature of Kostel1ko - the of his delegation.'
deputy chairman - was a forgery. He took the However, let us return to the summer of
letter and set off with it to the Moscow Council, 1987 - on 16th June, a couple of weeks after
where with the help of acquaintances he checked my meeting with Feldman, Dorfman ar-
by number the authenticity of the resolution, rived at the training camp in Zagulba.
and discovered, to his indescribable surprise, When I informed him of Feldman's visit, he
that it was genuine. Here Rzaev shuddered. behaved in a nervous fashion, and began
Astounded by the fact that practically within complaining that Feldman and his people
two days it was possible to obtain both a resi- had been following him in Moscow and
dential permit and a flat in Moscow (what sort Lvov, and had threatened to kill him if he
of forces were involved?), from the Azerbaijan would not pass on information. At the
Permanent Representative's office he promptly same time Dorfman denied passing on to
phoned me in Baku on the government line ... Karpov any information about my chess
Rzaev realised that it was a question of the preparations.
higher echelons of power, who were interested in Litvinov: 'One has to be an expert to handle
Dorfman providing information.' sllch C07lversations. After all, Dorfman could
Report (October 1985): 'Ill the course of not guess what we knew, and what we didn't
the conversion with the KGB officer, the delega- know. '
tion leader Rzaev emphasised that the situation Report (June 1987): 'With the aim of clari-
in the match was highly inflamed. Karpov's fying the situation, on 17th June Kasparov and
entourage and his highly-placed patrons were the deputy leader of his team Litvinov (who was
constantly putting pressure on Kasparov and responsible for security matters) held a conver-
the members of his delegation, aiming to U7lset- sation with grandmaster Dorfman. After a
tIe them. In this situation a leak of information difficult, exhausting 'conversation-cum-
about the "Feldman and Dorfman affair" could interrogation', in a state of extreme stress
lead to a scandal of unprecedented proportions Dorfman admitted that in the autumn of 1984
and the collapse of the match. In view of this in Moscow, during the first Karpov-Kasparov
Rzaev asked that the information be handed to match, Feldman had met him and had suggested
the Republic authorities and to the USSR KGB playing on the totalizer, and Dorfman had
for the taking of urgent and necessary measures. accepted this suggestion ... '
'The information received was immediately In many details Dorfman's admission
265
Kasparov vs. Karpov: 1986-1987
266
The Year between Matches
team. Why was Karpov the one who initiated In the previous volume Kasparov vs. Kar-
them? Because if he did not want to use such pov 1975-1985 (p.166) I have already men-
information, if he had no interest in it, it was tioned Tofik Dadashev, the Baku psycholo-
simply not needed by anyone ... How did such a gist, who from time to time helped me
situation come about? Since the time of the during the matches in Moscow and Lenin-
matches with Korchnoi the preparation for a grad. I called him 'my talisman'. I don't
sports competition passed into the field of state know whether a talisman can betray you,
politics. An entire team from the security but Dadashev did this.
authorities travelled to Baguio. And, naturally, A few days before my departure to
Karpov acquired a taste and, above all, a knowl- Seville I was informed that Liternaturnaya
edge of how to obtain information in a specific Gazeta was preparing to publish a sensa-
way. But this way could be guaranteed only by tional interview, in which Dadashev would
the KGB. In addition, the forces who protected describe how, with the help of a 'special
Karpov were enlisted. Thus from back in the power emanating from him', he had made
1970s he had experience of enlisting the work of Kasparov the champion. At almost the
the special services with their specific methods. ' same time, similar material entitled 'Reve-
In the spring of 2008 I was sent a chap- lation' was offered to Moscow News, but on
ter from a book KGB igraet v shakhmaty (The 10 January 1988 the newspaper stated that
KGB Plays Chess) which was being pre- 'the text of Dadashev's "revelation" had
pared for publication. KGB Colonel Vladi- been obtained from Karpov', and called it a
mir Popov, one of the custodians of Soviet non-chess move.
sport in the distant 1980s, recalls: During the Seville match this interview
'In September 1984 the unlimited match for was published with some changes in the
the world championship began between Kas- German newspaper Der Spiegel (No.50, 7
parov and the KGB protege Karpov. In advance December 1987), and in the Soviet press it
the 11th section of the Fifth USSR KGB Direc- appeared later - first in Vechernaya Kazan,
torate prepared a multi-page plan of agency and and then also in Literatumaya Rossiya (be-
operational measures ... In accordance with this fore my fifth match with Karpov in 1990).
plan, which was approved by F.D.Bobkol', the In his interview Dadashev openly stated
deputy director of the KGB, all Kasparov's that Karpov had contacted him not long
telephone conversations were intercepted, and before departing for Spain and had asked
the telephones used by his mother and trainers him not to interfere any more in our
were also monitored. The rest room, assigned to matches. At the same time the psychologist
Kasparov in the building of the Hall of Col- reported: 'Garry also tried to make contact
umns, was bugged. Thanks to the measures with me, but his enquiries went unanswered.'
taken, the KGB was quite well informed about Dadashev gave my opponent 'three pieces of
Kasparov's theoretical preparations, the moves advice on how to fight Kasparov'. He claimed
he was planning in the next game, and his that Karpov would win if he followed his
psychological and physical condition '. advice.
The value of his advice will become
Psychological blow clear, if it is remembered that this person -
My opponent struck me another unpleasant who was undoubtedly a gifted and subtle
and unexpected blow just before the match, psychologist - had helped me at the most
when my team had assembled in Moscow difficult period of the first match.
and was preparing to fly to Spain. Nikitin: 'Later Dadashev admitted that at
267
Kasparov vs. Karpov: 1986-1987
that time he had been summoned to the Central published by the world-famous parapsychologist
Committee of the Azerbaijan Communist Party Oadashev, who, it transpired, had been actively
and invited to fly urgently to the aid of Garry, helping Kasparov from the auditorium in all the
who was already losing 0-4. Nothing was said preceding matches. But not at all the games: he
to the trainers about the work and "miraculous would appear only at critical moments and
predictions" of Oadashev, and it was only a few would invariably contribute to a win for Kas-
years later (after the event!) that he reported in parov, directly influencing the results of deci-
an interview that his predictions had been sive games ...
correct. Well, in the art offoreseeing past events 'If (in the first match) I had gone in for
any of us, including Kasparov's chauffeur, sharp play (here Kasparov is correct in his
could have competed. But, one way or another, evaluation) I could have lost one game, or even
the job of the psychologist was to instil confi- two, but I would certainly have won a sixth,
dence in Garry.' and with it the match. And I myself realised
Dadashev had often talked with me and this, but for some reason (I find it hard to
he knew quite a lot about me: my strong explain why) at the board during the second half
and weak points, my mistrustfulness and of the match I drove away any thoughts of such
my belief in lucky numbers, but above all play. It is simplest to place the blame for this
he knew that I had faith in him. He quite obstinate passivity of thinking on Oadashev ...
often predicted the course of the play and He is a parapsychologist, a world-famous one,
gave me valuable psychological advice. moreover... And it transpires that this man
With the score standing at 0-5 he tried to came to Kasparov's aid when in the first match
convince me that I would not lose the the score became 4-0.
match. 'As regards those who were party to this ac-
When he began helping Karpov, he tion, one has to admire their skill in keeping it
could not have failed to realise what a blow secret: Oadashev took an active "part" in our
this would be to me when I found out. first three matches, but I learned about this only
Indeed, still believing in Dadashev's 'pow- from the man himself, when he approached me
ers', in Seville I often racked my brains over not long before the match in Seville and con-
the question: 'Why did he do this?' Such a fessed to interfering in our sports encounter.
gift to the opponent! This act of my former "'I did not cause you evil", he tried to con-
'talisman' could well be seen as another vince me. "I only helped Kasparov. You will
KGB intrigue, since my mother and I, as far understand that these are quite different things.
as we were aware, had not given him any To me Kasparov seemed so pure, inexperienced
grounds for offence. and nai·ve... He was so confused, and he so
Later in the book Sestra moya Kaissa (My needed support ... But now I see that this was
sister Caissa, 1990) Karpov finally ex- only a guise, because he uses his position not for
plained the veiled cause of his match the good, as I had hoped, but he spreads evil.
failures: But I did not harm you. Believe me! Our
'Unfortunately, in the Leningrad match professional code does not allow this."
(1986) I was unable to finish Kasparov off .. 'Possibly. But in one of our conversations he
And I would have thought - well, I was unable, himself told me how in childhood he discovered
that's how chess arranged it, but before our next his gift. It was at school, where he had disputes
match - now in Seville - there was discord in with one of the teachers. And then one day,
Kasparov's camp, certain sections began to fall when some inspectors came into the class, the
out, and then some staggering revelations were boy sensed that now he would be able to pay the
268
The Year between Matches
teacher back for all her malice. How? He did not on the topic 'psychology and chess'.
know how precisely, but he stared at her and Since the time of the Spassky-Fischer
began passionately suggesting: make a mistake! encounter, matches for the world crown
make a mistake! And she did. However, that boy ceased to be only competitive events and
did not yet know about any parapsychologists' acquired a political status. The course of the
code of honour. battle interested millions of people
'Dadashev brought me his manuscript, throughout the world, and chess appeared
which he called "Revelation". I was staggered on the front pages of leading newspapers,
by what I read. I could not imagine that this on radio and on television. When Karpov
really happened, that this was possible. But the became world champion and it transpired
details, which could not be made up, which were that his opponent in the 1978 match would
known only to a handful of people who were be the defector Korchnoi, the political factor
party to the innermost secrets of the match, was increased to absurd dimensions. The
supported the account. And photographs of Soviet authorities provided the young
Kasparov with fervent inscriptions such as "For champion with support on an unprece-
invaluable assistance in the match ", "For help dented scale, enlisting the KGB and various
and support in the 22nd game of the return experts, including professional psycholo-
match in Leningrad"... The tickets which gists.
Kasparov sent him were not normal entrance But the first to make use of a psycholo-
tickets, but free tickets stamped by the admini- gist was Korchnoi himself during the final
stration, which each of the teams received for Candidates match with KarpO\' (\10sco\\'
particular seats, stipulated beforehand. From 1974). The man in question \\'as Rudolf
these tickets I could now imagine where he was Zagainov, who was \\'ell known in the
sitting, and since with the years I developed a world of sport. At the wry start of the
habit of carefully surveying the auditorium, match KarpO\' began noticing that in the
especially the seats of the opponel1t's team, I opponent's box in the theatre there was a
remembered him. Yes, it was him. He was there man who tried to catch his glance: 'I en-
at those games. A penetrating, eI'e1l piercins qllired ,['ho this man was - it transpired that he
glance. I remember. Moreouer, I n01[, remember ems ,t'Orkillg with Korchnoi. No thoughts arose
especially clearly his intrusive, trollblesome concerni11g the employment against me of
presence at the last game of the second //latch - hypnotic action. I did not demand that this man
at that game after which I lost the title or (['odd be removed, but I decided to employ my own
champion. This means, Mr Professor, ~/Oll say psychologist. '
that you were only helping Kaspar011, and YOll And in the course of the match Doctor
tried not to hinder me? Well, that's (lery illter- of Sciences Vladimir Zukhar, an official of
esting: the Space Medicine Scientific Research
With present-day young players, such Institute, was invited to fulfil this role.
revelations may merely provoke a smile. Later he directed Karpov's psychological
The exploits of the 'famous parapsycholo- preparation for the unplayed match with
gists' have long since disappeared into the Fischer (1975) and at the match in Baguio
past - today the main problem, one which (1978), where he thoroughly unnerved
is not far-fetched, has become computer Korchnoi. This is described in detail in
prompting. Volume V of My Great Predecessors. Here I
Nevertheless I hope that the reader will will give only a few interesting testimonies:
be interested in a short historical excursion Karpov: 'In all sports teams - anywhere,
269
Kasparov vs. Karpov: 1986-1987
where there is a psychological battle - there are And for the first time in this match I turned to
psychologists, who create a fighting mood and Zukhar for help. It is not losses that torment
raise the tone of a competitor - there is nothing you, so much as missed opportunities. After all,
shameful in this... Zukhar helped me with if I had played normally in the last few games,
advice - how to relax, how to escape from the the match would already have been finished. I
prolonged match tension, rest, divert myself and didn't understand what was happening to me. I
engage in exercises with experts on physical analysed each action of mine, the course of my
preparation. How best to get to sleep, and in the thoughts - and failed to find a justification.
morning to emerge from this state - this is why True, I did not yet have any doubts about the
I needed Doctor Zukhar. ' outcome of the match, but dissatisfaction with
Tal: 'In Baguio the work of Doctor Zukhar my play was sufficient to unsettle me. I needed
occupied a significant time and was completely somehow to be diverted, to forget about the
excluded to any of the members of our delega- idiotic fruitless thoughts, I needed to sleep well,
tion. Karpov very much needed psychological calmly prepare for the next game and arrive for
support. Firstly, he is a person who has doubts, play in a confident mood. But above all - to
although, naturally, he does not like to admit sleep. But I couldn't sleep. I was tormented for
this. Secondly, he becomes tired both physically half the night and I phoned Zukhar. He tried his
and psychologically. This is especially noticeable sorcery on me - but in vain ... '
after about 25 games. In this period during play I don't know what disagreements arose
his mind goes blank. His decisions often become between Zukhar and Karpov, but for his
impulsive and inexplicable. Thus he can get third match with Korchnoi (1981) the
worked up and rattle off five middling moves at champion already had another psycholo-
blitz speed, or, on the contrary, he can stand gist, and Zukhar attended our first match
still. During this period his evaluation of (1984/85) merely as a spectator. Once, in the
position and calculation of variations deterio- winter of 1981, I met Zukhar and Spassky
rate sharply.' at the home of my trainer, Alexander
Karpov was not noted for his ability to Nikitin. We had an interesting chat, but no
withstand a long event. This is how he 'commercial' relations.
described his condition during the 11th (!) Spassky: 'Hypnotists and parapsychologists
game of the match in Baguio: 'I had a rea- became an important fixture of matches for the
sonable position with Black, but then something world championship. I know about this from my
inexplicable happened. I look at the board: it own experience, but I wouldn't say that they
would seem I can play this, and that will be just can have a decisive influence on the result of a
as good, I can carry out a combination, acquire match. All that can be said is that they are part
some initiative. No, I think, there's something of the overall strategy, conducted over the entire
about my condition that's not right today, I front and aimed at unnerving the opponent'
may miscalculate. And I make a neutral move. (France-Presse, 25 October 1985).
Then suddenly a decline sets in. I overlook an Nikitin recalls: 'At the very start of the
attack, then another one, and my position Kasparov-Smyslov final Candidates match
becomes difficult. I am losing. Yes, days occur (1984), Leonid Ostrovsky, a well-known chess
when apathy sets in, and then everything seems arbiter from Kiev, sent Garry's mother a letter,
to go wrong.' in which with staggering accuracy he predicted
After missing wins in the 18th, 20th and the result of each game. A month after the
22nd games, Karpov's state of health be- match I travelled to Kiev to talk with Ostrovsky
came critical. 'That was when I couldn't sleep. about preparations for the forthcoming encoun-
270
The Year between Matches
ter with Karpov, and in particular about possi- parapsychologist Dzhuna Davitashvili. She
ble extraneous influences on the play. And 1 presented him with an amulet with a depiction
heard the following: "As soon as Kasparov of herself, but in the end the amulet did not
arrives in Moscow, your home telephone will be work. Apparently, after losing faith in Dzhuna,
bugged, so be careful about what you say. And before the Seville match Karpov found support
as for parapsychologists and clairvoyants, bear in Dadashev, who earlier had helped Kasparov.
in mind that many of them are under the The switching of a psychologist to the oppo-
control of the KGB, participating in secret nent's camp is comparable only with treachery:
projects of state importance and looking for after all, for a competitor his psychologist is like
criminals. They live quite well, but they are a priest who receives a confession, and there is
restricted to a definite zone. Between them and no greater sin than revealing the secret of a
those parapsychologists, who do not advertise confession. But for Dadashev, switching from
their capabilities and do not work for the KGB, Kasparov to Karpov was a simple change of
stands a wall of alienation. 1 invited Garry's patient. "1 had no sympathies, but purely a
mother to provide psychological support for his competitive interest", he declared at that time in
entire team, but she did not reply." As we were an interview for Der Spiegel. "To neither of
parting, he said: "You will Will the match, but them did 1 give any commitments. They were all
in a fierce struggle". And indeed, the struggle the same to me ".'
lasted 72 games.' During the second match (1985) Karpov
My mother did not reply to Ostrovsky, was helped by Grigory Rozhnovsky from
because she did not want to involve me and Odessa, a psychiatrist, who was familiar
the team in non-chess machinations. At that with hypnosis and parapsychology. He
time it was not yet possible to imagine just would arrive for games in the same black
what was awaiting us ... 'Volga' as the champion.
A few words abou t the psychologists 'He was first seen in the 4th row of the stalls
who helped Karpov after Zukhar. In Merano in the Tchaikovsky Concert Hall when Karpov
(1981), when the expected blitzkrieg did not was leading, and since then he has been care-
work, Modest Kabanov, a prominent scien- fully observing each game. He openly fixes his
tist and Doctor of Medicine, flew to Kar- glance on the challenger, often using a large
pov's aid from Leningrad. From 1981-85 pair of binoculars. Simultaneously with his
Karpov's team included Mikhail Novikov - appearance, it was noticed that Kasparov
an expert on maintaining the efficient func- sharply curtailed the amount of time he spent
tioning of a training team. From 1986 along- on the stage. After making a move, he began
side the ex-champion was Igor Akimov - a immediately going off to his rest room' (France-
psychologist and journalist, the author of a Presse, 20 October 1985).
number of articles and books on questions of From an interview which Rozhnovsky
psychology, a sensational essay Osen na gave during the match to the Yugoslav
Kamennom os trove (Autumn on Kamenny journalist Dimitrije Bjelice: '1 am a hypnotist,
Island - cf. pp.l72 and 199) and ghost writer and 1 admit this... 1 could have hypnotised
of the book Sestra moya Kaissa (My sister Kasparov when he was on the stage, but the
Caissa). To all appearances, his help in the morals and ethics of a psychologist do not allow
middle of the Leningrad part of the 1986 this. What goes on at the chess board is an
match was effective. honest battle.'
Nikitin: 'Karpov several times went for re- Nikitin: 'At the 24th game of the second
vitalisation to the famous healer, astrologist and match, a young man, whom earlier 1 had often
271
Kasparov vs. Karpov: 1986-1987
seen with Rozhnovsky, sat down in one of the previous matches, despite the dramatic
front rows of the stalls and began looking changes in my group of trainers.
intently at Kasparov. I immediately remembered For Black, all our efforts were directed
the 32nd game in Baguio ... I had to approach a towards rehabilitating the Griinfeld De-
doctor acquaintance of the Karpov delegation fence. We looked a little at the King's
and invite him to a press conference for foreign Indian and, as usual, the Queen's Gambit,
journalists, which I was able to organise very but it was on the Griinfeld that we spent
quickly. It all became clear to him. When I most of our time. (For the 2000 match with
returned to the auditorium, the young man was Kramnik I also mainly prepared this open-
now standing far from the stage.' ing and only briefly the Queen's Gambit
In conclusion, I must once again men- Accepted. Alas, my predictability proved to
tion the St. Petersburg psychologist Rudolf be a mistake, which cost me dearly.) Al-
Zagainov, who after the 1974 match, like though in the 1986 match the Griinfeld had
Korchnoi, was out of favour. Later at vari- brought me nothing but disillusionment, I
ous times he helped the women's world put this down mainly to Karpov's unusual
champion Nona Gaprindashvili, the 'foresight', and I assumed that the dynamic
women Candidates Nana Alexandria and nature of the ensuing complicated positions
N ana Ioseliani, and the young grandmas- was basically far more in keeping with my
ters Boris Gelfand and Sergey Dolmatov. I style than that of my opponent.
also met him a couple of times during the Just in case we also prepared for 1 e4,
Candidates match with Beliavsky (1983). since we did not yet know that Karpov had
Zagainov presented me with a poster finally given up this move. Apart from the
bearing the well-known slogan 'If not you, Sicilian, we looked at the Pirc-Ufimtsev
then who?' (this was the motto of the Defence, against which Karpov played
trainer of the Japanese volleyball team, exclusively the set-up with ctJf3 and ..ie2,
which in 1980 broke the hegemony of the and not very confidently (he even lost to
USSR at volleyball). But we did not agree to Azmaiparashvili in the 50th USSR Champi-
work together... Later Doctor Zagainov onship, Moscow 1983).
helped Karpov in the Lyons half of our fifth With White, on the other hand, I wanted
match (1990) and in his semi-final Candi- to surprise Karpov in some way, and after
dates match with Short (1992), which he much thought I decided to try 1 c4, a new
described in his best-seller Porazhenie move for me. Initially this idea seemed
(Defeat), published soon afterwards. questionable, but gradually material was
In matches during the 1990s public in- gathered and it began to seem quite a good
terest in the work of chess psychologists surprise for my opponent, roughly equiva-
began to wane. People no longer saw lent to the set-up with g2-g3 against the
anything unusual in it, but the main thing Nimzo-Indian Defence in the 1985 match.
was that the world chess championship had But in the 1986 match Karpov had adapted
escaped from the close surveillance of the to this set-up, and 1 c4 was good if only for
KGB. the reason that it altogether excluded the
Nimzo-Indian: if 1...e6 2 ctJc3 ctJf6 we were
Before the start planning 3 e4.
My opening preparation for the battle in We also prepared another way of com-
Seville was serious enough, and in this bating this defence - the old variation 1 d4
respect we maintained the momentum of ctJf6 2 c4 e6 3 ctJc3 ..ib4 4 "IWc2. And here we
2 72
The Year between Matches
found a number of fresh ideas, which could seemed like the final victory.
have caused Karpov serious problems, such And in 1987, in addition to my continu-
as 4 ... 0-0 5 a3 .i.xc3+ 6 'iYxc3 b6 7 .i.g5 .i.b7 8 ing fight with the FIDE and the Sports
f3 (and if 8 ... d6, then 9 e4 c5 10 d5 Kas- Committee bureaucracy, I was busy with
parov-Hjartarson, Reykjavik 1988), which numerous other important things: work at
became one of the most popular lines, or the junior chess school (six weeks before
4 ... c5 5 dxc5 0-0 6 .i.f4 .i.xc5 7 ttJf3 ttJc6 8 e3 the match Botvinnik and I held our last
d5 9 a3 'iYa5 10 O-O-O!? (d. Game No.39, note joint session), trying to set up a Soviet-
to White's 4th move). There were also British chess publishing house and a chess
aggressive plans in the Queen's Gambit. .. magazine, the founding and leading of the
In fact none of this occurred in the match, GMA in Brussels and a junior computer
but our discoveries proved very long-lived, club in Moscow. Incidentally, the games in
and they later became firmly established in the first-ever 'Star chess' display, in which I
tournament play. played simultaneously with ten cities in
Less attention was devoted to 1 e-l. Here various parts of the world, were transmit-
we mainly looked at the Ruy Lopez, where ted from the junior computer club via a
as a matter of principle we decided not to telecommunications link.
avoid the complications such as those which Work on my autobiography Child of
occurred in the 14th and 16th games of the Change, which began at the end of 1986,
1986 match. These double-edged variations was another distraction. As the work
with various sacrifices caused us serious progressed, it was hard to shake off the
fears. Much of our analysis of that time was feeling that the main fight for the world
quite topical and it came in useful later in title was already behind me. Not long
the 1990 match. But in Seville, as earlier in before the match, the book was published
his match with Sokolo\', Karpov only played in London, and then in another five Euro-
1...c6, and I contented myself with a quiet pean languages, and in Seville I arranged
set-up suggested by Dorfman - at the time for it to be launched.
we did not have any other fruitful ideas Numerous problems, including those
against the Caro-Kann. As a result I twice involving Dorfman and Dadashev (d. the
'lost' the white pieces (Game Nos.37, 41), preceding sections) prevented me from
although quite playable positions were focusing on the main thing, which was
reached - if there had been a desire to fight! psychological preparation for the next
But for me the match in Seville was psy- match with Karpov. The forthcoming clash
chologically the most difficult of all my five was somehow very remote from my mind,
duels for the title with Karpov. And this and I couldn't help wishing that I didn't
despite the incredible tension of the first have to go through with it. Why did I have
match, when over a period of two months I play again? I'd already proved my superi-
had to go out on to the stage facing a final ority - I'd already won and then defended
defeat, the unparalleled emotional strain of my world title ... Although I realised that far
the second match, when the goal came so more was at stake this time than in the
searingly close, and the shock towards the previous matches, I really didn't want to
finish of the third match after the business play the match, and my whole being re-
with Vladimirov. Seville was the worst belled against the very thought that I must
ordeal of my life. To some extent, my win once again (for the umpteenth time!) mobi-
in the 1986 match had disarmed me, since it lise myself for battle.
273
Kasparov vs. Karpov: 1986-1987
On the whole, my preparations before win. The psychological battle had begun ... '
the match were of quite an adequate stan- Just think: this was our fourth match for
dard, but, alas, my playing mood did not the world championship during the last
match them. In the end my confused condi- three years! In 1984 I felt that I was tired of
tion generated a fear of defeat, which had a looking at Karpov, but now seeing his face
paralysing effect on me. every day became pure torture. But at least
Nikitin: 'As the match approached, Garry on this occasion I wasn't expecting any new
became increasingly nervous; he kept saying that machinations, and a win or draw in the
defending the title was far more difficult than match would free me until 1990 of the need
winning it, and that he no longer had the same to see before me every day this or any other
e7lergy that drove him to the top. In an interview challenger for the title. Perhaps it was the
he said he would arrange a first-class burial for realisation of this fact, that the next battle
his oppone7lt. I71is hackneyed method of intimi- for the crown would only take place three
dation, employed by professio7lal boxers, merely years later, that led to a stormy start to the
co7lfirmed his u7lcertail1ty ... just before the match match: in the first eight games we each
KarpOLl also stated that he was sure he would gained two wins.
274
Chapter Four I
Play with 'golden towers' squares there were also enormous multi-
We arrived at the site of the battle five days coloured pawns, arranged against a back-
before the start. Judging by the media ground of flower-beds and fountains. We
reports, 'not only Seville, but also the whole would arrive for the games separately, just
of Spain was awaiting the clash of the two a few minutes before the clocks were
strongest grandmasters on the planet'. The started at 16.30, entering through different
match was very well organised, especially doors, depending on which colour we were
for a city with no rich chess traditions. going to play: one door had a white pawn
We played in the Lope de Vega Theatre, on it, and the other a black one.
which was erected in the centre of Seville At the back of the stage hung a huge
for the international exhibition of 1929. The poster advertising the World Fair, which
theatre stands in a beautiful park and is was due to be held in Seville: 'EXPO-92'
built in the Spanish baroque style, resem- (the letter '0' being represented by an
bling a huge flower, dazzling white in the attractive orange). This explained the high
sun. Here, in a stylish hall with seating for prize fund: the city of Seville had provided
a thousand, plays, operas and musical half a million pesetas to finance the match!
concerts are normally performed, but now However, to enhance the advertising effect
the theatre was given over to chess. A few we had to play with rather unusual pieces:
hundred yards away was the old Alphonse our rooks resembled the famous Seville
XIII Hotel. It had been planned that both 'Golden tower' - the distinctive emblem of
teams would be staying there, but then a this ancient Spanish city.
modern hotel with the poetic name of The theatre's spacious foyer was
Seville Sol (Sun of Seville) was chosen dubbed by the spectators 'the casino', and
instead (the contestants also gave press with good reason: it was where bets were
conferences here). In addition, Karpov and placed on the outcome of the next game.
I each had the use of an out-of-town villa. The more bitter the fight on the stage
It was easy for visitors to Seville to find became, the greater the number of people
their way to the Lope de Vega Theatre: who rushed from the auditorium into the
drivers could follow placards with the 'casino'. But, of course, the laying of bets
match emblem, and at intersections and wasn't the only thing that went on in the
275
Kasparov vs. Karpov: 1986-1987
foyer. It was a place where you could chat, Defence until a competitively more propi-
discuss the game over a cup of coffee, play tious moment.' (Makarychev)
chess or buy chess books and souvenirs. In 3... c6
the middle of the foyer was an enormous
demonstration board, with the aid of which
grandmasters took it in turns to provide
commentary on the game in progress. All
around there were monitors, on which the
players and the position on the electronic
board could be seen. Here too, behind some
pillars, was the press centre, as well
equipped with telex machines and com-
puters as a conference on disarmament ... At
the start of the match some three hundred
journalists were present, and later many
more arrived.
On Saturday 10 October the opening As has already been mentioned, at that
ceremony and the drawing of lots look time I was not yet ready for the more
place. The chief arbiter Geurt Gijssen first dynamic, purely Griinfeld set-ups with
invited me (as the world champion) to 3 ... ~g7 and 4 ... d5. It would also have been
draw one of two bookmarks from an enor- risky to repeat the experimental 3 ... c5
mous folio. I drew the bookmark with the (Game No.25). The solid, symmetric position
number 2, and it was Karpov who gained seemed to me to be the correct choice for a
the right to draw for the colour of the world championship match, especially
pieces. After some hesitation Karpov ges- since a year earlier in this way I had twice
tured towards one of two dark-haired calmly equalised.
Spanish girls in black and white costumes; 4 tDf3 ..\tg7 5 ..\tg2 d5 6 cxd5 cxd5 7 tDC3 0-0
she began an intricate dance, opened her 8 tDe5 e6 9 0-0 tDfd7 10 f4 (of course, not 10
fan, and gracefully displayed a white piece tDf3 - Game No.3)
depicted on it. Two days later Karpov and I
met at the chess board.
Game 28
A.Karpov-G.Kasparov
World Championship Match,
1st Game, Seville 12.10.1987
GriJnfeld Defence 079
276
The Fourth Match: 1987
see a clear plus for White after 10 ... f6 (Game tage of the two bishops and seized space
No. 13), I decided to avoid repeating this with the pawn march e2-e4-eS. His plan
move and to surprise my opponent slightly includes a pawn offensive on the kingside.
with one of my two preparations (the However, it is not clear how quickly real
second, of higher quality, was to occur in threats will arise, although White does not
the 3rd game). have to rush with it, since he already has
11..ie3 some quite important trumps. For his part,
And again Black has a choice. Black has some counter-chances on the
queenside.
11 ... 4:Jb6
Karpov also had occasion to fight 15 .. ..::Hc8 (19)
against 11...4:JdxeS 12 fxeS f6 or 11 .. .£6 Little is changed by IS .. J::tac8 16 l:Ic1,
(which as a matter of principle I did not since all the same Black needs to double
want to play!), to which he replied both 12 rooks on the c-file. In 1992 Karpov won an
4:Jd3 and 12 4:Jf3, as in the 13th game of our excellent game in this variation against
1986 match. More details of these variations Kamsky: 16 ... a6 17 b3 (restricting the knight
can be found in the notes to it and also to on b6) 17 ... l:Ic7 18 "iYd2 Ufc8 19 g4! ..if8 20
Game Nos.lOG and 101 in Voilime V of My "iYe3! 4:Jc6 21 fS! etc. (Game No. lOG in Volume
Great Predecessors. V of My Great Predecessors). The only differ-
12..if2 ence made by the move in the game is the
'In the event of 12 4:Jxc6 bxc6 13 ..if2 as! creation of the ... ..if8 resource, of which I
Black would have succeeded in creating did in fact make use.
counterplay, by combining pressure on the 16l:Icl
a6-e2 diagonal and the b-file with the ... as- The attempt to attack on the kingside,
a4 advance.' (Makarychev) avoiding exchanges on the c-file - 16 l:Iel
12 ... ..id7?! 4:Jc4 (16 .. Jk7!?) 17 ~e2 l:Ic7 18 Uadl Uac8
This mechanical developing move al- 19 Ud3 (Illescas-Topalov, Madrid 1996)
lows White too comfortable a game. The leads after 19 ... bS! 20 g4 b4 (Bologan) to
immediate 12 ... 4:Je7 (Game No.3G) is more double-edged play.
accurate. 16 ... ..if8!? (33)
13 e4 (IS) 13 ... 4:Je7 (the dS-point must be In a rather unpleasant position I man-
held) 14 4:Jxd7! (08) 14 .. :~~Vxd7 15 e5 aged to find a move which disturbed my
As a result White has gained the advan- opponent. But the first alarm signal also
277
Kasparov vs. Karpov: 1986-1987
sounded: this and the next move could himself to play g3-g4. Besides, for him the
have been decided on more quickly. advance of the g-pawn probably evoked
unpleasant memories of the 24th game of
the 1985 match, where he played g2-g4-gS,
but at a critical moment avoided the sharp
f4-fS. However, later, after thoroughly
studying the nuances of the symmetric
position, the ex-champion nevertheless
adopted the plan with g3-g4 and f4-fS (d.
the note to Black's lSth move).
But if White had decided to manoeuvre,
avoiding all risk, it would have been better,
for example, to play 17 Mel Mc7 18 ~f1,
since, as we will now see, the insipid move
17 ~f3 also involves a tactical oversight.
17 ..Itf3?! (21) 17 ... MC7 (2S) 18 b3 (22) 18 ... Mac8 19 'iVd2
'A crucial moment. The pawn structure (all in the same unhurried manner)
suggests g3-g4 and f4-fS. With his last 19 ... ttJc6!
move Kasparov as though provoked the A surprise, which reveals the drawback
opponent into beginning a reckless offen- to the move 17 ~f3.
sive on the kingside - 17 g4!? Now 17 ... Mc7
18 ~h4! or 17 ... ~h6 18 ~h4! ttJc4 19 'iVe2
does not promise Black an easy life. True,
after 17 ... Mc4 it is no longer possible to play
18 ~h4 in view of the weakness of the d4-
pawn (18 ... ttJc6!), but the positional sacrifice
18 fS!? secures White the initiative. If
18 ... ~h6, then 19 f6! is very strong, while in
the event of 18 ... exfS 19 gxfS ttJ('iV)xfS? there
is the tactical stroke 20 ttJxdS!. The pawn
can be won only by the committal 18 ... gxfS
(say, 19 ~e3 Mac8 20 ~h3 with excellent
compensation. - G.K.). Sharp play could have
begun, in which a definite result for either With the bishop on g2 the ... ttJc6 ma-
side would have been more probable than a noeuvre would have been impossible
draw. But Karpov does not play like that, because of the reply ttJbS, and therefore
and this is mainly a question of individual Black would have had to waste a tempo on
style. He prefers a quiet solution to the ... a7-a6 (as Kamsky did in the afore-
position.' (Makarychev) mentioned game with Karpov). But here
I should add that after 17 g4!? Mc4 apart after 20 ttJbS? there is the counter-stroke
from the sharp 18 fS the solid 18 b3 Mc7 19 20 ... ttJxeS!, exploiting the undefended
a4 Mac8 20 'iVd2 is also good, with a small position of the bishop at £3.
but enduring positional advantage. 20 'iVb2
It is apparent that in the first game of But not 20 ..te2? ..ta3!. Instead of making
the match Karpov simply could not bring an energetic march to the kingside (as in
278
The Fourth Match: 1987
Karpov's game with Kamsky) the white proaches for his excessively peaceful ap-
queen is forced to remain on the queenside. proach.
20 ... a621..\ie2 2s ... ctJd7 26 Mxe7 Mxe7
It is not altogether clear what the bishop This is roughly the same position as be-
is doing here (controlling the d3-point?), fore, but without a pair of rooks. White still
but White must somehow disentangle his has no way to gain an advantage.
pieces. 27 ctJe3 ctJe6 28 ctJbl (19)
21 .. :tlVe7! (14) If 28 Mel 'lWa3 29 ctJd1 (hoping to trans-
pose into a slightly better endgame) there is
the reply 29 ... 'lWa5, and it is unclear what
White can achieve in view of his potentially
weak d4-pawn.
28 ...ctJb4 29 ctJe3 ctJe6 30 ctJbl ctJb4 Vz-Vz
279
Kasparov vs. Karpov: 1986-1987
ous system was not ready for the severe 3 ctJf3 ctJc6 4 g3 ..I¥..b4 5 ..I¥..g2
pressure of a world championship match. I had also studied the line with 5 ctJd5,
Whereas experience and intuition would in which Karpov had a wealth of experi-
sometimes suggest the correct solution, ence, but I was more interested in 5 ..I¥..g2,
some kind of general physical inertia and where I had discovered the possibility of
sluggishness of thought, and above all a creating interesting complications literally
complete lack of inspiration, made my play out of nothing. And indeed, the course of
ponderous and incoherent, forcing me to the 2nd and 4th games was much different
check variations again and again and to to the traditionally measured 'English'
end up in time-trouble. play.
5... 0-060-0 e4 (06)
For the moment Karpov follows his
Game 29 usual path. Later he was to prefer 6 ... .l:te8
G.Kasparov-A.Karpov (Game No.43).
World Championship Match,
2nd Game, Seville 14.10.1987
English Opening A29
1 c4
I made this move, a new one for me,
without any hesitation. Karpov sank into
thought, trying to guess my intentions.
Apparently White wants to avoid the
Nimzo-Indian Defence and invite the
'Carlsbad' version of the Queen's Gambit -
1...e6 2 ctJc3 d5 3 d4 ctJf6 4 cxd5 exd5 5 ..I¥..g5
or 3 ... ..I¥..e7 4 cxd5 exd5 5 ..I¥..f4 (Game Nos.7, 8, 7 ctJg5
39). But what had he prepared against the 'Again a surprise: this sortie had long
opponent's favourite reply ... e7-e5 ? gone out of serious use, since it gave White
1 ... ctJf6 (06) even less than the "anaemic" move 7 ctJe 1,
This was the move played, and not the leading after 7... ..I¥..xc3 8 dxc3 h6 9 ctJc2 to
immediate 1...e5, which for some reason is fairly insipid play' (Makarychev), for
given by Illformator and Chess Base. 'Kas- example: 9 ... ~e8 10 ctJe3 d6 11 'iYc2 as
parov's adoption of the English Opening in (Korchnoi-Mecking, 9th match game,
this match was not something that I ex- Augusta 1974; Karpov-Korchnoi, 6th match
pected. However, strangely enough, it was game, Baguio 1978) or 9 ... b6 10 ctJe3 ..I¥..b7 11
I who was able to spring a surprise at the ctJd5 ctJe5 (Korchnoi-Karpov, 9th match
start.' (Karpov) game, Moscow 1974).
2 ctJc3 e5 7 ••• ..IiLxc3 8 bXc3 ~e8 (03)
In the 2 ... e6 3 e4 variation Karpov once Similar positions with reversed colours
lost with Black to Korchnoi (29th match are often reached in the Sicilian - 1 e4 c5 2
game, Baguio 1978), while later with White ctJf3 ctJc6 3 ..I¥..b5 g6 4 ..I¥..xc6 bxc6 5 0-0 ~g7 6
I defeated A.Sokolov (Belfort 1988) and ~el ctJf6 7 e5 etc., but there White has an
Beliavsky (Linares 1991). extra tempo and he feels very comfortable.
280
The Fourth Match: 1987
281
Kasparov vs. Karpov: 1986-1987
282
The Fourth Match: 1987
(Karpov and Zaitsev give only the 'sharp' noni - a path is opened for the bishop on
13 .. ,cDxf4 14l:txf4, but here White's chances g2. At the same time a square is vacated for
are better) 14 ..ItxdS ..Itg4 (Grant- the retreat of the knight.
Timoshchenko, Pula 1997) or 13 .. .'iWf6!? 14 14 ... lDc6 (22)
..ItxdS ..Ite6 (14 ... lDe7!? and ... lDfS) 15 ~xe6 'This position is a difficult one for Black
l:txe6 16 fS !lee8, in every case with ap- to play, since any change in the pawn
proximate equality. configuration in the centre, blocking one
11 ...lDa5! (03) white bishop, will make the other ex-
Indirect concern for the e3-pawn. After tremely dangerous. If 14 ... ..Itg4 there fol-
1l ... d4 White could have set about activat- lows 15 lDf3 lDc6 16 h3!, and there is no
ing his bishops by 12 ..Itb2 or 12 f4. time for 16 .. .'ii'd7.' (Makarychev). For
12 'iWa3 c6 (12) example: 16 ... ..Itxf3 17 !lxf3 d4 18 ..Itb2
(Gulko-Korneev, Montreal 2006). Of course,
the exchange of bishop for knight is advan-
tageous to White, but the e3-pawn still
separates his two wings, and the situation
remains unclear - hence the current popu-
larity of 14 ... ..Itg4!?
15 l:tbl (08)
A logical move, somewhat hindering
Black's development (after this I was in-
tending ..Itb2, !lfd and c3-c4). After 15 ..Itb2
~g4 16 lDf3 'iYd7 the position is dynami-
cally balanced.
283
Kasparov vs. Karpov: 1986-1987
However, this is objectively the best re- all the trumps: 21...dxc4 (21...d4? 22 tiJxd4)
ply: 15 ... d4?! 16 Mel, and the white knight 22 j"xf6 gxf6 23 Mxc4 etc.
has obtained the e4-square, or 15 ... h6 16 As we see, in the event of 17 tiJf3! Black
tiJf3 1li'c7 17 j"b2 Md8?! 18 'iVc5!, pinning the would urgently have had to devise some
c6-knight and stealing up on the e3-pawn active play, as otherwise his position could
(Komarov-Ma.Tseitlin, St. Petersburg 1997). have become unpleasant, whereas after 17
In Informator and in his book of best games c4 the two sides' chances become equal
Karpov gave the 'drawing' variation and, moreover, it is now White who has to
15 ... j"g4(?) 16 Mxb7(!) j"xe2 17 tLlxf7(!) 'iVa5! play carefully.
18 iVd6(?) j"xfl 19 tiJh6+ ~h8 20 tLlf7+ with 17 ... dxc4! (10)
perpetual check (and therefore he recom- Absolutely correct: the position must be
mended the passive 16 ':e1), but after 18 opened up - White has 'everything hang-
':e1! li.xd3 19 :;jd6 ':eb8 20~~haS CDxa5 21 ing'. 'There is just no time for the capture
':xb8+ ':xb8 22 ~xe3 White has a won on e2' (Makarychev). And if 17 ... d4 (Dlugy)
endgame. I would simply have replied 18 tiJf3! l:lad8
16 ~b2 (16) 16 ... li.g4 (19) 19 i,a1 with the intention of Mb5, Mfb1 and
so on.
18 ~xf6 (10) 18 ... gxf6 (of course, not
18 ... ~xe2? 19 i,a1! and 'iVc3) 19 tiJe4 'it'g7!
(11)
It is unlikely that Karpov considered se-
riously the line given later in Informator:
19 ... Mxe4(?) 20 j"xe4 f5 21 j"f3 tiJd4, espe-
cially since this is bad because of 22 j"d5!
cxd3 23 Mxb7, winning.
17 c4?! (02)
An impulsive advance, which makes
things easier for Black: now he mainly has
to make the only and obvious moves.
Of course, the temptation to harass the
enemy king is very great... By calmly
continuing 17 ':fe1 White would have
retained good prospects. But 17 tiJf3! was
even better (I prepared this for the 4th
game) 17 ... iVd7 (Dlugy praised 17 ... j"xf3, The turning-point of the battle. If the
but after 18 j"xf3 all the same White plays black knight should jump to d4, the e3-
Mfd and c3-c4) 18 j"a1 with the idea of pawn will become astronomically strong!
18 ... b6 19 Mfd! Mad8 20 c4 d4 21 c5! Here Karpov had 42 minutes left for 21
(Makarychev). Karpov suggested 18 ... Me7 moves, while I had only 20, and I spent 15
19 Mfd j"h3 'with sufficient counterplay', of them on my next two moves ...
but after 20 c4! j"xg2 21 ~xg2 White holds 20 dXC4? (08)
284
The Fourth Match: 1987
A fatal error. This dreadful move - the a) 21 ctJxc4 iL.xe2. 'Now the d3-pawn is
worst of all the candidate moves! - demon- attacked' (Makarychev), but after 22 ctJxe3
strates my unreadiness for a large-scale White is not losing: 22 ... iL.xfl 23 ctJf5+ 'it'g8
battle (as too do my long think on the 10th 24 'it'xfl with full compensation for the
move and my forgetfulness on the 26th). exchange, or 22 ... Jbd3! 23 ctJd5 I:!.xd5 24
Let us briefly consider the others: iL.xd5 iL.xb1 25 I:!.xb1 with quite good com-
1) 20 h3?! (a cunning idea) 20 ... iLxe2, pensation for the pawn;
and now not 21 'iYb2 I:!.xe4! 22 \\Wxe2 I:!.e7 23 b) 21 ctJxb7 iL.xe2 22 ctJxdS I:!.xd8
dxc4 \\Wd6 (the e3-pawn causes White's (22 ... iL.xfl 23 ctJe6+! fxe6 24 'it'xfl is equal),
downfall), but the unexpected 21 c'LJxf6! when 23 I:!.fe1(?) cxd3! 24 \\Wc5
iL.xfl! 22 \\Wc3!, seemingly with a pretty win. (Makarychev) is inadequate in view of
But Makarychev discovered 22 ... .::e5!! - 'the 24 ... \\Wa5! 25 \\Wxc6 iL.g4 26 I:!.xe3 d2 27 I:!.f3
rook covers "with its body" the terrible d1\\W+ 28 I:!.xd1 I:!.xd1+ 29 I:!.fl \\Wxa2, but after
diagonal' (White loses after both 23 c'LJd5 23 dxc4 iL.xfl 24 .Jtxfl I:!.eS 25 \\Wd3 \\We7 26
\\Wa5, and 23 ctJh5+ 'it'fS!); I:!.e1 \\Wc5 27 'it'g2 White blocks the e-pawn
2) 20 ctJxf6?! (the time has not yet come and holds the position;
for such drastic measures) 20 ... ~xf6 21 I:!.b5 4) 20 \\Wc3. This obvious move was sug-
I:!.adS! 22 \\Wc3+ 'it'e7 23 ·~·xc-l '\,,'ith quite gested by many commentators.
good practical compensation for the piece'
(Makarychev), but after 23 ... ~f8 the advan-
tage is with Black, as also in the event of 21
h3 iL.e6 22 'iYb2+ (22 '~'c5 '::g8) 22 ... 'it'e7 23
dxc4 I:!.ab8 24 \\Wa3+ ~d8 25 '::fd 1+ ~c8 26
\\Wxe3 \\We7;
3) 20 ctJd6. 'A dangerous manoeuvre -
Black has to play carefully', writes Karpov,
but he gives only 20 ... .::e6(?) 21 ctJxc4 I:!.dS,
not noticing the simple 22 '::fe1 with an
obvious advantage for White. 20 ... I:!.edS is
'more careful', after which there are two
acceptable replies:
Analysis diagram
285
Kasparov vs. Karpov: 1986-1987
draw. He also does not equalise with 21 A simple and accurate reply. 20 ... Axe2?
'iVxc4 l:tb8 22 l:tfe1 CLld4. But 21 l:txb7! cxd3 would have been a mistake in view of 21
22 'iVxd3 relieves White of any problems: CLlxf6! 'it'xf6 22 ~c3+ 'it'e7 23 l:tfel. Now,
22 ... 'iVxd3 (22 ... CLld4 23 'iVxe3) 23 exd3 CLld4 however, Black has a strategically won
24 CLld6 with equality; position.
5) 20 m2!? (one of the most interesting 21l:tb3?! (07)
moves) 20 ... 'iVe7, and White has two re- Not good, but what can White do? He
sources: loses a piece after 21 'iVxe3? Af5, and his e2-
a) 21 dxc4 Af5! (this was pointed out by pawn in the event of 21 l:tb2 CLld4. 21 CLlc3
the commentators, but...) 22 CLlc3! Axb1 23 looks more tenacious: 21...CLld4 22 CLld5
CLld5 'iVe6 (23 .. .'iVd6 24 c5 'iVe6 25 CLlc7 with (Makarychev) 22 .. :iVxc4 23 CLlxe3 'iVxe2 24
equality) 24 l:txb1 l:tec8 (24 ... l:tac8 25 Ah3!) CLlxg4 'iVxg4 'with a tense situation on the
25 'iVc3 with the initiative for the exchange; board' (Karpov) - and indeed, after 25 l:tb2
b) 21 :tfel cxd3 22 ':c5! h6 23 exd3l:tad8 (25 l:txb7?! l:te2 is dangerous) White has
24 h3 .ic8 25 'ib'c3 or 24 ... Ae6 25 g4, fighting every chance of a draw. But 21...l:td2! 22
for the initiative; l:tfe1 CLld4 is much stronger, attacking the
c) 20 l:tfel!? This would also have given c4- and e2-pawns, or 22 CLld5 'iVa5! 23 'iVd6
good play. Since 20 ... Axe2? 21 CLlxf6! 'it'xf6 l:te6 24 'iVd7 Axe2, and Black nevertheless
22 'iVb2+ and ~xe2 is unfavourable for wins (25l:txb7 CLld8 etc.).
Black, and in the event of 20 ... cxd3 21 ~xd3 21 ... CLld4 (13) 22 l:txe3
Af5 (21...l:tad8?! 22 CLlxf6! 'it'xf6 23 1ib5!) 22 White would have lost ignominiously
'iVd5! White is out of danger, 20 ... 'iVe7! must after 22 ~2? CLlxe2+ 23 'it'h1 'uxe4! 24 Axe4
be played, leading after 21 ~2! to a dou- 'ud2 (or 24 ... ~xc4 25 l:txe3 l:td1!) 25 ~1
ble-edged position from the previous CLlxg3+! 26 hxg3 e2 27l:te1l:tdl.
variation.
Thus with resourceful play White could
have maintained the balance (and in more
than one way), and even forced his oppo-
nent to display accuracy in defence. Instead
of this I made a mistake, after which my
game went downhill...
20 ... l:tad8! (02)
22 ... 'ivxc4
'Here already many ways lead to the
goal: 22 ... CLlc2(?) 23 'iVc3 CLlxe3 24 'ii'xf6+ 'it'f8
25 ~6+ 'it'e7 26 'iYf6(?) 'it'd7, and the king
escapes from the pursuit', Karpov wrote in
2001, copying an old variation from Infor-
mator. However, after 26 'iVg5+! 'it'd7 27
CLlc5+ the king has nowhere to escape to
The Fourth Match: 1987
(27 ... 'it'c8? 28 .ltxb7+ 'it'b8 29 tZJa6+) and adopted in the USSR) he did not have the
things end in perpetual check: 27 ... 'it'd6 28 right to do this. Here are two press reports
tZJxb7+ 'it'd7 29 tZJc5+. on this dramatic episode with the clock:
23 'it'hl?! (01) 'Later a slow-motion sequence of this
Yet another lapse compared with the incident was shown dozens of times on
hopeless 23 'iVxa7 .ltxe2 24l:!.b1 .ltd3 25 'iVb6 television, and especially the shots showing
(25 1lVxb7 l:!.b8) 25 ... l:!.e6 and the more tena- how KasparO\', after making his move, is
cious 23 'iVb2! l:!.e6! 24 tZJd2 tZJxe2+ 25 'it'f2 thinking about the position but his clock is
~c5 26 tZJe4 ~a5 27 l:!.xe2 .ltxe2 28 'it'xe2 still running, and how, suddenly coming to
~de8, when Black wins. his senses, he desperately reaches out and
23 ... tZJf5?! (06) punches the clock button .. :
In the opponent's time-trouble Karpov 'This would not have happened, if Kas-
avoids 23 ... tZJc2! 24 'iVb2 tZJxe3, not wishing parov had been playing - Steinitz! In a
to allow even an illusion of counterplay: 25 game with him it was impossible to lose on
~xf6+ (if 25 tZJxf6 both the 'bold' 25 ... tZJxfl time: when the hand on the opponent's
and the cautious 25 ... 'it'g6 26 tZJxe8 ~xe8 27 clock was approaching the fateful mark,
l:!.c1 'iVxe2 are decisive) 25 ... 'it'g8 26 'iVg5+ Steinitz would begin displaying increased
'it'f8 27 'iYh6+ 'it'e7 28 'iVg5+ 'it'd7 29 tZJf6+ anxiety, constantly sitting forward in his
'it'c7 30 tZJxe8+ l:!.xe8, and the curtain comes chair and bringing his curly ginger head
down. close to the opponent's clock, obviouslv
24l:!.d3 .ltxe2 25 l:!.xd8 l:!.xd8 reminding him about his time.'
26 ... l:!.e8 (03)
26l:!.el
After quickly making this move, I forgot 27 'iVa5?
to press my clock button, and I noticed my Stepping over the precipice. 'If 27 tZJd6
mistake only a couple of minutes later, Black had prepared a simple mating con-
when Karpov, who had nearly 20 minutes struction: 27 ... tZJxd6 28 ~xd6 iLf3 29 ~xe8
in reserve, was still considering his reply, 'iVfl mate' (Karpov). However, I could have
and I had not more than 30 seconds left landed a tactical blow - 27 g4! ~xg4 28
before the fall of my flag. 'iVxa7 tZJh4 (or 28 ... tZJd6 29 'i'gl!) 29 'i¥f2
The arbiter did not draw my attention to tZJxg2 30 1lVxg2 'it'f8 31 ~xg4 ':xe4 32 l:!.d1,
the fact that my clock was running: accord- when the battle is not yet over. As we will
ing to the FIDE rules (in contrast to those see, Karpov now missed a direct win, but
287
Kasparov vs. Karpov: 1986-1987
with White in such time-trouble this does success with White. After all, he was in the
not play any significant role. lead, which had not happened since the
27 ... bS start of the second match (when my oppo-
A decisive advantage would more sim- nent won the Sth game). But the match was
ply have been gained by 27 .. .lhe4 28 ~xfS a long one, and Karpov remained true to
i.f3! and then 29 ~xe4 i.xe4 30 i.xe4 ~xa2 himself, deciding for the moment to 'digest'
31 i.xb7 as. his win. His choice of opening in the 3rd
28 tZ:ld2 ~d3 29 tZ:lb3 i.f3 30 i.xf3 ~xf3+ 31 game was something of a surprise: he again
'iitgll::txel+ 32 ~xel tZ:le3 opted for the quiet variation with g2-g3
against the Griinfeld.
Game 30
A.Karpov-G.Kasparov
World Championship Match,
3rd Game, Seville 16.10.1987
GriJnfeld Defence 079
1 d4 tZ:lf6 2 c4 g6 3 g3
Karpov obviously liked his position
from the 1st game, as he deferred until later
And White lost on time, without even the new weapon he had prepared, which
managing to resign (0-1). Times: 2.30-2.1S. was to cause me considerable trouble - the
variation with 12 ..I1I..xf7+ (Game Nos.32, 34,
It need hardly be said what a state of 36, 38). And indeed, why not try once again
shock I was in, having at the start repeated to gain an advantage without any particu-
Karpov's anti-record, established by him in lar risk? But on this occasion I accurately
the 8th game of the previous match. My solved my opening problems, and the
failure in the opening was also painful and move 3 g3 did not occur again in the match.
left me feeling desperately disillusioned. 3 ... c6 4 i.g2 dS 5 cxdS cxdS 6 tZ:lf3 i.g7 7
However, I urgently needed to come to and tZ:lC3 0-0 8 tZ:les e6 90-0 tZ:lfd7 10 f4 (10 tZ:lf3
compose myself for the next game. - Game No.3) 10 ... tZ:lc6 11 i.e3 tZ:lb6 (07)
From the press: 'At the first two games This plan still has the right to exist, but
there were, frankly speaking, few people in the 11 .. .£6 or 10 ... f6 (Game No. 13) is nevertheless
auditorium, but beginning with the third there better, driving away the knight, as Black
was a full house. There was no doubt that the mainly plays nowadays.
"blame" for this lay with the win by the ex- 12 ..I1I..f2 tZ:le7
champion, who employed a prepared line pre- An attempted improvement, directed
served in his bank of ideas since 1980, and the against the plan with e2-e4 (12 ... i.d7?! 13 e4
fact that in time-trouble the champion forgot to tZ:le7 14 tZ:lxd7! ~xd7 IS eS - Game No.28).
press his clock button.' But in principle Black is preparing the same
It could have been assumed that, after regrouping: ... i.d7-c6 and ... tZ:lb6-c8-d6.
his striking victory with Black, Karpov 13 a4 (08)
would immediately try to build on his The natural reaction, and we expected it.
288
The Fourth Match: 1987
There is no longer any point in playing 13 of 15 ctJxd7 ctJxd7 16 'iYxb7 I'tb8 17 'iYa6
e4, since after 13 ... dxe4 Black has solid I'txb2 (17 ... I'tb6 18 'iYd3 I'txb2 is also possi-
control of the d5-point. ble, Bareev-Svidler, Kazan 1995) Black
But later we decided that the immediate obtains precisely the position for which he
13 'iVb3!? was more promising - Azmai- is aiming: \-Vhite's two bishops do not play
parashvili played this against Alterman any role, and the b4-point is not at all less
(Struga 1995) and Eljanov (Ohrid 2001). If important than the bS-point.
13 ... .id7 White replies 14 I'tfdl .ic6 15 IS e4 dxe4 is also harmless, for example:
I'tac1, unhurriedly preparing e2-e4. 16 .ixe4 .ic6 17 ctJxc6 bxc6! (an idea known
since the times of Steinitz), or 16 ctJxd7
ctJxd7 17 ctJxe4 (17 .ixe4 ctJf6! Makarychev)
17 ... ctJb6, and Black's control of d5 relieves
him of any problems.
ls ... .ic6 (04) 16 ctJbs
Karpov has improved his position as
much as he has been able to by making all
the active moves.
289
Kasparov vs. Karpov: 1986-1987
290
The Fourth Match: 1987
1 c4
Without any doubts! We didn't prepare
1 c4 for the match, in order that after one
novelty by the opponent, prepared for his
battles with Korchnoi, I should be scared
away from this move. Of course, the strug-
gle with the black e-pawn contained a
considerable risk, but how otherwise to try
and win?
1 ... ct:Jf6 2 ct:JC3 e5
I think that Karpov was expecting a
repetition of the opening duel (in our
Draw agreed on Black's proposal (Yz-Yz): matches I always began by playing the
after 30 ct:Jd2 f5 the fighting resources are main system I had prepared with White)
exhausted. Times: 2.01-2.16. and he did not arrive for the game empty-
handed.
By confidently gaining a draw with 3 ct:Jf3 ct:Jc6 4 g3 ..tb4 5 ..tg2 0-0 (04) 6 0-0 e4
Black, I thought I had achieved my playing (10)
objective. I seemed to have overcome the Again, as in the 2nd game, this came af-
crisis which arose after the 2nd game. ter some hesitation - perhaps 6 ... Me8 (Game
Karpov also did not look upset, since he No.43) should have been played?
was still leading. However, he must once 7 ct:Jg5 ..txc3 8 bxc3 Me8 (04) 9 f3! exf3?! (03)
again have realised that he was no longer In the press centre this exchange pro-
superior to me in quiet positional play. This voked a lively debate: many were perplexed
had been shown in the previous match, and as to why Black should avoid the new - and
in the present one Karpov's traditional better! - move 9 ... e3, which had been suc-
trumps - endgame technique, subtle ma- cessfully tried in the 2nd game. It seems to
noeuvring, skilful handling of simple me that the sharp position after 10 d3 was
positions - altogether ceased to bring him not to Karpov's taste. He did not want to
dividends (and it was no accident that in check what improvement the opponent had
this style I won against him the decisive, prepared (d. Game No.29, the note to White's
24th game). 17th move), and already in the course of the
On the Saturday and Sunday I relaxed match, with limited time for preparation, he
and prepared, tuning myself up for a large- found an interesting palliative (although he
scale battle in the 4th game. And my expec- did not play 9... exf3 immediately, and
tations were justified. somehow not very confidently).
291
Kasparov vs. Karpov: 1986-1987
292
The Fourth Match: 1987
293
Kasparov vs. Karpov: 1986-1987
17... d6 18 ct:Jxg7! .l:!.f8(e7) 19 ct:Je8! etc.). It This means that the queen must keep
appears that 15 ... d5 16 exd5 ~e2 solves the knight at f6 covered. However, 13 ... ~d8
Black's problems, but after 17 ct:Jh6+ 'it'f8 18 leaves the formidable white bishop alive -
~xe2 .l:!.xe2 19 g4! ct:Jd6 20 c4! even without 14 Jia3! (Dlugy) 14 ... d6 15 'iVe2 ct:Jc5 16 d3
the queens White's two bishops ensure him with an obvious advantage. If 13 ... ~e6
a dangerous initiative; there is the strong reply 14 'iVe2! ct:Jxc1
b) 13 ... ~e6 (here Makarychev 'did not (14 ... ~xc4? 15 ct:Jh6+ 'it'f8 16 .l:i.xf6 gxf6 17
discover any decisive threats') 14 d3 ct:Jd6 Jif1! Najdorf) 15 .l:!.axc1, when White's
15 ct:Jd4 ~e5 (15 ... ~e7? 16 e4 c5 17 ct:Jf3 is position is even better than in the game
worse) 16 Jid2 with excellent compensation (15 ... d5?! leads to the loss of a pawn: 16
for the pawn, but consideration should be cxd5 ct:Jxd5 17 ~g4 'YlUg6 18 ~xg6 hxg6 19
given to 14 ... ct:Je5!? 15 d4 tLlg6 or 15 e4 ct:Jxg7!). And if 13 ... ~e5, then 14 ~c2! ct:Jxc1
(Dlugy) 15 ... dS!, although after 16 d4! White 15 .l:!.axc1 (simpler than Makarychev's
retains somewhat the better prospects. interposition 15 d4!?) 15 ... d5 (15 ... g6 16 ct:Jd4
The quiet 12 ... d6!? 13 d3 c5 was also or 15 ... d6 16 .l:!.f4 does not help) 16 cxd5
possible with the idea of 14 tLlfS gxf5 15 ct:Jxd5 17 e4 ct:Je7 18 d4, and under the cover
.l:!.xf5 d5! 16 cxd5 c4. Even so, after 14 tLle2!?, of his powerful pawn centre White begins
intending ct:Jf4, e3-e4 and h2-h3, White's preparing an attack on the kingside.
chances are rather better. 13 ... ct:JXCl (02) 14 .l:!.axcl d6 15 .l:!.f4! (10)
Instead of this Karpov took a risky deci- White doubles rooks on the f-file, hop-
sion, leading to the loss of precious time. ing to prevent the knight from moving
from f6, and with the combined attack of all
his forces to gradually destroy the black
king's fortress.
13 ~e2 (34)
At the board the obvious attack 13 ct:Jf5!?
seemed premature to me, although it was
also unpleasant for Black. Where should 15 ... c6 (02)
the queen go? Many commentators sug- Preparing the development of the c8-
gested 13 ... ~c5(?), overlooking 14 ct:Jh6+!, bishop. 'For carrying out a regrouping
when the alternatives are all bad for Black - Black is all the time one tempo short: if
14 ... gxh6 15 .l:!.xf6, 14 ... 'it'h8 15 ct:Jxf7+, or 15 ... ct:Jd7 there follows 16 ct:Jf5, and there is
14 ... ~f8 15 .l:!.xf6! gxf6 16 gd5 with an no better reply than 16 ... ~f8' (Makarychev).
irresistible attack (16 ... c6 17 ~f3! 'it'g7 18 And then 17 .l:!.cfl with increasing pressure.
'ii'h5). In the press centre 15 ... c5?! 16 ct:Jf5 Jixf5
294
The Fourth Match: 1987
17 l:txf5 was also studied, but in the end it l:txf7 'iVbl+ 24 i.f1, and Black faces a battle
was deemed that 'the weakening of the d5- for a draw.
square is fatal' (Andersson). In Illjormntor Dorfman suggested
16l:tcfl (04) 16 ... 'iUes (18) 17 ... l:tf8!? with the idea of 18 ttJf5 i.xf5 19
Prophylaxis against the capture on f6. If :xf5 'iVe7 and ... ttJd7. Then White could
16 ... i.e6 there could follow 17 l:txf6! gxf6 18 have chosen between the standard ex-
i.e4! with a dangerous attack: 18 .. .f5 (but change sacrifice 20 l:txf6!? gxf6 21 l:tf4 with
not 18 ... i.xc4? 19 i.xh7+! or 18 ... 'iUf8?! 19 quite good compensation, and 20 l:t1£4!, in
l:tf4!) 19 i.xf5 i.xf5 20 ttJxf5 'iUf6 21 ~'h5 order after 20 ... ttJd7 to retain the initiative
(Dorfman). by 21 l:te4 ttJe5 22 c5!. In addition, 18 a4 is
In the event of 16 ... i.d7 nothing is given possible, not hurrying with ttJf5.
by 17l:txf6 gxf6 18 'iVh5 in view of 18 ... 'it'h8! 18 ttJfS (06) 18 ... i.xfs (02) 19l:txfS
19 l:tf4 l:tg8, but 17 ttJf5 i.xf5 18 l:txf5 is
good, as is Makarychev's suggestion 17
g4!? with the threat of g4-g5 (exploiting the
fact that the bishop at d7 is occupying the
knight's retreat square).
19 .. :iVe6
A quick reply. This move seemed to me
to be a mistake, aggravating Black's diffi-
culties. Like many of the commentators, I
examined 19 ... 'iUe7 20 'iUd4 and after 20 ... l:tf8
17 'iUd3! (14) with the idea of ... ttJd7 (but not 20 ... 'iUf8?! 21
The queen joins the attack. To judge by g4! h6 22 g5 hxg5 23 l:txg5) I did not see any
the time spent on his reply, Karpov had particular advantage for White:
underestimated this manoeuvre. It turns 1) 21l:txf6 gxf6 22 l:txf6l:tae8 23 .l:.xd6 (23
out that Black faces difficult problems. i..h3 'iUe4!) 23 ... l:td8 24 c5 b6! with equality,
17 ... i.d7 (17) or 23 'iUg4+ 'it'h8 24 'iUf4 'iUe5 (d. variation
Black completes his development. 2);
17 ... 'iVh5 is too risky - 18l:txf6 gxf6 19l:txf6 2) 21 'iVh4 .l:.ae8 22 l:txf6 (after Suetin's
merely gives White solid compensation for move 22 l:t1£4 'with an appreciable advan-
the exchange, but after 18 ttJf5! ..i.xf5 19 tage' Black equalises by 22 ... ttJd7! 23 'iUxe7
l:txf5 it becomes uncomfortable for the l:txe7 24 l:td4 ttJe5) 22 ... gxf6 23 .:::l.xf6 (23 .:::l.f5
queen on the kingside: 19 .. .'iVg4 20 e4! or 'it'g7! Makarychev) 23 ... 'it'h8 24 ~'f4 '~e5 25
19 ... 'iUg6 20 'iUxd6 (20 'iUd4 l:te6 21 a4 is 'iVh6 'iUe7! 26 l:txd6 l:tg8 27 c5 (27 '~'f4 .:::l.g4!)
steadier) 20 ... ttJe4 (not 20 ... l:tad8? 21 'iUf4 27 ... l:tg6 28 'iUf4 'it'g7 29 ~f1 'iVe5 30 l:td7
l:txd2 22 l:tg5) 21 'iUf4 ttJxd2 22 l:tf2 l:tad8 23 l:te7, and Black should be able to hold out.
295
Kasparov vs. Karpov: 1986-1987
However, instead of 20 "iVd4 there is the iVe5 27 1ixd6, but 23 ... ~f8! 24 MfS iVg7 25
much stronger 20 .l:tlf4!. Now 20 .. .lH8? is ~g5 "i'ih8 26 "ilif4 .l:i.e5 - G.K.). After heated
not possible on account of 21 .l:txf6 gxf6 22 debates it was decided that here it would
Ile4! "YWd7 23 .l:tM f5 24 ii.h3 and ii.xf5, have been very hard for White to convert
winning, nor is 20 ... 'iVf8? in view of 21 .l:tM! his advantage. Kasparov's reply is much
and .l:i.xf6. And in the event of 20 ... h6 White stronger.' (Suetin)
continues to build up the pressure: 21 "YWe2
(21 iVf1!?) 21...'it'h7 (21...4Jd7 22 lhf7 can
hardly be correct) 22 e4 .l:tad8 23 d4 b6 24
'iVe3 etc.
20 "iVd4! (04)
21 'iVh4! (08)
It would appear that Karpov had only
expected the exchange sacrifice. Now,
however, White has decisive threats.
21 ... 4Jd7 (24)
20 ...l:te7? (07) The result of an agonising think. There
With the vain hope of solving Black's is nothing better: if 21...4Je8, then here too
problems by ... 4Jd7. It was psychologically 22 ii.h3 (22 ... h6 23 ~b5!); after 21.. ..l:tae8
difficult to decide on a second successive apart from the simple 22 .l:tlf4 White now
queen move, but 20 .. :iWe7 was necessary. wins by 22 ~xf6 gxf6 23 .l:txf6 "iVe5 24 ii.h3!;
Since the white queen has left the d3- it is also hopeless to play 21...4Je4 22 ii.xe4
square, 21 .l:tlf4 is no longer so terrible, 'iVxe4 23 'iVxe4 .l:txe4 24 .l:txf7 l:tb8 25 l:tc7
although even here the threat of ~xf6 l:txc4 26 l:tff7 l:tg4 27 l:tfd7 d5 28 'it'g2 a6 29
condemns Black to a difficult defence. The h3 l:tg6 30 l:txb7 etc.
immediate 21 .l:txf6!? gxf6 22 .l:txf6 is also 22 ii.h3! (03) 22 ... 4Jf8 23l:t5f3! (04)
interesting, for example: 22 ....l:tad8 23 ii.f3! 'An accurate and rational decision -
or 22 .. .'i'e5 23 ·~·g4+! 'it'h8 (23 ... 'it'f8 24 .l:th6) White transposes into a winning endgame,
24 .l:txf7 .l:te7 25 .::i.f5 tlg7 26 .i'f3 with excel- although the mass of tempting possibilities
lent play for the exchange - but not yet a was simply dazzling!' (Suetin). And in-
win. deed, the spectacular 23 l:tb5!? 4Jg6! 24
'After 20 ... .l:te7 the press centre began ii.xe6 4JxM 25 l:txf7! suggested itself, with a
buzzing like a beehive. The oldest grand- simple win after 25 ... l:txe6? 26 l:tbxb7 4Jg6
master Najdorf insistently suggested 21 27 l:txg7+ 'it'f8 28 l:tbf7+ 'it>e8 29 l:txh7, but
.l:txf6?! gxf6 22 l:txf6 'iVe5 23 .l:txd6 (if 23 25 ... .l:txf7! is more tenacious: 26 ii.xf7+ 'it'f8!
"VJfIg4+, then not Dorfman's variation 23.Aith8 27 l:!.xb7 4Jf3+ 28 'it'g2 4Jxd2 (B.Kristensen)
24 nfS "VJfIe6? because of 25 "YWd4+! f6 26 .l:txf6 29 .l:tc7!, when White nevertheless retains
296
The Fourth Match: 1987
28 Mxb7? (05)
It was a pity that in impending time-
trouble I did not take my idea to its logical
conclusion (yet another sign of my indiffer-
ent form!). 28 c5! dxc5 29 d6 was correct,
27 dS!? (02) with a straightforward win: 29 ... Md8
This leads to a clear win, but 27 ~xb7! (29 ... Md3? 30 Mxf8+ or 29 ... g6 30 d7 also
would also have been quickl~- decisive. The loses) 30 Me7! Md3 31 d7! tiJxd7 32 i.e6+
commentators unanimousl\' condemned 'it>h8 33 i.xd7. Now, however, the battle
this move in view of 27 ... :xc3 28 :ff7 g6, continues.
thereby committing no fewer than three 28 ... cxdS 29 cxdS lI3e7! 30 Mfbl
mistakes! First, White wins by 29 ~f1! Ma3
30 i.g2 or 29 ... Mc1 30 :f6. Second,
28 .. Jhc4!? 29 Mxg7+ ~h8 is more tenacious,
although here too after 30 ~f5! :xd4 31
i.xh7 it is probable that Black is unable to
save himself. But the main thing is that
White has the fantastic resource 28 :c7!
Mxc4 29 Mcf7!! - the leap of this particular
rook to f7 is hard for a human to spot'
Black's resignation is in order: there is no
defence against Mxf8+ and i.e6+. Things are
also not helped by 28 ... Md3 29 :xc6 :xd4
on account of 30 i.g2! with the threat of
i.d5+ and Mxd6. 30 ... hS? (02)
27 ... Mae8 (06) 30 ... 'it>f7! was more tenacious, with the
27 ... Mxc3? 28 Mxf8+ was completely bad, idea of 31 'it>f2 'it>f6! - here Andersson
but Black would also not have saved him- assessed the chances of White winning and
self with 27 ... cxd5 28 i.g2!, 27 ... c5 28 Mxb7 of Black drawing as 'fifty-fifty'. Of course,
Mxc3 29 Mff7, or 27 ...Mb8 28 c5! dxcS 29 d6 White has numerous technical difficulties
g5 (29 .. .l:td8 30 Me7! - cf. the following note) to overcome, but after 31 a4!? Mxb7 32
30 Mc7 lId8 31 d7 ~g7 32 Mc8. Mxb7+ Me7 33 Mb8 Mc7 34 as he would have
297
Kasparov vs. Karpov: 1986-1987
the leaders of the two delegations. They dis- my opponent would employ against the
cussed the possibility of deferring the start of a Griinfeld. It appeared that I was beginning
game in the event of "the unforeseen delay of to run into form, and I wanted to aim for
one of the contestants". The point was that it counterplay.
had become difficult for Karpov to reach the 3 CDC3 d5 4 cxd5
Lope de Vega Theatre. At the villa where he was A match novelty - the main variation!
staying an aerial telephone line had been set lip Up till then Karpov had played only 4 ..Iii.f4
for a direct link with Moscow, and the inhabi- or 4 CDf3 (which we looked at in the first
tants of the local village, on learning about this instance), as well as, a move earlier, 3 CDf3
from the press, rebelled - such a thing had Heuer or 3 g3.
happened there before! They began writing 4 ... CDxd5 5 e4 CDxC3 6 ..Iii.xC3 ..Iii.g7 7 ..Iii.c4 C5 8
protest letters, including some to Karpov, and CDe2
above all they threatened to picket his villa, to
try and stop him leaving for a game. They were
unexpectedly supported by the workers at a
tobacco factory close to the playing venue: they
were intending to organise pickets directly by
the entrance to the theatre ... Fortunately, it did
not come to that.'
'Round about that time Miguel Najdorf
said: "For the moment one of the contestants is
playing badly, but the other is playing even
worse! If Kasparov can keep the scores level in
the first half, he will easily win the match." The
first remark was close to the truth, but with his
forecast the venerable grandmaster was wrong.' We analysed this line in our prepara-
Suetin called the extremely tense and tions for the third match, but we thought
nervy 5th game 'really eccentric, and with an that it was not the most appropriate
interesting psychological implication, which weapon for Karpov: the positions that
became apparent in the opening'. Alas, on result here are too complicated and unclear.
encountering another surprise by the Besides, as Makarychev aptly commented,
opponent, I again ended up in terrible time- 'tm'>'ards the mid-1970s the variation with 7
trouble, which decided the outcome of the ~c4 and 8 CZJe2 went out of fashion for a
battle. long time, its status remaining unchanged.'
8 .. .tt:Jc6 9 ..Iii.e3 0-0100-0 ..Iii.g4
A worthy alternative to Smyslov's plan
Game 32 with 10 .. .'iVc7 (Game Nos.91, 99 in Volume II
A.Karpov-G.Kasparov and Game Nos.70, 79 in Volume III of My
World Championship Match, Great Predecessors.).
5th Game, Seville 23.10.1987 The idea of the immediate bishop move
GriJnfeld Defence 087 to g4 belongs to Timman - earlier they
played 10 ... cxd4 11 cxd4 ..Iii.g4 12 f3 CDa5,
and after 13 ..Iii.d3 ..Iii.e6 one of the main
1 d4 CDf6 2 c4 g6 tabiyas was reached, while the rare con-
I was curious to know what variation tinuation 13 ..Iii.xf7+ .:txf7 14 fxg4 was con-
299
Kasparov vs. Karpov: 1986-1987
sidered harmless (d. the note to White's 'does not give any advantage', and 'Black's
12th move). The presence on the board of position fully compensates for the sacri-
the c5- and c3-pawns increases Black's ficed pawn'. The same verdict was given by
defensive resources somewhat and reduces Botvinnik and Estrin in their monograph
White's attacking potential. on the Grunfeld Defence (1979) and by
11 f3 (I would have been quite happy with Karpov himself in the Yugoslav Encyclopae-
the gambit variation 11 d5 ctJa5 12 .I1i.d3 c4 dia of Chess Openings (1976). And now at the
13 .I1i.c2 .I1i.xc3 14 ~bl Polugayevsky- board I was forced to try and understand
Timman, 6th match game, Breda 1979) why Karpov had nevertheless captured on
11...ctJa5 f7.
12 ... ~xf7 13 fxg4 ~xf1+ (03) 14 'it'xf1 (03)
Here I 'went to sleep' for more than an
hour - the second longest think in the
match after my 83-minute record in the 2nd
game. And again I managed to find the best
plan!
14... 'iVd6 (64)
A new direction. I decided fairly quickly
not to relieve the pawn tension in the
centre, but a mass of time was taken up by
all kinds of doubts and hesitations, typical
of an unfamiliar situation.
14 ... cxd4 15 cxd4 could have led to a po-
12.11i.xf7+!? sition that had been known for a long time.
But here is a surprise - the old 'harmless
continuation', revived in a new version by
Igor Zaitsev.
Previously only 12 .I1i.d5 .I1i.d7 had oc-
curred, as in Timman's games with Hort
(Niksic 1978), Spassky (Montreal 1979) and
Polugayevsky (Tilburg 1985), or 12 .I1i.d3
cxd4 13 cxd4 .I1i.e6, transposing into the
main lines. Here we had looked at both the
exchange sacrifice - 14 d5 .I1i.xal 15 'iVxal f6
(Game No.55 in Voilime II of My Great Prede-
cessors) and the sacrifice of the a2-pawn -14
lid .I1i.xa2 15 'iVa4 .I1i.e6 16 d5 .I1i.d7 17 'iVb4 e6
(these analyses later came in useful in my Analysis diagram
games with Beliavsky and Yusupov in the
55th USSR Championship, Moscow 1988). At one time they played 15 ... 'iVd7 (think-
The capture on f7 surprised me, of ing about how to deviate from theory, I also
course. All of us Soviet players grew up on examined both the currently fashionable
Kurs Debyutov by Panov and Estrin, where 15 ... e5, and 15 ... 'iVb6 - Game No.36) 16 h3
in black on white it was written that 13 'iVe6 17 'iVd3 'iVc4 18 'iVd2 'iVa6! 19 'iVc2 ctJc4
.I1i.xf7+ (after the exchange of pawns on d4) 20 'iVb3 'it'h8 with good counter-chances for
300
The Fourth Match: 1987
Black (Spassky-Korchnoi, 22nd USSR Later Karpov also introduced 15 Wgl (Game
Championship, Moscow 1955) or 18 No.38). 15 'iYa4 'iYxh2!? (15 ... 'iYa6 is steadier)
'iYxc4+(!) iLlxc4 19 iL.g5 e6 20 :adl b5 with 16 'iYxa51:!.f8+ 17 WeI 'iVhl+ 18 iLlgl 'iVxg2 or
equality (Encyclopaedia of Chess Openings). 18 ... iL.h6 (Khenkin) would have been exces-
But this last assessment seems question- sively sharp.
able: after 20 e5!? or 20 :abl!? with the idea
of e4-e5 it is not to easy for Black to find
sufficient compensation for the pawn.
19 ... e5 20 d5 h6 21 iL.c1 iLld6 22 CZJg3 :ac8
(Dlugy-Nickoloff, Toronto 1989) is hardly
any better in view of 23 ~a3! and :acl. And
besides, instead of 16 h3 there is the logical
move 16 g5!?, which had already occurred
but was not yet favoured by theory.
After the immediate 14 ... 'iYd7 (or
14 ...'iYc8 Portisch-Korchnoi, Reykjavik 1988)
again 15 g5 is not bad, for example:
15 ... 'iYe6 16 e5 (shutting in the bishop at g7)
16 ...'iYc4 17 Wgl :ad8 18 '~el with a compli- lS ... li'ds (10)
cated game (Karpov-Gavrikov, Gijon In Informator Zaitsev recommended
(rapid) 1988). 15 .. .'~!Ve6(?!) with the idea of 16 g5 CZJc4, but
Today, casting my eye over the entire 16 CZJf4 l:tf8 (16 ... li'c4+ 17 li'e2) 17 Wgl li'c4
tree of variations which developed after 12 18 g5! is stronger. But now after 16 iLlf4
iL.xf7+ became popular, I come to the con- Black has the equalising 16 ... l:tf8 17 'ti;>gl
clusion that the best reply was' articulated' li'e4!. The black queen controls the impor-
in this, the very first published game. tant light squares, and I was intending to
'The world champion devised a plan exploit this factor, although as yet I had
which exploits an important nuance - the only a vague impression of what would
presence of the c-pawns (for the moment happen next: the position is very unclear.
the exchange on d4 has not been included). 16 .if2 (10)
With the position more closed, Black was By our fifth match (1990) this variation
able to keep control of all the points of had already been thoroughly studied, but it
potential counterplay for the opponent, did not in fact occur there, although against
thus retaining excellent compensation for Beliavsky (Linares 1992) I was able to
the pawn. And although many commenta- demonstrate a clear-cut draw after 16 g5
tors saw the reason for Kasparm"s defeat to li'e4! 17 iL.f2 l:tf8 18 CZJgl (18 Wgl 'iVf5!)
be his hour-long think, I find it hard to 18 ... CZJc4 19 iLlf3 CZJe3+ 20 iL.xe3 'iVxe3 21
suggest that the loser acted unwisely. The 'iVb3+ 'it'h8 22 l:tel 'iVxg5 23 'iYxb7 'iVd2! 24
favourable character of the subsequent dxc5 (24 li'xe7 li'xc3 with equality, Onis-
middlegame was worth the boldly sacri- chuk-Grischuk, Poikovsky 2005) 24 ....~xc3
ficed time.' (Makarychev) 25 'iVd5 iL.h6 etc.
is eS (11) 16 ... l:i.f8 (12)
The opponent's hesitation was probably Probably the strongest move. However,
caused precisely by the fact that the c5- and during the match we had doubts about this,
c3-pawns have remained on the board. and I switched to 16 ... l:td8 (Game No.34).
301
Kasparov vs. Karpov: 1986-1987
17 'it'gl (05) 17 ... ~h6! (01) during my preparations for the 7th game, it
'Strangely enough, Black's 14th move is possible that the entire variation with 12
determined the actions of the two sides up ~xf7+ would have had a different fate.
to the 20th move. White was threatening to 18 ... 'iVf7 (02) 19 ~g3
shut the g7-bishop out of the game by g4- The further activation of the enemy
g5. In defending against this threat, Black, pieces has to be allowed: in the event of 19
in turn, threatens to achieve domination in 'iVfl? (into a pin!) 19 ... ~d2! 20 dxc5 tiJc6 or
the centre. Thus 18 'iVd3 is weak because of 20 Mdl tiJc4 21 tiJg3 tiJe3 22 ~xe3 kxe3+ 23
18 ... tiJc4, intending ... tiJxe5. Therefore, by ~h2 'iVxa2 Black simply dominates.
renewing his previous threats, White 19 ... ke3+ 20 ~h2
temporarily forces his opponent to move
away from the centre, i.e. to carry out an
unusual, active defence.' (Makarychev)
302
The Fourth Match: 1987
nature, to create a complicated position on equal game after 22 ... .S!.xcS 23 ctJd4 ~xa2
the board. I think that the main thing for (and if 24 hS, then 24 .. :ti'dS!), or 22 ... .S!.xcS
Karpov was to "drive" his opponent into 23 ctJgl 'iYe4! (but not Zaitsev's recommen-
time-trouble. There was a clash between dation 23 ... .S!.f2? because of 24 .S!.xf2 ~xf2 2S
two fundamentally different approaches to ctJh3 ~f8 26 iVd7 and wins) 24 ctJf3 ctJc4. But
the fight - the classical and the purely although the 'psychological' rook move
competitive. At the same time Karpov was maintains the intensity of the struggle, it
relying on his defensive skill and, of course, does not take into account the dynamics of
was taking a risk (in fact he was in the the position and places White on the verge
"danger zone" almost to the very end of the of defeat. I think that Karpov underesti-
game). But he was very consistent in the mated the strength of my reply.
implementation of this strategy.' (Suetin) 22 ... iVdS! (08) 23 iVd3 ctJC4 24 ~bl
21 ... b6 (OS) The critical moment of the game and of
Of course, not 21..."iVxa2? 22 ~al '\'jb3 23 the entire initial stage of the match.
"iVd3 ctJc4 24 dxcS when White \vins. And
after 21...~d8 22 "iVel! all the same Black has
to play 22 ... b6, defending his knight.
303
Kasparov vs. Karpov: 1986-1987
reply 25 ':d1! with the sequel 25 ... gxh4 26 ttJd2 39 'iih3+ White should be able to save
~xh4 ~f2 27 'iih3! "iVe4! 28 ~xe7! (but not the game.
28 ttJg3 "iVf4! 29 ~xe7 ttJe3!, winning). In an interview after the match I claimed
that 24 ... g5 would have won the game. This
was probably an exaggeration, but, as is
evident from the variations given, the
resulting difficult problems would have
demanded enormous inventiveness on
White's part. In my opinion, it would be
unrealistic to expect to find all these moves
with the clock ticking away.
25 'it>h3?! (17)
It seems that Karpov 'took me at my
word' and decided that 25 ':xb5 would not
do. In fact it was a sure way to equalise,
since 25 ... ttJxe5 26 ':xc5! (of course, not 26
Analysis diagram ~xe5? ':f2) 26 ... ttJxg4+ (and not 26 .. :~xc5?
27 ~xe5) 27 'it>h3 "iVe6 28 'ii'c4 (28 '>/lib 1 ?
Here in Informator Zaitsev gave the very ~gl!) 28 ... ttJf2+ leads to perpetual check.
interesting variation 28 .. J:H7 29 ~f6 ttJe3! 30 After the game move, the chances are again
g5! (threatening perpetual check - 31 "iVc8+ on Black's side.
':f8 32 "iVe6+) 30 ... ttJg4+! 31 'it>h1 ~xd4! 32
ttJg3! "iVf4 33 'iih5! "iVxg3 34 cxd4 ttJf2+ 35
'it>gl ttJxd1 36 "iVxd1 "iVe3+ 37 'it>h1 'iYxd4 38
'iih5(?) "iVd2! 39 g6 ':d7 'with advantage to
Black'. But the computer suggests that after
38 "iVf3! 'iih4+ 39 'it>gl or 38 ... ':f8 39 "iVe6 the
menacing e-pawn guarantees White a
draw.
This variation can be avoided by
28 ... ':e8, suggested by L.Silaev in the
magazine Shakhmaty v SSSR (1988, No.3).
After 29 ':£1 ttJxe5 30 dxe5 "iVxe5+ 31 ttJg3
~xg3+ 32 '>/iixg3 ':xe7 Black is effectively a
pawn up in the rook ending, but his win- 25 ... a6
ning chances are slight. A continuation of the light-square strat-
28 .. .'ihe2!? also comes into considera- egy. Georgadze in the match bulletin and
tion. The forced 29 'i116! :f7 30 e6 '>/iixd1 31 Suetin in Shakhmaty v SSSR recommended
exf7+ 'it>xf7 32 "iVf8+ 'it>e6 33 '>/iixf2 'it>xe7 34 the 'energetic' 25 ... b4, but after 26 cxb4 cxd4
'iih4+ 'it>e6 35 "iVh6+ 'it>d5 36 ''iin5+ 'it>e4! 27 ':e1 White's defences would have held
(36 ... 'it>d6 37 'iih6+) 37 "iVxh7+ 'it>e3 leads to a by keeping an eye on the newly-created
position which I have analysed a great deal. passed pawn on d4.
After 38 dxc5 "iVxg4 39 c6 "iVf4+ 40 'it>h3 ttJd6 26 ttJgl! (04)
Black succeeds in converting his extra piece Probably the best decision - an attempt
(41 c7? ttJf5), but with the accurate 38 g5! to create at least some counterplay and 'an
304
The Fourth Match: 1987
unexpected resource, particularly in view we each had less than 10 minutes for 12
of the lack of time for the opponent, who of moves.
course wanted to refute this strategy.' 29 'ivxc3 'iVe6! (02)
(Makarychev) A very strong move, with which Black
retains hopes of an advantage. I had no
desire to switch on to a drawing path by
29 ... 'iVd3.
305
Kasparov vs. Karpov: 1986-1987
White can feel comfortable. pion has the right to lose his bearings'
(Makarychev). Which is exactly what
happened! In any sort of time-trouble Black
should have played 33 ... h5!, easily main-
taining the balance: 34 "iVc4 (34 .i.e 1 ?! Md5!;
34 "iVaI tLld2) 34 ... "iVxc4 35 ~xc4 ct:Jxg3 36
'it'xg3 a5.
'The press centre froze: everyone saw
that White had 34 ~xa6 "iVxa6 35 "iVb3+ 'it'g7
36 "iVxdl tLlxg3 37 'it'xg3 with an extra pawn
and quite good winning chances. To uni-
versal surprise, Karpov played differently.'
(Suetin). The ex-champion also rejected
another endgame with an extra pawn -
31 ... tLld2 after 34 .i.f4 (winning the g5-square for the
'Black had based his tactical operation knight) 34 ... .i.d2! 35 .i.xd2 tLlxd2 36 tLlg5
on this manoeuvre. The threat is 32 ... tLlf1 + "iVb3 37 "iVxb3+ ct:Jxb3 38 Mxa6.
33 'it'h3 h5. But for the moment the g4- It is probable that this advantage
pawn is defended by the rook along the 4th seemed insufficient to Karpov, and - leav-
rank.' (Suetin) ing himself with just a couple of minutes
32 ~xa4 tLlfl+?! (01) for 6 moves! - he made a move that was far
I should have reconciled myself to a more dangerous and unpleasant for Black,
draw - 32 ... tLlxf3+ 33 gxf3 .i.d2 34 "iVc4 "iVxc4 exploiting the 'hanging' state of his pieces.
35 ~xc4 a5 or 34 "iVaI a5 35 ~d4 ~xd4 36 34 "iVa! (07) 34... MCl (34 ... Md8 35 "iVe2!)
~xd4 .i.b4, but I was loathe to deviate from
the planned course.
33 'it'h3
35 "iVe2?
A typical time-trouble error. White
could have gained an almost decisive
33 ...~dl? advantage by 35 "iVd3! h5 36 .i.el g5! (there
Again a rapid reply 7 moves before the is nothing else) 37 hxg5 ~al! 38 "iVc4!,
time control, and a completely senseless transposing into an endgame with an extra
move, which simply blunders a pawn. 'In pawn, which, moreover, is passed.
severe time-trouble even the world cham- 35 ... h5 36 .i.el
306
The Fourth Match: 1987
307
Kasparov vs. Karpov: 1986-1987
Game 33
G.Kasparov-A.Karpov
World Championship Match,
6th Game, Seville 26.10.1987
English Opening A25
1 c4e5
The time for 1...e6 (Game Nos.39, 45, 47, 3 ... g6
51) had not yet arrived. In the light of what follows, it should be
308
The Fourth Match: 1987
309
Kasparov vs. Karpov: 1986-1987
ttJxe6 ttJxe6 16 ~a3 f5 the knight at e3 clearly apparent.' (Suetin). But the time had
comes under attack by .. .£5-f4 come to choose the optimal plan.
(Dzindzichashvili-Barua, New York 1987). 15 'iWa4 (09)
The knight retreat to c3 (with the bishop 15 'i¥b3 (but not 15 ttJe4 ttJc8 and ... f7-f5)
on f5) had not occurred previously, and it 15 ... f5 16 ttJd2 ttJf7 17 'i¥b7 l:tad8 looked
seemingly came as a surprise to Karpov. unclear, and I decided not to exchange the
Whereas before this he had spent only 20 queens - also partly under the influence of
minutes (to my 40), here he thought for a the game with Mnatsakanian, which was
long time. Evidently Black's position ceased won with the queens on.
to appeal to him. 1s ... fs
11 ... ttJe7 (24) The key moment of the game. 16 ... e4 is
threatened.
12 ~a3 (03)
An important move, preventing both 16 ttJd2?!
... d6-d5, and 12 ... 0-0 13 ttJf3 ~h3 (the d- This move, planned in advance and
pawn is hanging). quickly played, proved to be a serious
12 ... 0-0 (03) 13 ttJf3 h6 (04) inaccuracy. As all the commentators re-
Black has no better plan than ... ~e6, marked, 16 l:tb3! was correct, defending the
.. .£7-f5 and ... ttJf7, but he does not want to knight and seizing control of the b-file.
play 13 ... ~e6 immediately because of 14 Black would have had to switch to passive
ttJg5. 13 ... f6 (all the same ... ~e6 and .. .£6-f5 defence. Since 16 ... ttJf7?! exposes the b7-
will have to be played) 14 0-0 ttJf7 (Stohl- square, in Informator I recommended
Pribyl, Stary Smokovec 1988) is somewhat 16 ...'iWc7, but here too the universal reply 17
passive in view of 15 'i¥b3!, and the control l:tfbl is good. 16 ... l:tc8! 17 l:tfbl l:tc7 is more
of the open b-file plus pressure on the c6- solid, guarding the invasion square, al-
and d6-pawns gives White an enduring though after 18 'iWa6 ~f7 19 e3 ttJe6 20 ttJd2
initiative. White undoubtedly retains the initiative.
14 0-0 (03) 14... ~e6 16 ... ttJf7 (22)
Remembering my game with Mnatsa- Now Black succeeds in exchanging one
kanian, I was feeling optimistic: I had pair of rooks. The opening of the game by
managed to win even in that position, and 16 ... e4?! 17 'iWc2 exd3 (if 17 ... d5, then 18 ttJa4
in this one I definitely had better prospects. ttJb7 19 ttJb3 is unpleasant) 18 exd3 is most
'White's initiative on the queenside is probably to White's advantage.
310
The Fourth Match: 1987
311
Kasparov vs. Karpov: 1986-1987
this, and the move 22 ... ~f7). serious. The surprise again caused me to
25 e3 (02) think for a painfully long time in the open-
Some remnants of the former pressure ing, although before the game I had felt
would still have been retained by 25 ct:Jb2 that I was now almost completely alright.
i.c8 26 'iVaI ct:Je6 27 b4, and if 27 ... d5, then
28 ct:Ja4 d4 29 i.b2.
25 •.. i.c8 (14) 26 'iVa5 'iVxa5 27 i.xa5 ct:Je6 Game 34
(now Black has a very comfortable ending) A.Karpov-G.Kasparov
28 i.b4?! (20) World Championship Match,
Again a dubious move! The king should 7th Game, Seville 30.10.1987
have been brought up - 28 ~f1 ~d7 29 b4 Griinfeld Defence 087
g5 30 ~e2.
312
The Fourth Match: 1987
17 .. :'i:lie4?! (03)
;'\Jot the best reply: by chasing after the
g-l-pawn Black wastes time and loses
control of the light squares. Later 17 ... tDc4
(17 ... ith6?! 18 M) 18 g5 b5 19 'it'gl occurred
(Naumkin-Yasnikovsky, Moscow 1991), but
before the 11 th game of the match in Seville
we prepared 17... b5! 18 'itfgl 'iVc4 with the
17 'iVel (03) idea of ... tDc6 and ... b5-b4, undermining the
This solid and rather passive move was d4-pawn (a new argument against the
one that we had hardly looked at, as was exchange 14 ... cxd4 15 cxd4). This, the only
also the case with 17 tDf4?!, to which Black clear-cut way of equalising known to me,
can reply simply 17 ...'iVc4+ 18 'i'd3 'iVxd3+ was, alas, not in fact carried out in practice.
19 tDxd3 cxd4 20 cxd4 tDc6 and ... tDxd4 18 gs (06) 18 ... 'iVfS? (17)
(pointed out in the match bulletin by Geor- An even more dubious undertaking.
313
Kasparov vs. Karpov: 1986-1987
18 .. :iVg4! was correct, and after 19 'Llgl! (if tage. But after other moves there now
19 h4 'Llc4 20 'it'gl, then 20 ... h6!, releasing follows 21 'it'gl with the idea of 'Lle4. It is
the bishop on g7) 19 .. :iVxgS 20 'Llf3 ~fS 21 important that after 20 .. :YWf4 21 Wgl Black
~1! White's chances are only slightly can play neither 21...~xh4? 22 ~e2! and
better. 'Lle4, nor 21...cxd4 22 cxd4 'iVxh4 23 ~4!
19 h4 (12) 'LlxeS 24 ~xe7 'Llc6 (24 ... 'Llf7 2S 'Lle4!) 2S
An obvious and solid move. The sharp ~e6+ Wh8 26 'LlfS! ~f4 27 'Llxg7 Wxg7 28 dS
sacrifice of the gS- and h2-pawns was also l:i:d6 29 ~c8 ct:JeS 30 l:i:el, when White wins.
interesting - 19 'Llg3!? (I underestimated 20 ... ~g4!
this manoeuvre, thinking only about 19 Of course, not 20 ... h6? 21 'Llg3.
'Llgl ~d3+) 19 ... ~xgS?! 20 'Lle4 ~f4 21
'LlxcS! ~xh2 22 ''li:Ve4 'iVhl+ 23 ~gl, and the
powerful centralisation of the white pieces
condemns Black to a difficult defence:
23 .. .'i!<'h5 (othenvise 24 ~e2!) 24 ct:Je6 ct:Jc4 2S
g4! 'i!<'h3+ 26 ~e2 '::e8 27 'i'f3 'i'xf3+ 28 'it'xf3
with a clearly better endgame. But things
are by no means so clear after 19 ... ~d3+
(Makarychev) 20 'it'gl ct:Jc6!, trying to create
counterplay typical of this variation.
21 a4! (07)
'A profound idea' (Suetin). True, after
his mistake on the 20th move White no
longer has much choice: 21 'Llg3?! is weak
in view of 21...cxd4 22 cxd4 l:i:xd4!, while
the standard 21 l:i:bl runs into 21...i.xeS!
(but not 21...b6?! 22 'Llg3!) 22 dxeS ct:Jd2 23
l:i:dl 'Llf3+ 24 'it'f1 'Llh2+ with perpetual
check, or 22 'it'hl cxd4 23 cxd4 i.d6 24l:i:xb7
l:i:f8 and ... a7-aS with full compensation for
19 ... ct:JC4 (12) the pawn (2Sl:i:xa7? 'Lle3! and wins).
The knight Joms the battle, making 21 ... h6! (12) 22l:i:a2 (IS)
White's task more difficult. The undermin- Consistently continuing the plan begun
ing 19 ... h6?! is insufficient in view of 20 with a2-a4. In the event of 22 gxh6? i.xh6
~1! or 20 gxh6 ~xh6 21 ~1!, and if the black bishop breaks free and White's
21...~g4, then 22 'Llgl and ct:Jf3. position becomes dangerous: 23 ... 'Lle3! is
20 'it'gl? (01) threatened, and if 23 ct:Jg3, then 23 ... cxd4 24
But this natural move loses White his cxd4l:i:xd4!.
advantage, which he could have consoli- 22 ... hxgs (13)
dated with the accurate 20 'Llg3!. Now if After 22 ... ~e4?! 23 ~c1! White would
20 ... ~g4?! White has 21 ~e2! - the ex- have succeeded in clinging on 'with his
change of queens is clearly to his ad van- teeth' to the gS-pawn.
314
The Fourth Match: 1987
315
Kasparov vs. Karpov: 1986-1987
Various commentators have suggested (29 ... hxg2?! 30 "iYxg2 "iYxg2+ 31 ~xg2). But
27 ... ~h7(?!), for example: 28 "iYe2 (28 i..xh4 with the queens on Black has nothing to
cxd4 and ...'iYbl + is unfavourable for White) fear, and the accurate 29 ... ~g8! (29 ... b6!?)
28 ... h3!? 29 ~c2 (Zaitsev) 29 ... cxd4 30 cxd4 would have retained equal chances: 30
hxg2 31 "iYxg6+ ~xg6 32 ~xg2 .l:!.c8 or 28 .l:!.xb7 cxd4 31 cxd4 (31 i..xd4 .l:!.f8! 32 "iYe2
J::tb2 h3 29 "iYf1 b6, in each case with equal- .l:!.f4) 31...i..xe5 32 J::txe7 ..Itxd4 33 i..xd4 J::txd4
ity. But with the king on h7 White has the or 10 'iith2 cxd4 31 cxd4 hxg2 32 "iYxg2 "iVh6+
strong move 28 'iith2! - here 28 ... .l:!.g8? is not 33 "iVh3 "iYf4+ 34 Wg2 .l:!.f8 35 'i¥b3+ .l:!.f7 etc.
possible on account of the simple 29 i..xh4, 29 cxd4
and Black fails to equalise with either 'In the press centre they were rather ex-
28 ... b6 29 .l:!.e2 or 28 ... h3 29 g3. pecting 29 ..Itxd4.' (Suetin). But this would
have broken up White's pawn chain, and
after 29 ... h3 Black would no longer have
had to avoid the exchange of queens.
28 .l:!.b2!? (25)
If 28 ~h2 there is now the quiet move
28 ... .l:!.g8! (an ambush!), when 29 i..xh4? is
bad because of 29 ... "iYg4! 30 g3 i..xe5. A 29 ..."iYg4?!
quick draw could have resulted from 28 'Black, probably underestimating the
i..xh4 'in>1+ 29 ~h2 cxd4 30 cxd4 "iYe4 opponent's reply, persistently seeks ways
(threatening ... ..Itxe5+) 31 i..g3 .l:!.xd4 (or to strengthen his position. The normal
31..."iVh7+) 32 "iYc8+ ~h7 33 "iVh3+ (Zaitsev). continuation 29 ... h3 would have guaran-
28 "iYe2 is little better: 28 ... "iVbl + 29 ~h2 "iYf5, teed his safety' (Makarychev). Especially in
and if 30 "iYf3 "iYxf3 31 gxf3 .l:!.f8! with a the event of 30 g3 "iYe4! 31 'iff1 i..xe5 32
drawn endgame. 'iUxh3+ ~g7 and ... .l:!.xd4. White has the
Instead of this Karpov calmly covers the stronger 30 ~f1! ~f5, but here too Black
bl-square with his rook and attacks the b7- holds on after both 31 g3 b6 32 'iUe2!? (Zait-
pawn. Without taking any risks, White sev) 32 ... .l:!.c8, and 31 'ife2!? hxg2 or 31...i..h6
continues playing for a win. (32 .l:!.xb7?! .l:!.c8!).
28 ... cxd4 (05) At this point I had 13 minutes left to the
Slightly premature. Many analysts sug- time control, while Karpov had 37.
gested 28 ... h3!? 29 "iYf1!, avoiding a trap (29 30 'iUf7! (07)
g3?! i..xe5! 30 dxe5? .l:!.dl+ 31 ~h2 .l:!.hl+! Very strong! 'An unexpected manoeu-
with mate) and offering an exchange of vre, especially effective with the opponent
queens which is advantageous to White in time-trouble. Everything is 'hanging'.
316
The Fourth Match: 1987
How does Black cope with the numerous even worse) 35 ~h3 .i.f6 (35 ... 'iYxa4? 36
problems?' (Suetin) 'ud3!, winning) 36 Vi'g4+ (36 'uxb7 is also
30 'iYe2 was much more harmless: possible) 36 ... 'iYxg4+ 37 'it'xg4 b6 38 'ud3 ~f7
30 .. ..lhd4! 31 .i.xd4 'li'xd4+ 32 'li'f2 (32 'it'h1 39 'ud7 with a won endgame .
.i.xe5) 32 ... 'iYd1+ 33 'ifn ~d4+ 34 'uf2 .i.xe5 32 ,Uf2 'iYxes
with a draw.
33,UfS?
30 ...,Uxd4? (07) In the heat of the battle and in the sub-
The trouble is that now if 30 ... h3? there sequent commentaries no one noticed the
is the reply 31 Vi'f3!, and after 31.. .·..'xf3 32 strong move 33 '~'f4! with the persistent
gxf3 'uf8 33 'ub3 Black loses his h3-pawn, idea of exchanging the queens: 33 ... ~5
while after 31...,Uxd4 32 ii..xd.t ·..·xd.t+ 33 (33 .. :'i·d4 34 ~h1!) 34 'iYb8+! (but not 34 'ud2
'uf2! he has insufficient compensation for ~f6 35 l:d7 'iiVa5 36 'iYe3 ~g7 37 'uxb7 'li'xa4
the exchange. with equality) 34 ... ~h7 35 Vi'c7! (the pawn
The only way to maintain equality was is 'poisoned': 35 'iYxa7 'iYd1 + 36 ,Un itd4+),
by 30 ... .i.h6!, a hard-to-find mO\'e sug- and White, by attacking the pawns and the
gested by Zaitsev: 31 'iYf3 (31 ':xb7" ':g8) king, achieves his aim - 35 ... 'iYg5(g4) 36
31...,Uxd4! 32 .i.xd4 'li'xd4+ 33 ':f2 1..e3 or 31 'iiVf4!, 35 ... 'iYd1 + 36 'it'h2 'iYxa4 37 'uf4 'iYb5 38
~xe7 ~d1 + 32 'it'h2 .i.f4+, and \ \'hite gains 'iiVxe7 'iYe5 39 'iVxh4+ ith6 40 g3 'it'g6 41 'ue4
no advantage with either 33 g3 J.xg3+ 'iYf5 42 'iYg4+ or the desperate 35 ... h3 36
(33 ... hxg3+? 34 'it'g2!) 34 ~xg3 hxg3+ 35 'iYxe7 hxg2 37 'iYe4+ ~h6 38 'uxg2 'iYd1+ 39
'it'g2 'iYxd4 36 'iYf6+ 'it'g8, or 33 ~h3 ii..gS! (a ~f2 'iYd2+ 40 ~f3 'iYd1+ 41 'it'g3 'iYd6+ 42
tactical trick, which saves Black literally by 'iVf4+.
a miracle) 34 'iYxg5 (34 '~ixb7? ':g8!) The various versions of the ending 'rook
34 ... 'iVh1+ 35 ~g4 'ug8 or 34 '>j'f7 '''h1+ 35 and pawn against bishop and pawns'
~g4 'ifxg2+ 36 ~h5 'ug8 37 .i.e3! (but not 37 arising after 33 'iYf4! would have promised
'uxb7? .i.e7! or 37 ~xb7? ~3, winning) White real winning chances. Now, how-
37.. .'iVxb2 38 .i.xg5 'ug7! 39 ~'f8+ ~h7 40 ever, Black keeps the queens on.
'li'f5+ ~g8 41 d5 'iYg2! 42 d6 'iYc6. H ... 'iVe1+?! (03)
31 .i.xd4 (03) 31 ...'iYxd4+ I had less than six minutes left, and I
Zaitsev's recommendation 31...''''d1+ 32 spent half of them on this inaccurate move.
~h2 ~xd4 33 ~5+ ~g8 does not help on Here too no one noticed 33 .. :iVa1+! 34 ,Un
account of 34 'ub3! itxe5+ (34 .. .'~;xe5+ is (34 ~h2 .i.e5+) 34 ... 'iiVd4+ 35 ~h1 .i.f6,
317
Kasparov vs. Karpov: 1986-1987
reaching a drawn position similar to that from 35 'iVf4! White could also have won by
which occurred in the game. 37 ~f3!, as Karpov stated several times in
34 ~f1 'iYe5 interviews'. Is this so? After 37... .i.f6 38
'iYb8+ <;t>g7 39 'iYxa7 'iYd5 40 'iYa6 (40 'iYxb6
'iYd1 + and ... 'iYxa4 with a draw) 40 ... 'iYd4 41
'iYb5 'iYe4! the centralised queen would
have prevented White from strengthening
his position.
However, better practical chances were
given by 37 ~e1! .i.f6 (37 ... e5? 38 'iYe4) 38
'iYb8+ <;t>h7 39 'iYxa7 - here I would have
had to choose between the risky opening of
the position by 39 ... h3 40 'iYd7! hxg2++ 41
<;t>xg2 and the relatively quiet 39 ... 'iYe8 40
'iYa6 (40 'iYxb6 'iYxa4 with a draw) 40 ...'iYc6
41 'iYb5 'iYc2(d6), although this position
35 <;t>h1?! (02) looks more pleasant for White than in the
Again missing the chance 35 'iYf4! variation with 37 ~f3.
(pointed out by Karpov in an interview 37 ...'iYe2 38 ~c1.i.f6
after the game). After 35 ... 'iYxf4? 36 ~xf4 The outcome is that Black, with his flag
.i.f6 37 ~c4 <;t>g7 38 ~c7 b6 39 <;t>fl or about to fall, has managed to set up a
35 ... 'iYd4+? 36 'iYxd4 .i.xd4+ 37 <;t>h2 <;t>g7 38 fortress, albeit a less favourable version
<;t>h3 .i.f6 39 ~c1 Black loses, since he is not than after 33 ... 'iYa1 +!.
able to get rid of the white a-pawn. 39 'iYg6 'iYe6 40 ~d1 (04) 40 ...'iYc8 41 ~f1
But here (in contrast to the position after (04) 41 ... 'iYd7 (04)
33 'iYf4!) Black could have fought for a draw Here the game was adjourned, and
by 35 ...'iYc5+ 36 <;t>h1 .i.f6 37 'iYb8+ <;t>g7 38 White sealed his next move.
'iYxb7 'iYc4 39 'iYb5 'iYe4 (Zaitsev) - I think
that my task would have been feasible: 40
~f3 'iYe1+ (40 ... h3?! is worse: 41 'iYfl! 'iYxa4
42 .uxh3 as 43 ~g3+ <;t>f7 44 'iYf5) 41 'iYfl
'iYd2 or 40 'iYb3 as 41 ~c1 'iYg4 42 'iYc4 'iYf5
43 ~fl 'iYd7 44 'iYe4 'iYc8 45 ~d1 'iYc5, and
White appears to have no way of convert-
ing his advantage.
35 ... b6?! (yet another time-trouble error;
35 ... iJ6! would have equalised) 36 'iYf4 (11)
36 ...'iYh5! (of course, not 36 ... 'iYxf4? 37 ~xf4
.i.f6 38 ~c4 etc.) 37 'iYf5?!
In the event of 37 'iYb8+ <;t>h7 38 'iYxa7 h3!
because of the threat of perpetual check 'Our analysis lasted until 3.30 a.m., then
White would have lost his last pawn: 39 it was resumed in the morning, and by three
'iYxe7 hxg2+ 40 <;t>xg2 'iYg4+ 41 <;t>h2 (41 <;t>f2 0' clock in the afternoon a unanimous opin-
'iYf4+!) 41...'iYxa4 with a draw. ion had been reached: the position was a
As Makarychev reported in 64, 'apart definite draw.' (Nikitin). In the press centre
318
The Fourth Match: 1987
319
Kasparov vs. Karpov: 1986-1987
.i.a5! 62 l:ha5 bxa5 63 'it'xa5 'it'c6! with a The culminating point of the adjourn-
drawn pawn endgame: 64 'it'b4 Wb6 65 'it>c4 ment session, which was not noticed by the
Wa5 66 Wd5 Wxa4 67 We6 Wb5 68 Wxe7 commentators.
Wc5! (but not 68 ... Wc6? 69 We6 or 68 ... Wc4?
69 Wf6) 69 Wf6 (69 We6 Wd4) 69 ... 'it'd6 70
Wg5 h3!.
55l:.C4 (04)
Black's position has markedly deterio-
rated: the b6-pawn is immobile, and his
bishop can move only between g5 and f6.
58 We3? (02)
An error in reply, allowing Black never-
theless to set up a fortress. In my view, 58
We2! would have been decisive. Here are
some sample variations:
1) 58 ... .i.c5 59 l:.g5+! (not 59 l:th7 b5!)
59 ... Wd4 (59 ... We6 60 l:.h5 or 59 ... Wc4 60 ~g4
55 ... e5 is even worse) 60 l:.g6! (not 60 l:.g4 b5!)
Also quickly played - and rather riskily, 60 .. .'it'e5 61l:.h6, winning one of the pawns,
since the e7-pawn, protected by the bishop, and with it the game;
looks to be the basis of the fortress. But I no 2) 58 ... .i.f6 59 l:.f7! (59 l:.h7 'it'c5 60 'uc7+
longer believed in the fortress, thinking that 'it'b4! 61 l:.c6 .i.g5 62 .l:tg6 .i.el 63 'it'd1 .i.a3
the white king would inevitably reach the 64 Wc2 Wxa4 65 l:.g4 Wb5! 66 l:.xe4 Wc6 67
denuded b5-square, and therefore that l:.xM .i.c5 is unclear) 59 ... .i.e5 60 .l:If5 'it'd4
Black should try to create active play. From 61l:.h5 .i.f6 62l:.b5 .i.d8 63l:!.b1 We5 64l:!.d1
afar I did not see any refutation of my idea. .i.g5 65 l:.d7, and Black has no defence (say,
56l:.g4 .i.e7 65 ... .i.f6 66 l:.b7 .id8 67 l:!.b8 .i.c7 68 l:.h8
Here we reached the second time con- and l:.xM).
trol, and Karpov sank into thought. Apparently Karpov thought that 58 We3
57l:.g7 (14) would also win: he wanted to divert the
57 l:.g6!? was also good. According to bishop from the defence of the h4-pawn,
Makarychev, in this way 'Karpov could but he overlooked the study-like draw
have prevented the switching of the bishop involving ... b6-b5.
to c5-d4; one of the pawns (b6, e5 or M) 58 ... .i.c5+ 59 We2 (07)
would inevitably be lost, and White would If 59 'it'f4 Black prettily saves himself
win.' But, as we will see from the note to with 59 ... b5! 60 axb5 e3 61 'it'f3 Wc4 62 ~g4+
White's 58th move, playing his bishop to c5 Wxb5 63l:.xM a4, for example: 64l:.g4 a3 65
would not have helped Black. l:.g8 Wc4 or 64 'it'e2 a3 65 Wd3 a2 66 l:.h1
57 ... e4+ (04) Wb4.
320
The Fourth Match: 1987
321
Kasparov vs. Karpov: 1986-1987
76 'it'g4 'iVxg2+ with an obvious draw.' decisive advantage) he missed his chances
(Makarychev) in five games, and he won the 23rd only
73 ...'it>xd4 74 b7 aliV 75 b8iV iVa6+ 76 ~f2 thanks to a blunder by me ...
iVf6+! 77 'it>gl 'it>e4 78 iVb4+ (18) 'It can be imagined how the ex-world
78 'iVb7+ ~f4 79 iVf3+ ~gS 80 iVxf6+ champion was feeling when he arrived for
'it>xf6 81 'it>h2 'it>g6! with a draw (Zaitsev) the 8th game. After all, in the preceding
was more interesting. game the bird of fortune, as though teasing
78 ...~f5 79 iVel ~d4+ (04) him, had time after time flown into his
hands, but he was unable to grasp it.'
(Makarychev)
Game 35
G.Kasparov-A.Karpov
World Championship Match,
8th Game, Seville 02/03.11.1987
English Opening A36
1 (4 e5 2 ctJ(3 d6
Instead of the usual 2 ... ctJf6 (Game
Draw agreed (Yz-Yz). A grandiose battle! Nos.29, 31, 43) or 2 ... ctJc6 (Game No.33),
Times: 4.50-3.48. Karpov springs a surprise - a move from
my 'Black' repertoire. He himself had made
Of course, I was very pleased to have this move only once in his early youth.
held out in this tense and difficult struggle: 3 g3 (02)
after all, had I lost, the match score would I don't known what would have fol-
already have been 'plus two' in Karpov's lowed after 3 ctJf3 (logically, 3 .. .fS or 3 ... cS),
favour. but in the event of 3 d4 my opponent could
As it later transpired, this adjournment have been using the following game as a
session was the start of a bad run for my guideline: 3 ... exd4 4 iVxd4 ctJf6 S g3 ctJc6 6
opponent. In Seville there was an unusually iVd2 ~e6 7 ctJdS ctJeS 8 b3 ctJe4 9 iVe3 ctJcS!?
large number of adjourned games - 11, out (Hiibner-Kasparov, 1st match game, Ham-
of which eight were resumed. This means burg 1985).
that in a third of the games the dispute was 3 ... (5?! (01)
settled after the 40th move, and in the 'A rare and highly questionable con-
recent past I do not recall this happening in tinuation: Black sets up a powerful pawn
a match for the world championship. centre, but weakens the dS-point' (Suetin).
Earlier Karpov skilfully took account of In general this idea is not new: Botvinnik
psychological nuances, and when adjourn- tried it both with White against Smyslov
ing a game he used to choose the best move (Game No.109 in Volume II of My Great
or find unexpected ideas. But here it all Predecessors), and with Black against Pach-
turned out differently: out of six games (if man (Moscow 19S6) - 1 c4 cS 2 ctJc3 g6 3 g3
one disregards the 8th and 11 th games, ~g7 4 ~g2 ctJc6, followed by ... e7-eS,
where it was a question of converting a ... ctJge7, ... 0-0 and only then ... d7-d6. Much
322
The Fourth Match: 1987
earlier, Karpov had also played it success- blockading S... aS. Makarychev called it,
fully - against Donchenko (Leningrad 'quite possible' but 'clearly unaesthetic',
1969), Barcza (Caracas 1970) and Ve- and pointed out that in this case White
selovsky (Rostov-on-Don 1971). would have gained a tempo compared with
But after 2 ... d6 this set-up is positionally the variation S .l:i.bl g6 6 a3 as. Indeed, after
dubious: this move order allows White S... aS White is no longer obliged to play
immediately, without .l:i.bl and with his .l:i.bl and proceed with b2-b4. Black has
knight on gl, to begin an offensive on the acquired a second 'hole' - the bS-point, and
queenside, emphasising the weakness of it is possible to continue simply 6 e3 g6 7
the dS-point (2 ... cS and 3 ... ct:Jc6 really is tDge2 .iL.g7 8 d3 and 0-0 with a slight but
better). But the main thing is that, being a enduring advantage.
player of classical style, Karpov did not Karpov did not want to give himself
much like playing ... g7-g6 and ... .iL.g7 - additional weaknesses, and he refrained
these positions were alien to him and he from action on the queenside. But this
handled them uncertainly. Not without game showed that one should not play
reason did he avoid both 3 ... g6 (Levy- English-cum-King's Indian set-ups using
Karpov, Groningen 1967/68), and 'my' plan the criteria of the Queen's Gambit.
3 ... fS 4 .iL.g2 ct:Jf6 (not 4 ... tDc6 Karpov- 6 b4 (03)
Korchnoi, 26th match game, Baguio 1978)
followed by ... g7-g6, ... .iL.g7, ... 0-0 and ... c7-
c6 (the source game: Golombek-Botvinnik,
Budapest 19S2), and after the 8th game he
altogether ceased such experiments.
4 .iL.g2 ct:Jc6
6... .iL.g7
Black sticks to his solid but passive plan.
6 ... cxb4?! 7 axb4 ct:Jxb4 was dangerous, as
after this White has two good continua-
tions: 8 Vi'a4+ ct:Jc6 9 .iL.xc6+ bxc6 10 Vi'xc6+
.iL.d7 11 ·t!ib7, regaining the pawn with
5 a3! (06) unpleasant pressure (11 ... .l:i.c8 12 .l:i.xa7 .l:i.xc4
'The game Uhlmann-Pietzsch (Groditz 13 ct:Jf3 Dorfman), and the gambit line 8 ct:Jf3
1976) went S e3 .iL.fS 6 a3 ct:Jf6 7 d3 Vi'd7 8 h3 or 8 i..a3 ct:Jc6 9 ct:Jf3, when the weakness of
g6 9 ct:Jge2 ~g7 10 .l:i.bl 0-0 11 b4 with the the d6-pawn prevents Black from develop-
initiative for White. The world champion ing his bishop at g7.
chooses a more active plan.' (Suetin) 7 .l:i.bl (02)
5 ... g6 (OS) 'Now, when the pawn has already ad-
Many commentators recommended the vanced to b4, this move, taking control of
323
Kasparov vs. Karpov: 1986-1987
the b4-point and, "by X-ray", the b7-pawn, able to prevent ... d6-dS by 11 bS ttJaS 12
is quite appropriate.' (Makarychev) ~a3.
7. ..tbge7 10 ttJge2 ~e6 (09)
After this move the two players began In Suetin's opinion, '10 ... a6 came into
spending a lot of time on the opening, consideration, and if 11 0-0, then 11...cxb4
which is not surprising: this non-standard 12 axb4 bS'. However, 11 bxcS! dxcS 12 0-0
position was unfamiliar to both of us, and ~fS 13 ttJe4 is stronger - White has an extra
we were trying to play 'by logic' on the pawn in the centre and the better prospects:
basis of general considerations. 13 ... b6 14 ~c2 ~d7 IS ttJ2c3 etc.
11 bS?! (10)
Georgadze and Makarychev attached an
exclamation mark to this move: 'In this way
(rather than 11 ttJdS) White defends against
... d6-dS'. But I have come to the conclusion
that the cool-headed 11 0-0 was more
accurate, when Karpov would hardly have
ventured 11...cxb4 12 axb4 dS, since after 13
bS ttJaS the better chances are promised by
all of 14 cS, 14 cxdS and 14 ~a3 dxc4 IS
dxc4 'ue8 16 cS ttJc4 17 ~a4.
As for 11 bS, although it does not
squander White's entire advantage, it gives
8 e3 (24) Black a simple and understandable game.
If 8 ttJf3 there could have followed But, paradoxically, this move, although not
8 .. .£S!, hindering White's development (9 the best, wins the game! Here we are al-
d3? e4), while 8 d3 0-0 9 ttJf3 (with the idea ready in the realms of psychology, which is
of 9 .. .£S 10 ~d2) would allow the freeing not directly related to the quality of moves:
9 ... cxb4 10 axb4 dS. And, so that Black Karpov did not like playing with bad
should not be able to free himself (for pieces. The knight on as became a curse for
Karpov this was the most unpleasant him and it vegetated out of play right to the
thing!), I decided not to weaken my control moment of capitulation ...
of dS and to develop my king's knight at l1...ttJas
e2.
8 ... 0-0 (OS) 9 d3! (OS)
'After 9 ttJge2 cxb4 (or first 9... ~e6 -
G.K.) 10 axb4 ~e6 White would have had to
reckon with numerous tactical possibilities
for the opponent, for example: 11 bS ttJaS 12
d3 dS or 11 ttJdS bS!? Kasparov does not
intend to occupy dS with his knight, which
is a very unusual way of playing this set-
up.' (Makarychev)
9 ...,Ub8 (21)
Now in the event of 9 ... cxb4?! 10 axb4
~e6 the c4-pawn is defended and White is
324
The Fourth Match: 1987
325
Kasparov vs. Karpov: 1986-1987
19 ... e4! which disrupts all White's plans. I 'The rook replaces the queen in guard-
would have had to try and demonstrate an ing the b3-square, via which the black
advantage for White by 19 ttJd5 ~xd5 20 knight dreams of jumping to d4.'
~xd5 ttJd8 21 ttJc3 ttJe6 22 ~xe6 ~xe6 23 f3 (Makarychev)
and ttJd5, but in this case Black would have 27 ...~f7 28 ~c3 (08) 28 ...~d8 (04) 29 ~a2
got rid of his hapless knight and retained ~h6 (03) 30 ttJgs ~ff8 31 ~e2 ~g7 32 ~C2
counterplay on the kingside. ~de8 33 ttJe3 ~h6 (04)
19 ttJdS 'iVd7 20 ~d2 ttJas (07) 33 ... h6 34 ttJe4 ~e6 35 ttJc3 and ~e4
According to Sue tin, 'this move is the would not have changed the character of
cause of Black's later difficulties - 20 ... ttJd8 the play.
suggested itself'. Dorfman was of the same 34 ~dS (03) 34... ~g7 (04)
opinion, indicating the variation 21 a4 ttJf7 The exchange 34 ... ~xg5? 35 hxg5 would
22 as ~h6 with the idea of ... ttJg5-h3+. have substantially weakened the black
However, 22 ttJec3! ~h6 23 ~d1 ttJg5 24 M king's defences and facilitated the f2-f4
ttJh3+ 25 '.t>h2 is more accurate, killing advance (35 ... ~e7?! 36 ttJxf5 and 37 f4!).
Black's initiative on the kingside. Then 3S~d1
White carries out a4-a5 and breaks through White's pieces are harmoniously placed,
on the a-file. and the time for decisive action is ap-
Instead of this Karpov took the bold de- proaching.
cision to 'stand still', endeavouring not to
allow the opening of the position. Indeed,
although White is completely dominant in
the centre, the opportunities for breaking
through are very restricted. He needs to
play f2-f4, but this demands serious prepa-
ration, since it activates the bishop on g7. A
phase of prolonged manoeuvring begins -
White gradually strengthens his position.
21 ttJec3 (03) 21 ... ~be8 22 ttJe4 (05)
22 ... ttJb7 23 a4 ttJas (05) 24 h4 (03) 24 ... ttJb7
(04) 2S '.t>h2 ~b8 26 ~a1 (08) 26 ... ttJas 27
~a3
3S ... h6
, A serious weakening of the light
squares, probably provoked by Karpov's
desire to change the position before the
adjournment.' (Makarychev). Dlugy was
also doubtful about this move, although he
noticed the strong threat of 36 ~hl, '.t>g2
and M-h5.
36 ttJe4 ~d8 (04)
After 36 ... h5 37 ttJxf5! ~xf5 (37 ... gxf5? 38
~b2 with the threat of ~xh5+) 38 ttJc3 ~ff8
39 ~e4 Black has no defence: 39 ... ~e6 40
ttJd5 ~c8(b8) 41 f4 etc.
The Fourth Match: 1987
41 ...'it'g7? (05)
'Black should have played 41...i..b7!
with the aim of eliminating the knight on
d5: (Suetin). This would have made things
much harder for me, since the immediate
42 f4?! no longer achieves its goal on ac-
count of 42 ... i..xd5 43 i..xd5 .l::i.ef6 44 .l::i.af2
exf4 45 gxf4 'iVd7.
Of course, here Karpov should have re-
membered about the 'Botvinnik rule' -
have a good think, find the move 41...i..b7,
and seal it in the envelope, taking on one's
A necessary regrouping: the bishop goes clock the time remaining. In this case there
327
Kasparov vs. Karpov: 1986-1987
would have been a lengthy adjournment a) 43 fS!? .i.xdS (43 ...11ee8? 44 f6+!, while
session in prospect: no direct way for White 43 ... gxfS? 44 .i.xfS leads to the loss of the
to win is apparent, and I would again have hS-pawn) 44 fxe6 ~xfl (44 ... .i.xe6 4S ~af2)
had to manoeuvre, preparing a break- 4S 'lWxfl .i.xe6 46 oUf2 (46 oUb2!?) 46 ... ct:Jb3 47
through (for example, 42 f3 ~g7 43 11af2 'lWeI ct:Jd4 48 ct:Je3 with good chances of
.i.xdS 44 .i.xdS 11ef6 4S .i.e4 'iUd7 46 f4, converting the exchange advantage;
although this is not yet the end of things). b) 43 g4! (this is even stronger: g4-gS is
threatened) 43 ... hxg4?! 44 'lWxg4 .i.xdS 4S
cxdS oUef6 46 fxeS dxeS 47 hS and wins, or
43 ... .i.xf4+ 44 ct:Jdxf4 exf4 (44 ... MXf4 4S
oUxf4!) 4S .i.xb7 ct:Jxb7 46 gxhS gxhS 47 iYxhS
Mh6 (47 ... Mef6? 48 ct:Jxf4!) 48 'iUdS 'iUe7 49
Maf2 and oUxf4 with an extra pawn and
winning prospects.
The difficult defence over the entire
game had evidently so tired Karpov that he
decided 'simply' to get rid of the threat of
f4-fS, by playing 42 ... exf4. But he went from
the frying pan into the fire ...
43 ct:Jgxf4 oUe5
42 f4! (11) 43 ... .i.xf4 is also hopeless: 44 ct:Jxf4 (but
My opponent was probably hoping that not Dlugy's move 44 Mxf4? because of
I would not enter into an open battle before 44 ... ct:Jxc4!) 44 ... OUef6 4S oUaf2 and now, say,
the adjournment. But I had no doubts about 4S ... 'lWe7 46 'lWc2 .i.g4 47 ~g2 M8f7 48 ct:Jxg6
the move chosen: Black's position is on the etc.
verge of cracking up - such a chance could
not be missed! And I made the move when
the five hours of play had almost expired
and the time had corne to adjourn the
game. Karpov had to find the best reply on
the stage, with his clock ticking away.
42 .•. exf4? (07)
The sealed move, for which I was not
even hoping: now Black loses very quickly.
He had two ways to prolong resistance:
1) 42 ...11ee8 43 11af2 (with the threat of
f4-fS) 43 ... .i.fS (43 ... .i.g4 44 "iWc2) 44 .i.xfS
:xfS 4S ct:Jge3 11f7 46 fS .i.xe3 (46 ... 11ef8 47
f6+ ~h8 48 ct:Je7 or 47 ... ~h7 48 ct:JfS! is no 44 ct:Jxg6! (OS)
better) 47 f6+ ~h7 48 ct:Jxe3 e4 49 Mf4 Me6 SO This combinative solution was accu-
ct:JdS, and White must win, although not as rately calculated during analysis, although
easily as in the game; 44 oUaf2 would also have won easily.
2) 42 ... .i.b7! - alas, with a delay of one 44 ... OUxfl (02) 45 'lWxfl oUxe4 46 dxe4 ~xg6
move. White has two tempting continua- 47 oUf2 'iUe8 (12)
tions: The only move: if 47 ... .i.g7, then 48 oUf7!.
328
The Fourth Match: 1987
Game 36
A.Karpov-G.Kasparov
World Championship Match,
9th Game, Seville 04/05.11.1987
Grunfeld Defence 088
329
Kasparov vs. Karpov: 1986-1987
Black's 14th move). In addition, back in the 22 jL,xd4 4Jxd4 23 .l:.xb7 ~f7 with a draw)
1st edition of the Encyclopaedia of Chess 18 ... ~6 19 4Je2 4Jc6! 20 .l:.b1 4Jxd4 21 .l:.xb6
Openings (1976) Karpov recommended 4Jf3+ 22 'it'f2 .l:.xd1 23 .l:.xb7 4Jxh2 with
15 ... e5!? 16 d5 4Jc4 - this line, which came equality.
into use after Seville (Kir.Georgiev-Ivan-
chuk, Reggio Emilia 1989/90), proved more
promising (Black tries to make use of the
c4-point and his extra pawn on the queen-
side), but I was able to play this only
against Kramnik (Linares 1999).
17 ~d3! (02)
Another discovery by Igor Zaitsev -
White sacrifices his g4-pawn for the sake of
retaining his d4-e4 mobile pawn pair.
17 •.. ~xg4
17 ... .l:.d8 18 g5! 4Jc4 (Chernin-
16 ~gl! (11) Malisuaskas, Lvov 1987) is inadequate
The most accurate reply. I was sure that because of 19 a4!, when the bishop at g7
Karpov would not risk playing 16 'iVd3 remains shut in. 17 .. .'iVc4 18 ~d2! ~a6 19
because of 16 .. .'iVb2! 17 .l:.c1 ~xa2: although ~c2 is also dangerous for Black - compared
this is dangerous for Black, he also has his with the 'theoretical' game Spas sky-
trumps - a pair of connected passed pawns. Korchnoi (1955) White has an extra tempo:
In the only previous game Uhlmann- instead of h2-h3 he has made the far more
Milev (Halle 1954), which few knew of important move ~gl, and after 19 ... 4Jc4 20
before the appearance of ChessBase, White jL,f2 or 20 ~3 ~h8 21 g5 he can count on
gained an advantage by 16 ~c2 ~e6 an advantage (whereas without the ex-
(16 ... e5!?) 17 .l:.c1 (17 g5 gives Black a tempo change on d4, after 14 ... 'iVd6 15 'it'gl 'iVe6 16
for 17 ... .l:.c8 18 ~d3 ~c4! with full compen- 'iYd3 the move 16 ... 'iYc4! would be good - d.
sation for the pawn) 17.. .'ii'xg4 18 ~gl, but the 11th game).
17... b5! and ... 4Jc4 is stronger, with good 18 .l:.fl! (04)
counterplay. Of course, not 18 .l:.c1 .l:.c8!. White is
16 ...~e6 (02) planning to build up an attack, exploiting
Transposing into a position from the his extra pawn in the centre and the 'hole'
afore-mentioned correspondence game, on e6. At this point Black's position began
which was published in Informator Vo1.39 to give me serious cause for alarm.
(1985). It continued 17 4Jg3 .l:.d8! 18 .l:.c1 (18 18 ... Me8 (08)
~a4 ~6!, while if 18 .l:.b1, then 18 ... ~xa2 'After 18 ... .l:.f8?! 19 .§.xf8+ jL,xf8 (or
19 .l:.a1 ~3 20 ~xb3+ 4Jxb3 21 .l:.xa7 jL,xd4 19 ... ~xf8 20 d5 b6 21 4Jf4!) 20 d5 b6 21 4Jd4
330
The Fourth Match: 1987
White's initiative snowballs (21..:Wics 22 In my view, this would also have given
Gt:Je6 etc. - G.K.), and the 'isolated' knight at White a dangerous initiative.
as (one involuntarily remembers the previ- 19 ... 'iid7 20 dS Gt:JC4 (IS)
ous game) cannot be quickly included in According to Suetin, 'Black has good
the play.' (Makarychev) counterplay with his pieces and every
reason to count on equality'. But in fact the
picture is far more gloomy.
21 ..id4! (23)
The exchange of bishops will expose the
black king even more. All other moves
were weaker:
1) 21 ..ixa7 b6 22 ..ibS (suggested in In-
formator by Karpov and Zaitsev instead of
the co-operative 22 J::tc1?? 'iVxa7 23 J::txc4
bS+) 22 ... bS! 23 ..ig3 'i'a7+ 24 rtih2 'iixa2 2S
Gt:Jf4 'iiVd2, and Black is alright: 26 'iVf3l:tf8 27
'i'g4 eS! 2S .l:!.f2 exf4! 29 'iVe6+ 'it>h8 30 ~xd2
fxg3+ 31 'it'xg3 Gt:Jxd2 with equality;
19 h3 (13) 2) 21 Gt:Jf4 Gt:Jxe3 22 'iixe3 ..ih6 (Suetin) 23
The immediate 19 dS?! would have been 'iVxa7 J::tfS! 24 ltJe2 (24 ltJd3 J:i.xfl + and
parried by 19 ... Gt:Jc4! 20 ..ixa7 (if 20 J::tf4, then ... 'iYbS) 24 .. .'ifbS 2S lhfS+ ..ixfS 26 ltJf4 (26
20 .. :ilVd7 21 ..id4 eS or 20 ... Gt:JeS 21 'ifbs 'iid7 'iVe3 e6!) 26 .. :~bl + 27 ~f2 iixe4 with equal-
is good) 20 ... Gt:JeS 21 ~S 'iVxe4 22 'iVxb7 (22 ity;
Gt:Jg3 'iVg4 23 'iVxb7 'iVd7! with equality) 3) 21 ltJd4 Gt:Jxe3 22 'iixe3 ~c4!, and after
22 ... J::tfS 23 J::txfS+ ..ixfS 24 ~S (but not 23 ltJe6 the commentators unanimously
Makarychev's move 24 ~2? because of suggested 23 ... 'i'xe6! 24 dxe6 ..id4 with
24 ... Gt:Jc4, winning) 24 .. .'~Vc2! - thanks to the equality. The modest retreat 23 ltJf3 is more
domination of his queen in the centre, Black dangerous, although the accurate 23 ... h6!
regains the pawn and maintains the bal- counters White's feeble efforts.
ance. Now, however, Black faces a difficult
But 19 Gt:Jf4!? Gt:Jc4 20 ..ic1 carne into con- choice.
sideration, with the threat of d4-dS and
Gt:Je6, for example:
1) 20 ... 'iVd7 21 dS 'iVa4 22 Gt:Je6 Gt:Jd6 23
Gt:Jxg7 rtixg7, and apart from the obvious 24
eS White also has the strong 24 'iVd2!? rtigS
(24 ... 'iVxe4? 2S ~2+ and ..ih6) 2S eS Gt:JfS 26
d6! etc.;
2) 20 ... bS 21 h3! (21 dS Gt:Jd6! 22 h3 'ilVh4 is
unclear) 21...'iVd7 22 dS with the unavoid-
able Gt:Je6;
3) 20 ... Gt:Jd6 21 ..ia3 'iVd7 (21...J::tfS 22 h3!
'iid7 23 eS) 22 eS Gt:Jc4 23 iLcs bS 24 e6 'iVdS
2S rtihl! or 22 ... Gt:JfS 23 ..ib2 with the inten-
tion of d4-dS (and if 23 ... 'iVa4, then 24 g4!). 21 ... eS?! (3S)
Kasparov vs. Karpov: 1986-1987
An extremely risky step, dictated by the (23 ...1:Ic4? 24 1lVb2) 24 1:Ic1 or 24 ..ic3 and
threat of 22 ..ixg7 and ctJd4-e6. In the press ctJd4-e6.
centre the more solid 21...ctJe5 was actively It appeared that the radical 21...e5
discussed, with the following interesting would cut the Gordian knot of problems
variations: facing Black ...
22 dxe6 (14)
'Refusing to capture en pass ant would
have transformed the pride of White's
position - his knight on e2 - into a useless
piece.' (Makarychev). After 22 ..ixa7? b6
Black would not have won, but would have
forced a drawn ending: 23 d6 'iVxd6
(23 ... 'iVxa7? 24 'iVd5+ 'it'h8 25 .i:i:f7) 24 "iYxd6
ctJxd6 25 i.xb6 ctJxe4 26 a4 .i:i:c2 271:Iel.
22 ... 'iVxe6 23 ..ixg7 'it'xg7 24 ctJf4 'iVd6 2S
'iVc3+ (03) 2S ...'it'h6!
An unusual move, which confused Kar-
pov. Black would have lost after 25 ... 'iVe5?
Analysis diagram 26 ctJe6+ 'it'h6 27 'iVc1 + or 26 ... 'it'h8(g8) 27
~xc4!, while the routine 25 ... 'it'g8 would
1) 22 ..ixe5 ..ixe5 23 ctJd4 ..ixd4+(?!) 24 have led after 26 ctJd5! ctJb6 27 ctJf6+ 'it'h8 28
1lVxd4 b6 25 e5 'iVc7 ('and then ...'iVc5 with 1lVb2 ctJc4 (28 ...'iVc5+ 29 'it'hl! 'iVc3 30 ctJe8!!
equality' - Suetin) 26 'iVg4 (!! - Georgadze) or 29 ... 'iVe7 30 e5 etc. is hopeless) 29 'iVxb7
26 ... l::tf8!, maintaining the balance. In my 1lVb6+ 30 'iVxb6 axb6 31 Mbl to an ending
opinion, it is better to play 25 'it'hl! 'iVd6 26 with an extra pawn and winning chances
e5 'iVc5 27 "iYg4 1:If8 28 1:Idl with the threat for White.
of creating a passed pawn. Therefore I
would have preferred 23 ... ..ib8!? with the
idea of 24 ctJf3 'iVd6 25 e5 'iVc5+ and ...'iVc4;
2) 22 1lVb3! (after 22 'iVg3 'iVd6 White also
has to play 23 1lVb3, since if 23 ctJf4 there is
23 ... ctJf3+!), 'and White retains enduring
pressure, avoiding risk and premature
simplification' (Makarychev). But after
22 ... b5 no one has demonstrated this: 23
1lVb2 ctJc4 24 'iVaI ..ih6! with the threat of
... ctJd2 (variation from the match bulletin),
or 23 'iVg3 'iVd6 24 ..ixa7 ctJc4, regaining the
pawn with equality, while in the event of
23 ctJf4 ctJf3+! 24 'iVxf3 ..ixd4+ 25 'it'hl Black 26 ctJdS? (07)
is saved by 25 ...1:If8! 26 'iVe2 i.e5 27 ctJe6 Through inertia White makes the same
1:Ixfl+ 28 'iVxfl 'iVc8 with equality. move as after 25 ... 'it'g8. This important
In my opinion, only 23 1:Idl! with the moment was not covered by the commenta-
threat of ctJf4-e6 would have retained the tors, and only after the match in an obser-
better chances, for example: 23 .. J:H8 vation 'What remained in the shadow ... '
332
The Fourth Match: 1987
did Makarychev reveal the latent possibil- Fortunately for me, all this remained 'in
ity 26 ttJd3!!, 'suggested by grandmaster theory', and 'in practice' after the obvious
Dolmatov - one of the world champion's 26 ttJdS Black escaped with a slight fright ...
seconds - who had returned from Seville. 26 ...'iVes (03) 27 'iVd3 (08)
In anticipation of White's move, the
grandmasters gathered in the press centre
were enthusiastically discussing 27 'iVb4(?!),
intending 27 ... b6(?) 28 cJ;;hl! with threats to
the black king, or 27 ... ttJd6(?!) 28 ttJf6 "iYcS+
29 'iVxcs .l:i.xcS 30 eS! (30 .l:i.dl .l:i.c6! Geor-
gadze) 30 ... ttJfS(?) 31 e6 (Gufeld), although
this is unclear after 31...ttJd6!, but there is a
win by 31 .l:i.xfS! gxfS 32 e6 .l:i.eS 33 e7 .l:i.xe7 34
ttJg8+ and ttJxe7. But Black could have held
out after 30 ...ttJc8 31 .l:i.dl .l:i.c7 32 e6 ttJe7 33
.l:i.d7 .l:i.xd7 34 exd7 ttJc6 (Makarychev).
Everyone rejected the simple reply
Analysis diagram 27 ... 'iVxe4! because of 28 ttJf6 'with an
irresistible attack', but later in Infarmatar
'Surprisingly, the threat of ttJf2-g4+ can- Karpov and Zaitsev suggested the interme-
not be successfully parried, since 26 ... ttJb6 diate move 28 ... aS! (I saw it at the board)
27 'iVd2+ cJ;;g7 28 'iVb2+ cJ;;g8 merely defers it, with the drawing variation 29 iVbS(?) 'iVd4+
the knight at b6 being unable to come 30 .l:i.f2 (30 cJ;;hl ttJe3!) 30 .. :~al + 31 .l:i.fl
quickly to the aid of its king'. 'iVd4+. However, in my view, after 30 ... ttJe3!
Indeed, after 29 'iVb3+ cJ;;h8?! 30 .l:i.f7 Black has the advantage, and therefore
Black has no defence, but 29 ... ttJc4 30 'iVxb7 White should play 29 ttJxe4 axb4 30 .l:i.bl
'iVb6+ 31 'iVxb6 axb6 is more tenacious, .l:i.e8 31 ttJg3 with equality.
trying to save himself in an endgame a
pawn down. Therefore after 26 ... ttJb6 White
should also consider 27 'iVb3!? with the idea
of exploiting the position of the king at h6,
for example: 27 ... ttJc4 28 ttJf2! or 27 ... .l:i.f8 28
.l:i.xf8 'iVxf8 29 'iVc3 'iVe7 30 ttJf2 etc.
26 ttJd3 is also strong in that it prevents
Black from blocking the long diagonal by
... 'iVeS. And other attempts to exchange the
queens do not bring him any relief: 26 ... 'iVa3
27 'iVxa3 ttJxa3 28 .l:i.f7! with a clearly better
endgame, or 26 ...'iVb6+ 27 ttJf2! 'iVe3 28 'iVf6!
'iVb6 29 'iVf7! 'iVe3 (29 ... ttJe3 30 .l:i.el or
29 ... 'iVcS 30 .l:i.c1 bS 31 cJ;;hl and ttJg4+ is 27 ...cJ;;g7! (12)
worse) 30 'iVxb7 'iVcs (30 ... .l:i.f8 31 'iVb4) 31 Accurate, neutralising White's activity.
"iYf7 with an extra pawn and a powerful 28 ttJf6 (04) 28 .. :~d6! (02)
initiative (if 31...ttJe3 White has the strong Another accurate move. Black would
32 h4!, when 32 ... ttJxfl? fails to 33 g4!). have lost ignominiously after 28 ... cJ;;h8? 29
333
Kasparov vs. Karpov: 1986-1987
'iYd7 'iYc7 30 ctJe8!! 'iY'xd7 31 .l::i.f8 mate (Tim- create some problems for myself out of
man), or 28 ... .l::i.c7? 29 'iVd8 (Makarychev). nothing.
29 'iY'c3 (11) 29 ...'iY'e5 30 'iY'd3 (06) 30 ... 'iVd6 36 .l::i.d7+ (03) 36 ... Wh6 37 'iY'e2 'iVC5+ (18)
31 'iVC3 'iY'e5 32 'iY'b3 An unnecessary check - it was simpler
After repeating moves to gain time on to play 37... 'iVf4! 38 e5 (38 .l::i.xa7? 'iY'c1 + 39
the clock, Karpov plays on. But White has Wh2 .l::i.f1) 38 ... .l::i.c6 39 .l::i.dl .l::i.e6 40 .l::i.e1 (Kar-
no advantage. pov, Zaitsev) 40 ... b4.
32 .. J~c7 (09) 38 Wh2 (03) 38 ...'iVe5+?!
A sound plan, although it was also pos- 38 ... 'Yi'c1 39 .l::i.d2 'iY'c5 40 .l::i.d5 'iYc1! or
sible to play 32 ... 'iVb2 33ctJe8+ .l::i.xe8 34 'iVxc4 38 ... .l::i.f2 39 ~e1 a6 40 a3 (40 e5 .l::i.xg2+)
Wh6 or 33 ~xb2 ctJxb2 34 e5 ctJc4 35 e6 ctJd6 40 ... g5 was better, although one does not
with a simple draw. very much want to make such a weakening
33 'iVd3 (19) move.
'If 33 ctJd5 there would have followed 39 g3
33 ... ctJd2.' (Suetin) Now White's chances are somewhat bet-
n ....l::i.f7! (02) 34 'iVxC4 (07) 34 ....l::i.xf6 35 .l::i.d1 ter: his weak e4-pawn may let it be known
b5 (08) that it is passed.
39 ... 'iY'c3 (10) 40 Wg2 (03) 40 ...'iVc4
40 ... .l::i.c6 41 e5 'iVc2 42 .l::i.d2 'iY'c3! would
have equalised more quickly, for example:
43 g4 .l::i.c7 44 e6 .l::i.e7 45 g5+ Wg7 46i:td6 h6.
334
The Fourth Match: 1987
50~a8
At least winning a pawn. After 50 ~a4
the simplest is 50 ... ~c5!? 51 ttc2 ~e7 52 ttc6
'it'g7, although it is also possible to play
50 ... ttxe5 51 a3 tte1! 52 axb4 ttb1 53 ~a6+
'it'g7 54 ~a5 h6 55 ~c7+ 'it'h8 56 ~c2 ttxb4!
57 ttd2 ttc4 58 ~a2 tta4 with a draw (Kar-
pov, Zaitsev), or 54 ~7+ 'it'h6 55 ~e7!
ttxb4 - here we overlooked 56 h4!, but after
46 ... a5! 56 ... gxh4! 57 ~e6+ 'it'h5 58 g4+ 'it'g5 59 ~g8+
'Active defence. Black sacrifices a pawn, 'it'f4 60 °iVf7+ 'it'g5 61 ~5+ 'it'f6 62 ~h6+ 'it'f7
335
Kasparov vs. Karpov: 1986-1987
63 'iYxh7+ Wf8 Black holds on. Wh6 69 Wg2 (69 a7 Wg7) 69 .•. ~a3 70 Wf2
50 ... 'iYxe5! (05) 70 ... Wg7 (01)
Things are more difficult for Black after
50 .. Jhe5 51 'iYf8+ Wg6 52 M! 'iYg4! (but not
52 ...'iYc5? 53 h5+! Wxh5 54 'iYf7+ with a
winning attack).
51 'iYf8+ Wg6 52 'iYxb4
336
The Fourth Match: 1987
The 3 e5 ..\tf5 4 ct:Jc3 e6 5 g4 ..\tg6 6 ct:Jge2 The set-up with 7 c3 ..\tg4 (7 ... ..\tf5?! 8
variation (Sokolov-Karpov, 9th match ct:Je5 e6 9 g4 ..\tg6 10 h4! - d. the note to
game, Linares 1987) was to gain in popular- Black's 8th move) 8 h3 was drawn to my
ity in the 1990s and later it occurred in a attention by Dorfman. I liked the fact that
game of mine with Karpov (Linares 2001). here White immediately gains the advan-
This and all our other duels in the Caro- tage of the two bishops, retains some
Kann will be covered in detail in the next dominance in the centre and can try to play
volume. for a win without any risk. But for this a
3... dxe4 4 ct:Jxe4 ct:Jd7 thirst for battle was needed - and it was
A sound variation, taken up by the ex- this that I did not have!
champion instead of 4 ... ..\tf5, which at one 7... ..\tg4 (13) 8 h3
time he successfully tried in his match with 8 ..\te2 is harmless: 8 ... e6 9 h3 ..\th5 10 0-0
Spassky (Game Nos.60, 62 in Volume V of My ..\td6 11 ct:Je5 ..\txe2 12 'iUxe2 0-0 13 ..\tg5 .te7
Great Predecessors). 14 Mfdl 'iUc7 (Anand-Karpov, Frankfurt
5 ct:Jf3 (rapid) 1997).
Later I employed against Karpov the 8 ... .txf3
more topical plan with 5 ct:Jg5 ct:Jgf6 6 ..\td3 Played without hesitation. 8 ... ..\th5?! is
(Amsterdam 1988, 4th round) or 6 ..\tc4 e6 7 dangerous: 9 g4 .tg6 10 ct:Je5 e6 11 h4! ..\td6
'iUe2ct:Jb6 8 ..\tb3 (Linares 1992 and 1994). 12 'iUe2 with an attack (Karpov-A.Zaitsev,
S... ct:Jgf6 6 ct:Jxf6+ (little is promised by the Kuybyshev 1970) or 10 ... ct:Jd711ct:Jxg6 hxg6
old line 6 ct:Jg3 e6 7 oltd3 Kasparov-Karpov, 12 d5!, clearing diagonals for the bishops
Amsterdam 1988, 2nd round) 6... ct:Jxf6 (Serper-Ruzjale, Sverdlovsk 1987). After
8 ... ..\txf3 Black has a more solid position.
9 'iUxf3 e6 (03)
In the 14th game Karpov was to im-
prove Black's play with 9 ... 'iUd5!.
7 c3
A rare move in contrast to 7 ct:Je5 to
which the best reply is 7 ... ct:Jd7, for exam-
ple: 8 ct:Jd3 g6! (Karpov-Sosonko, Amster-
dam 1980; Timman-Korchnoi, Montpellier 10 .tC4!? (03)
Candidates 1985), 8 ..\te3 ct:Jxe5 9 dxe5 ..\tf5 White has also played 10 g3 'iUd5! 11
(Sokolov-Karpov, 5th match game, Linares ..\tg2, unsuccessfully trying to demonstrate
1987) or 8 ~f4 ct:Jxe5 9 oltxe5 'iUd5 10 c4 the advantage of his two bishops in an
'iUa5+ 11 'iUd2 'iUxd2+ 12 Wxd2 ..\tf5 ending (Bogoljubow-Eliskases, Bad Harz-
(Sutovsky-Karpov, Tilburg 1996). burg 1938). 10 ... .te7 is more lively: 11 .tg2
337
Kasparov vs. Karpov: 1986-1987
15 i..b3
If IS ..iel there would have followed
IS ... i..f4! (for example: 16 i..xdS i..xel 17
i..xe6 'iVxb2! 18 'iVc4 i..d2 with equality), but
even so I should have withdrawn the
bishop.
12 i..e3?! ls ... ttJxe3 16 fxe3 cs!
A rather ugly move, allowing the ex- Here I realised that only my position
change of the bishop. The correct continua- could be worse: on the kingside White's
tion was 12 'iVg4 (with the idea of 12 ... 0-0 13 dark squares are weakened.
i..h6 i..f6 14 oUadl and .!tel!) or 12 g3!?, 17 oUf3 oUae8 18 oUdfl oUe7 19 'iVf2 'iVc7 20
when 12 ... i..gS is unfavourable because of 'iVh4
13 'iWg4 h6 14 f4 i..f6 IS fS with the initia- Draw agreed on White's proposal
tive, while if 12 ... 0-0, then 13 M. It is not (Yz-Yz). Black could have played on, by
surprising that in the 14th game Karpov advancing his queenside pawns and meet-
decided to avoid this position. ing e3-e4 with ... cS-c4 and ... e6-eS, but his
12 ••• iVb6 chances of gaining a real advantage were
Now White does not manage to pre- short-lived, and Karpov decided not to
serve his bishop by oUadl and i..el. expend effort in vain. Times: 1.3S-1.30.
338
The Fourth Match: 1987
339
Kasparov vs. Karpov: 1986-1987
340
The Fourth Match: 1987
341
Kasparov vs. Karpov: 1986-1987
move): 29 'it'f2 'it'e7 30 'it'e3 ttJc4+ 31 Wd3 as equal after 31 ~d3 1:!.g4!, hindering ttJf1-e3, but
32 ttJf1 ttJb2+ 33 'it'd2 ttJa4 34 ttJh2 ttJcS 3S not 31 ... .l::!xh4?! 32 ttJf5! - G.K.). What could
We3 ttJd7 (or 3S ... b4 36 ttJf3 a4 37 ttJxeS b3 have stopped Kasparov? The hypothesis put
38 axb3 a3! 39 ttJc6+ 'it'd6 40 ttJb4 'it'eS with a forward by Yusupov looks the most likely:
draw) 36 ttJg4 'it'f7, and Black has a fortress. by 29 ttJf1 ttJd6 30 .l::!c6! ttJxe4 31 ttJg3! White
24l:tc2! (13) retains pressure in a position where material
'A typical Karpov move! The ex-world is equal, but there is no longer counterplay
champion often resorts to such useful with ... ..ib6.' (Makarychev)
waiting manoeuvres, as if the opponent is In Informator Karpov and Zaitsev gave
in zugzwang.' (Taimanov) this variation by Yusupov with even a '±'
evaluation, although after 31...ttJc3! Black
gains a draw without any particular prob-
lems: 32 ..ie1 ttJe2+ 33 ttJxe2 ..ixel.
With 28 ... a6 Black would effectively
have set up a fortress, since nothing is
given by either 29 ..ics ..ie1! 30 'it'h2 .l::!c8 or
29 'it'h2 .l::!f8 30 'it'h3 .l::!f4 31 ..ics h6! 32 ..ie7
(32 gxh6 ..id8!) 32 ... hS and ... 'it'f7.
But by blocking the f-file for my rook
(and all because of the fixed idea ... ..ib6!), I
allowed the serious - and probably decisive
- activation of the white pieces. But... 'The
most surprising thing is that, against all
24...ttJd6?! (02) expectations, in the end these risky tactics
An unnecessary retreat (since after 2S proved justified!' (Taimanov)
ttJg3 it is weak to play 2S ... ..ib6? because of 29 ttJf1 (16)
26 .l::!c6! ttJc4 27 gS). 24 ... a6!? was better - I Another repetition of moves to gain
was again afraid of the reply 2S ttJc1 and time on the clock.
ttJd3(b3), but there was no reason for this: 29 ... ttJd6 30 ttJg3 (13) 30 ...ttJC4 31 'it'f1
after 2S ... ttJd6 26 .l::!e2 .l::!c8 27 ttJd3 ..ic3 If White should succeed in playing 'it'e2-
White would not have achieved anything. d3 and bringing his knight into play,
25 ttJg3 ttJC4 26 ttJf1 (08) 26 ... ttJd6 (IS) 27 Black's fortress will begin to crack.
ttJg3 (09) 27 ... ttJC4 28 g5! (OS)
After repeating moves, Karpov plays on
- after all, he is a pawn up!
28 ... 'it'f7? (37)
But this is an obvious mistake. 'Where is
the king heading for? This is a great secret!
The purely outward impression made by the
... ~g8-f7-e7-f7-g7 manoeuvre is almost
depressing. However, knowing what hap-
pened just a few moves later, one cannot
help being staggered by the world cham-
pion's intuition. 28 ... a6 suggested itself, and
if 29 'it'f1, then 29 .. J:1f8 30 'it'e2 .l::!f4 (Black is
342
The Fourth Match: 1987
33 ~f2+?
'The advantage would have been kept by
33 'it>e2 .ili,b6 34 .ili,xb6 axb6 35 ct:Jf1, and if
35 ... ~a3, then 36 ct:Jd2 (with the king on e7 Black has at last carried out his long-
there would have been the defence ... 'it>d7!; also planned regrouping (with the idea of 35
35 ... ~a4 36 ct:Jh2! is no better - G.K.), or 33 ... a6 ~xb6 axb6!) and he can breathe a sigh of
(with a delay!) 34 ct:Jf1 ~c8 35 .ili,f2 ct:Jd6 36 relief... There was no severe time-trouble
.l:txc8 ct:Jxc8 37 ct:Jh2, when the possibility of that day: we each had about 14 minutes left
the ct:Jg4-f6 breakthrough leaves White with for 6 moves.
very good chances.' (Makarychev) 35 ~c6?? (01)
343
Kasparov vs. Karpov: 1986-1987
Blundering the exchange. I had a a4) 45 ... 'it>xg4 46 'it'xe4 a4 47 'it>d3 'it'xM 48
strange premonition that this move would 'i¥ic4 b3 49 axb3 a3! 50 'i¥ic3 ~a8 or 48 i.d6
be played, and Karpov made it very confi- 'it>xg5 49 'i¥ic4 b3 50 axb3 a3 51 i.xa3 ~xc7+
dently! Thus the rook, with the distinctive 52 i.c5 h5 53 b4 M 54 b5 h3 55 b6 ~b7,
shape of the ancient Seville tower, played a winning by literally a single tempo. Be-
particular role in the Seville match. sides, 40 ... 'i¥ig6!? would have led more
'It would appear that Karpov believed quickly to the goal: 41 i.xe5 (41 i.a5 'i¥if5
his formidable opponent had gone "mad"; and ... 'it>e6!) 41...'i¥if5 42 i.h2 a5 or 42 i.d6
he quickly made three rook moves, and, .l::!.c8 43 c7 a5.
falling into a psychological trap, committed 38 ...~f7 39 i.d6 ~d7 40 i.cs (on his five
a fatal oversight.' (Makarychev) 'post-blunder' moves Karpov spent just 4
'A terrible mistake, which the Spanish minutes!) 40 ... tbas 41 tbf1 (12)
press labelled "historic". Paradoxically, Here the game was adjourned.
nearly all the grandmasters in the press
centre fell under the spell of this tempting
manoeuvre, thinking that it would lead to a
quick win for Karpov! Of course, 35 i.f2
was essential, still retaining some advan-
tage for White.' (Taimanov). But 35 .. J:I£8 36
.l::!.xf8 'i¥ixf8 37 i.xb6 tbxb6 would have led to
a drawn knight endgame (already familiar
to us from the note to 23 ... b5): 38 'i¥ie2 'i¥ie7
39 'it>d3 'i¥id6 (or immediately 39 ... tbd7 and
... a7-a5) 40 tbh1 tbd7 41 tbf2 a5 42 tbg4 'i¥ie7
43 'it>e3(c3) 'it'f7.
3s ... tbas!
I replied instantly, still not believing my 41 ... ~C7! (32)
luck. 'It is evident that a knight on the rim I spent a long time choosing between
is not always bad.' (Taimanov). The evalua- this, the sealed move, and 41...tbc4, which
tion of the position changes sharply in was also good enough to win: 42 tbe3 tbxe3
Black's favour. 43 'i¥ixe3 'i¥if7 44 'i¥id3 a5 45 i.b6 'i¥ie8 46 i.f2
36 i.xb6 (02) 36 ...tbxc6 37 i.C7 ~f8+?! (06) ~c7 (46 ... b4!?) 47 i.g3 ~c4 48 i.xe5 ~a4 49
There was a simpler win by 37... ~c8! 38 i.f6 ~a3+ 50 'i¥id4 ~xa2 51 e5 ~d2+ 52 'i¥ic5
i.d6 (38 d6 'i¥if7) 38 ... ~d8 39 i.c5 tba5 40 b4 53 e6 a4 54 'i¥ixb4 (54 d6 ~xd6! 55 'i¥ixd6
i.xa7 ~a8 41 .i.f2 tbc4 or 41 i.c5 tbb7 and a3! and wins) 54 ... ~xd5 55 'i¥ixa4 ~d6 56 e7
... ~xa2. But I was absorbed in the variation ~e6 (Karpov, Zaitsev) or 56 ... h5 57 ~b4 ~e6
with ... ~f7. 58 'i¥ic5 ~e2 59 g3 ~e3, and White is unable
38 'it>e2 to set up a fortress.
The best practical chance would also not Analysis at home also confirmed my
have helped - 38 tbf5+!? gxf5 39 dxc6 fxe4+ evaluation of 41...~c7 - this decides matters
40 'it>e2. Here Karpov and Zaitsev recom- more quickly and forcefully, since White's
mended 40 ... ~c8 41 i.xe5+ 'it>g6 42 c7 a5(?), passed pawns are unable to promote. I
allowing 43 g4! b4 44 ~e3 a4 45 'it'xe4 b3 46 think Karpov did not expect that I would
axb3 axb3 47 'i¥id3 with a draw. 42 ... 'it>f5! is seal such a forceful and seemingly risky
correct: 43 i.h2 a5 44 'i¥ie3 b4 45 g4+ (45 h5 move.
344
The Fourth Match: 1987
42 ~d6 ':c2+ 43 ~d3 ':xa2 44 ttJe3 was transformed into torture, and my
44 ~xe5+ is equally hopeless: 44 ... ~f7 45 dream was that it would all end as soon as
~f6 (in Taimanov's variation 45 ttJe3 ':a3+ possible. The typical syndrome of a return
46 'it'e2 the best move is also 46 ... ttJc4!) match! It is hardly surprising that the
45 ... ttJc4 46 e5 ':xg2 47 ~d4 ~e8 or 47 ... l:!.gl psychological initiative soon passed to
48 ttJh2 ':el and ... b5-b4 . Karpov ...
44 ... 'it'f7 45 ttJg4 ttJC4 46 ttJxe5+ ttJxe5+ 47 After the 11th game, a stressful one for
itxe5 b4 48 itf6 (05) 48 ... b3 (all in accor- both players, for some reason neither of us
dance with our analysis, although 48 ... .:xg2 wanted to take a time-out (a rare occur-
49 ~c4 ':g4 was also strong) 49 e5 rence!), and its role was fulfilled by the 12th
'If 49 ~c3, then simply 49 ... .:e2.' (Tai- game - largely because of my indecisive-
manov). But now White's downfall is ness.
caused by the position of his king at d3: the
b-pawn promotes with check.
49 ....:xg2 50 e6+ Game 39
G.Kasparov-A.Karpov
World Championship Match,
12th Game, Seville 11.11.1987
Queen's Gambit 031
1 C4 e6
Switching to the last line of defence - an
indication of the problems which Karpov
was encountering in the English Opening
with ... e7-e5 (to which he would risk re-
turning only once - and successfully! - in
the 16th game).
50 ...~f8! 2 ttJC3 d5 3 d4 ~e7 4 cxd5 (01)
And in view of 51 d6 b2 White resigned The first indication of White's unreadi-
(0-1). Times: 2.45-2.58. ness for a fierce struggle. In such a favour-
able competitive situation it made sense to
An undeserved win, no doubt, but there try a more aggressive weapon - for exam-
have been many of them in encounters for ple, 4 ttJf3 ttJf6 5 ~f4 0-0 6 e3 c5 7 dxc5 ~xc5
the world championship - for example, it is 8 'li'c2 ttJc6 9 a3 'li'a5 (a tabiya, which also
also hard to say that Karpov deserved to arises in the Nimzo-Indian Defence with 4
win the 6th game of our first match 'li'c2), and here not 10 J:tdl, which was
(1984/85), or the 18th game of our third employed by Korchnoi in his matches with
match (1986). Karpov (Game No.30 in Volume V of My
I think that this 'gift' did me a bad ser- Great Predecessors), but the new plan with
vice: after finally taking the lead (6-5) I 10 O-O-O!? It was already prepared, but it
completely 'fell asleep'. I was seized by a did not in fact occur in Seville and it was
sub-conscious desire to curtail the play and 'fired' only later in the game M.Gurevich-
eliminate any risk. The opponent needs to A.Sokolov (55th USSR Championship,
win two games - let him come at me. Play Moscow 1988).
345
Kasparov vs. Karpov: 1986-1987
4 •.. exds 5 .i.f4 tt:lf6 (S ... c6 - Game No.7) 6 e3 cal nuance: Karpov would have been
.i.fS!? (01) unlikely to venture 12 ... d4 (or 9 ... cS at all -
In the 8th game of the previous match at the time this was still virgin territory). At
Karpov chose the modest 6 ... 0-0 7 .i.d3! (if 7 any event, Ubilava later played 12 ... .i.d6,
tt:lf3, then 7 ... .i.fS - Game No.74 in Kasparov but in this complicated position with an
vs. Karpov 1975-1985) 7 ... cS 8 tt:lf3 tt:lc6. But 'isolani' White has more chances in the
now he challenges me to a theoretical duel fight for an advantage (examples: Kas-
in the sharp variation named after Ubilava parov-Yusupov, Linares 1993; Gelfand-
(one of his helpers). Kasparov, Linares 1994).
8 ... c6 (11)
Now 8 ... cS involves a pawn sacrifice - 9
dxcS .i.xcs 10 tt:lxdS etc.
9 tt:lg3 (01)
Here the plan with 9 h3 tt:lbd7 10 g4
(analogous to the 22nd game of the 1985
match) does not bring White any divi-
dends, if only in view of 10 ... .i.e4 11 tt:lxe4
tt:lxe4.
7 tt:lge2 (06)
A voiding the critical 7 'iVb3 tt:lc6 - an-
other minor psychological concession. In
numerous analyses - both at training
sessions and during the match - we did not
find any advantage after 8 'iVxb7 tt:lb4 9 l:rc1
0-0 (this was played three times by Ubilava)
or 9 .i.bS+ 'it'f8 10 'it'd2 (Salov-
Timoshchenko, Irkutsk 1986), but we
overlooked the solid continuation 8 a3 tt:laS 9 ... .i.e6 (08)
(Gavrikov-Ubilava, Tbilisi 1983) 9 'iVa2!?, 9 ... ..\tg6 is not so safe - after 10 M! h6 11
which Timoshchenko had suggested back hS .i.h7 12 ..td3 .i.xd3 13 'iVxd3 White
in 1986. secures a post for his knight at fS and gains
7 ... 0-0 (OS) 8l:rcl (03) the better prospects (in connection with
Pointless prophylaxis. The immediate 8 tt:lfS, f2-£3 and g2-g4).
tt:lg3 .i.e6 (8 ... .i.g6 9 M!) 9 .i.d3 is more 10 .i.d3 (seizing the diagonal) 10 ... l:re8 (20)
natural, and in the event of 9 ... c6 10 'iVc2 For the moment Karpov waits, not de-
(Kasparov-Thorsteins, Saint John (blitz) termining the situation prematurely: he
1988) White still has the plan with 0-0-0. I does not want to give up bishop for knight
did not want to allow 9 ... cS 10 dxcS .i.xcs 11 after 10 ... tt:lbd7 11 tt:lfS and he does not
0-0 tt:lc6 12 l:rc1 d4!? with sharp play (Lau- hurry with 10 ... g6 because of 11 M!.
tier-Ivanchuk, Moscow Olympiad 1994), 11 ~b3 (11)
but I did not take account of a psychologi- I also decided to wait. If 11 0-0, then
346
The Fourth Match: 1987
Black now plays 11...g6, restricting the Black manoeuvres, awaiting White's re-
knight at g3. action and not hurrying with the committal
11 .. :iYb6 (02) 12 'iYC2 (the tempo-gaining advance ... c6-cS. After 14 ... Mac8 IS ttJge2 cS
ttJa4 is now possible) 12 ... ttJbd7 (03) (Ljubojevic, Taimanov) there could have
Here too if 12 ... g6 White has the un- followed 16 'iYd2, when 16 ... c4 17 i.c2 is
pleasant 13 h4!, for example: 13 ... ttJbd7 14 rather to White's advantage.
hS ttJf8 IS hxg6 hxg6 16 'iitf1 with some 15 ttJge2 (16) 15 ... M3C8 (20) 16 'iVd2 (09)
initiative thanks to the open h-file. If 16 ttJa4, then 16 .. .'iYd8! with equality
(Dorfman).
130-0
'Nothing was given by the tempting 16 ... ttJh5! (04) 17 i.h2 ttJg7
manoeuvre 13 ttJfS ~f8 14 ttJd6, if only In order to carry out Black's main
because of 14 ... ~xd6 IS ~xd6 cS' (Tai- 'Carlsbad' idea - ... i.fS with the exchange
manov). However, after 16 dxcS ttJxcS 17 of the light-square bishops. Gufeld recom-
~xcS 'iYxcs 18 ttJbS! (18 0-0 is equal - Geor- mended 17... fS, but Taimanov rightly
gadze) 18 ... 'iYe7 19 0-0 Mac8 20 'iYbl Black retorted that 'this risky advance is not in
would have been left in a somewhat infe- the spirit of Karpov's restrained strategy'.
rior position with an 'isolani'. 13 ... i.xfS 14 18 g4! (06)
i.xfS ttJf8 (Dorfman) IS 0-0 ttJg6 is more An important moment. For some reason
solid, with the idea of 16 i.g3 ttJhS, but here in Informator Dorfman attached a question
White could continue to fight for an advan- mark to this move (perhaps it was a mis-
tage by 16 i.eS. Therefore 13 ttJfS!? never- print?). However, it was dictated by cir-
theless deserved consideration. cumstances: otherwise, by playing ... i.fS,
13 ... g6 (of course!) 14 h3 (23) Black would even have gained slightly the
14 a3 (Ljubojevic) was unpromising: more promising position (say, 18 ttJa4 'iVd8
14 ... 'iYd8 IS b4?! as (Dorfman), as was 14 19 b4 i.fS 20 ttJcS ttJxcS 21 bxcS i.xd3 22
ttJa4 'iYd8 IS ttJcS ttJxcS 16 dxcS 'iYxd3 b6!). By preventing the exchange of
(Makarychev) 16 ... ttJd7 17 b4 as 18 a3 axb4 bishops, White retains the tension.
19 axb4 i.f6 20 ttJe2 ttJeS with equality. I 18 ...'iYd8! (OS) 19 f3 (06)
devised another plan, typical of this type of In an interview after the game I con-
'Carlsbad' structure: after the preparatory demned this 'weakening' move and sug-
ttJge2 and i.h2 to play f2-f3 and e3-e4. gested 19 'iitg2. The commentators' sugges-
14... ~f8 (12) tion 19 ttJf4 i.d6 20 ttJce2 was also possible.
347
Kasparov vs. Karpov: 1986-1987
But all the same White cannot get by with- a certain caution. At any event, the white
out f2-f3 (in combination with 'it>g2 and pieces are well placed, at g2 the king feels
ttJf4). safe enough, and there is still much play to
19... ttJb6 (02) come. In reality White would not have been
taking much of a risk, and there was no
direct reason to offer a draw.
I think that Karpov was very happy
with this outcome to the 12th game. Even
more importantly - being a mature match
fighter, he sensed that I was playing only to
maintain the current score and that I was
psychologically not ready for a full-blooded
fight for victory. This meant that the situa-
tion in the match was still far from hope-
less! On picking up this signal, the ex-
champion began playing more boldly, and
almost the entire second half of the Seville
20 b3 (01) match was dictated by him.
But this is indeed a weakening and un- And so, as in London a year earlier, af-
necessary move, since the threat of ... ttJc4 is ter twelve games I was leading by 'plus
not dangerous for White. 20 'it>g2 was good, one' (61J2-5 1h), but on this occasion my play
and if 20 ... ttJc4, then 21 i.xc4 dxc4 22 e4 b5 left much to be desired. There was perhaps
23 ttJd1 and ttJf2, relying on the powerful only one thing that cheered me - the suc-
pawn centre and eyeing the 'hole' at f6, or cessful solving of Black's opening problems
20 ttJf4!? i.d7 (20 ... c5 21 'iVf2) 21 e4 with in the Griinfeld Defence. No exception was
chances of developing the initiative. the 13th game, which was played on 13th
20 ••. i.a3 (02) 21l:tc2 (10) November, and on a Friday!
Draw agreed on White's proposal (Yz-
Yz). Times: 1.40-1.50.
Game 40
The only instance in my games with A.Karpov-G.Kasparov
Karpov where a draw was agreed without World Championship Match,
the exchange of even a single piece! This 13th Game, Seville 13.11.1987
result, as Taimanov put it, 'reflected not so GriJnfeld Defence 085
much the character of the position, as the
state of mind of the two players:
There is no doubt that, from both the 1 d4 ttJf6 2 c4 g6 3 ttJC3 ds 4 ttJf3
chess and the psychological point of view, A fully expected 'change of tune'. Ap-
taking into account the match situation, I parently the opponent's team had not yet
should have played on, and after 21...i.d7 found any way of strengthening White's
replied, if not immediately 22 e4, then 22 play in the variation with 4 cxd5 and 12
ttJf4 'iVe7 (22 .. .f5 involves an obvious posi- i.xf7+ (Game Nos.32, 34, 35, 38). Later it was
tional risk) 23 e4 ttJe6 24 ttJxe6 'iVxe6 25 'it>g2 to occur in our tournament games (Am-
(and if 25 ... c5, then 26 i.g1), or 22 'it>g2 'iVe7 sterdam and Belfort 1988), but never again
23 e4 ttJe6 24 i.g1, and Black has to display in matches.
348
The Fourth Match: 1987
14,Ufd1
It is also possible to play 14 'ufc1 ~b7 15
~b4 'ud8 16 ~b5 ~h6! (the simplest) or 14
'ubc1 ~b7 15 ~b4 ~d8 16 ~b5 ~a6 17 a4
i.xb5 18 axb5 a6 etc. But Karpov chose a
plan which is more dangerous for Black -
the creation of a passed d-pawn. He also
did this in the variation with 5 ~3 (in the
15th and 21st games, and the 19th game of
the 1986 match). Apparent in this is a
definite psychological reckoning: I played
such positions very successfully with White
11 'iVd2 (the passed d-pawn seemed to me to be
At that time the pawn sacrifice 11 ~d2! enormously strong and it inspired me), and
'iVxa2 12 0-0 (Game No.S2 in Revolution in the it could be assumed that it would be un-
70s, where an entire picture of the variation pleasant for me to fight' against myself'.
is given) was not yet being considered 14... ~b7 15 d5 exd5 16 exd5
seriously, and Karpov would not have On the last 7-8 moves the two players
given up a pawn 'just like that', especially spent nearly 25 minutes, refreshing the
in a match for the world championship. variations in their memory and getting
11 ... 'iVxd2+ 12 ~xd2 e6 their 'teeth' into the position. 'The main
349
Kasparov vs. Karpov: 1986-1987
350
The Fourth Match: 1987
Over-protection of the bishop on f6. 23 ... ~xf3 24 gxf3 .l:!.ad8 25 .l:!.c7 Me5 26 ~c4
19 ... Mc5 came into consideration: 20 ~b5 (26 ~b7 'De8) 26 ... .l:!.d7, gaining a draw, or
~c6 21 ~xc6 .l:!.xc6 22 g4!? (22 ~xf6 'Dxf6 23 21 ~xc6 .l:!.xc6 22 g4 h6! 23 h4 ~xe7 24 dxe7
'DeS ':c5 24 d7 .l:!.d8 25 .l:!.bc1, recommended .l:!.c7 25 ':bc1 Mb7 26 'DeS (26 'Dd4 'DeS 27
by me in Informator, does not win because 'Dc6 .l:!.e8 with the idea of ... .l:!.c7 and ... 'Db7
of 25 ... 'De4!) 22 ... ~g7 with chances of with equality) 26 ... 'Dxe5 27 .l:!.d8 .l:!.xe7 28
gradually equalising. .l:!.xa8 'Dxg4, and Black, the exchange down,
should be able to hold the position;
2) 20 'Dd4!? 'Dc5 (in Informator I sug-
gested 20 ... ~e4 21 .l:!.b4 as with the sequel
22 'De6+ fxe6 23 .l:!.xe4 ~xe7 24 dxe7 'DeS 25
.l:!.ed4 'it'f6 26 .l:!.d6 .l:!.ab8, equalising, but after
22 .l:!.b2! White has the advantage: 22 ... ~xd4
23 .l:!.xd4 .l:!.c1+ 24 .l:!.d1 .l:!.ac8 25 f3 .l:!.xd1+ 26
~xd1 .l:!.c1 27 .l:!.d2 ~c2 28 .l:!.xc2 .l:!.xd1+ 29
'it'f2 'Dc5 30 'it'e2 .l:!.d5 31 .l:!.d2 .l:!.xd2+ 32
'it'xd2 f6 33 'it'c3 etc.) 21 ~b5 (it is no better
to play 21 ~g4 .l:!.e8 22 SLxf6+ 'it'xf6 23 'Db5
.l:!.ed8 or 21 'Db5 SLxe7 22 dxe7 .l:!.e8 23 'Dc7
.l:!.xe7 24 'Dxa8 .l:!.xe2 25 'Dc7 .l:!.xa2 26 'De8+
The critical moment. It was Karpov's 'it'f8 with sufficient compensation for the
turn to solve problems at the board. exchange) 21...a6 (21...'De4 22 'De2!? is
20 .l:!.e1?! (50) dangerous for Black) 22 d7 SLxe7 23 dxc8"iV
'After thinking for nearly an hour, I was .l:!.xc8 - this was recommended by me in
unable to find anything that was genuinely Informator, but after 24 SLe2 b5 25 SLf3 SLxf3
dangerous for Black. 20 ~b5 or 20 'Dd4 26 'Dxf3 Black still has to demonstrate that
looks more resolute, but then Black defends his bishop and pawn are no weaker than
with two or three accurate moves and later the rook.
equalises by exchanges. With 20 .l:!.e1 I tried It was these variations that induced me
to provoke the activation of one of the black to look for a replacement for 16 ... 'Dd7. But
bishops - 20 ... ~e4 or 20 ... ~c3, which would after the slow move 20 .l:!.e1 Black easily
have led to complicated manoeuvring. But solves all his problems.
Kasparov chose the most solid and also
probably the simplest reply.' (Karpov)
It would appear that the ex-champion
was influenced by his home preparation
(dreaming of 20 ... ~xe7? 21 dxe7 'Df6 22
'Dd4 - d. the note to Black's 18th move).
Both of the moves suggested by him would
have given White far more chances than the
move in the game (although in the match
bulletin Georgadze even attached an ex-
clamation mark to it!):
1) 20 ~b5 ~c6 (20 ... 'Dc5 21 .l:!.b4) 21 ~a6
.l:!.e8 22 ~xf6+ 'Dxf6!? 23 .l:!.bc1 (23 'Dd4 ~d5)
351
Kasparov vs. Karpov: 1986-1987
352
The Fourth Match: 1987
Analysis diagram
353
Kasparov vs. Karpov: 1986-1987
liant reply: all White's plans collapse be- played for a win, forcing the opponent to
cause of the threat of .. Jhg4+') 29 ct:Jxg5(?), avoid various 'mined' squares.
and in Informator I gave 29 .. Jhg4+ 30 'it'f1
.i.e5! (of course, not Taimanov's variation
30 ... .l:I.xg5? 31 .l:I.xg5+ fxg5 32 .l:I.xc3) 31 ct:Je6+
'it'f7 32 ct:Jd8+ 'it'g6 33 ct:Jc6 .l:I.h8 with a sym-
bolic advantage for Black, but the variation
29 ... fxg5! 30 .l:I.xg5+ 'it'f7 31 .l:I.f5+ 'it'g8 32
.l:I.g5+ .i.g7!, given in 64 by Makarychev, is
better. True, he concluded that 'White's
counterplay is most probably sufficient for
equality', but I have doubts about that (for
example, if 33 .i.d8, then 33 ... .l:I.c5! - such a
move was not easy to see in the pre-
computer era).
28 ... .i.xd4 29 .l:I.xd4 30 .l:I.e1! (11)
An important prophylactic move - de-
fending the back rank. 'A voiding 30 f4?
.l:I.c1+ 31 'it'g2 .l:I.c2+ 32 'it'g3 (32 'it'gl! .l:I.hh2 33
fxg5 is necessary - G.K.) 32 ... .l:I.ch2! 33 fxg5
.l:I.8h3+ 34 'it'f4 .l:I.f2+ 35 'it'e4 ct:Jc5+ 36 'it'd5
.l:I.xe3' (Makarychev). '30 .l:I.e2 is also dan-
gerous in view of 30 ... .l:I.c1+ 31 'it'g2 l::.hh1'
(Karpov). But now f2-f4 cannot be pre-
vented.
30 ... .l:I.c2
In 111formator I recommended 30 ... .l:I.c3! 31
'it'g2 (31 f4? .l:I.g3+) 31...'it'f7, but after 32 f4
Mc2+ 33 'it'g3 gxf4 34 l::.xf4 .l:I.hh2 35 .l:I.g1
With the intention of f2-f4. This is also with the idea of .i.xf6 or g4-g5 it is hard for
consistent play: White continues fighting Black to increase his advantage. Even so,
for the liberation of his bishop and at the 30 ... .l:I.c3 was stronger, since it would have
same time exposes the black king, aiming to set White more practical problems.
penetrate with his rook on e7. 31 a4
29 ....l:I.h8?! (14) The immediate 31 f4!? .l:I.hh2 32 fxg5
'29 ... .l:I.c1+ 30 'it'g2 l:!.c2!, preventing f2-f4, .l:I.cg2+ 33 'it'f1 .l:I.xa2 34 'it'gl was more
seemed to me to be more dangerous' (Kar- resolute - Black's rooks look well placed,
pov). And this is correct! In 111formator I but this does not give him anything real.
then gave 31 a4 .l:I.ac8 32 as! b5 33 a6! 'with 31 ... aS?!
counterplay', but after 33 ... .l:I.8c4 34 .l:I.xc4 , A loss of time; 31...'it'g6!? would have
.l:I.xc4 35 f3 'it'g6 White has an unpleasant retained the initiative: (Taimanov). And
position, resembling the ending from the indeed, 32 f4? .l:I.hh2 33 fxg5 is now bad
6th game of the Spassky-Karpov match. because of 33 ... 'it'xg5, while if White 'stands
The two sides' chances are almost equal - still', then Black will endeavour to ex-
but only almost! Black could have calmly change a pair of rooks and bring up his
354
The Fourth Match: 1987
king to e6. After 32 a5 b5 33 a6 the position 35 ...l:i.hc8 36l:i.df4! (with a series of accurate
is apparently drawn, but here it is clear that moves White has created the threat of
White is the weaker side. l:i.xf6+) 36 ... l:i.xf2 (06)
But now Karpov, who had left himself Draw agreed on Black's proposal (Yz-Yz):
13 minutes for 8 moves, advances f2-f4 in there is too little material left on the board.
the most comfortable situation and quickly Times: 2.29-2.17.
forces a draw.
32 f4! (06) Of course, a draw with Black was a
'Very bold. I did not have time to check good result, but I again failed to exploit
the variations thoroughly, but I realised some favourable opportunities. Usually this
that this courageous and crucial decision would seriously annoy me, but here noth-
had to be taken now - on the next move it ing of the sort occurred. I felt increasingly
would already be too late.' (Karpov) apathetic, and simply could not force
32 ... 'it>g6 (15) myself to play flat out. This tendency
If 32".gxf4 33 l:i.xf4 l:i.hh2, then 34 Ji.xf6+! became increasingly evident in the very
tLlxf6 35 l:i.e7+ 'it>g6 36 l:i.e6, regaining the next game, played after a weekend.
piece with a draw. Before this game Karpov arrived at the
33 fxg5 theatre before me, but, strangely enough,
'Also possible was 33 f5+ 'it>f7 34 Ji.xf6! he appeared on the stage five minutes later
tLlxf6 35 l:i.e7+ 'it>f8 36 l:i.e6 tLld7 37 l:i.e7 l:i.hh2 than I did. 'Is this part of the psychological
38 l:i.xd7, and Black is obliged to force a warfare?' Leontxo Garcia asked him in an
draw.' (Karpov) interview for Spanish television. The ex-
H ... 'it>xg5 34l:i.f1! champion explained: 'I arrived at the
Nullifying the effect of Black doubling Casino earlier than usual because the Police
rooks on the 2nd rank. cleared the way. Then something unusual
happened. I went to relax in my rest room
and lost any notion of time. I closed my
eyes and switched off, as chess players say.
When I opened my eyes, I saw that the
reporters and Kasparov were on the stage
and that the clock was running. Then I left
the rest room immediately'.
Game 41
G.Kasparov-A.Karpov
World Championship Match,
14th Game, Seville 16.11.1987
34... 'it>g6 Caro-Kann Defence 817
I A last trick' (Karpov). After 34".l:i.hh2
35 l:i.xf6! tLlxf6 36 d7 l:i.cg2+ 37 'it>f1 l:i.f2+ 38
'it>gl (but not 38 'it>el?? l:i.c2! 39 ii.xf6+ 'it>g6!) 1 e4
38".l:i.hg2+ things would again have ended The avoidance of 1 c4 was probably as-
in perpetual check. sociated with my psychological condition. I
35 l:i.f2! (of course, not 35 l:i.df4? l:i.hh2) was satisfied with the slightly better posi-
355
Kasparov vs. Karpov: 1986-1987
tion with the two bishops which arose in 'To play 13 ... 'iVd6, provoking the ad-
the 10th game. vance 14 g3, which all the same comes into
1 .•. c6 2 d4 d5 3 CDC3 dxe4 4 CDxe4 CDd7 5 White's plans, would have been inadvis-
CDf3 CDgf6 6 CDxf6+ CDxf6 7 c3 ~g4 8 h3 able.' (Makarychev). Especially since after
~xf3 9 'iVxf3 14 ... h5 (Georgadze) 15 M! 0-0-0 16 ~g5
This variation is not the most aggres- Black would have been unable to build up
sive, but it is perfectly suitable for fighting any real attack on the kingside.
for an advantage, including when you are 14l:!.dl (06) 14... 0-0 (07) 15 C4 (13)
leading 'plus one' in a match for the world
championship. You simply have to be in
the mood to play such positions with
White. But it was precisely that mood
which I lacked, and the consequences were
soon seen.
356
The Fourth Match: 1987
idea of 20 ... ~c7 21 'iVc3 itb6 22 b4. ~e2 ~c6 (giving up the exchange to avoid
2o ... h6 (02) 21 a3 (11) 29 ... aS 30 ~xaS) 30 ~bS ~xcS 31 ~xcS ~c8
Up to this point I had played correctly 32 'iVdS e3 33 ~e2 exf2+ 34 ~xf2 'iVf6+ 3S
and had gained a quite promising position, ~g2 White has an obvious advantage.
but here I committed an irreparable psy- It stands to reason that Karpov would
chological mistake - I offered a draw ... not have made the move 21...CUe8, which is
'against the general rules'; moreover, he
did not even consider it! I should also have
guessed this at the board ... But what should
Black play? If 21...a6, then 22 'iVb6! is
unpleasant, for example: 22 ... eS (22 ... CUe8 23
dS) 23 dxeS ~xeS 24 ~xd7 ~xd7 2S ~xd7
'iVxd7 26 ~g2, and the white bishops have
become much more active.
Therefore Karpov would most probably
have begun 'marking time' - say, with
21...'iVf8. But then White could have played
22 'iVc2 with the idea of b2-b4, 'iVb3, a3-a4
and b4-bS (he only needs to make sure that
'Black has no grounds for refusal, since the possible ... e6-eS advance does not
with the queen on b3 he is unable to play equalise).
... e6-eS: 21...eS? 22 dxeS ~xeS (22 .. Jhd2 23 I think that in any case Black would
~xd2 ~xd2 24 exf6!) 23 ~xd7 ~xd7 24 ~xd7 have faced a rather unpleasant defence.
'iVxd7 2S ~xa7, winning a pawn.' Karpov did not think for along about
(Makarychev). And if 2S ... bS, then 26 ~b6 my offer - less than half a minute passed
'iVe6 27 a4! bxc4 28 'iVc2 etc. before he agreed to a draw (Yz-Yz). Times:
'But what am I to play after 21...CUe8 1.34-1.14.
with the idea of ... cud6-fS?' I asked my
astonished trainers after the game. And The day after this game, the two con-
indeed, the attempt 22 'iVc2 cud6 23 ~g4 is testants were interviewed again by the
parried by 23 ... 'iVf6!, for example: 24 ~f4 g6 Spanish TV correspondent Leontxo Garcia,
2S ~eS 'iVe7 26 ~d3 hS 27 ~f3 cufS with who in particular informed them that 'the
equality, or 24 dS cxdS 2S ~d4 'iVe7 26 cxdS spectators are already becoming tired with
~c7 27 'iVb3 fS 28 ~f3 eS also with a roughly this lack of fighting spirit'.
equal position, but one that is strategically Kasparov: 'I have indicated already that
dangerous for White - his passed dS-pawn the end of this Championship is still far
is securely blockaded by the knight, ahead and that there are sports targets
whereas the black e-pawn is mobile. which are prevailing. This is why it makes
However, my feeling for dynamic play no sense to fight mercilessly in each game.
let me down! It transpired that 22 dS! cxdS Besides, much energy is used up already.
23 cxdS eS 24 d6!, sacrificing the 'trump' Therefore, if a position is complicated, you
pawn, was much stronger. After both play; if it is simple, a draw is signed ... Of
24 ... 'iVe6 2S 'iVxe6 fxe6 26 a4 CUxd6 27 as ~f8 course, it would not have been bad if
28 ~cS ~e7 29 b4, and 24 ... CUxd6 2S ~g4 Karpov had attacked. It is him who has to
~c7 26 'iVb4 ~c6 27 ~f3 ~cc8 28 ~cS e4 29 push. I can wait. It depends on the devel-
357
Kasparov vs. Karpov: 1986-1987
opment of the game. Karpov plays more preparation for the match.
solid openings than I do. It is difficult for 5... dxC4 6 'Yi'xC4 0-07 e4 tt:la6 (08)
me to play to win. I give him better chances I did not even consider 7... iL.g4 (Game
with Black. I don't know. We shall see what Nos.IS, 17), but I wondered what surprise
happens in the next games, how Karpov the opponent's team had prepared for me
plays with White and how he does with after 7... tt:la6. The wounds from my failure
Black. It depends more on him than on in the 19th game of the 1986 match had
me ... I believe that further on the blood now healed: I twice employed this variation
will be spilled which the fans have been in the spring of 1987 and on neither occa-
waiting for.' sion did I experience any problems.
Karpov: 'If I had received the present I 8 iL.e2 c5 9 d5 e6 100-0 exd5 11 exd5 iL.f5
made Kasparov in the eleventh game, this 11...'Yi'b6 (Stahlberg-Najdorf, Saltsjo-
point which is as big as a blessing from baden Interzonal 1948) is premature be-
heaven, I would certainly cling to it as cause of 12 .if4! .if5 (12 ... 'Yi'xb2? 13 .l:tab1
Kasparov is doing now. But there is still a 'Yi'a3 14 .l:tb5 with the threat of tt:lb1) 13 iL.e5!,
long and very intense fight ahead. An when White's chances are better; this
advantage of one point is not decisive. This bishop is hard to drive away (there is not
advantage is not too stable and cannot the move ... tt:lb8-d7), and the exchange of
guarantee a quiet life'. the dark-square bishops (after ... tt:lf6-d7) is
The short and outwardly inconspicuous not something Black wants.
14th game had an important psychological
outcome. My refusal to fight for a win even
in a safe position induced Karpov to play
more sharply, and in the next two games he
permitted himself strategically risky open-
ing experiments as in the 1986 match. But I,
in order somehow to rouse myself, was
forced before the 15th game to take a sec-
ond time-out.
Game 42
A.Karpov-G.Kasparov
World Championship Match, 12 .l:td1 (12)
15th Game, Seville 20.11.1987 The prelude to a surprise - at the time
Grunfeld Defence 097 this move was much more rare than 12 iL.e3
or 12 iL.f4, as Karpov played in the 19th
game of the previous match (the notes to
1 d4 tt:lf6 2 C4 g6 3 tt:lc3 d5 4 tt:lf3 iL.g7 5 'Yi'b3 that game explain what we had prepared
A return to the dispute of the previous against 12 iL.e3 and why 12 iL.f4 lost its
year. This was the fourth variation Karpov popularity).
had tried against the Griinfeld Defence, 12 ... Me8 (13)
each time demonstrating new and topical Here I tried to guess my opponent's
ideas which developed contemporary idea behind immediately determining the
theory. This testifies to his high level of position of the £I-rook (at the time this was
358
The Fourth Match: 1987
considered less accurate than 12 ~f4(e3) replies 14 ~g5, forcing the play thanks to
and Madl), and I decided to make a useful the attack on the c5-pawn.' (Makarychev)
rook move, intending ... iLle4.
But later it transpired that 12 ...'iVb6! is
stronger, negating the effect of d5-d6. For
example: 13 d6?! MadS (now the d-pawn is
immediately under attack) 14 iLla4 ~c6! 15
~e3 iLle4 16 'iVb5 ~d7! with advantage to
Black, or 13 ~4 ..ic2! 14 Md2 ~f5 (creating
discomfort for White) 15 Mdl ~c2 16 Md2
with a draw (as in two rapid Anand-
Kasparov games, Frankfurt 1999).
In the end it was 12 ...'iVb6 that neutral-
ised 12 Mdl. Practice helped to establish a
clear-cut rule: the ...'iVb6 manoeuvre is
always good if White does not manage to My modest move of the rook's pawn
play ~c1-f4-e5 (d. the note to Black's 11 th (Black as though invites the opponent to
move). Here if 12 ~f4?! there is 12 ... ~xb2! declare his intentions) carne as a surprise to
(Muir-Atalik, Batumi 1999), when 13 Mabl? Karpov: he thought for a long time and
~xbl 14 Mxbl is bad because of 14 ... b5! - a made a poor reply.
typical move in this variation. 14 h3?! (29)
13 d6! (04) Whereas for Black the move ... h7-h6 was
A novelty which kept Black under pres- vitally necessary, at the moment there is no
sure for nearly a whole decade. 'This is specific reason for White to play h2-h3.
White's idea! The passed pawn is assigned Only 14 ~f4 (Game No.48) gives chances of
the role of a battering-ram, and the space an advantage.
gained can be used to expand the sphere of Karpov's sluggish move can be ex-
influence of the pieces.' (Taimanov) plained by the fact that the character of the
Indeed, if the passed pawn can advance position was not in accordance with either
towards the queening square, why his style or his chess philosophy. As had
shouldn't it do so? Especially since Black occurred both in the previous match and in
cannot attack it directly, and all his ideas a number of games from the present one,
for counterplay - exploiting the advanced he tried to change the course of the play in
position of the white queen, and the ma- an unusual and double-edged set-up,
noeuvres ... iLla6-b4-c2 and ... iLlf6-e4 or where the value of every move was very
... iLlf6-d7-b6 - are now rather dubious. high - the d-pawn may be instantly trans-
After some thought, at the board I intui- formed from a strength into a weakness! -
tively found the best reply ... and at the board it was difficult to find the
13 ... h6! (23) most effective solution. But in complica-
Essential prophylaxis. 'It is not easy for tions of this sort (as in the labyrinth of the
Black to activate his pieces. It is this aim Scheveningen Sicilian with g2-g4) he was
that is pursued by Kasparov's far from not in his element, and therefore, even in
obvious move. If 13 ... iLle4? White has the my 'drowsy' condition, I was able to solve
very unpleasant 14 d7! Me7 15 iLlxe4 Mxe4 my problems. It was not without reason
16 ~g5!, while to the immediate 13 ... iLlb4 he that for our fifth match (1990) Karpov
359
Kasparov vs. Karpov: 1986-1987
prepared a clearer and more balanced set- i.xc2 - an interesting queen sacrifice,
up against the Griinfeld: 4 cxd5 CDxd5 5 e4 although after 21...ii.b6! 22 ii.xh6 .l:i.e2 23
CDxc3 6 bxc3 ii.g7 7 ii.e3 c5 8 'iVd2 (about ii.g5 'iVe8 Black has good counterplay.
which see the next volume).
360
'IS a4? would have been met by the from the 14th and 16th games of the previ-
very strong reply Is .. .ebc2!, and if 19 l:txc2, ous match, where the opening was - the
then 19 ... c4!' (Georgadze). 'The tempting IS Ruy Lopez! But, of course, this similarity is
a3 would have left the queen a loophole at purely superficial: (Taimanov)
a2, but the "Kasparov" knight would have
leapt directly on to "its" square - IS ... ct:Jd3!
19 .i.xd3 .i.xd3 20 l:txd3? c4: (Makarychev)
18 ... c4 (05) 19 a4! (15)
Very strong: Karpov realised that he
needed urgently to create counterplay, as
otherwise Black's activity would become
threatening. 19 ct:Jd4?! was much weaker:
19 ... .i.xd4 20 l:txd4 ct:Jc2 21 .i.f3 (21 g4 g5 or
21..:iiVf6) 21.. ..i.d3! (but not 21...l:ta7(?), as I
suggested in Informator, because of 22 l:tc1!
l:tel+ 23 'iVxel ct:Jxel 24l:txel with sufficient
compensation for the queen) 22 .i.xaS 'iVxaS
23 .i.e3 'iVc6, and White is in trouble. 20 ... ct:Jbd3 (OS)
'In some circles a black knight, estab-
lished in the middlegame at d3, became
known as a "Kasparov knight" - following
the 16th game of the second match (19S5).
Therefore one should not be surprised by
the actions of Karpov, who resolutely (and
at any price) eliminated the opponent's
piece, which induced such unpleasant
memories: (Makarychev)
The seemingly tempting 'fork'
20 ... ct:Jb3?! (the wrong square!) would have
given the opponent too wide a choice:
1) 21 d7 (my recommendation in Infor-
19 ... ct:JcS (04) mator) 21...l:i.e7 (21...11fS 22 .i.d6) 22 .i.c7
The most natural and energetic move. In 'iVxc7 23 dSiV+ l:!.xdS 24 l:!.xdS+ ~h7
the press centre the more cautious 19 ... l:tbS (Makarychev) 25 l:ta3 (25 l:ta4 or even 25
(Georgadze) was suggested, 'but in this ~xa6 is also not bad) 25 ... .i.c2! 26 b6 'iVxb6
case too after 20 axb5 axb5 21 l:ta7 White 27 ~d6 'iVxd6 2S l:txd6 with a sharp multi-
has his trumps' (Taimanov). 19 ... 'iVb6 was piece ending;
also possible, for example: 20 .i.e3 l:i.xe3 21 2) 21 l:ta4!? (the preference of most of the
fxe3 l:teS with compensation for the ex- experts) 21...ct:Jxd2 22 l:1.xb4 ct:Je4, and now
change (my recommendation in Informator) not Georgadze's move 23 ct:Jd5(?) because of
or 20 axb5 axb5 21 l:txaS l:txaS 22 .i.e3 'iVbs 23 ... axb5! (24 l:txb5 ct:Jxd6 or 24 ct:Jc7 .i.d7!),
with complicated play. but simply 23 .i.xc4 with excellent compen-
20 axbs (04) sation for the exchange. And in the event of
'The contours of the position bear an 21...axb5 22 l:txb4 ~al 23 'iVxal ct:Jxal 24
amazing resemblance to the middle game l:txb5 White has quite sufficient compensa-
361
Kasparov vs. Karpov: 1986-1987
362
The Fourth Match: 1987
(even slightly unharmonious) move but under the influence of the tense com-
29 ... iLf8!. After 30 'iVd7 (it turns out that 30 petitive and psychological situation in the
d7? is bad because of 30 ... l:i.al+) 30 ... 'it'g8! match we continued playing and even
(30 ... 'iVe6 leads only to a draw: 31 'iVxe6 fxe6 adjourned the game.
32 d7 iLd6 33 g3 l:i.d8 34 ttJd4 l:i.xd7 35
ttJxe6) 31 'iVc7 l:i.al+ 32 'it'h2 l:i.a6! White
would have lost his d-pawn and been
forced to conduct an unpleasant defence,
without good compensation for the ex-
change.
29 ... l:i.al+!? 30 'it'h2 l:i.dl was also more
interesting than the move in the game -
after 31 d7 iLe5+ 32 'iVxe5 'iVxe5+ 33 ttJxe5
l:i.xd2 34 ttJxf7 l:i.xd7 35 ttJe5 l:i.d2 Black
would have won the b-pawn and for a long
time could have tormented White.
30 'iVcS! (03)
After this accurate reply I realised that I 37 'it'g2 iLb2 38 ttJe7+ 'it'f6 39 ttJc6 l:i.d7 40
had missed my chance. Now Black does not ttJb8 l:i.d8 (02)
win the d6-pawn, but only the b2-pawn. 40 ... l:i.b7! 41 ttJc6 l:i.d7 or 41 ttJa6 iLa3 42
ttJc7 l:i.b3 was slightly more accurate.
41 d7 'it'e6 42 'it'f3 (02) 42 ... iLa3 (02)
The simplest. Here the game was ad-
journed, and Karpov sealed his move,
which, as it later transpired, was 43 iLc7
(06). A draw was agreed (Yz-Yz) without
resumption in view of the obvious
43 ... l:i.xd7 44 ttJxd7 'it'xd7 45 iLf4 'it'e6 46 'it'e4
f5+ 47 gxf5+ 'it'f7 (Georgadze) or 45 iLb6
'it'e6 46 'it'f4 f5! 47 g5 iLd6+ and ... 'it'f7-g6
(Makarychev). Times: 2.43 -2.19.
363
Kasparov vs. Karpov: 1986-1987
364
The Fourth Match: 1987
even if in a dangerous but well-studied ~xf6 14 cxd5 ~xd4 15l:tc1 ~g416 'iVb2 ~e2
position, also taking account of the fact 17l:tfel ~xd3 18 ~xb7 l:tab8 the chances are
that, captivated by the magic of numbers, I equal.
might overstep the mark at some point. More often Black immediately prevents
7 d3 (02) the pin on his knight by 10 ... h6 (the source
Before the match we analysed this criti- game: Pfleger-Korchnoi, Hastings 1971/72).
cal line quite deeply and found a number of Nowadays this is the main tabiya of the
fresh ideas. I avoided the old 7 ct:Jd5 variation, occurring in the games
(Rubinstein-Duras, Carlsbad 1911) because M.Gurevich-Salov (Linares 1991), Karpov-
of 7... ct:Jxd5 8 cxd5 ct:Jd4 9 ct:Jel (9 ct:Jxd4 exd4 Anand (Frankfurt (rapid) 1997), Tkachiev-
10 e4 c6 or 10 e3 ~c5 is no better) 9 ... c6 10 Anand (Moscow 2001), Ivanchuk-Kramnik
e3 ct:Jb5! with a comfortable game for Black (Turin Olympiad 2006), and others. In the
(Petrosian-Kuzmin, Moscow 1974). 1990s I was aiming here for the unexplored
7 ... ~xC3 (03) 8 bXc3 e4 9 ct:Jd4 move 11 ct:Jc2!?
9 ct:Jg5 leads to a position from the varia- The immediate 9 ... h6 seems tempting
tion 6 ... e4 7 ct:Jg5 ~xc3 8 bxc3 l:te8, where (why hurry with the exchange of the e4-
instead of the sharp 9 f3! (2nd and 4th pawn, which is cramping the opponent's
games), the insipid move 9 d3 has been position?), but White acquires an additional
made. possibility in the fight for an opening
advantage.
10 dxe4! (03)
An important novelty, which was pre-
pared for the match (previously 10 c5, 10
l:tb1, 10 ct:Jxc6 and 10 ~c2 had each oc-
curred once). Although this capture (sug-
gested, I think, by Dorfman) looks unaes-
thetic, it effectively refutes 9 ... h6. However,
Karpov, as he himself put it, 'was not
caught unawares'.
10 ... ct:Jxe4 11 ~C2
9 ... h6?!
9 ... exd3 10 exd3 is better. The source
game Botvinnik-Basman (Hastings 1966/67)
continued 10 ... ct:Jxd4 11 cxd4 d5 12 ~e3 ~e6
13 l:tc1 h6 14 h3 c6 with a solid position for
Black. This assessment is also confirmed by
12 l:tbl h6 13 ~f4 b6 14 ~e5 c6 15 h3 ~e6
(Aronian-Gelfand, Dresden Olympiad
2008). In 1987 we considered 12 ~g5 to be
more promising, but later analyses and
practice showed that after 12 ... h6! (12 ... c6 13
cxd5 cxd5 14 'iVb3 is less good for Black, We have reached a similar type of posi-
Gulko-Hubner, San Francisco 1995) 13 ~xf6 tion as in the Nimzo-Indian Defence with
365
Kasparov vs. Karpov: 1986-1987
g2-g3, where Karpov suffered in the previ- .itxe4 dxe4 IS .itf4(e3) iYc6 is harmless,
ous matches (Game Nos.2, 4). White has the while 14 e3!? .itfS IS cxdS ttJxg3 16 'iVb2 ttJe4
same bad pawns, but potentially powerful 17 f3 followed by e3-e4 and f3-f4 leads to
bishops; in addition, the enemy knight complicated, double-edged play.
cannot be maintained at e4 - 11...iYe7 12 12 ... iYxds
ttJbS!, while after 11...ttJcS 12 .ita3 Black I have to admit that this swift and con-
begins to experience some discomfort. fident response by the opponent to my
After assessing the position as favouring novelty unsettled me, although objectively
White, my trainers and I cut short our White still retains an opening advantage.
analysis. Therefore Karpov's reply came as 13 e3?! (16)
a surprise to me. Alas, a second-rate move. The solid 13
1l... dS! (01) .itf4(e3) was good, but it was even better to
'The crux of Black's defensive strategy. play 13 ~dl .itfS, when Karpov was count-
As Nimzowitsch commented long ago, ing on 14 ttJxfS (14 f3? ttJf2!) 14 .. :~xfS IS
doubled pawns possess enormous static .itf4(e3) ttJxg3, but after 14 'iVb2!? White has
strength. I would venture to suggest that appreciable pressure - and in addition after
the secret of this strength is largely psycho- 14 ... ttJxd4 IS cxd4 he gets rid of his weak
logical. One wants to win such pawns, not c3-pawn (Portisch-Salov, Tilburg 1994).
exchange them. After home analysis Kar- Karpov did not play this again, but in
pov acts in contrary fashion.' (Makarychev) the present game his psychological reckon-
'In our preparations for the game, this ing proved fully justified. I became agitated
move was suggested by Mikhail Podgaets, and was unable to work out the variations
one of my seconds. Black gains excellent accurately, and so I chose the 'clearest'
counterplay. It was no accident that Kas- continuation, in the hope that the two
parov thought for nearly 40 minutes over bishops would enable White to fight for the
his reply. He had to examine a variety of initiative. But now Black too can reveal his
possibilities - a number of spectacular strategic trumps.
variations remained off-stage.' (Karpov)
12 cxdS (36)
12 ttJbS? was weak: 12 ... .itfS! 13 cxdS
ttJxg3 14 e4 ttJxfl IS dxc6 (if IS exfS, then
IS ... ~el! 16 .itxfl iYxdS), and now not
IS ... bxc6(?) 16 ttJd4 .itxe4 17 .itxe4 iYh4 18 f3
ttJxh2 (Makarychev) 19 iYf2! or IS ... .itxe4(?)
16 .itxe4 iYh4 (Karpov) 17 f3! ttJxh2 18 iYf2,
but immediately IS ... iYh4! with a powerful
attack: 16 f3 iYxh2+ 17 'it>xfl .ith3!
(H.Olafsson-Danielsen, Hillerod 1995), or
16 cxb7 ~ad8! 17 ttJd4 ..Itxe4 18 .itxe4 ~xe4
19 h3 ~de8 20 iYd3 ~el, and in each case
Black wins. 13 ... ttJas! (39)
12 ~dl ttJxd4 13 ~xd4 iYf6! would have This is the correct decision - to retain
given Black a more or less tolerable posi- the knights and occupy the c4-point. If
tion -14 .ite3 cS! (Makarychev), 14 .itf4 .itfS! 13 ... .itfS, then 14 'iVb3!. Georgadze (in the
(Hertneck-Epishin, Aschach 1997), or 14 match bulletin) and Karpov himself (in
366
The Fourth Match: 1987
Injormator and his book of best games) counter-attacking resource ... c7-c5: (Tai-
suggested that '13 .. .iDxd4 14 cxd4 c6 15 manov)
..ib2(?!) ..if5 16 f3(?) tDxg3 17 e4 tDxe4(?) 18 Indeed, Black is ready to drive the
fxe4 ..ixe4 came into consideration', over- knight from d4 by ... c7-c5. But on the other
looking the simple 17 ... tDxfl, winning. But hand, his minor pieces are stuck on the
after 15 'iVb2! with the idea of f2-f3 and e3- queenside and White can concentrate his
e4 White's chances are better. forces for a breakthrough on the kingside.
14 13! (24) The prospect of a fierce battle brought me
I was wrong to think for so long: this into a state of nervous excitement: despite
logical move, leading to the advance of the all the opponent's trickery, White has
pawn chain, should have been made im- nevertheless obtained a promising position!
mediately! But there was a major drawback, which I
14 ... tDd6 (05) 15 e4 'iVcs (02) 16 ..ie3 (07) appreciated too late: the position was
16 ... tDdc4?! (09) suited to Karpov's style. Perhaps even
It was better to move the knight from slightly dubious, but 'his' type of position
the edge of the board - 16 ... tDac4! 17 ..if2 nonetheless - with normal development of
iVh5, for example: 18 g4 'iVg6 19 h3 h5 20 the pieces and customary positional land-
~adl hxg4 21 hxg4 tDe5 22 tDf5 ..ixf5 23 marks ... And here the 'phenomenon of one
exf5 iVh6 with complicated play, in which move' again showed up: firstly, I moved
Black relies on his centralised knights and the wrong rook, and secondly, it was to the
White on the potential of his bishops. wrong square.
17 ..if2 'iVe7
18 ~ad1? (06)
'It is time to sum up the results of the 'This natural developing move was
opening. It is obvious that White has man- unanimously criticised by the experts: it
aged to achieve certain gains: he has a gives Black grounds for counterplay. 18 f4
spatial advantage, a dynamic pawn centre, was hardly good in view of 18 ... c5 19 tDb3
and two (potentially) strong bishops, while b6, but after 18 ~fel!? (many also recom-
for the moment the black knight at a5 is out mended 18 fiae1 - G.K.) 18 ... c5 19 tDfS White
of play. But Black also has his achieve- would have retained chances of developing
ments: the invulnerability of his pawn his initiative' (Taimanov). For example:
ranks, the possibility of freely mobilising 19 ...'iVc7 20 f4 ..ie6 21 'fWdl!? ~ad8 (2l...~h7?
his forces, a powerful outpost at c4 and the 22 'ii'h5) 22 'iVg4, and although Black's
367
Kasparov vs. Karpov: 1986-1987
24 fS?! (09)
'Sensing that in his "lawful" 16th game
the initiative is slipping away, Kasparov
decides on extreme measures.' (Taimanov).
'This was the moment when the number 16
20 ••• ..ig4! (1S) turned against its favourite, by inducing
An important move, exploiting the tem- him to take a highly critical, intuitive
porary weakening of the light squares. decision in a complicated tactical position.'
'This is where the drawbacks to the posi- (Makarychev)
tion of the rook at dl are seen. By gaining a I had at least two ways of maintaining
tempo, Black gains the opportunity to approximate equality:
368
The Fourth Match: 1987
1) 24 l:!.e1 cS 2S ctJf3 (2S ctJb3 b6), and 27 ... ctJec4 28 jL,xd4 ctJc6 or 28 ... ctJd2 29 'iYeS
now not the greedy 2S ... jL,xf3?! 26 jL,xf3 l:!.d2 l:!.xd4 and 30 ... ctJxfl there is no danger and
27 ~e4 l:!.xa2 28 fS (Georgadze) or 28 .ii.hS the chances are equal.
with an attack, but 2S ... jL,fS 26 'iYc1 b6 - 25 'iYe4 (02) 25 ... cxd4 (OS)
here I did not like the fact that Black had Of course, not 2S ... ctJd2? 26 'iYxg4 ctJxfl
some pressure on the light squares after 27 27 e6! cxd4 28 f6 g6 29 exf7+ 'it'h7 (29 ... 'it'xf7
ctJh4 jL,e6 or 27 l:!.d 1 jL,e4; 30 jL,dS+! Taimanov) 30 ~xc8 l:!.xc8 31
2) 24 jL,e4!? (fighting for the light iLxd4, and Black has no defence.
squares) 24 ... cS 2S ctJfS 'iYd7 (2S ... 'it'h8(f8) is 26 'iYxg4 (01)
also good, but if 2S ... 'iYc7 White has the
interesting 26 h3!? .ii.xh3 27 l:!.d1 b6 28 'iYe2
with the initiative for the pawn) 26 .ii.xcs
ctJd2! 27 ctJe7+(?!) 'it'h8 28 jL,dS(?) iYbS!
(Georg adze), or 28l:!.f2 ctJxe4 29 'iYxe4 .ii.h3!,
and therefore the computer move 27
l:!.a1(c1) is necessary, with equality. Or
24 ... b6 2S l:!.e1 cS 26 ctJfS (26 ctJf3 'iYd7,
maintaining the blockade) 26 ... 'iYe6 27 ctJe3
ctJxe3 28 l:!.xe3 with equality (28 ... l:!.d1+ 29
Il.e1).
In any event, I should have reconciled
myself to the loss of the initiative and
gradually adjusted my play. True, even 26 ...ctJxe5 (14)
after the sharp 24 fS the position remains The correct practical choice. 26 ... d3? 27
within the bounds of equality - as yet I had jL,d4 followed by the inevitable eS-e6 was
overstepped the mark purely in the psycho- bad for Black. According to Karpov,
logical sense ... 26 ... ctJe3(?) was 'interesting': 27 jL,xe3 dxe3
28 e6 f6 29 'iYf3(?) 'iYxc3 'with sharp play',
but after 29 iYb4! White has an obvious
advantage.
Karpov laments that shortage of time
prevented him from finding 'the most
clear-cut way - 26 ... dxc3 27 e6 (27 jL,d4?
ctJxeS!) 27 ... ctJeS followed by ... £7-f6 with
advantage to Black'. However, here too
White would have maintained the balance
without particular trouble, and in more
than one way:
1) 28 exf7+ ctJxf7 29 .ii.d4 ctJgS 30 jL,e3! c2
31 iYhs with the threat of fS-f6, or 30 ... ctJc4
24 ... c5 (07) 31 jL,xgS hxgS 32 'iYxgS also with sufficient
The natural move. The commentators counterplay, while if 30 ...'iYc4, then 31 'iYxc4
unanimously condemned 24 ... ctJxeS(!?) 2S ctJxc4 32 jL,xgS hxgS 33 l:!.c1 with equality;
h3 cS 26 hxg4 cxd4 27 ~e4 'with a danger- 2) 28 'iYe4 f6 29 ~a4! (intending l:!.c1)
ous attack for White', although after 29 ... b6 30 jL,e4 l:!.d2 31 l:!.c1 'iYc4 32 iVxc4
369
Kasparov vs. Karpov: 1986-1987
A cold shower. For some reason I over- edged play' (Makarychev). In the endgame
looked this simple move. The sharp change after 33 'iVb5 'iYe2 34 'iYxe2 ~xe2 35 fxg7
of scene completely confused me, and I did White would still have had some saving
not even try to save myself in an endgame chances.
a pawn down with rook and bishop against 33 fxg7 ct:JC4
rook and knight. Black completely dominates and the bat-
31 'iYb2? (14) tle is at an end. What a pitiful spectacle I
Missing the last chance. Makarychev had managed to make of White's recent
recommended 31 ~f2 ~xf6 32 'iYxf6 gxf6 33 good position ...
:l:!.xf6, but 'after 33 ... :l:!.d1+ 34 .ifl ct:Jc4 White 34 'iYc2 ~xg7 (02) 35 .ids ct:Jd6 (04) 36 'iYb2
is lost' (Karpov). Things are also difficult (01) 36 ... 'iYes (01) 37 .ib3 as 38 'iYf2 fS 39
for him after 34 ~f2 :l:!.d2+ or 34 l::tfl l:!xfl + 'iYb2 (01) 39 ... bS 40 a3 ~g6 (01)
35 ~xfl ~g7 36 ~e2 ~f6 37 .id5 ct:Jc6, Here the game was adjourned.
planning ... ct:JdS-e6-c7 etc. In addition Black
has the useful move 33 ... ~g7!? with the
idea of 34 :l:!.f5 ~d1+!, while if 34 ~fl, then
34 ... b5.
However, it was still possible to fight for
a draw by 31 "iVxe6! fxe6 and now either 32
~e1 gxf6 33 ~xe6 ~f7 34 ~e2! (Taimanov's
recommendation 34 ~e4 is weaker in view
of 34 ... ~xe4 35 .ixe4 ~e6 and ... ~e5), or 32
.ih3! l:!d6 (Karpov) 33 l:!e1 gxf6 34 .ixe6+
'iit>fS 35 ..if5 or 32 ... e5 33 .ie6+ ~fS 34 fxg7+
Wxg7 35 ~f7+ ~g6 36 ~f5 with hopes of
exploiting the energy of the rook and
bishop. 41 'iYf2 (02)
31 ... 'iYe3+ 32 ~hl The sealed move. The next day White
resigned (0-1) without resuming (the
simplest is 41...~d3). Times: 2.32-2.2S.
Drawing series
The last third of the match had begun, and
here for me the question loomed large:
32 ... b6! (02) what to do next - aim to win the match or
'The right way! Now the black knight continue seeking salvation in draws?
switches to an ideal position, whereas the After the blow suffered in the 16th
obvious 32 ... ~d2 would have led to double- game, for a short time I 'woke up'. And in
371
Kasparov vs. Karpov: 1986-1987
Game 44
A.Karpov-G.Kasparov
World Championship Match,
17th Game, Seville 25/26.11.1987
King's Indian Defence E97
9 ttJd2
At that time this was one of the most
1 ttJf3 fashionable continuations: in anticipation of
This move already requires a comment: ... f7-f5 White reinforces his e4-point and
as in the first match (1984/85), Karpov plans b2-b4, c4-c5 and ttJc4. 9 ~d2 (Game
suddenly abandons 1 d4, avoiding a dis- Nos.90, 92 in Volume IV of My Great Prede-
cussion in the Griinfeld Defence. cessors) was much more rarely played,
1 ...ttJf6 (07) while the one-time main move 9 ttJe1 (Game
Natural hesitation: should I not play Nos.29, 64 in Volume IV of My Great Prede-
1...d5 and after 2 d4 ttJf6 3 c4 e6 (Game cessors) was revived closer to the end of the
No.46) transpose into the good old Queen's century by Korchnoi (Game Nos.39-41 in
Gambit? Volume V of My Great Predecessors).
2 C4g6 But after wins by Kramnik during 1996-
In principle I was also ready to play the 98 the direct 9 b4 became the most popular
variation with 2 ... c5 3 ttJc3 d5, which we move, since the skirmish on the queenside
had analysed quite extensively (Game after 9 ... a5 10 ~a3 axb4 11 ~xb4 ttJd7 (or
No.SO). 11 ... b6) 12 a4 favours White. Therefore
3 ttJC3 ~g7 Black mainly replies 9 ... ttJh5, and then not
Here 3 ... d5 appealed to me less. While 10 g3 f5, 10 c5?! ttJf4 or 10 ttJd2 ttJf4 (Game
thinking about 1...ttJf6, I placed my choice No.13 in Volume III of My Great Predeces-
on the King's Indian. In the fifth match sors), but 10 .l:!.e1! - this modern plan also
(1990) it was to become my main opening, brought Karpov success against Kamsky
but here it came as a surprise to Karpov. (7th match game, Elista 1996). However, in
This was a good psychological move. the 21st century Black increasingly often
4 e4 d6 S d4 0-0 6 ~e2 (03) 6 ... eS 7 0-0 ttJc6 achieves a good game here.
8 dS (11) 9 ... aS
Clearly Karpov spent this time not A typical blockading move. After
thinking about d4-d5, but about which 9 ... ttJd7(e8) 10 b4 White immediately cre-
variation to choose on the next move. In ates pressure on the queenside (if 10 ... a5,
any case White intends to attack on the then 11 bxa5 .l:!.xa5 12 a4 with a growing
queenside, and Black on the kingside. initiative). As for the immediate 9 ... c5 with
8 ... ttJe7 the idea of 10 .l:!.b1 a5 11 a3 ttJe8 12 b4 axb4
372
The Fourth Match: 1987
13 axb4 b6 and ... f7-f5 (Taimanov-Kavalek, position, beginning with the Korchnoi-
Wijk aan Zee 1970) or 10 ... ttJe8 11 b4 b6 12 Geller game, far more often 10 ... ttJd7 11
a4 f5 13 a5 ttJf6 14 'ii'a4 .td7 15 'YiIVa3 .th6 .ta3! followed by 12 b4 axb4 13 .txb4 was
with equality (Larsen-Fischer, 4th match played. In principle this is not bad for
game, Denver 1971), many King's Indian Black: he effectively has an extra tempo
players were frightened off by a variation compared with the variation 9 b4 a5 10 .ta3
which appeared in 1975: 10 dxc6!? bxc6 11 axb4 11 .txb4. However, in my view, the
b4 (too sharp a change of scene!). blockading 10 ... c5 is a simpler way to
10 b3?! equalise. As far as I recall, after this move
A slow move - on encountering a sur- Karpov became slightly disconcerted, on
prise, Karpov played as in the well-known apparently realising that White had
source game Korchnoi-Geller (1st match achieved not quite what he wanted.
game, Moscow 1971). 11 a3 (07)
10 a3 is more energetic, and in the event Here 11 dxc6 is unattractive in view of
of 10 ... .td7 - 11 b3 (preventing ... a5-a4), 11 ... bxc6 12 .ta3 (12 l:!,el Ij2_1/2 Stohl-
after which Geller's amusing manoeuvre Schmidt, Polanica Zdroj 1985) 12 ... c5! with
11...ttJc812l:!,bl ttJb6 is possible, or l1l:!,bl!? the intention of ... ttJc6-d4. But after 11 a3
(Sosonko-Kavalek, Wijk aan Zee 1975) with White is simply a tempo down compared
the idea of 11...a4 12 b4 axb3 13 ttJ(l:!,)xb3, with the afore-mentioned variation 9 ... c5 10
first tried in 1987 by Lputian (slightly l:!,bl a5 11 a3, where b2-b4 is carried out in
earlier he also introduced IIl:!,a2!?). one go.
But I always replied 10 ... ttJd7, when 11 11..• ttJe8 (02) 12l:!,b1
l:!,bl f5 12 b4 led to a variation which rap- In the event of 12 .tb2 f5 13 b4 b6 14
idly developed only after Seville, with the ttJb5 ttJf6 15 exf5 ttJxf5 (but not 15 ... gxf5?! 16
games Gavrikov-Kasparov and Kasparov- f4! Ehlvest-Hellers, Biel Interzonal 1993)
Smirin (55th USSR Championship, Moscow Black also has a good game.
1988). Karpov also twice played this against 12 ... fS 13 b4 axb4 14 axb4 b6
me with White (Skelleftea 1989; Tilburg
1991). These and all our other King's Indian
duels will be annotated in the next volume.
373
Kasparov vs. Karpov: 1986-1987
began spending a considerable time in was that Black had a considerable choice.
thought. He has solved his opening problems, but
15 'ii'b3 (09) 15 ... tiJf6 (17) 16 ..id3 (08) the question is what he wants to do next.
Karpov does not want to weaken him- The move 17 ... ~al causes White a certain
self with f2-f3 and he makes a consolidating discomfort and induces him to simplify.
move, aiming to maintain his e4-point with The typical 17 ... tiJh5!?, threatening an
pieces alone. After the standard 16 bxc5 invasion at f4, also looked quite good.
bxc5 17 'ii'b6 Black can play 17... 'ii'd7, when However, after 18 ~el tiJf4?! 19 ..ifl fxe420
the pressure on the e4-pawn is already tiJdxe4 tiJf5 there is the unpleasant move 21
appreciable. tiJb5. Therefore 18 ... fxe4 19 tiJdxe4 ..ixc1 20
16 ... ..ih6! (09) ~xc1 tiJf5 is more accurate, with a comfort-
Another typical King's Indian move: at able game.
g7 the bishop is not doing anything, and it 18 'ii'c2 (22) 18 ... ..if4 (11)
finds employment on another diagonal - it Provoking the weakening g2-g3. Ac-
presses on the d2-knight and thereby on the cording to Makarychev, King's Indian
centre, and at a convenient moment it is supporters occupied the 'commanding
ready to be exchanged. heights' in the press centre, and they
17 ~b2 (10) unanimously recommended 18 ... ..ixd2(!?)
'If White has to resort to such unprepos- 19 ..ixd2 ~xfl + 20 'It>xfl f4 followed by ... g6-
sessing manoeuvres, his position cannot g5 and ... 'ii'e8-h5, disregarding possible loss
really be very promising,' (Taimanov). In of material. This same variation with an
my opinion, this is a waiting move typical 'unclear' assessment was given in 1l1forma-
of Karpov's style - rather 'creeping' and tor by Karpov and Zaitsev. But to me the
mysterious. plan with ... f5-f4 seemed dubious, whereas
20 ... fxe4! 21 tiJxe4 tiJxe4 22 ..ixe4 tiJf5 with
the ideas of ... 'ii'h4 and ... tiJd4 was quite
safe for Black.
Karpov and Zaitsev suggested 19 'ii'xd2
with the idea after 19 .. .£4(?!) of trapping the
rook by 20 tiJa2(?!), ~b3 and 'ii'b2 (although
after 20 ... f3! Black has strong counterplay).
But here too 19 ... fxe4! 20 tiJxe4 tiJxe4 21
..ixe4 tiJf5 is safer, with the same ideas of
... 'ii'h4 and ... tiJd4, equalising.
19 tiJf3 (10)
Forcibly neutralising the opponent's ac-
tivity. 'White sounds the retreat. But are
17 ...~al!? (43) there any other tempting plans?' (Tai-
,An exceptionally strong move' manov). If 19 g3, then 19 ... ..ih6 is good, as is
(Gufeld). 'An excellent idea: not hurrying 19 ... ..ixd2 20 ..ixd2 (20 'ii'xd2 fxe4 21 tiJxe4
to force events, Black activates his forces' tiJf5 with equality) 20 ... ~xfl + 21 ..ixfl tiJxe4
(Taimanov), and 'together with the bishops, 22 tiJxe4 fxe4 etc. After 19 tiJb3 (the knight
he also wants to exchange a pair of rooks' heads in the wrong direction) 19 ... ..ixc1
(Makarychev) . (Tal's pretty joke 19 ... tiJexd5?! 20 tiJxd5
The reason I thought for so long here ..ixh2+ 21 'It>xh2 tiJg4+ is parried by 22 'It>g3)
374
The Fourth Match: 1987
20 .l:!.xc1 .l:!.xc1 + 21 'iYxc1 fxe4 22 liJxe4 liJxe4 Practically forcing a draw, but not alto-
23 ..ixe4 liJfS (Taimanov) or 23 ... cxb4 Black gether appropriate. I remembered that this
is also alright. Karpov and Zaitsev recom- was my match objective... 24 ... 'iYc7! fol-
mended 19 bxcS bxcS 20 .l:!.b1 .l:!.xb1 21 lowed by ... liJd4, maintaining the tension,
liJdxb1, but after 21...fxe4 and ... liJfS White would have been psychologically more
has not even a hint of an advantage. unpleasant for Karpov.
Therefore 19 liJf3 is the correct move, 25 'iYxd8liJxf3+
provoking mass exchanges. I shouldn't have been in a hurry to make
this additional exchange, since the knight
on d4 could have become a problem for
White. After 2S ... .l:!.xd8!? 26 bxcS bxcS 27
.l:!.b8 ..tf8 Black's defences are solid: 28 liJgS
h6! 29 liJe6+ liJxe6 30 dxe6 <tIe7 31 ..idS ..ia6
(the simplest, although there is also 31...hS,
as well as the sharp exchange sacrifice
31.. ...ixe6!? 32 .l:!.xd8 ..ixdS 33 .l:!.h8 ..ixc4 34
.l:!.xh6 ..td3) 32 .l:!.b3 ..ic8 with equality.
26 i.xf3 .l:!.xd8 27 bXc5 bXc5 28 .l:!.b8 .l:!.f8
After 28 ... hS?! 29 g4! h4 30 gS Black
could have ended up in an unpleasant
position.
19 ... fxe4 (13) 20 liJxe4 liJxe4 21 ..ixe4 .l:!.xc1 29 .l:!.b6 .l:!.f6 (06) 30 .l:!.b8 (20) 30 ... .l:!.f8 31.l:!.b6
(03) 22 .l:!.XC1..tXC1 23 'iYxc1liJf5 24 'iYg5 (21) .l:!.f6 32 ..ie4 (OS) 32 ... ..if5 33 ..ixf5 .l:!.xf5
24 liJgS liJd4 2S bxcS bxcS 26 h4 h6 27
liJe6 ..ixe6! 28 dxe6 'iYxh4 is double-edged
(Karpov, Zaitsev). Therefore White offers
the exchange of queens: in the endgame he
will at least be in no danger. 'Everyone
observing the game at that moment
thought that play would go on for a maxi-
mum of half an hour.' (Makarychev)
375
Kasparov vs. Karpov: 1986-1987
with a draw. But here my pride got the .l::te6 'it>f7 with a draw.
better of me: 'Why should I give up my d6- Here the game was finally adjourned.
pawn? After all, in any case it's a dead
draw!'
35 h4 h6 (03)
Again 35 ... .l::tf3! 36 .l::txd6 .l::tc3 and ... .l::txc4
was possible.
36 'it>g2 'it>g7 37 f3 'it>g8 (05)
There was more risk in 37... g5 38 hxg5
hxg5 39 'it>h3 .l::txf3 40 .l::txd6 .l::tc3 41 'it>g4
(Makarychev), but here too after 41.. ..l::txc4+
42 'it>xg5 (42 'it>f5 e4!) 42 ... 'it>f7 Black would
have gained a draw.
38 'it>f2 g5 (38 ... 'it>g7 was steadier) 39 hxg5
hxg5 40 'it>e3 'it>g7 41 .l::tb8 (05)
The control was reached, but the time 42 .•. 'it>g7?! (04)
for the session had not yet expired, and A rapidly sealed move, with which I de-
Karpov did not want to seal a move. cided not to change anything, but rely on
my adjournment analysis. 42 ... g4! would
again have led to a draw: 43 f4 (43 fxg4
'it>h6) 43 ... exf4+ 44 gxf4 g3 45 .l::ta8 (45 ~f3
.l::txf4+!) 45 ... g2 46 .l::ta1 .l::tg6 47 l:tg1 l:tg3+! 48
'it>f2 .l::tc3 (Makarychev).
An event which affected us both oc-
curred the following day. In the morning
we awaited the offer of a draw, which
could have come any time before twelve
o'clock. We considered the adjourned
position to be absolutely drawn, and eve-
ryone thought that there would be no
resumption. This somehow weighed on me,
41 ... 'it>h7 (01) and possibly also on my opponent... But
To this day I cannot understand why, noon arrived, and Karpov did not in fact
after playing 38 ... g5, I decided against offer a draw. Why was this?
41...g4! with an elementary draw: 42 fxg4 At one o'clock I went out for a walk, at
'it>h6 43 .l::th8+ 'it>g5 44 .l::td8 (44 .l::tg8+ .l::tg6 three we sat down for a meal, and although
with a draw) 44 ... 'it>xg4 45 .l::tg8+ 'it>h3 I was sure that the resumption would
(45 ... 'it>f5?? 46 g4 mate) 46 g4 .l::tf4 47 g5 'it>g4 conclude after a few simple moves (which
48 g6 Wg5 or 42 f4 exf4+ 43 gxf4 .l::tf7 44 'it>e4 is what in fact happened, but only after a
'it>f6 45 f5 .l::tg7 etc. great deal of anxiety!), my helpers and I
42 .l::td8 (06) decided: let's look at the adjourned position
Other attempts were no better: 42 .l::tb7+ once more. And suddenly at 3.15, an hour
~g6 43 .l::td7 (Karpov, Zaitsev) 43 ... g4 44 and fifteen minutes before the clocks were
fxg4 'it>g5 45 .l::tg7+ .l::tg6 or 42 'it>e4 (Karpov, due to be restarted, we discovered a dan-
Zaitsev) 42 ... 'it>g7 43 .l::te8 g4 44 fxg4 'it>g6 45 gerous manoeuvre that my opponent might
376
The Fourth Match: 1987
employ. This was terrible! We faced a This was a chance which White acquired
tough choice: to go into a queen endgame thanks to my marking time. We will con-
with an extra pawn for Karpov, or seek a sider two possible king moves:
draw in a difficult rook endgame. We 1) 43 .. .'it'g8? (this leads to the afore-
worked in great haste - of course, there was mentioned queen endgame) 44 'it>e4 'it>h8
no longer any time to finish the meal. We (after 44 ... 'it>f8 45 l:th7! 'it>g8 46 l:th5 an
seemed to find a literally study-like way of important pawn is lost) 45 l:te7 'it>g8 46 l:te6
saving the game, but in such 'time-trouble' 'it>f7 47 l:txf6+ 'it>xf6 48 g4 'it>g6 49 'it>d3 'it> f6
there was not complete confidence in the 50 'it>c3 e4 51 fxe4 'it>e5 52 'it>b3 'it>xe4 53 'it>a4,
analysis - suppose we had overlooked and the black king is at the cross-roads:
something?! I arrived for the resumption a) 53 ... 'it>f4? 54 'it>b5 'it>xg4 55 'it>c6 'it>f5 56
with the darkest thoughts, 15 minutes late 'it>xd6 g4 57 'it>c6 g3 58 d6 g2 59 d7 gl'if 60
moreover, and Karpov could have thought d8'if, and 'Black faces a gruelling defence
that this was some psychological challenge with insignificant saving chances' (Tai-
- that I had decided in this way to demon- manov), whereas the modern endgame
strate my confidence in a draw and to program Tablebase concludes precisely that
'repay' him for the incident with the re- White wins - in 61 moves;
sumption of the 15th game ... b) 53 ... 'it>d4! 54 'it>b5 'it>c3 55 'it>c6 'it>xc4 56
Nikitin: 'Agitated and hungry, Garry set 'it>xd6 'it>b4 57 'it>e6 c4 58 d6 c3 59 d7 c2 60
off to the Lope de Vega Theatre, not know- d8'if c1 'if 61 'ifd4+ 'it>b3 62 'it>f5, winning the
ing where he was heading - to a game or to g5-pawn, but by no means the game, in the
the scaffold. When the car drove out of the opinion of the all-seeing Tablebase. How-
gates of our villa, we returned to our analy- ever, as I stated back in Seville, such end-
sis and in a now calm atmosphere we ings are practically always lost. 'Of course,
ascertained that our planned system of you can defend for an evening or two, but
defence was correct and would guarantee a then because of this you lose the match as a
draw. But Garry learned of this only after whole ... '
the game.' 2) 43 ... 'it>g6! (the wise Botvinnik was a
43l:ta8 long way from us, but by telephone he
The first surprise. 43 l:td7+! was far more advised that the king should be kept on the
dangerous for Black. 6th rank, and this heartened me) 44 g4! 'it>h6
45 'it>e2 'it>g6 46 'it>d2 'it>h6 47 'it>c2 l:txf3 48
l:txd6+ 'it>g7 - it was this position that I
feared: we were quite unable to give it an
exact evaluation. But later a 'cast-iron'
draw was found: 49 l:tc6 l:tf4 50 l:txc5 l:txg4
and ... ~d4!, or 49 l:te6 l:tf4 50 'it>b3 l:txg4 51
l:txe5 'it>f6 52 l:te6+ 'it>f7 53 l:tc6 l:td 4 54 l:txc5
g4 55 l:ta5 (55 l:tc7+ 'it>f6 56 l:tc8 'it>f7 57 'it>c3
g3!) 55 ... g3 56 l:ta1 g2 57 'it>b4 l:tg4 58 l:tg1
<it>e7.
I have to admit that, when 43 l:ta8 was
played, the thought occurred to me that my
opponent was merely delaying matters,
Analysis diagram and that a move later he would most
377
Kasparov vs. Karpov: 1986-1987
probably return to the main continuation. adhering to since the second match - not to
43 ...'it>f7 analyse with Karpov immediately after a
It was again simpler to play 43 ... g4! 44 f4 game. Feeling embarrassed about being late
exf4+ 45 gxf4 ~f7!, when 'the check on e7 for the start of play, I decided to explain the
has been transformed into a real threat, and reason to him and I began feverishly show-
the king is tied to the defence of the f-pawn, ing my opponent the main lines of our
while 46 ~a6 ~e7+ 47 'it>f2 'it>g6 48 ~xd6+ crazy analysis. To all appearances, Karpov
'it>f5 leads to a forced draw.' (Makarychev) and his trainers had not seen any of this ...
The ex-champion was clearly upset and the
following day he took a time-out.
The respite also did me no harm, but it
did nothing to improve my fighting spirit.
After this there followed a further five
draws. I was already counting the days to
the end of the match and was dreaming
only of a 12-12 score, which would enable
me to retain the title of champion. I made
every effort to avoid the slightest risk, and
my play with White became altogether
toothless.
44 'it>e4?!
Here I breathed a sigh of relief: it was Game 45
evident that the Karpov team had not G.Kasparov-A.Karpov
found the strongest plan - now White World Championship Match,
cannot revert to it. In the event of 44 ~a7+ 18th Game, Seville 30.11.1987
'it>g6 nothing is given by either 45 g4 ~f7 46 Queen's Gambit D58
~a8 ~f4 47 ~d8 ~xc4 48 ~xd6+ 'it>f7 49 ~e6
~d4 50 ~xe5 'it>f6 51 ~f5+ 'it>g6, or 45 ~e7 g4!
(as I would now have played - there is no 1 c4 e6
other way) 46 f4 exf4+ 47 gxf4 g3 48 ~e8 The English Opening (Game Nos.29, 31,
'it>f7 49 ~a8 g2 50 ~a1 ~g6 51 ~gl ~g3+. 33, 35, 43) is finally abandoned.
But by 44 ~d8! 'it>g7 45 ~d7+ it was pos- 2 ct:JC3 d5 3 d4 ~e7 4 ct:Jf3 (04)
sible to set Black the same choice as on the On this occasion, instead of 4 cxd5
43rd move: to where should he move his (Game Nos.7, 8, 39) I chose a thoroughly
king? analysed variation, in which after the first
44 ... 'it>g7 45 ~a7+ 'it>g6 46 ~e7 (or 46 ~d7 two Karpov-Kasparov matches (1984/85
g4!, while if 46 ~a6, then simply 46 ... 'it>h6) and 1985) there were hardly any 'blank
46 ... g4! spots' remaining. I needed somehow to
Draw agreed on Black's proposal (Yz-Yz): hold things up: I had not yet come to after
47 fxg4 'it>g5 48 ~g7+ ~g6 or 47 f4 exf4 48 the dramatic adjournment session of the
gxf4 g3 etc. Times: 3.00-2.48. 17th game.
4 ... ct:Jf6 5 ~g5 h6 6 ~h4 (6 ~xf6 - Game
At that moment I could not restrain my- Nos. 10, 12, 22) 6 ... 0-0 7 e3 b6 8 ..ie2 ..ib7
self, and I broke a rule which I had been (05) 9 ..ixf6 ..ixf6 10 cxd5 exd5 (02) 11 b4
378
The Fourth Match: 1987
379
Kasparov vs. Karpov: 1986-1987
(Margate 1935; Game No.2 in Volume IV of say, 23 "iVb7 MXc1 + 24 ct:Jxc1 "iYa3! 25 "iYxd7
My Great Predecessors), where Black realised "iYxc1 + 26 'it>h2 "iYc5.
too late that the strength of his passed c4- Even so, by 23 ct:Jf4! ct:Jb6 24 M! I could
pawn was largely an illusion. And al- have set my opponent an unpleasant
though objectively the position is still psychological problem. In the event of
bordering on equal, there is no need to play 24 ... MC4 25 h5 g5 26 ct:Je2 "iVb4 27 ct:Jfd4!
... c5-c4. White acquires the f5-point, while 24 ... h5!
19 ~xc6 (41) demands a certain courage, since it allows a
My long think was caused by a search tempting piece sacrifice - 25 Mxc8 "iYxc8 26
for some small advantage. After 19 a4 there ct:Jxg6!? fxg6 27 llYxg6+ ~g7 28 ct:Jg5, which,
is again 19 .. .'iVd6 (d. the note to White's however, gives no more than a draw:
18th move), but Black can also consider 28 ... "iYd7 29 'iYh7+! 'it>f8 30 "iYxh5 ~f6! (after
19 ... c4 20 "iYc2 ~b7 21 as with double-edged 30 ... ct:Jc4 31 "iYg6 'it>g8 32 h5 ct:Je5 33 "iVh7+
play. But that day both contestants endeav- 'it>f8 34 f4 ct:Jf7 35 ct:Jxf7 "iYxf7 36 h6 ~f6 37
oured to play very cautiously ... 'iVf5 ~h8 38 "iYg4 White has some
19 ... MXb1 (06) 20 "iYxb1 "iYxc6 21 dXc5 "iYxC5 advantage) 31 ct:Jh7+ 'it>g7 32 ct:Jxf6 'it>xf6 33
(07) "iYg5+ 'it>f7 34 h5 "iYe6 35 g4 ct:Jc4 36 ~f4+ ~f6
21...'~Jxc5 22 ct:Je2 "iYd6!? would also have 13.~.el)
equalised. Of course! 'Further simplification, fa-
22 ct:Je2 "iYf8 vourable for Black, is now inevitable'
'Thus we have a variation on a familiar (Taimanov). The tension is reduced to zero.
theme: White has the better pawn forma- 24 ct:Jxe5 (04) 24 ... ~xe5 (02) 25 Md1 (04)
tion, while Black has free piece play.' If 25 h5, then 25 ... Mxc1 +, for example: 26
(Taimanov). 'The inconvenience for Black is ct:Jxc1 "iVb8 27 ct:Jb3 'it>g7 28 hxg6 fxg6 29 g3
only temporary: the great amount of space "iYc7 or 26 "iYxc1 "iVb4 27 hxg6 fxg6 28 iVc8+
on the queenside always allows him to 'it>h7 29 "iYd7+ i.g7 with a draw.
counter any threats, maintaining material 25 ... "iYC5 (02) 26 h5 (04)
equality. However, the degree of technical
difficulty depends on White's plan.'
(Makarychev)
380
The Fourth Match: 1987
lems, despite his king being cut off. called the 'turncoat' or 'aping' variation.
32 Ma7 a4 (01) 33 g3 (03) Taimanov: 'History repeats itself! As in the
33 g4 g5 34 Wg2 a3 35 Wg3 Ma 1 36 Wf3 first match, the contestants decided to
Ma2 would not have changed anything. compete in the ability to play one and the
33 ... h5 (08) 34 Wg2 (02) 34... a3 (02) 35 e4 same critical position for both sides. And
(03) 35 ... g5! (07) 36 Wf3 (05) this with the finishing straight in sight! And
The greedy 36 Ma5?! g4 37 Mxh5? (37 so, following in the footsteps of the previ-
Ma7 draws) leads to disaster: after 37 ... MC2 ous game .. .'
38 1:I.a5 a2 the white king is shut in, and
Black's proceeds across to the queenside.
36 ... g4+ 37 We3 Ma1 38 Wf4 Mf1 39 Wg5 Game 46
(06) A.Karpov-G.Kasparov
Or 39 Mxa3 Mxf2+ 40 Wg5 Mf3! with a World Championship Match,
draw. 19th Game, Seville 02-03.12.1987
39 ...MXf2 40 Wxh5 Me2 Queen's Gambit 058
1 lLlf3 d5 (03)
I also decided to follow the motto of
maximum safety, by avoiding 1...lLlf6 2 c4
(Game No. 44).
2 d4 lLlf6 3 c4 e6 4 lLlc3 i..e7 5 i..g5 (05)
5... 0-0 6 e3 (02) 6 ... h6 7 i..h4 b6 (02) 8 i..e2
(01) 8 ... i..b7 (02) 9 i..xf6 i..xf6 10 cxd5 exd5
110-0
A small nuance: usually I played 11 b4
(Game No.45).
11 ...lLld7 (01)
Draw agreed (Yz-Yz). If 41 Wg6, then Other possibilities for Black are exam-
41...Wf8, while after 41 Wxg4 Mxe4+ 42 Wf3 ined in the notes to the 11 th game of the
Mb4 43 Mxa3 Wg7 White's extra pawn is Karpov-Spassky match (Game No.65 in
completely useless. Times: 2.14-2.29. Volume V of My Great Predecessors).
12 b4 (03)
Again I committed an inaccuracy and If 12 'i¥b3, then 12 ... c6 and ... Me8 is per-
failed to exploit my opportunity, albeit a fectly safe (d. Volume III of My Great Prede-
not very serious one - but in such encoun- cessors, p.326).
ters any trifle may unexpectedly produce a 12 •.. c5 13 bxc5 bxc5
result. Alas, throughout the second half of 'What, not again the tiresome 14 Mbl
the Seville match (with the exception of the i.c6 15 i..b5? No! On this occasion, as
last, 24th game) I overlooked similar oppor- Gufeld aptly put it, we have a turncoat
tunities ... with a surprise!' (Taimanov)
In the 19th game Karpov and I, having 14 iYb3!? (01)
exchanged places, again chose the Tarta- New for our matches, although also an
kower-Makagonov-Bondarevsky Variation, already known continuation. But I had not
which back in 1985 the commentators had studied it seriously, and therefore I was
381
Kasparov vs. Karpov: 1986-1987
forced to solve problems at the board. play and complete his development nor-
14... cxd4 (19) lSlDxd4 mally; White tries to prevent this and seize
IS exd4 is harmless: Is ... lDb6 16 l:tfel the initiative.
J::!.e8 17 i..bS l:txel+ 18 l:txel Wlic7 with equal- 17 ... l:tb8 (12) 18 as (08) 18 ... lDc4! (23)
ity (Lukacs-Lengyel, Budapest 1985). I felt that this was the correct move. 'A
bold decision. By sacrificing a pawn, Black
greatly activates his pieces' (Taimanov).
The passive 18 ... lDd7(a8) 19 Wlia3 'would
have let to an unpleasant position, which
White could have played in accordance
with his tastes' (Makarychev) and without
any risk.
19 i..xc4 (20) 19 ... dxc4 20 Wlixc4
The section of the game between the
17th and 21st moves cost the two players
nearly one half of their allotted time. The
last three half-moves came to them with
particular difficulty - important decisions
lS ... i..xd4! (03) had to be taken in a situation that had
Probably the simplest solution, although become unexpectedly complicated after the
after Is ... lDcs 16 Wlib4 Wlib6 (Lukacs- pawn sacrifice. I thought over my 20th
Staniszewski, Naleczow 1985) or IS ... l:tb8 move for nearly half an hour, sensing that
16 ~abl (16 Mfdl WliaS, while if 16 lDxdS?! the position contained certain dynamic
there is the tactical resource 16 ... i..xd4 17 features which promised Black sufficient
exd4 WligS 18 i..f3 i..a6 winning the ex- compensation for the pawn ...
change) 16 ... i..xd4 17 exd4 Wlif6 18 Wlibs lDb6
the situation is roughly balanced. The
existence of several acceptable replies
testifies to the solidity of Black's position.
16 exd4lDb6
382
The Fourth Match: 1987
idea of .. ..l:'c5 and ... ~xd5) with equality, as The length of time spent indicates that
occurred in a game that was unknown to the opponent's move came as a surprise to
me, Farago-Geller (Baden-Baden 19S5); me.
2) 21l:.fbll:.fdS 22 'iVc5 'iVg6 23 g3l:.bcS!, 22 dxcs l:.bc8 23 a6! (OS)
suddenly creating unpleasant pressure: 24 'Of course, after 23 ttJa4 ~c6 24 l:.fc1
'iVxa7 (24 'iVe5 ~aS!) 24 ... ~f3 25 'iVb6 'iVf5 26 l:.fdS it is difficult for White to strengthen
ttJdl ~aS, and Black's attack more than his position' (Taimanov). Therefore 23 a6 is
compensates for the material deficit; the best chance: the pawn advances still
3) 21 ~fel ~xg2! (I hadn't seen this) 22 further. However, even here Black should
'it>xg2 'iVg6+! (it is important to interpose most probably be able to gain a draw.
this check: the queen takes control of the 23 ... ~a8 (01)
d3-square) 23 'it>f1l:.fcS, regaining the piece
with a draw: 24 ttJd5 l:.xc4 25 ttJe7+ 'it>h7 26
ttJxg6 'it>xg6;
4) 21 'iVc5 ~fcS (21...l:.bcS 22 'iVxa7 ~xg2!
23 'it>xg2 l:.xc3 24 'iVb7 l:.c6! is also not bad)
22 ttJd5 ~xd5 23 'iVxd5 l:.dS 24 'iVe5 l:.b4! or
22 'iVxa7 'iVc6 23 d5 'iVxc3 24 l:.abl 'iVe5 25
l:.xb7 l:.aS 26 'iVb6 (26 'iVe3 'iVxd5) 26 ... 'iVxd5
27 a6l:.c6 with a draw.
Thus 20 ... 'iVf6 would have led to simpli-
fication and a quick draw. I thought that
20 ... 'iVd6 was no worse, but Karpov found a
flaw in this move.
21 'iVcs! (22) 24 ttJbs?! (09)
Forcing the exchange of queens. 'In this The most tempting move, but not the
way Karpov deprives the opponent of any best. 'After driving the bishop into the
tactical resources and "pins him to the corner square, designated for the rook,
ropes". Now the only question is whether White embarks on the most committal part
or not White's advantage is sufficient for a of his plan, which involves the temporary
win.' (Taimanov) imprisonment of his own knight.'
(Makarychev)
'24 l:.a5 was passive' (Taimanov). But in
this way White would have retained more
chances of converting his extra pawn! Black
would have had to reply 24 ... l:.fdS!
(24 ... l:.feS 25 l:.dl is less good) with the
hope of holding out thanks to the activity of
his pieces and the vulnerability of the white
pawns: 25 l:.dl?! l:.xdl+ 26 ttJxdl l:.bS 27
l:.a1 l:.b5 2S l:.c1 ~c6 and ... l:.a5 with equal-
ity, or 25 f3! l:.d3 26 l:.c1 ~c6 27 ct:Je4 (but
not 27 ct:Jb5?! ~xb5 2S l:.xb5 l:.a3 and ... l:.xa6
with equality) 27 ... l:.bS 2S l:.a2l:.bb3.
21 ... 'iVxcs (15) 24 ... ~XCs (05) 25 ct:Jxa7 (01)
383
Kasparov vs. Karpov: 1986-1987
384
The Fourth Match: 1987
other I cut off the enemy king and easily 'it'e6 39 ~fa5 ~b6 40 ~e5+ (02) 40 ... cJ;;f6
gained a draw. Here the game was adjourned.
41 ~ea5 (07)
On this occasion Karpov immediately
sealed a move, pinning his hopes on horne
analysis. But it transpired that none of the
breakthroughs bring White success.
41 ... cJ;;e6 42 ~a1 (03) 42 ...~c6 43 ~e5+ (02)
43 ... cJ;;f6 44 ~f5+ cJ;;e6 (01) 45 ~e5+ (02)
45 ...cJ;;f6 46 ~ea5 cJ;;e6 47lha2 ~b6
Analysis diagram
385
Kasparov vs. Karpov: 1986-1987
comes of it: 53 ... l:te6 54 We3 .l:!.d6 55 .l:!.a3 Mc6 51l:txg5 Wf7 52l:tga5 We6, and in both cases
56 Wd4l:td7+! 57 We5l:te6+ 58 Wf4l:ta7 with Black has an impregnable fortress.
a draw. 49 ... l:tc6 50 l:tb2!? (01)
White gives up his a6-pawn, in order to
break through on to the 7th rank with his
rook and capture the g7-pawn, and then
also approach the h6-pawn. But Black is
fully prepared for this plan.
50 ... l:tcxa6 51 l:tb6+ (02) 51 ... l:txb6 52 l:txa7
Mb1 53 l:txg7 (threatening .l:!.g6 and .l:!.xh6)
53 ... .l:!.f1+ 54 We3 .l:!.e1+ 55 Wf3 .l:!.f1+ 56 We2
.l:!.f4 57 We3
48 ...f6!
The correct drawing path, although
Black would also not have lost after 48 ... Wf6
49 h5 l:tc6 50 l:tf5+ We6 51 l:te5+ Wf6 52 l:te8
(52 l:tea5 We6), and here not 52 ... l:tc4? 53
g5+! (53 l:tc2?? g5+) 53 ... hxg5 54 Wg4 g6 55
h6 and wins, but 52 ... g5+ (Karpov).
It would appear that my opponent pro-
voked me into playing .. .f7-f6, assuming
that now, by breaking through on the 57 ... We5!!
kingside, White would obtain a rook end- A startling move - 'A wonderful, purely
game with winning chances. study-like idea!' (Taimanov). This genu-
49 h5 (01) inely rare method of defence was pointed
out in analysis by Dolmatov.
58l:te7+ (14)
Karpov's long think indicates that my
king move carne as a revelation to him:
usually - such is chess psychology! - in
analysis you do not look at going into a
pawn endgame a pawn down, considering
it to be hopeless.
58 ... Wd6 59 .l:!.e6+ (or 59 .l:!.h7 .l:!.xg4 60 .l:!.xh6
We5 with a draw) 59 ... Wxe6 60 Wxf4 We7
Draw agreed (Yz-Yz): 61 e5 We6 or 61
Wf5 Wf7. Times: 3.16-2.31.
After 49 l:t2a4 there is now 49 ... g5+! 50 A difficult game, memorable for its in-
hxg5 hxg5+ 51 We3 l:tc6, while if 49 g5 teresting endgame. But in the opening, after
(Taimanov), then 49 ... hxg5+ 50 hxg5 fxg5+ correctly sacrificing a pawn, I again failed
386
The Fourth Match: 1987
to make a last accurate move (20 ...'iVf6!), (as I played in the 19th game) allows the
and ended up having to spend a mass of reply 6 'iVc2, giving White additional re-
energy, merely in order to beat off the sources, so why not begin with 5 'iVc2, in
opponent's pressure. order after 5 ... 0-0 6 ~g5 to transpose into
Nevertheless, I was heartened by my this variation? However, now Black's
successful defence. In the 20th game, played resources are also expanded.
the next day, I decided to avoid the thor- 5 ... 0-0 (04)
oughly analysed Queen's Gambit set-ups The natural move. Black has also tried
and to set Karpov a new opening problem. 5 ... b6 (back in the 19th century!), 5 ... c6,
5 ... ttJbd7, 5 ... dxc4, 5 ... ttJc6 or 5 ... ttJa6 (Topa-
lov-Karpov, Dos Hermanas 1999; this
Game 47 reaction to 'iVc2 became typical after the 4th
G.Kasparov-A.Karpov game of the 1985 match), but the most
World Championship Match, energetic reply is considered to be the
20th Game, Seville 04.12.1987 gambit 5 ... c5!? 6 dxc5 d4! (the developing
Queen's Gambit D53 ... ttJa6(c6) does not give Black full equality)
7 ttJa4 b6! with very complicated play
(Krasenkow-P.Nielsen, Copenhagen 2003)
1 c4 e6 2 ttJC3 d5 3 d4 ~e7 4 ttJf3 ttJf6 5 'iVc2 or 7 ttJb5 e5! (the source game: Dreev-
Vaganian, Budapest 1996) 8 ttJxe5 a6 9 ttJa3
O-O! 10 g3 'iVa5+ 11 ~d2 'iVxc5 12 ttJd3 'iVc6
13 IIg1 ~f5 with equality (Khalifman-
Topalov, Dubai (rapid) 2002).
However, in 1987 the theory of the set-
up with 5 'iVc2 was only in its infancy, and
the sound reply 5 ... 0-0 was undoubtedly
the correct choice.
6~g5
Of course, not 6 e4?! dxe4 7 ttJxe4 ttJc6!
(Steinitz-Janowski, Nuremberg 1896). The
modern treatment is far more interesting -
6 cxd5 ttJxd5 (6 ... exd5 7 ~g5 leads to a
Instead of the usual 5 ~g5 (Game Nos. 10, Carlsbad structure) 7 g3 (Eingom-
12, 22, 45, 46) or 5 ~f4 (Game No.49), White Vaganian, Sochi 1986) or 7 ~d2 (Razuvaev-
chooses a rare continuation, aimed at Timoshchenko, Irkutsk 1986).
confusing the opponent and 'giving greater 6 ... c5 (05)
opportunities for imagination at the board.' A natural move, understandable to Kar-
(Taimanov). 'By beginning with this move, pov, and in addition one which' allows the
White reduces the effectiveness of ... h7-h6 opponent the narrowest choice' (Makary-
in reply to ~c1-g5.' (Makarychev) chev). 6 ... dxc4 7 e4 c5 (7 ... ttJc6!? Torre-
Today 5 'iVc2 has long been a regular Kurajica, Novi Sad 1984) would transpose.
guest in tournaments of any standard. The 6 ... b6 did not become popular: 7 ~xf6 ~xf6
idea of the move is simple: if 5 ~g5, then 8 e4, and if 8 ... ttJc6, then 9 0-0-0 (Petrosian-
5 ... h6 is good, determining the fate of the Hubner, 6th match game, Seville 1971).
bishop (6 ~h4 or 6 ~xf6), whereas 5 ... 0-0 After the standard 6 ... h6 7 ~xf6 ~xf6 it
387
Kasparov vs. Karpov: 1986-1987
388
The Fourth Match: 1987
14 a3 (11)
Alas, this is effectively a waste of time.
On the whole, the advance of the queenside
pawns does not give anything: Black ex-
ploits the weakening of the light squares to
create counterplay.
'14 .l:!,fdl, .l:!,d6 and .l:!,adl suggested itself.
It would not have been easy for Black to
defend, although his position would have
12 ... LDc6 (13) remained solid and, as they say, tolerable.'
Now 12 ... ii.xgS?! 13 LDxgS LDc6 would (Makarychev). Indeed, the natural 14
have run into 14 LDxh7! .l:!,d8 (14 ... ~xh7 IS .l:!,fdl!? would have consolidated White's
ii.d3+) IS LDf6+! gxf6 16 'iYg3+ ~f8 17 exf6 gains, for example: 14 ... Ad7 IS .l:!,d6 .l:!,fc8 16
with a very strong attack. In Informator .l:!,adl ii.e8 17 'iYe3 LDaS 18 ii.d3, beginning
Dorfman gave 12 ... b6 13 .l:!,ac1 ii.b7 with an an attack on the king, or 14 ... .l:!,d8 IS .l:!,xd8+
'=' assessment, but, in my view, after 13 LDxd8 16 LDd2 ii.d7 17LDe4 .l:!,c8 18 LDd6 .l:!,c7
389
Kasparov vs. Karpov: 1986-1987
19 kbl (04)
The variation 19 Mfel kc6 20 ct:Jd4 (20
ct:JgS? Mxd3) did not appeal to me - the
\.<..\\\'E,\\\ \\c.s ~\\Q.~Q. u.p '0\\ \\\~ \l'J1:'0\\'E, squ.a1:~.
16 ... a6 (16) And I decided to take urgent measures.
A defence against the cramping b4-bS. 19 ... .i.c6 (03)
All the commentators approved of Gufeld's There was no point in playing 19 ... kbS
recommendation 16 ... bS!? with the idea of 20 Mfel. It is useful for Black to keep the e4-
17 .i.xbS ct:Jxb4 18 .i.c4 (18 .i.xd7 ct:JdS) square under control.
18 ... ct:Jc6 or 17 .i.d3 as! 'with easy equality', 20 ct:Jg5 (02) 20 ... h6 (03)
although after 18 'iVel! Black would have
had to sacrifice a piece - 18 ... axb4 (18 ... .i.e8
19 'iVe4) 19 'iVe4 g6 20 jLxbS ct:JxeS 21 'iVxeS
jLxbS 22 'iVxbS bxa3, and although the
passed a-pawn is very strong, nevertheless
White has a knight for two pawns. And if
no concrete play is apparent, Karpov re-
frains from such tactical excesses.
17 'iVe3 (04) 17 ... .i.e8 (OS) 18 .i.d3 (06)
'Nothing was given by 18 'iVb6 on ac-
count of 18 .. .l:tac8 and ...'iVc7.' (Makary-
chev)
18 ... ct:Ja7! (08)
390
The Fourth Match: 1987
391
Kasparov vs. Karpov: 1986-1987
392
The Fourth Match: 1987
'i'xg4 i¥b3 would have been parried by 38 instead of the 'mild' 1 etJf3 - Game Nos.44,
'i'f4+ 'it'g8 39 'i'c1 'it'h7 40 'i'c6 with a draw. 46) 1 ... etJf6 2 c4 g6 3 etJc3 d5
37 'i'd 7+ 'it'f8 Of course, I realised that at such an im-
portant moment it was risky to return to
the Griinfeld Defence, but this was also a
matter of principle. Despite all the fears, the
character of the play in the main lines of
this opening was more in keeping with my
style than that of Karpov.
4 etJf3 ii.g7 5 'i'b3 (as in the 15th game)
5... dxc4 6 'i'xC4 0-0 7 e4 etJa6 (02) 8 ii.e2
(01) 8 ... c5 9 d5 e6 100-0 exd5 11 exd5 ii.f5
12 ~d1 (12 ii.f4 - Game No.19) 12 ...~e8 (02)
As has already been mentioned,
12 ... i¥b6! is stronger.
13 d6! (01) 13 ... h6 (01)
Draw agreed (Yz-Yz). Times: 2.26-2.28.
393
Kasparov vs. Karpov: 1986-1987
394
The Fourth Match: 1987
An amazing oversight after a long think essential, with drawing chances) 23 1IVxb4
- in the opinion of many experts, caused by a5 24 ~xb7l;'i.xd2 25 d7 .i:txb2 2611Vd5l;'i.b5 27
the fact that the d3-square is not controlled .l::td1 ~f8 28 ~d6 ~xd6 29 1IVxd6 .i:tab8
by a single black piece. But in my view, also (earlier I had been planning 29 ...1IVf8, but I
by the fact that Karpov did not find any had missed the deadly 30 4Je4! followed by
advantage after other continuations and 30 ... 1IVxd6 31 4Jxd6 .l::tbb8 32 .i:tc1 .i:td8 33 .i:tc8
was perplexed by the dynamic nature of 'it'g7 34 .l:txa8 J:txa8 35 4Jc8) 30 h3, and
the resulting positions: White won.
1) 19 ~g3 4Jd3(?!) 20 4Jb5 c4 21 1IVa3. 19 ... 4Jxa4! 20 'iVxa4 4Jc6 is much stronger.
This variation with an 'unclear' assessment
was given in Infarmatar by Karpov and
Zaitsev (21...4Jd5 is not bad), but a queen
sacrifice is more dangerous for Black - 20
1IVxe6!? fxe6 21 l:hd3 4Jc4 22 b3 4Jb2 23 .l::te3
~d4 24 4Je4 with powerful compensation.
In our preparations for the 23rd game we
also looked at 19 ... a5!?, but in the end we
preferred the safe 19 ... 1IVd7 20 a3 4Jc6 21
'iVb5 .l::tc8! - in the game Rao-Kasparov from
a simul' with the USA junior team (New
York 1988) after 22 l;'i.ad1 (22 1IVxc5 4Je5!)
22 ... ~xc3! 23 bxc3 4Je5 24 1IVxd7 4Jxf3+ 25
gxf3 4Jxd7 White immediately found Analysis diagram
himself in an inferior endgame, but also
after 24 .l::tb2! (trying to avoid a blockade) If 21 .i:tc1 g5 22 ~g3 there is the reply
24 ... 4Jxf3+ 25 gxf3 .l::td8 26 1IVxd7 .i:txd7 27 22 ...1IVd7! with unclear play - perhaps
.l::tdb1 g5 Black has no reason for complaint; White's chances may be slightly better (for
2) 19 4Jb5 .i:te4! 20 ~e3 4Jc4 21 ~xc5 example, 23 h4 g4 24 4Jh2 ~d4 25 b4 etc.),
4Jxd2 22 4Jxd2 .l::tg4!? (instead of 22 ... l;'i.e2, but the position is very dynamic and its
the move that suggests itself) 23 ~xb4 1IVg5 assessment is not straightforward. Piket
24 4Jc7 .l::txg2+ 25 'it'h1 .l::txf2 with an unex- also recommended 21 d7 (in analysis I
pectedly dangerous counterattack, or 20 reached an '=' assessment in this line), as
~g3 1IVd7 21 4Jc7 .l::td8 with unclear play - if well as 21 'iVb5!? - this has occurred in
22 a3 4Jc6 23 1IVc2 there is the good reply practice and it leaves White with an insig-
23 .. .£5, when the knight at c7 is out of play; nificant plus.
3) 19 4Ja4!. This simplifying move, rec- Although after 19 4Ja4 Black's position
ommended by Tal, is undoubtedly the best: looks dangerous, an extensive analysis,
as the minor pieces are exchanged, the including with a computer, shows that it
tactical complexity of the position dimin- can nevertheless be held. But why suffer
ishes and it is more difficult for Black to such torment? After the game with Piket, I
block the d6-pawn, the influence of which began seeking an improvement for Black
increases. This is what Piket played against somewhere earlier - and I found one
me (Amsterdam 1995), and I responded (12 ... 'iVb6!).
badly: 19 ... l;'i.e4? 20 ~g3 4Jc4 21 4Jxc5 4Jxd2 19 ...4Jd3! (02)
22 4Jxd2 .l::te2? (22 ... .l::td4 23 4Jf3 .i:txd6! was 'Now Black seizes the initiative: if 20
395
Kasparov vs. Karpov: 1986-1987
l:i:xd3? there follows a "fork" - 20 ... c4.' (Tai- d3, and in search of compensation he
manov) prepares to nibble at Black's defences by
20 ..Iig3 (04) 20 ..• c4 (03) 21 "iVc2 (02) h4-h5.
23 ... fS (16)
I was unhappy with this move, and
some commentators criticised it for the
'premature weakening of the important a2-
g8 diagonal'. Nevertheless Black retains the
advantage! Whereas 23 ... l:i:c5, which many
recommended, would have led after 24
"iVb1 with the idea of tiJe1 (Karpov, Zaitsev)
or 24 l:i:e2 to complicated, double-edged
play.
But 23 ... l:i:ee8!?, a far from obvious move
not suggested by anyone, was also interest-
ing - Black maintains the tension and
21 ... l:i:c8?! (14) threatens a possible .. :iVg4. For example: 24
A plausible idea - to prevent b2-b3 and l:i:b 1 f5 25 tiJe2 l:i:e4 26 b3 tiJc5 or 24 l:i:e2
maintain the knight on d3. But no one l:i:xe2 25 "iVxe2 "iVe6, and the endgame after
noticed that 21..."iVd7! was better, since if 22 26 "iVxe6 fxe6 27 l:i:d2 tiJxb2 28 l:i:xb2 .Jtxc3 29
b3 there is the unpleasant 22 ... tiJc5! 23 bxc4 l:i:c2 ~g7 is worse for White, while if 26
tiJxc4 24 l:i:dd1 (24 l:i:d5 l:i:c8) 24 ... tiJxd6, and "iVd2, then 26 ... ctJd7! 27 tiJe1 tiJ7c5, retaining
otherwise 22 ... l:i:ae8! is strong, with the the outpost at d3.
threats of ... tiJe1 and .. .f7-f5-f4, for example:
22 a4 l:i:ae8 23 h3 "iVc6!, and Black domi-
nates, or 22 l:i:ad1 l:i:ae8 23 h3 f5, and if 24
l:i:xd3 cxd3 25 "iVxd3, then 25 ... g5 26 tiJb5 f4
27 .Jth2 l:i:d8 with an obvious advantage.
22l:i:ad1 (11) 22 ..."iVd7 (04)
24l:i:xd3 (16)
Necessary: a more convenient opportu-
nity may not present itself.
24 ... cxd3 2S "iVxd3 (02) 2S ... tiJC4 (19)
With the obvious threat of ... tiJxb2. I as-
sumed that Black was quite alright here,
23 h4! (04) but my opponent's reply gave me a fright...
Karpov correctly realises that he cannot 26 "iVdS! (05) 26 ...tiJb6
get by without sacrificing the exchange on A safe move. 'It was very dangerous to
396
The Fourth Match: 1987
play 26 ... tiJxb2 27 I:te1 I:tce8 28 I:txe6 I:txe6 and tiJc5) 31 tiJe7 ~xe7 (now forced) 32
(28 ... 'iVxe6? 29 d7!) 29 tiJb5 <;1;>h7 30 tiJe5 dxe7 'iVxe7 33 h5, and White has at least an
.itxe5 31 .itxe5 tiJc4 32 f4! (the only move!) equal game.
32 ... tiJb6 (32 ... a6? 33 tiJd4) 33 'iVc5.' 28 'iVds (03)
(Makarychev). True, after 33 ...tiJc8 Black's
chances are slightly better, and therefore 33
'iVd3! is more accurate, with sufficient
compensation for the exchange. After
32 ... tiJe3 33 'iVd3 tiJg4 34 'iVd5 tiJf6 (34 ... I:te8
35 tiJc7 ~b8 36 g3 is no better) 35 'iVc5 tiJe8
36 'iVd5 a6 37 tiJd4 I:txd6 38 ~xd6 tiJxd6 39
h5 things also head for a draw.
In addition, the immediate 30 h5!? was
quite good, 'when the black king is in
danger' (Taimanov), for example: 30 ... tiJa4
31 tiJh4 tiJb6 32 hxg6+ I:txg6 33 'iVd3 I:tg5 34
tiJxa7 'iVe6 35 f4 I:th5 36 'iVd1 with unclear,
dynamic play (the main role is played by 28 ... tiJb6
the d6-pawn). Black's king is exposed, and A tacit peace invitation (whereas
with the queens on the board his exchange 28 ... <;1;>h8! would still have set White diffi-
advantage does not tell. cult problems). After 12 minutes' thought,
The commentators also condemned without making the move 29 'iVd3, Karpov
26 ... <;1;>h8(!), although this cool-headed move agreed to a draw (Yz-Yz). Times: 2.19-2.04.
would have led after 27 tiJb5 tiJb6!
(27 ... tiJxb2? 28 I:tb1!) 28 'iVb3 I:tc5 to a com- To the question 'why in a better position
plicated position with somewhat the better did the world champion force a draw?',
chances for Black: 29 tiJfd4 I:te4 30 f3 .itxd4+ asked in the traditional TV interview after
31 tiJxd4 I:te8 32 i.f4 J::td5 or 29 tiJxa7 tiJc4 30 the game, I replied: 'As soon as the danger
'iVb4 b6 31 b3 tiJxd6 32 'iVxb6 tiJe8 33 I:txd7 came to an end, I stopped fighting - a
I:txb6 34 .itf4 l:!.e6 35 b4 l:!.c4 36 g3 tiJf6 37 draw, probably a draw ... ' Yes, all my
l:!.d8+ <;1;>h7 38 tiJb5 l:!.e2, continuing to play thoughts had already turned to how to
for a win without any particular risk. hold out in the last 'Black' game - the 23rd.
27 'iVd3 (11) For a time, between the 18th and 21st
Karpov sensibly judged that 27 'iVb3 was games, the course of the match was favour-
pointless in view of 27 ... 'iVf7! (in the heat of able for me and Karpov endured several
the moment I was intending 27 ... <;1;>h7, but I anxious moments, but my inaccuracies
could well have changed my mind on allowed him to escape the worst ... Before
account of 28 tiJd5 tiJxd5 29 'iVxd5 i.xb2 30 the 22nd game I took my third and last
h5), when after 28 tiJd5! everyone sug- time-out. If at that moment either of us had
gested 28 ... l:!.d8, but I prefer 28 ... l:!.c5!, when been concerned about the magic of num-
only White has problems. bers, it was more probably Karpov: he
27 ... tiJC4 could have remembered the 22nd game of
Nothing was given by 27 ... l:!.c5 28 tiJd4 the 1986 match. But I was in a much differ-
I:te8 in view of 29 tiJb3! l:!.cc8 30 tiJd5 <;1;>h7 ent state to that of a year earlier.
(30 ... tiJxd5 is worse in view of 31 'iVxd5+ Taimanov: 'The time-out claimed by the
397
Kasparov vs. Karpov: 1986-1987
10 ... d4!?
In my game with Korchnoi I transposed
by 10 ... ~e6 11 0-0 into a well-known tabiya
of the variation 8 cxd5 exd5 9 ~e2 ~xc5 10
0-0 ~e6 11 Uc1 and after 11...Uc8 (if
11 ... ~b6, then 12 'iVa4! ~d7 13 Ufd1 ctJd4 14
A rare move: White avoids the variation 'iVb4 is good, Karpov-Beliavsky, Moscow
8 cxd5 ctJxd5!? 9 ctJxd5 exd5 (for example, 1981) 12 ctJb5 - it is no better to play 12 ctJe5
10 a3 ctJc6 11 ~d3 ~b6 12 0-0 ~g4 13 h3 ~d6 (Tal-Keres, 26th USSR Championship,
~h5 14 ~e2 ~xf3 15 ~xf3 d4 with equality, Tbilisi 1959) or 12 a3 h6 (Fischer-Spassky,
398
The Fourth Match: 1987
14th match game, Reykjavik 1972) - 16 i..d2 i..xd2+ 17 'it>xd2 l:tb8 18 l:thdl ~d8+
12 .. .'~Je4 13 ct:ld2 ct:lxd2 (13 ... i..e7!? Lilien- 19 'it>e3 ~xdl 20 l:txdl i..d7 21 i..f3 g5, and
thal-Bondarevsky, Moscow 1945) 14 'iVxd2 16 'it>f1 i..xc7 17 i..xc7 i..e6 18 i..f3!? ct:ld5! 19
i..b4 I soon gained a draw, although I i..a5 l:tac8, forcing 20 'it>e2 ct:lf4+ 21 'it>e3
experienced some problems. ct:ld5+, or 16 b4 i..xc7 (but not 16 ... i..xb4+?
On this occasion I was intending to im- 17 'it>f1l:tb8 18 ct:ld5) 17 i..xc7l:te8 18 f3 ct:ld5!
prove White's play, but the main reckoning 19 a3 (19 b5 ct:lb4) 19 ... l:te3! etc.
was that Karpov did not like positions with 15 a3 was also possible, with the idea of
an isolated pawn. We didn't particularly 15 ... i..e6 16 ct:lc7, although after 16 ... l:tad8 17
look at 10 ... d4, by which Black immediately ct:lxe6 fxe6 18 0-0 ct:ld5 19 i..g3 ct:lf4 Black
gets rid of it: during the game with holds the position. 15 ... i..d7 16 ct:ld6 i..c6
Korchnoi I had the feeling that this was to was also not bad.
White's advantage. Alas, such 'feelings at ls ... i..e6 16 a3 (after 16 b3 ct:ld5 17 i..g3 it is
the board' often hinder later objective not worth Black sacrificing the exchange -
analysis after a game. 17 ... ct:lb4?! 18 i..d6 ct:lxa2 19 l:tal etc.;
11 exd4 ct:lxd4 12 ct:lxd4 'iVxd4 13 'iVxd4 17... l:tfd8 18 i..f3 ~d7 19l:tfdl ~ad8 is more
i..xd4 14 ct:lbS! i..b6 solid, with approximate equality) 16 .. J:tfd8
Of course, not 14 ... i..xb2? 15 l:tc2 i..al(?)
160-0 i..f5 17 l:tc5 and wins (from my notes
to the game with Korchnoi in Infarmatar
No.42) or 15 ... ct:ld5 16 i..d6 (Makarychev)
16 ... i..f6 17 i..xf8 'it>xf8 18 0-0, and Black
does not have sufficient compensation for
the exchange.
17 ct:ld6?!
A completely insipid move - yet an-
other 'Seville error'! After the cool-headed
17 i..f3! Black would have had to choose
from various inferior endings, none of them
very pleasant for Karpov. The position
would still have been not altogether
150-0 straightforward:
15 ct:lc7 was condemned by the commen- 1) 17 ... i..d5 - then 18 i..c7! is unpleasant,
tators because of 15 ... i..xc7 16 i..xc7 l:te8, and after 18 ... i..xf3 19 i..xd8 i..e2 20 i..xb6
but after 17 f3 i..e6 18 a3 ct:ld5 19 i..g3! i..xb5 21 i..d4 i..xfl 22 'it>xfl Black stands
(19 ... f5 20 i..e5) White has an excellent worse;
ending - two bishops and the more central- 2) 17... ct:ld5 18 i..g3l:td7 19l:tfdl ~ad8 20
ised king. Therefore Black would have had ct:ld4 ct:lf4 21 ct:lxe6l:txdl+ 22l:txdl .l:txdl+ 23
to find 15 ... i..a5+!, with equality after both i..xdl ct:lxe6 24 i..f3 i..d4 25 b3, and al-
399
Kasparov vs. Karpov: 1986-1987
though the position is almost drawn, the this prevented them from demonstrating the
two bishops still allow White some hopes; best aspects of their play.'
3) 17 ... J:!.d7 (evidently the best move) 18 Lev Psakhis: 'Last year's match was, in
Ci:Jc7 ..txc7 19 J:!.xc7 J:!.xc7 20 ..txc7 ..td5, but every respect, one of the most outstanding
here too after 21 ..te2 or 21 ..tdl (in order to contests for the world title, but this one will not
retain the two bishops: 21...J:!.e8 22 ..ta5) leave any significant mark in chess. Of course,
White would have remained the stronger one could pick out the 4th game, which the
side. world champion won confidently, and the 7th,
In short, after 17 ..tf3 there would still full of sharp drama. But the number of mistakes
have been lot of fight in the position - if I and oversights exceeds what we have seen in
had possessed the desire to fight! previous matches.'
17 ... J:!.d7 18 ..tbS J:!.e7 19 J:!.fel Yuri Balashov: 'It is obvious that, after
Draw agreed on White's proposal spending so long together at the board, it is
(Yz-Yz). Times: 1.15-1.25. becoming increasingly difficult for Karpov and
Kasparov to play against each other. Knowing
'Goalless draw' - this was how Bron- so much about the opponent makes it increas-
stein headed his newspaper report about ingly difficult to discover a chink in his armour.
this depressing game. And yet Karpov And then, having eliminated their shortcom-
must have been afraid of how it would turn ings, they suddenly stumble on easy ground ...
out! When after it we went off to the tradi- The psychological burden has increased
tional television interview, he looked sharply.'
extremely tired. I too was in a terrible state, Naum Rashkovsky: 'This match shows
and yet I should have tried to set my oppo- clearly that many of FIDE's decisions in recent
nent at least some problems: our matches years have proved to be unfruitful. This lengthy
had demonstrated that under pressure rivalry between the world's two strongest
Karpov would begin to feel uncertain. I was players is adversely affecting the quality of their
wrong not to play for a win - but I no play. In the present encounter it is psychologi-
longer had any nervous energy. cal factors that have particularly come to the
This short and colourless draw conclu- fore.'
sively showed that, irrespective of the There is indeed no doubt that the qual-
outcome of the match, neither player was at ity of the games in the Seville match was
his best. True, Karpov thought that he significantly inferior to all the previous
played well in Seville, that he had his matches. For this we have to thank those
creative achievements, and in general 'in who thought up this system and forced us
the creative sense the event was at a high to play annually. Even such a momentous
level'. But I thought that the creative con- event in a person's life as a world champi-
tent of the match was clearly inferior to that onship match can become humdrum if you
of the previous ones. This was not only my have to do it every year...
opinion, but also the opinion of many
grandmasters. A fantastic finish
Vitaly Tseshkovsky: 'In the competitive The expected full-blooded battle in Seville
sense, it was an uncompromising fight which may not have happened, but the conclusion
excited the fans. But the creative standard of the of the match left no one indifferent. It was
games was not so impressive. Both players were an incredible, simply fantastic finish!
under some kind of psychological pressure, and Sergey Makarychev: 'The intensity of the
400
fight in the two unprecedently dramatic games
at the finish exceeded anything observed in all
the preceding matches for the world champion- Came 50
ship. We witnessed a two-act drama, in which A.Karpov-G.Kasparov
the contestants had in turn to act in one and the World Championship Match,
same manner, by playing the next day the role 23rd Game, Seville 16/17.12.1987
performed by the other. ' English Opening A34
Thus, with the score standing at 11-11
the fate of the chess crown was to be de-
cided by the last two games. Of course, a 1 c4 (06)
drawn match was not the convincing As far as I recall, Karpov was not late
victory with which I would have wanted to for the game, but did indeed ponder over
conclude our marathon. But I did not have his first move 'a la Bronstein'. And he
the emotional strength to fight for a win, wisely decided that at this critical moment
while from Karpov's play it was not appar- of the match he should not go in for sharp
ent how he could change the character of and 'alien' positions from the Griinfeld
the fight, and two draws at the finish Defence, and that the variation with g2-g3
seemed inevitable. Later it transpired that was too passive. After 1 ct:Jf3 there could
this was also what the members of my have followed 1...dS (Game No.46), whereas
analytical team thought. Azmaiparashvili if 1 c4 I would not have wanted to play the
and Dorfman bet between them on the 'Karpov-style' 1...e6 2 ct:Jc3 dS 3 d4 iLe7
outcome of these two games, and after any (Game Nos.39, 45, 47), but I could have gone
result other than two draws Dorfman stood 1...eS, since before the match I had studied
to make a substantial profit. It would have this line for White. Apparently, these were
been immeasurably easier for me if the reasons for my opponent's hesitation.
Dorfman had lost his bet, but, as it turned 1 ... c5
out, the drawing limit in this match had I replied within 30 seconds, having de-
already been exhausted. cided to play a variation which I under-
Before the 23rd game Karpov took his stood and which we had prepared for
last time-out. I have to admit that during Black.
those days I was gripped by a kind of 2 ct:Jf3 ct:Jf6 3 ct:Jc3 d5
superstitious horror. I suddenly realised 'Kasparov avoids a pure English set-up
that my rejection of a fight in the 22nd and invites a transposition into the Griin-
game was a psychological mistake: I had feld Defence. This move order leads to a
demonstrated to my opponent that I was structure that was completely new for the
dreaming only of a draw, and hence he games between us.' (Karpov)
could act more forcibly and determinedly 4 cxd5 ct:Jxd5 5 d4 (01)
against me. This was what we expected, although
Karpov: 'In the last game of this match we also looked at S e4 ct:Jb4 or S... ct:Jxc3 6
where I had White, I had to make every effort to bxc3 g6, as well as S e3 ct:Jxc3 6 bxc3 g6 7
try and win. And I managed to do this, al- iLbS+ iLd7, and if 8 ~3 (the main line is 8
though only after numerous adventures. The a4 iLg7 Bareev-Kasparov, Cannes (rapid)
game proved to be one of the most tense and 2001), then 8 ... iLg7! 9 iLc4 0-0 10 llVxb7 ct:Jc6
fascinating in the match, with a dramatic with an excellent game.
conclusion. ' 5...ct:Jxc3 6 bxc3 g6
Kasparov vs. Karpov: 1986-1987
402
The Fourth Match: 1987
(Bugojno 1978) after 10 ... b6 11 ~e4 ~b7 12 1977). White is better after 10 ...tbd7 11 e4 e5
dxc5 'ilc7 13 'ilc2 .l::i.fc8 Black gained suffi- 12 i.a3 .l::i.d8 13 'ilb3 exd4 14 cxd4 cxd4 15
cient compensation for the pawn, in the .l::i.fc1 'ilf4 16 ~c4 (Karpov-Timman, Am-
event of 13 'ila4!? tba5 14 .l::i.abl .l::i.fd8 15 .l::i.fc1 sterdam 1985); 'however, here a draw was
or 13 tbd4!? bxc5 14 ~xc5 tbe5 15 ~xb7 agreed, because this result guaranteed me
'ilxb7 16 'ila4 White's chances are rather victory in the tournament.' (Karpov)
better. To go in for an unclear pawn sacri- 11 'ile2 (05)
fice in the opening of such an important 'If 11 e4 there could follow 11 ... ~g4, but
game was not something that I wanted to Karpov is not aiming for an early confron-
do. tation.' (Taimanov). Nevertheless this
deserved some consideration, for example:
12 h3 ~xf3 13 'ilxf3 cxd4 14 cxd4 ~xd4 15
~f4 ~e5, and nothing terrible for Black is
apparent.
11... .l::i.d8 (12) 12 ~e4 (18)
In an unfamiliar situation Karpov
played largely by intuition, not knowing
for sure which move here was best. He
could hardly have been attracted by 12 e4
~g4 13 ~c4, although this is a very inter-
esting position: 13 ... tbc6!? 14 i.xf7+ ~h8!
with unclear complications.
Even so, the developing move 12 .l::i.dl
10 .l::i.b1 (02) was more logical, after which we had
If 10 ~a3, then 10 ... tbd7 is possible analysed both 12 ... tbc6 (Bagirov-Tal, 46th
(Spassky-Karpov, Moscow 1973; Portisch- USSR Championship, Tbilisi 1978), and
Tal, Milan 1975), but 10 ... b6!? is better 12 ... ~b7?! (Ribli-Tal, Reykjavik 1988).
(Furman-Timoshchenko, Moscow 1977),
and if 11 ~e4, then 11 ... ~b7 12 ~xb7 'ilxb7
13 dxc5 ~xc3 14.l::i.c1 ~f6 with equality.
However, the main continuation was 10
'ile2, to which Black used to reply 10 ... b6 or
10 ...tbc6, but we had prepared 10 ... .l::i.d8,
assuming that the flexible set-up with the
delayed development of the b8-knight
would give Black a normal game: ll.l::i.dl b6
12 ~b2 tbc6 13 .l::i.ac1 ii.b7 14 e4 e6 (Portisch-
Kasparov, Madrid (rapid) 1988; Kramnik-
Topalov, Wijk aan Zee 1998).
10 ... b6 (02)
A new move, although also a typical 12 ... i.a6 (27)
one. It would perhaps have been more The normal reply, but there were also
accurate to play 10 ... .l::i.d8 (not allowing the the moves 12 ... tbc6 and 12 ... ~b7, which, in
sharp 11 e4) 11 'ile2 b6!? (instead of the Karpov's opinion, were 'insufficient for
earlier 11 ... tbc6 Tal-Vaganian, Leningrad equality':
403
Kasparov vs. Karpov: 1986-1987
404
The Fourth Match: 1987
recommendation 18 e4? h6!) 18 ... eS! 19 'Y.lUc2 'It is possible that in his preliminary cal-
h6, or 16 ... 'Y.lUd6 17 l:txb4 cxb4 18 ct:Je6 .i.c3 culations Kasparov underestimated the
with approximate equality. importance of this move. Black does not
It is no accident that White does not have complete equality' (Makarychev).
employ 12 .i.e4 any more - Black has too 'Preference must be given to White's posi-
many good possibilities. By playing tion' (Karpov). Is this so?
14 ... ct:Jb4 or IS .. .tt:Jb4, I could have forced
simplification and more quickly ap-
proached the cherished goal - a draw. The
showy IS ... eS appealed to me, but now the
position becomes stabilised and a compli-
cated strategic struggle commences.
16 e4 (14)
The move ... eS-e4 cannot be allowed: 16
ct:JgS? e4 (17 .i.c2 l:txdS!) or 16 dxc6?! e4 17
.i.xe4 fxe4 18 ct:Jd2 (to Taimanov's move 18
ct:JgS(?) there is the strong reply 18 ... 'iYe7!)
18 ... 'Y.lUxc6 etc.
405
Kasparov vs. Karpov: 1986-1987
2) 20 f3 i.f8 21 a4. Here Black has two 'iVf7 24 i.e7 'iVxc4 2S lIxc4 lIfb8! 26 lIbel
good alternatives: ~f7 27 lIxcS ~e6 Black holds on. Karpov
a) 21...i.c8 (with the positional threat of rightly suggested 20 f3!? (compared with
... a7-aS) 22 as i.d7 23 lIa1 h6 24 i.h4 i.cs the 18 ... lIe8! continuation, White has at
2S i.e1 gS 26 lIc2 ~h7 27 lIca2 lIab8 28 least an extra tempo), and 'after 20 ... i.f6 21
axb6 axb6, and White has no breakthrough i.xf6 lIxf6 22 'iVd1! and 'iVa4, preparing c4-
on the queenside (the bishop at cS is like a cS, White has a strong initiative', although
rock!), whereas Black has play with ... gS-g4; after 20 ... I:!.f7! and ... i.f8 Black's defences
b) 21...i.cS 22 as bxaS!? (now White has are solid.
to lose time regaining the pawn) 23 'iVa2 20 i.d2 (04) 20 ... I:!.f7 (08) 21 a4 (OS)
~g7 24 i.h4 lIeb8 2S i.e1 lIxb1 26 lIxb1 'The loss of time by Black makes itself
i.b6 27 'iVa4 lIc8 28 i.d2 'iVe7 with stable felt. Karpov has gained a concrete initiative
equality. on the queenside.' (Taimanov). The blame
During the play I remembered my game for this lies with the obstinacy of the black
with Petrosian (Moscow 1981), where there rooks, which have taken away squares
was a similar pawn structure with the same from their bishops.
material, and I contrived to lose with White 21 ... fxe4 (20)
from a far more favourable position (Game I At the cost of some concessions (the e4
No.45 in Volume III of My Great Predeces- outpost!) Black opens a second front.'
sors). The happy experience of 'iron Tigran' (Taimanov). 21...i.f8? 22 exfS gxfS 23 i.xfS
made me optimistic. However, the nervy I:!.xfS 24 'iVg4+ was bad for him. In Informa-
reply 18 ... lIf8 spoiled the idea ... tor Karpov and Zaitsev recommended
191Ifc1 (09) 21...f4(?!), but after 22 as! i.f8 23 I:!.a1 i.b7
24 a6! i.a8 2S 'iVg4 ~g7 26 i.e2 White has
an overwhelming position: with his bishop
on a8 Black is practically playing a piece
down, and the opening of the position on
the kingside will be ruinous for him.
22 'iVxe4 (01) 22 ... I:!.cf8 (06) 23 f3 (01)
23 ... i.c8 (02)
406
The Fourth Match: 1987
407
Kasparov vs. Karpov: 1986-1987
Mc7 'iVf6 41 h5 'iVf5 is correct, with equality. who remarked that Karpov was the best in
But White too can play more strongly - 35 the world at extinguishing a temporary
d6! Mc6 36 i..b4 with an obvious advantage. initiative for the opponent and then exploit-
After 33 'iVxe4 many commentators sug- ing the defects of his position. After
gested the sharp reply 33 ... i..a3 34 Mel Mxc4 34 ...'iVc7! White's position is merely slightly
35 f4 (I think that 35 axb6 axb6 36 Mxb6 'iVf5 more pleasant, for example: 35 axb6 axb6 36
is also acceptable for Black) 35 ... Mc2 36 Mfl Mef8 37 Mxf7+ Mxf7 38 h4 h5 39 'iVg5
'iVd3(?!) 'iVf5! etc. But I did not like 34 Mfl!? 'iVd6 etc. With the heavy pieces on the
Mxc4 35 axb6 axb6 36 Mxb6, when 36 ... 'iVf5? board, Black's strong d4-pawn and his
is now bad because of 37 d6! (37 .. .'iVxe438 bishop on c5 hold his defences together.
fxe4 and wins), or 34 ... bxa5 35 f4! with a 35 i..xa5 (02) 35 ... Mf4 (01) 36 Mel! (03)
dangerous initiative: 35 ... 'iVf5?! 36 'iVxf5 gxf5 An important move, although Karpov
37 fxe5 or 35 ... i..d6 36 fxe5 (but not 36 f5 thought that 'the rook did not stand badly
Mf7!) 36 ... i..xe5 37 i..f4 etc. on c1; 1 should have immediately played
Therefore I would have played 33 .. .'iWf5! my queen to d3'. However, here, in my
and if 34 Mel (there is nothing better: 34 view, after 36 ... Mef8! Black would have
'iVe1? e4!) - 34 .. .'iVxe4 35 Mxe4 bxa5! 36 i..xa5 gained counterplay, neutralising the defects
i..b6 with simplification and a quick draw: of his position.
37 i..xb6 axb6 38 f4 Mxc4 39 ~gl b5 etc. 36 ...'iVa6 (05)
33 ... Mf7 (05) 34 'iVg3 (05)
Karpov attaches an exclamation mark to
this move: White prevents the doubling of
the black rooks on the f-file, deprives the
black queen of the g4-square, and himself
prepares to occupy the f-file.
408
The Fourth Match: 1987
for White: he can play over the entire 1) 41...'iVe7 42 ~el M2f7 43 Ma6 'iYgS
board, combining threats on both wings. (43 .. J!c8 44 Me6 or 43 ... i..d6 44 Mc6 is no
The next moves took place in slight better) 44 Mc6 Mfl + 4S 'it'h2, and the desper-
time-trouble for me: for the four moves to ate 4S ... M8f3 does not help in view of 46
the control I had less than five minutes, 'YlUxf3 Mxf3 47 gxf3 'iYf4+ 48 'it'g2 'iVgS+ 49
whereas Karpov had roughly eleven. 'it'hl etc.;
37 i..d2 (01) 37 ... Mf7 38 'YlUd3 (03) 2) 41...hS 42 i..el Mfl+ 43 'it'h2 gS 44
White has managed to retain the poten- i..d2! (44 MaS?! is inaccurate in view of
tial threat of breaching the blockade and 44 ... h4!) 44 ... Mxbl (44 ... Mf2?! 4S MbS 'iYe7 46
activating his passed pawns. MXCS! and 'iYg3) 4S Mxbl, and it is doubtful
38 ... Mef8 39 h3 (02) 39 ... Mf2 40 Mal (but whether Black can defend the numerous
not 40 MbS 'YlUa3 - Karpov) 40 ...'iWf6 (02) holes in his position.
Here the game was adjourned. But after 41 Mgl Black is just in time to
create adequate counterplay. We devoted
our main attention to this continuation, and
as a result the quality of our adjournment
analysis was again superior to that of
Karpov's. We were able to find a mass of
interesting nuances, including the refuta-
tion of the combination with ... M7f3.
41 ... hS!
A very important tempo: now 42 l:gbl
h4 leads to the creation of mating construc-
tions around the white king (43 i..el? is not
possible because of 43 ... Mfl+ 44 'it'h2 'iYf4+).
41 Mgl? (06)
The sealed move. 'Not the best continua-
tion: two moves later the rook goes to bl,
and it could have occupied this square
immediately, without loss of time.' (Karpov)
Karpov sealed his move quite quickly,
and I was almost 100% sure that it would
be the prophylactic 41 Mgl - at first sight,
White retains the advantage. However, as
home analysis showed (and as all the
commentators remarked), 41 Mebl! was
much stronger - this would have refuted
the misguided conception with 34 ... bxaS. 42 Mas (02)
Black would have lost after 41...Mxg2? 42 42 i..elleads nowhere in view of 42 ... Mb2
'it'xg2 'YlUf2+ 43 'it'hl Mf3 44 Mfl (Karpov) or 43 MaS (43 i..g3 h4 - Karpov) 43 ... 'iVb6 or
41...'YlUh4? 42 i..el! Mfl+ 43 'it'h2 (Taimanov). 42 ... l::!.fl 43 i..g3 Mxgl+ 44 'it'xgl (44 Mxgl as!
The difficult problems facing him are - the march of the a-pawn changes the
illustrated by the following sample varia- assessment of the position) 44 ... h4 4S i..el
tions: 'iWgS(f4) with sufficient counterplay.
409
Kasparov vs. Karpov: 1986-1987
410
The Fourth Match: 1987
that the position held certain dangers for capture on e5 was possible'. And indeed, in
White, so therefore he did not take any time-trouble this exchange sacrifice sud-
resolute action and psychologically he was denly - for an instant! - seemed dangerous
already almost prepared for a repetition of to me, and I thought that Karpov was
moves. repeating moves merely to gain time on the
47 1:1e6 (17) 47 ... ~h7! 48 ~e1 (05) clock. Now I am sure that, on reflection, he
'This move was dictated by a desire to would not have tempted fate and would
maintain the tension and gain time on the again have played 50 ~el.
clock' (Karpov). Especially since 48 1:1b1?! To be fair, it should be said that from
a5! was dubious for White (49 ~xa5? 1:1fl+ the purely chess point of view 49 ... ~c5 does
50 ~h2 ~g5 and wins), and in addition, '48 not deserve to be criticised. In itself it is not
1:1xe5 was risky - after 48 ... ~d6 49 1:1h5+ (49 bad, but it allows White to continue the
1:1e6? 1:17f3!) 49 ... gxh5 50 e5+ <it>g8 51 exd6 fight.
'iWxd6 52 'iWxd4 'iWf6 the chances are only 50 1:1c6 (01)
with Black' (Taimanov). The exchange of A forced reply. 'Here a draw is the most
queens is dangerous for White, but 53 'iWd3 likely outcome, but for the moment I did
'iWf5 54 'iWe3 1:1fl 55 1:1xfl 'iWxfl + 56 <it>h2 not intend to begin peace negotiations - I
would have led to a draw - the black king was simply checking my opponent's inten-
is too open. tions. And now Kasparov cracked under
the tension and embarked on a faulty
combination.' (Karpov)
411
Kasparov vs. Karpov: 1986-1987
As Nikitin later reminded me, we had chanically replied and turned away from
also examined the best reply - 50 ... .ib4! the board and the auditorium ... ' (Makary-
with the idea of 51 .ixb4 .l::i.xg1 + 52 ~xg1 chev)
'iVxb4 53 'iVd1 .l::i.f4! or 51 .ig5 .ie1! 52 .l::i.a6 53 ... l:txd3 (01) 54 .ixf8 .l::i.xh3+ 55 ~g2 .l::i.g3+
l:txg1+ 53 ~xg1 .if2+ 54 ~h1, and here 56 ~h2l:txgl 57 .ixC5 d3 1-0
Black can immediately force a draw -
54 ... .ie1 55 ~gl .if2+ or even fight for an
advantage by 54 ... .ig3!? 55 .l::i.a1 as. Karpov
writes: 'However, now 51 l:ta6! is good,
maintaining the pressure'. And in Informa-
tor together with Zaitsev he specifies after
51.. ..ic5. But, in my opinion, it is simpler to
play 51.. ..ixd2 52 'iVxd2 'iVc8 53 .l::i.a4 'iVc5,
when Black's chances are not worse: 54
'iVd3l:tlf2 55 .l::i.ga1 ~h6 with equality.
Thus 49 ... .ic5 was not a dubious move
(in view of 50 ... .ib4), and I attached a'?!'
sign to it, only because it brought me
completely unnecessary problems: after 50 Black resigned, without waiting for his
.l::i.c6 I suddenly became feverish, and I opponent's reply. 'After 58 .ie3 White
stopped understanding what I should do gives up his bishop for the pawn, when the
next. The result of this time-trouble panic passed c- and d-pawns cannot be stopped'
was a 'powerful tactical stroke' ... (Karpov). Times: 3.28-3.27.
51 gxf3 (03) 51 ... l:txf3 52 .l::i.C7+ ~h8 A shocking end to the game! But, ap-
parently, the match intrigue demanded
such a blunder from me.
Nikitin: 'I stayed in the villa and watched
the adjournment session on television. Things
were inexorably heading for a draw. The only
worrying thing was that the players were
spending a long time on their moves, gradually
drifting into time-trouble, although the play
was exactly following our analysis. When there
were about five minutes left to the time control,
there followed a flurry of moves, begun by
Kasparov. Then the clocks were stopped and the
two players instantly left the stage. The televi-
53.ih6! sion screen went blank. I had no idea what had
'A spectacular counter, which decides happened and I phoned the press centre, but
the game. But the most surprising thing is there too they didn't yet know how the game
that Kasparov knew about this refutation, had ended... Soon the champion arrived, pale
having analysed it previously together with and depressed. In time-trouble he had decided
his seconds, but - he simply forgot about on a faulty combination and had lost a rook. I
it!' (Taimanov) silently showed him the record of our analysis,
'The world champion froze, then me- indicating this losing combination and together
412
The Fourth Match: 1987
with it a simple way to a draw from a position ahead (12-11), and only a draw separated
of strength. There was no need to say any- him from the cherished title. The day after
thing ... ' this disaster I had to play White in the final
Makarychev: 'Even before the conclusion of 24th game, and only a win would do! The
the dramatic 23rd game, I heard some of Kas- chess goddess Caissa had punished me for
parov's fans expressing their bewilderment and my excessively cautious play, which was
even criticism of the unusual way their idol was contrary to my character. Now it was
playing. But I think that what influences a possible that I would lose my title without
player most in his choice of tactics is not his winning a single game in the second half of
conscious decision as to whether he is going to the match.
play this way or that, but his subconscious Thus I had one remaining chance. Just
feelings, and here the player will sometimes one! Perhaps one in a thousand, perhaps
begin to deceive even himself using hindsight to one in a million, but I still had a chance,
make what has happened on the board fit the although it is well known that if you des-
right answer. This is true of any master or perately need to win the final game, it
grandmaster: not even the world's top players never happens. Previously only once in
are free from the pressure of objective psycho- chess history had the champion gained a
logical circumstances. And it seems to me that win in the last, decisive game, and retained
Kasparov, throughout this game, was con- his title. Lasker did this in his match with
stantly struggling with his own inner self Schlechter in 1910. But it is possible that
trying to hold him back. That's why at crucial Lasker would also have been satisfied with
moments in the game he made moves such as a draw: Schlechter's unusually aggressive
12 ... ii.a6, 34 ... bxa5 and 50 .. .'~7J3. Yes, no one play in that game (Game No.58 in Volume I
can now accuse the 13th world champion of dry of My Great Predecessors) suggested to some
pragmatism, but the split between his mind as historians that by the match regulations the
Kasparov the competitor and Kasparov the challenger had to win by a two-point
creative player was brilliantly exploited by margin in order to gain the title.
Karpov. In a television report Mikhail Tal The situation at the finish in Seville was
remarked, not without reason, that in Seville the a mirror reflection of the conclusion of the
competitive factor clearly prevailed over all 1985 match. There before the 24th game I
others - it was a battle for results, for the chess held a one-point lead, and to retain his title
crown.' Karpov needed to win. He also had White,
Yes, that's right. A match for the world and he launched an all-out assault, but at a
championship is above all of competitive critical moment through force of habit he
importance. But the two players' objectives chose a more cautious line, then went
are different: the challenger's objective is to wrong, and in the end lost the game.
win the match, while the champion's is to When I was preparing for the decisive
retain his title. As you can see, it all boils battle, I remembered that turning-point.
down to the world title. After the 23rd What strategy should be chosen for White,
game it became obvious that I'd almost lost when only a win would do? We were not
it. Even now I shudder when I recall those talking simply about the 24th game -
first few minutes after the adjournment overall it was already our 120th meeting!
session, the drive back, and the state of An inconceivable number of chess games
doom into which I sank. over a period of just three years and three
Unexpectedly, Karpov was now a point months! I had the feeling that it all was one
413
Kasparov vs. Karpov: 1986-1987
incredibly long match, beginning in Sep- order to draw the opponent into a slow
tember 1984 and only now, in December battle of manoeuvre, in the hope that
1987, approaching its climax. The plan for somewhere, in his striving for simplifica-
the last game had not only to take account tion, Karpov would falter.
of my own personal preferences, but also Before the last game of the 1985 match,
set my opponent the most difficult prob- which brought me the world title, what I
lems. And what could be more unpleasant most feared was precisely this kind of
for Karpov than play in his own style? viscous, protracted play, which would have
Had I not played 119 games with Kar- delayed my triumph. Fortunately, in that
pov, I would have been unable to win the game Karpov launched an all-out assault,
decisive 120th. The defeat in the 23rd game forcing me to defend and find the only
of the Seville match would have too trau- correct replies; in other words, he set me
matic, and I would have been unable to specific tasks, thereby lifting from me the
compose myself for a last fight for the title, burden of psychological problems. I wasn't
with less than twenty-four hours remain- going to make such a mistake, and psycho-
ing. What was the 'secret' of my prepara- logically I placed Karpov in a most un-
tion? pleasant position. Throughout the game he
For a start, I had to restore my equanim- was forced to agonise over whether it was
ity. And that same ill-starred evening I set better to make the strongest moves, or the
off with Nikitin and Litvinov to the Seville safest ones. Karpov chose the safe course;
Sol Hotel, where my team was staying. I his position worsened, but it was still quite
went in smiling, forcing myself to remain tenable ...
outwardly calm. For a few hours I sat with The day of the final trial arrived. That
my team in their room, playing cards, morning grandmaster Milunka Lazarevic
joking and laughing. We all tried to banish phoned my mother from Switzerland and
dark thoughts, and to relax and not think of said: 'Clara, tell Garry that only two people
the following day. Well after midnight I could win a decisive game to order: Alekhine,
returned to the villa and slept like a log for who failed to do this, and Fischer, who didn't
a good five to six hours until after day- need to. Tell Garry that he has to do it.' But my
break. mother didn't tell me this until after the
By then the overall score in our champi- game. Evidently, even she did not really
onship marathon was 16-16 with 87 draws, believe in a miracle ...
and a win in the 120th game would signify On the way to the 24th game I saw an
victory not only in the match, but also in enormous crowd outside the Lope de Vega
the entire unlimited duel. Why, then, did I Theatre. Even on ordinary days the match
play cards and have a good sleep, instead drew big crowds, but never so many peo-
of doing opening preparation? The point ple as on that day. I said sadly to Litvinov:
was that, after playing 119 games with 'I expect they've come to see me buried.'
Karpov, in a few hours we would have But no sooner had I got out of the car, when
been unable to discover anything new. I heard shouts of 'Suerte!' ('Good luck'),
Therefore it was more important to restore 'Suerte, Kasparov!' This support from the
at least partially my nerves and strength. people of Seville gave me added confi-
We merely outlined the general strategy. It dence. This noisy crowd of Spanish well-
was decided to playas calmly as possible, wishers, accompanying me to my decisive
without forcing events in the opening, in battle, reminded me of the millions of fans
414
The Fourth Match: 1987
415
Kasparov vs. Karpov: 1986-1987
noeuvre patiently and await a convenient (12 ... exdS 13 d4) 13 CLlxdS lIxdS 14 d4 cxd4
moment for active play. After all, an oppo- IS CLlxd4 CLlxd4 16 ~xd4 (Vaganian-Karpov,
nent who is aiming for a draw is eager to 39th USSR Championship, Leningrad 1971)
achieve it as soon as possible.' (Taimanov) White has some advantage, and under the
However, there was no guarantee that influence of the 8th Korchnoi-Spassky
the psychological pressure on Karpov match game (Kiev 1968) the main line
would force him to commit a decisive became 9 ... dxc4!? 10 bxc4 CLlc6.
mistake. It was ironic, but I had to play for Instead of this Black chose a more re-
a win in a comparatively simple position, strained and flexible continuation (the most
objectively less favourable than those 'Queen's Indian-like'!), which, however,
reached in several drawn games from the does not prevent White from deploying his
match (such as the 14th). pieces comfortably and obtaining the
4 ... ~e7 (07) freerer position.
A hopeful sign: from the very first 9 CLlc3 (IS)
moves my opponent began thinking. 9 'iVe2!? was probably more accurate,
5 ~g2 0-0 (03) 60-0 (01) 6 ... b6 (OS) with the idea of 9 ... aS 10 CLlc3 CLle4 11 lIfdl
The alternative is 6 ... cS 7 ~b2 CLlc6 and ~f6 12 d4, avoiding the exchange of bish-
then 8 cxdS CLlxdS 9 CLlc3 (in this way I once ops (Botvinnik-Stahlberg, Moscow 19S6), or
defeated Govashelishvili, 1978) or the more 9 ... cS 10 d3 and then CLlc3 (Polugayevsky-
complex 8 e3, as I played against Toro Petrosian, 1st match game, Moscow 1970).
Sanchez (Dortmund 1980) and Sosonko 9... CLle4! (11)
(Brussels (blitz) 1987). But my opponent Threatening to exchange a pair of
preferred to fianchetto his light-square knights, and then also the bishops (after
bishop. ... ~f6).
7 ~b2 ~b7 8 e3 10 CLle2 (03)
In order to retain as many pieces as pos-
sible, I avoided 10 'iVc2 CLlxc3 (Geller-Keres,
Riga 1968) or 10 cxdS exdS!? (after 10 ... CLlxc3
11 ~xc3 ~xdS - 11...exdS 12 b4!? - 12 'iVe2
CLlf6 13 lIfdl 'iVc8 14 lIac1 as IS d3 White is
slightly better, Geller-Kholmov, 37th USSR
Championship, Moscow 1969).
416
The Fourth Match: 1987
10 ... as?! (02) him both for his play, and for the stagger-
A not very Karpov-like move, creating a ing persistence displayed in such a situa-
potential weakness for Black on b6, which tion.' (Makarychev)
in time may (and will!) become perceptible. 15 d4 (35)
'Karpov waits for White to play d2-d3, in My hesitation was caused by the attrac-
order to definitely exchange the dark- tiveness of the variation 15 ttJf4!? ~b7 16 d4
square bishops. In a different competitive (16 ttJh5 is weaker because of 16 .. .f6!).
situation he would possibly have preferred Apparently this was more accurate, since it
10 ... ..tf6 11 d4 c5 with an excellent game' would have narrowed Black's choice, for
(Makarychev). Or 'the immediate 10 ... c5, example: 16 ... c5 17 ~ac1 'iYe7 (17 ... cxd4 18
and if 11 d3, then 11.. ...tf6 12 'iYc2 ..txb2 13 ~fdl!) 18 ttJd3 ~ac8 19 dxc5 or 18 ... ttJe4 19
'iYxb2 'iYf6' (Taimanov). ttJfe5 etc.
11 d3 (18) 11 ... ..tf6 (11) 12 'iYc2 (10) lS ... cS (17)
12 ... ..txb2 (02) 13 'iYxb2 ttJd6
The retreat of the knight to d6 creates a
definite lack of harmony in the placing of
the black pieces. But here 13 ... 'iYf6 is no
longer so good on account of 14 'iYc2 ttJec5
15 cxd5 (not 15 d4? because of 15 ... dxc4)
15 ... exd5 16 ttJf4 with a slight advantage to
White.
14 cxdS (01) 14 ... ..txdS (05)
The correct decision, maintaining a cer-
tain symmetry (in order to meet d3-d4 with
... c7-c5). '14 ... exd5 would have led to more
complicated play, but Karpov was not
interested in this' (Taimanov). Indeed, after 16 ~fdl
15 d4 c5 16 dxc5 bxc5 17 ~fdl White would 'Outwardly Black's position looks quite
retain some pressure. safe, but he has not yet achieved full equal-
ity: there are obvious defects in his queen-
side pawn structure, and on the d-file there
is a perceptible overloading of his pieces.'
(Taimanov)
16 ... ~c8 (22)
Black reconciles himself to the exchange
of his light-square bishop for a knight.
16 ... 'iYf6? 17 ttJe5 ~fd8 18 Jtxd5 exd5 19 ttJf4
was bad for him. However, 16 ... 'iYe7 sug-
gests itself - in Infarmatar I recommended
answering this with 17 ttJc3 Jtb7 18 ttJa4,
but after 18 ... c4! 19 bxc4 (19 ttJc3 ~fc8)
19 ... ttJxc4 20 'iVb5 ttJd6 Black has a good
'I think at this moment few of the world game. Therefore it is better to play 17 ttJf4
champion's supporters believed that a ..tb7 (Makarychev) 18 ~ac1 ~fd8 19 ttJd3,
miracle could occur. All the more credit to when White's chances are slightly better.
417
Kasparov vs. Karpov: 1986-1987
It is probable that Karpov also consid- strictly classical style'. And Makarychev
ered 16 ... c4!? - this is justified in the event even attached an exclamation mark to this
of 17 bxc4 ttJxc4 18 'ilfbS ttJd6 or 17 ttJf4 bS quiet developing move: 'A wise decision!
18 ttJxdS exdS 19 ttJeS ttJf6 20 a4 b4 21 bxc4 The world champion found the strength to
ttJxc4, but after 17 ttJc3!? White has the refrain from the active 19 ttJhS. This would
better pawn structure (if 17 ... bS, then 18 probably have been answered by 19 ... c4,
bxc4). when Black's position becomes more har-
Therefore I would not especially criticise monious. White also has nothing to boast of
16 ... .:!.c8. Black did not want to complicate after 19 dxcS ttJxcS 20 ttJhS eS with the
the play and he was seeking clear equality - threat of ... g7-g6.'
but it wasn't there, and the minor problems Nevertheless 19 dxcS!? was interesting,
gradually accumulated. The doubts which since after 19 ... ttJxcS (19 ... .:!.xcS?! 20 ttJd3!
tormented Karpov on almost every move .:!.cc8 21 .:!.ac1) 20 'ilfeS! (20 b4?! ttJc4!) con-
cost him time and created discomfort. That siderable resourcefulness would have been
is, White's chance was above all a psycho- demanded of Black: 20 ... .:!.fd8 21 ttJdS 'iVa7
logical one. 22 b4! ttJa4! 23 .:!.ac1 ttJe8! 24 .:!.xc8 .:!.xc8 2S
17 ttJf4 (IS) 17 ... .i.xf3 (OS) .i.g4 1lfb8 26 'iVxb8 .:!.xb8 27 ttJe7+ 'it'f8 28
Now this has to be played: if 17... .i.e4? ttJc6 .:!.c8 29 .i.f3 ttJc3 with equality.
there would have followed 18 dxcS ttJxcS 19 All is also not so clear after 19 ttJhS!? c4
'ilfeS ttJcb7 20 ttJhS and wins, while after (as pointed out by Taimanov, 19 ... f6 is
17... c4?! 18 ttJxdS exdS 19 bxc4 ttJxc4 weaker in view of 20 ttJf4, 'intending the
(19 ... dxc4 20 ttJeS) 20 1lfbs ttJf6 21 ttJeS breakthrough d4-dS' or simply strengthen-
White's position is simply better. ing the position) 20 .:!.ac1 - after all, Black
18.i.xf3 has weakened himself and a different game
'A long-range bishop against a knight begins, for example: 20 ... .:!.c7 21 .i.e2 bS 22
lacking a support - this is already a serious, bxc4 bxc4 23 'iVc3(a3) g6 24 'iVxaS ':!'fc8 2S
permanent factor' (Taimanov). It is hardly a ttJf4 ttJf6 with some - but apparently not
great achievement by White, but he un- altogether sufficient - compensation for the
doubtedly has a small plus. pawn (26 ':!'c2 etc.). I think that, from the
18 ...'ilfe7 purely practical point of view, the optimal
decision was in fact 19 ttJhS.
19 ... .:!.fd8 (01)
19 .:!.ac1 (23)
Taimanov: 'Kasparov manoeuvres in
418
The Fourth Match: 1987
20 dxc5 (01) 20 ... ttJxC5 24".ttJf6 25 ttJe5 ttJd5 26 ~xd5! exd5 27 ttJd3
Perhaps Black should have exchanged a or 24".ttJxd3 25 ttxc8 J:::t.xc8 26 :txd3 etc. The
pair of rooks by 20 .. Jhc5!? In the event of position of the knight at f5 has its plus
21 ttJd3 ttxc1 22 ttxc1 e5 he has nothing to point - its proximity to the important d6-
complain of, although after 21 ttxc5 ttJxc5 square.
22 b4 axb4 23 'i'xb4 he would still have had 24 ttb1 (02)
to display accuracy and patience. Here too 24 ttJd3 was possible, but I did
21 b4 (01) not want to hurry with the exchange of my
What else can be done? White at least agile knight. 24 ttbl, at the least, forces
discloses the weakness of the b6-pawn, by Black to go in for another minor concession
which he creates additional discomfort for - to give up the d-file.
the opponent. 24 ... ttxd1+ (03) 25 ttxd1 (01) 25 ...'i'C7 (02)
21 ... axb4 (14) Preparing further simplification. If
'In the event of 21...ttJce4 there is the 25".'i'a5, then simply 26 ~g2.
unpleasant 22 ttxc8 ttxc8 23 'i'd4!' (Tai- 26 ttJd3! (05)
manov). After 23".ttJg5 (23".e5? 24 ttJd5) 24 'A cunning plan. White intends a gen-
SLg2 ttJb5 25 'i'xb6 ttJc3 26 'iVb7! 'i'xb7 27 eral exchange on c5, in order to obtain a
SLxb7 Black would have had to fight for a passed a-pawn against a c-pawn (already a
draw in an ending a pawn down. real trump! - G.K.). Although this is still a
After thinking over this move, Karpov long way off, the very fact of offering an
overtook me in time spent on the clock, and exchange of knights, sharply changing the
for the last 19 moves to the control he had character of the play, is psychologically
just 22 minutes left against my 26. very effective.' (Taimanov). And indeed, in
22 'i'xb4 'i'a7 23 a3 the ensuing time scramble Karpov began to
overestimate the dangers threatening him
and to grow nervous.
419
Kasparov vs. Karpov: 1986-1987
square bishop I would have been unable to been under pressure after 28 .. :~!ia7 29 l:i.c3
exploit the weakening of the f6-square, and (but not 29 ttJxc5 'iVxa3! with equality)
the position would have remained almost 29 ... ttJf5 30 ttJe5 ttJd6 31 'iVb4.
equal: 27 'it>g2 ttJxd3 28 l:i.xd3 l:i.d8 29 l:i.c3 29 a4 (04) 29 ...ttJd6 (01) 30 'iYb1
'iVd6. With the threat of a4-a5. 'Purely time-
27l:i.C1 (02) trouble play, but very elegant!:
(Makarychev)
30 ...'iVa7 (01)
30 ... 'iVd8 31 ttJxc5 bxc5 (31...l:i.xc5? 32
l:i.xc5 bxc5 33 a5! and wins) 32 l:i.d1 would
have led to the desired position, dangerous
for Black, with a passed a-pawn for White ..
27 ...ttJe7 (03)
27 ... ttJxd3? loses to 28 l:i.xc7 l:i.xc7 29
'iVd2! etc. In Infarmatar I recommended
27 ... ttJd6, but then 28 ttJe5 is good, and if
28 ... f6?! (a new weakening!), then not 29
ttJc6? ttJd3! or 29 ttJd3 (Makarychev)
29 ... ttJxd3 30 l:i.xc7 ttJxb4 31 l:i.xc8+ ttJxc8 32 31 ttJes!
axb4 ttJd6 with equality, but 29 ttJg6! ttJf7 30 'An excellent manoeuvre, which with
h4 or 29 .. .f5 30 ttJf4 'iVb8 31 h4 etc. Black in time-trouble sets him difficult
28 'iVbs (02) practical problems. In the first instance 32
ttJc6 is threatened. How should Black
defend?' (Taimanov)
This is indeed a critical position. Karpov
had already used a lot of time - here he had
just three minutes left (I had twelve), but he
failed to find a safe way to draw.
31 ...ttJxa4? (01)
Allowing a decisive combination. 'A fa-
tal mistake. By playing 31..:iha4 32 ~xb6
'iVa3!, Black would have emerged un-
scathed: 33 l:i.d1 ttJce4! 34 ttJxf7!? (34 iLxe4
ttJxe4 35 'iVb7 ttJd6! 36l:i.xd6l:i.c1+) 34 ... 'iYa2!!
35 l:i.fl l:i.c1! 36 iLxe4 l:i.xfl + 37 'it>xfl ttJxe4'
28 ... ttJfS (02) (Makarychev). After 38 'iYb8+ Wh7! White is
In time-trouble Karpov decides to bring unable to convert his extra pawn because of
his knight back to d6. He would still have the insecurity of his king. And if 34 l::td4,
420
The Fourth Match: 1987
then 34 .. .''ilVc1 + 35 'it>g2 ttJxf2! or immedi- play, I had forgotten to write the moves
ately 34 ... ttJxf2! 35 l:ixd6 'iYa2!!. However, l:ixc8+ and ... ttJxc8 on my scoresheet. Of
with your flag about to fall it would be course, the arbiter had to remind me of the
unrealistic to find all these sharp moves. need to observe the rules, but what a mo-
It was also acceptable to play 'again ment for this to occur! Had it all turned out
31...ttJf5' (Taimanov), for example: 32 ttJc6 differently, this slap on the shoulder could
(32 'it>g2 'iYa6) 32 ... 'iYxa4 33 'iYxb6 ttJd3! 34 have become a blow of fate. By following
l:ifl 'iYa3 35 ..I1i.e4 'it>h8, holding the position. too strictly the letter of the law, Gijssen
But Karpov instinctively captured the almost changed the course of chess his-
pawn with his knight, retaining his infan- tory ...
tryman on b6. In such situations he always
'just in case' grabbed material, if he did not
see an immediate loss (remember, for
example, 26 ... ttJxa2 in the 6th game of the
first match).
33 'iYdl??
My nervous reaction was to instantly
pick up my queen and ... place it on the
wrong square! I overlooked a clever defen-
sive resource for Black and did not manage
32 l:ixcS+ ttJxcS to calculate the consequences of the obvi-
By sacrificing a pawn, I sensed that I ous 33 "iVb5!. As it transpired, this would
could develop a dangerous attack: Black's have won by force. After 33 ... ttJd6 34 "iVc6
pieces lack coordination, and his king is in ttJf5 35 'iYe8+ 'it>h7 there are two strong
a vulnerable position. I need to penetrate continuations: 36 ttJxf7 (Makarychev) and
with my queen into enemy territory, but 36 ttJd7 with the idea of 36 ... ttJc5 37 ttJf8+
the question is where it should now go to. 'it>g8 38 ttJg6+ 'it>h7 39 ..I1i.h5! - a spectacular
Karpov tensely awaited my move, prepar- mating finish! It is no better to play 33 ... Wf8
ing to reply almost instantly: otherwise he 34 ttJc6 "iVa8 35 'iYd3! g6 36 'iVd8+ Wg7 37
would simply not have managed to make 'iVd4+ Wh7 38 "iVd7 or 33 ... 'it>h7 34 ttJc6 (34
the remaining eight moves to the time 'iVe8!? ttJd6 35 'iYd8) 34 .. .'~a8 35 '*'d3+ (35
control. 'it>g2!?) 35 ... f5 (35 ... g6 36 'iYd7) 36 'ii'd8
I would have stopped to think, but here (threatening ttJe7) 36 ... ttJc5 37 "iVe8!, when
the chief arbiter Gijssen suddenly tapped Black is lost: 37.. .'~a1+ 38 'it>g2 ttJd6 39 'iVb8
me on the shoulder. Bending over me, he or 37 ... f4 38 Wg2 "iVb7 39 ..I1i.h5 ttJe7 40 'it'gl!
said: 'Mr Kasparov, you must record the '*'d7 41 ttJxe7.
moves'. And indeed, carried away by the In Taimanov's opinion, 'the energetic 33
421
Kasparov vs. Karpov: 1986-1987
il.h5!? would also have won the game of this chance. Later he remembered: 'I was
quickly, for example: 33 ... ctJd6 34 'iVdl! or presented with an excellent opportunity to
33 ... g6 34 il.xg6!'. 33 ... f6 34 il.f7+ 'it>f8 is far seize the initiative - one accurate move!
more tenacious, and in the event of 35 il.xe6 And I saw it, but for some reason I consid-
fxe5 36 'iVf5+ 'it>e8 37 'iVxe5 (37 il.xc8 'iVe7 is ered it to be impossible - and after it White
not so clear) 37... 'iVe7 38 'iVb5+ 'it>d8 39 il.xc8 would have had to fight for a draw. And I
'it>xc8 40 'iVxa4 White 'merely' reaches a would again have become world champion.
queen endgame with an extra pawn, but he But I had no time, I miscalculated, and I
has the more energetic 35 'iVh7! fxe5 36 chose an incorrect plan ... '
'iVg8+ 'it>e7 37 'iVe8+ 'it>d6 38 'iVxe6+ 'it>c7 39
'iVxe5+ 'it>d8 40 'iVe8+ 'it>c7 41 il.e6 b5 42
'iVxc8+ 'it>b6 43 il.d7, and although the play
still retains some sharpness, it is most
probable that Black cannot hold the posi-
tion.
The only virtue of 33 'iVdl was that it
came as a surprise to Karpov.
34 'iVd8+?!
'34 il.h5 was very strong, but Kas-
parov's move is also good enough' (Tai-
manov). Alas, perhaps not good enough!
Whereas 34 il.h5! would have given White
every chance of winning: 34 ... ctJc5 35 il.xf7+
'it>h7 36 M! 'iVc7 37 'iVaI (or 37 f4) and M-h5
with a fearfully strong attack. But I played
33 ... ctJe7?? (01) more safely.
'Black misses a chance opportunity to 34 ... 'it>h7 35 ctJxf7 (01)
retain an advantage in this titanic duel: Here after 35 il.h5?! ctJc5! (but not
33 ... ctJc5! 34 'iVd8+ 'it>h7 would have relieved Makarychev's variation 35 ... ctJg6? 36 il.xg6+
him of his problems (35 'iVxc8? 'iVal+ and fxg6 because of 37 ctJd7!, winning) 36 ctJxf7
... 'iVxe5). As sometimes happens, the chess ctJg8 37 M 'iVa2 White's advantage is
throne is shaky when its fate is decided by smaller than in the game.
seconds ... ' (Taimanov). 35 il.dl ctJe7 36 35 ... ctJg6 36 'iVe8! (01)
ctJxf7 ctJg6! (Makarychev) or 35 ... f5 36 'iVxc8 Undoubtedly the strongest. 36 ctJd6?
'iVaI 37 'iVd8! 'iVxe5 38 'iVxb6, and also 35 was incorrect in view of 36 ... 'iVe7 37 'iVb8 (37
'it>g2 f6! 36 'iVxc8 fxe5 37 'iVc6 should have 'iVa8 ctJc5) 37 ... 'iVf8!, transposing into a
led to a draw. drawn ending.
For an instant Karpov was one step 36 ...'iVe7!
away from regaining the title of champion, In desperate time-trouble Karpov rises to
but his hasty reply promptly deprived him the occasion - 36 ... ctJc5? 37 il.h5! was hope-
422
The Fourth Match: 1987
less for Black, for example: 37 ...'iVa1+ various possibilities of attack and defence,
(37 ...'iVa2 38 M) 38 'it>g2 'iVf6 39 f4 with the but we were unable to reach a clear verdict.
deadly threat of 40 .i.xg6+ 'it>xg6 41 ct:JgS+, The chances of a win and a draw seemed
while if 37 ... 'iVa4, then 38 ct:JgS+ hxgS 39 roughly equal, although in practice it is far
'iYxg6+ 'it>g8 40 'iVxgS or 38 'iVxa4 ct:Jxa4 39 from easy to save such an endgame against
.i.d1! 'it>g8 (39 ... ct:JcS 40 .i.c2 and h2-M-hS) 40 a persistent opponent. It later transpired
ct:Jxh6+ gxh6 41 i..xa4 with a won endgame. that Karpov had lost his belief in a success-
37 '*'xa4 'iYxf7 38 .i.e4 'it>g8 ful outcome and he assessed his chances
more pessimistically - 30% for a draw
against 70% for a loss. Thus I also had a
psychological advantage.
39 'iVbS! (01)
An accurate reply. The transposition
into a queen endgame with an extra pawn
suggested itself - 39 .i.xg6 'iVxg6 40 'iVb3 42 'it>g2 (09)
'iYg4 41 'iVxb6, which could then have been The sealed move. It was psychologically
'shuffled endlessly' (41...'iVd1+ 42 'it>g2 important that I did not play 42 M imme-
'iVdS+ 43 f3 'iVd2+ 44 'it>h3 etc.). But I intui- diately - for the moment it is better not to
tively realised that the bishop should be advance the pawns and to see what Black
retained: although it is 'of the wrong col- will do. Unhurried waiting strategy became
our' (the h8-square!), it may well come in one of the winning factors.
useful for an attack on the weak e6-pawn, Although analysis showed no clear way
cut off from base. to win, I went along to the resumption with
39 ...ct:Jf8 40 'iVxb6 'iVf6 41 'iVbs (02) 41 ...'iVe7 one sole desire - victory. The ovation which
'The time scramble came to an end, and greeted me when I walked on to the stage
here the game was adjourned. White's extra convinced me that my spectators were very
pawn promises him winning chances. But much in the majority. The possibility of
is it possible to breach Black's defences lengthy manoeuvring, with the aim of
with the play on a narrow sector of the provoking some mistake by my opponent,
board?' (Taimanov) also made me hopeful. Karpov was de-
Before the start of this decisive game I pressed by the prospect of a gruelling
did have some sleep, but before the re- defence: I could read this in his eyes when
sumption I didn't even have a nap: the fate soon after me he came on to the stage
of the world title was still in the balance. (obviously weighing on his mind was the
All night my trainers and I studied the thought that at one point he could have
423
Kasparov vs. Karpov: 1986-1987
4S ... hS?
I remember being very surprised when
Karpov weakened his defences so seri-
ously. The point of this move is not clear.
Now Black's pawn structure has lost its
flexibility and it contains more targets. The
'Play was renewed, and with bated degree of uncertainty has been reduced,
breath the chess world followed the devel- and the play has become more concrete.
opment of events. The first few moves were Sometimes the most difficult problem in a
made quickly. Karpov avoided 42 ... g5, tense situation is to maintain the tension! A
although most of the commentators player who is under positional pressure has
thought that this active plan of defence was a reflex desire to take any decision, even
the most promising' (Taimanov). But then one that is by no means best, if only to 'lift
Black would have run into 43 f4!, which is the burden'. That is what Karpov did,
not good for him - say, 43 ... 'iYf6 44 'it'h3 avoiding the 'main lines' of our analysis,
gxf4 45 exf4 etc. but this was sheer capitulation, because
And if 42 ... 'iVf6 43 h4 g5 we were intend- White was not yet threatening anything
ing 44 h5!? (unexpectedly the pawn moves and after the approximate 45 ... 'iYf6 46 'iVa7
to a square of the colour of its bishop) and (46 h5 g5) 46 ... g5 47 hxg5 hxg5 48 f3 tbg6 it
then the same plan with e3-e4-e5. True, I was possible to put up a tenacious defence.
am still not sure whether this was good 'The arrangement of the pawns on g6
enough to win. But after the game Karpov and h5 is the most vulnerable: now the
said that he was afraid of the exchange 44 exchange of queens is unacceptable for
hxg5 hxg5 and the appearance of a weak Black, and this factor plays a decisive role.
g5-pawn. Therefore he took the decision to Thanks to it White gains the opportunity to
424
The Fourth Match: 1987
56 'iVb7!
'The mechanism of threatening to ex-
change queens operates faultlessly!' (Tai-
manov). Here I sensed that the win was not
far off.
56 ... 'iVd8
425
Kasparov vs. Karpov: 1986-1987
The resistance would have been pro- the placing of his pieces, and Karpov
longed, but the assessment of the position conceded defeat. (1-0) The great battle was
would not have changed, after 56 .. .'iVxb7 57 crowned by an unforgettable finale!' (Tai-
i.xb7 e5 58 f4, when the white king em- manov). It was pointed out that 13 pieces
barks on an out-flanking manoeuvre - remained on the board - the last winning
Black's downfall is caused by the weakness factor!
of his g6- and h5-pawns.
426
The Fourth Match: 1987
of what to do if the session were to last its similar situation: I was leading 12-11, and
full six hours and the game were again Karpov needed to win the last game. It
adjourned. After all, the following morning never occurred to me or anyone on my side
all the interested parties were due to depart that if the match were drawn, at least I
for home ... The problem seemed a very real could be content with a moral victory. And
one (especially up to the move 45 ... h5), but I don't think Karpov was in the least inter-
both crises - chess and organisational - ested in the moral problem of whether he
were resolved in an instant, when someone retained his title by drawing or winning the
rushed into the room where the meeting match. So I have retained my title by effec-
was taking place, and exclaimed: 'Karpov tively playing, for the first time in history,
has resigned!' two return matches against one and the
The ovation was undoubtedly the loud- same person. And I didn't have a return
est and most prolonged (roughly 20 min- match behind me! A draw has always been
utes) I had ever been awarded outside of regarded as a victory for the world cham-
my own country. The theatre walls were pion, and now this is his only remaining
shaking, and Spanish TV interrupted the privilege. And the way the match ended (I
broadcast of a football match to switch to have in mind the 23rd and 24th games)
the conclusion of our duel. I had done that makes it a very convincing victory. I won
which Karpov failed to do in 1985: by the last game to order, against such a
winning the last game, I had drawn the player as Karpov. But you can't win the last
match (12-12) and retained my title. Now I game on psychology alone; you have to be
could enjoy three quiet years at the chess a better chess player, and also believe that
summit. you are a better player ... '
This adjournment session was one of the Karpov: 'The match in Seville would
most memorable moments in my life. The have been entirely good, if I had not con-
ex-champion's defeat, dashing his almost trived to lose the last game; I lost it stupidly
accomplished dream of a return to the and feebly, and even during the adjourn-
summit, was like a psychological knock- ment session I missed chances to escape
down. To Karpov's credit, he found the with a draw and thereby win the match.
strength to appear at the closing ceremony But there are days when you are not in the
two hours later and watch dispassionately mood for chess.'
as his 'eternal' rival received his award. But Karpov was also not helped by the ad-
Campomanes and the other FIDE officials, vice given to him before the match by my
almost all of whom had converged on former psychologist Tofik Dadashev, who
Seville, were a sorry sight. With their hopes promised the ex-champion he would win.
of Karpov's return lying in ruins, they were In this case Dadashev stood to benefit
plunged into despair, and it was written all enormously: his fame as a parapsychologist
over their faces. and champion-maker would have become
'Karpov has had another chance to legendary. After game 23, it seemed that
show that he plays chess better than me, Dadashev's prediction had come true. But
but he didn't take it', I declared at a press he was out of luck. The 'miracle' of game 24
conference in Seville. 'And for a player who delivered chess from the snare of parapsy-
hasn't regained his title, it must be cold chology.
comfort that the match as a whole ended in My win in this game drew a line under
a draw. Two years ago we were in a very our four-match duel. In Seville I played
427
Kasparov vs. Karpov: 1986-1987
much worse than previously (apart from done a great deal of preparatory work and
the start of the first match), and I was on the whole had achieved a new and
incredibly nervous, although for the first qualitatively higher level of opening prepa-
time I did not lose any weight at all during ration. I have to admit that I had underesti-
the match. Yes, I was clearly in poor form, mated Karpov's ability to repair and revive
but even so Karpov not enjoy any advan- his play so efficiently.
tage over me, and it is not at all clear why Another significant fact about our
he should have won this 'battle of equals'. fourth match is that it took place entirely
In my view, therefore, the overall score of outside the USSR. For many people this
our four matches - 17-16 with 87 draws - symbolised the 'era of perestroika' and a
can be considered quite in order. major change in policy by the Sports Com-
The Seville match clearly showed that mittee. I think that the main role here was
chess, like any other creative art, requires a played by commercial considerations: there
person to give his all. In 1987 I was possibly was a huge prize in hard currency, which
distracted too much by other matters, I did we had to hand over to the Sports Commit-
not allocate my energy properly, and I was tee. A special government decree was even
unable to concentrate properly on prepara- issued for this purpose! With support from
tions for the match. I am talking about such high quarters, the Sports Committee
psychological preparation, since I had no easily won its financial duel with the
complaints about my chess preparations. world's two top chess players: Of the
What's more, we were able to eliminate 2,280,000 Swiss francs we were awarded,
shortcomings in our analysis of adjourned each of us was actually given 137,000 francs
games and to achieve a substantial superi- plus the equivalent sum in Soviet currency
ority in this aspect of the game. Among our (about 60,000 roubles). Campomanes's
undoubted achievements were the defen- interests were also observed: over 700,000
sive plans worked out in the 9th and 19th francs went into FIDE's coffers, of which
games. For the first time in all the matches I more than 300,000 were for so-called organ-
gained creative pleasure from the harmoni- isational expenses.
ous work of my training team. As I left the stage of the Lope de Vega
As for the openings employed in the Theatre, I embraced Litvinov and joyfully
match, mention should be made of my good cried: 'Three years! I've got three years!'
result in the Griinfeld Defence (+1-1=6). Alas, at such moments time does not stop,
When he encountered it for the first time in however much we would like it to. The
the 1986 match, Karpov kept finding chinks three years until the following title match
in my preparation and he achieved an flew quickly past, but they also contained a
impressive score against this opening (+3- mass of fascinating events, including my
0=6). As we see, just a year later, in Seville, intense rivalry with Karpov in the 55th
my opponent had lost his deep strategic USSR Championship (1988) and the World
anticipation of my ideas, although he had Cup tournaments (1988-89) ...
428
Index of Openings I
Caro-Kann Defence
B17 336, 355
Catalan Opening
E05201
English Opening
A14415
A25308
A29 280, 291, 364
A30223
A31247
A34401
A36322
Grunfeld Defence
07939,113,276,288
08252
085348
087299,312,329,339
09221
09387,98
097187,358,393
098129,163
429
Kasparov vs. Karpov: 1986-1987
Nimzo-Indian Defence
E20 26, 44
Petroff Defence
(4258
Ruy Lopez
(92120,136
430
Index of Games I
431
Kasparov vs. Karpov: 1986-1987
432
GARRY KASPAROV
MODERN CHESS
PART THREE
KASPAROVvs
KARPOV 1986 -1987
Garry Kasparov and Anatoly Karpov are unquestionably the protagonists who
featured in the greatest ever chess rivalry. Between 1984 and 1990 they contested five
long matches for the World Championship . This third volume of the Garry Kasparov on
Modern Chess series concentrates on the third and fourth matches in this sequence:
London/Leningrad 1986 and Seville 1987. The two matches were tremendously
exciting. hard fought and both produced chess of an extremely high level.
The 1986 clash was groundbreaking in that it was the first World Championship match
between two Soviets to take place outside Moscow. It was split between London and
Leningrad with twelve games being played at both venues. The defending champion
was now Kasparov (having won the 1985 match) and he leapt into an apparently
decisive three point lead. However. this sensationally dissolved when a crisis broke
out in the Kasparov camp. Karpov exploited th is and pulled off the remarkable feat of
winning three games in a row. Kasparov finally regained his composure and eventually
clinched the match with a late victory.
The 1987 match was notable for it's sensational finale . Kasparov approached the final
game with a one point deficit. knowing that only a win would enable him to retain the
title. When the game was adjourned overnight in a position where Kasparov had to
win to stay champion. Spanish TV cleared its entire schedule so that the nail-biting
conclusion could be watched live. A pre- internet global audience of millions was
glued to their TV screens as Kasparov ground out his historic victory.
In this volume Garry Kasparov (world champion between 1985 and 2000 and generally
regarded as the greatest player ever) analyses in depth the clashes from 1986 and 1987.
giving his opinion on the background to the matches as well as the games themselves .
EVERYMAN CHESS
www.everymanchess .com
US $45 UK £30