Professional Documents
Culture Documents
Best Games
Anatoly Karpov
ISBN 0 7 1 34 7843 8
I have also written new commentaries for the older games, or at least re
worked the opening part, as theory has advanced so much since then.
In an overwhelming number of games in this collection I am playing
White. This is not surprising, as beautiful victories are most often gained
with the white pieces.
Let us quickly run through the composition of this collection. Which of
my victories have got into it?
The first game, dated 1969, was played when I was still a master, but the
next one is taken from a strong grandmaster tournament. Then my pro
gress towards the Mount Olympus of chess is illustrated - here the reader
will find one game from the Interzonal tournament and all the Candidates'
matches. My match with Fischer did not take place, through no fault of
mine, and in 1975 I became the 1 2th World Champion. Then there are
some games from prominent international tournaments, and then two en
counters from my duels with Korchnoi in Baguio and Merano. You could
say that my victory in our anniversary national championship ends the
first part of the book.
In 1984 a new era began in chess, linked with Karpov-Kasparov con
frontation. Here the reader will find ten of my victories from my first four
duels with Kasparov. I also bring you some games played in between
these 'epoch-making' matches. After our battle in Seville, our dispute was
interrupted for three years. In this part of the book you will find my victo
ries in the World Cup, our 55th national championships, and other con
tests. After 15 years I again had to climb the Candidates ladder, and most
of the steps are illustrated with a victory, as well as games from my next
duel with Kasparov. This is where you could say the second part of my
book ends.
Soon I was joining the regular cycle of battles for the crown, which also
means regular games from new Candidates matches. Unfortunately, my
encounter with Short did not end happily for me, and, alas, my possible
sixth meeting with Kasparov did not take place. Therefore there are sev
eral wins from tournaments in the early 1990s. In 1993 a schism occurred
in the chess world (it is scarcely worth dwelling on that question in this
book), as a result of which I was able to play another match for the chess
crown. Having won my match against Timman, I was again declared
FIDE World Champion (Kasparov and Short had played for the title of
PCA World Champion).
8 Introduction
I end the book with two victories from the 'super-tournament' at Li
nares in 1994. Although after that I have won other interesting tournament
games, I think that the full stop in this book should be placed precisely af
ter Linares, where I achieved one of the most convincing victories in the
entire history of chess.
In conclusion it remains for me to thank the chess master and writer
Evgeny Gik for his help in preparing the manuscript.
Anatoly Karpov
Karpov - Gik 9
B
18 g4 lLlf6
19 gS lLlhS (D)
w
16 bxc3 'if:xc3
The black queen here is occupy
ing an ideal position to generate
threats to the white king, and it is
difficult to believe that this move
can already be a decisive mistake.
Either 16 . . .lLlf6 or 1 6 ....:r.c8 was
necessary.
17 lLle2! (D)
In the event of 1 7 ..t>b1 a5 Black
obtains fair counterchances. How
ever, the modest knight retreat to w
e2 is extremely unpleasant for 20 .:r.xhS!
Black. The knight deals with the There is no time to lose. 20 lLlg3,
problem of ousting the queen beau which I almost played at the board,
tifully, and simultaneously joins in had to be rejected because at the
the attack on the kingside. last moment I spotted the effective
17 ••• 'iWcS 20... .i.g4 ! , which would exclude
Alas, after 1 7 ... lLld3+ 1 8 .:r.xd3 the white queen from the game.
'iWa1+ 19 ..t>d2 'iVxh1 20 g4 lLlg3 2 1 20 ••• gxhS
'iVxh 1 lLlxh 1 2 2 �e3 ! and 2 3 .:r.d 1 21 .:r.h1 'iWe3+
Black's knight is lost. 22 �bl!
Karpov - Gik 1 1
In the Dragon even a slight inac for White here was only revealed
curacy can spoil the whole game. ten years after this game. The key
Thus, for example, 22 �b2? would move for White was ignored by the
have given Black at least a draw: world press, as many commentar
22 . . . �d3+ 23 cxd3 (23 �bl loses ies only examined, after 24 "ii'xh7+
after 23 . . .'ifxf3 ! ) 23 ... 'ifxe2+ 24 'ittf8, 25 .::th6 and 25 'fi'h6+:
�al 'ifxd3 and Black is guaran a) 25 .::t h6 e6 26 .::t xg6 fxg6 27
teed a perpetual check. it'xd7 "ii'xe2 28 "ii'xd6+ �g7 29
22••• 'i'xf3 "ii'e7+ 'itth8 30 "ii'f6+ �h7 31 "ii'f7+
The white knight is untouch �h8 32 "ii'xg6 "ii'd l + 33 �b2
able: 22... 'ifxe2 23 'fi'xh5 e6 24 'ii'd4+ 34 c3 "ii'd2+ 35 .i.c2 'it'd7 !
'ifxh7+� 25 'fi'h8+ �e7 26 'iff6+ and Black maintains equality.
�e8 27 .::t h8#. 22 ... e6 also does not b) 25 il'h6+ was also suggested
work: 23 'ifxh5 'ifxf3 (23 . . .�g6 24 in my notes, with the following
'ifxh7+ �f8 25 �g3 and 26 �f5) variation: 25 . . .�e8 26 il'h8+ �f8
24 'ifxh7+ �f8 25 �d4. 27 .::th7 and now:
23 .::txhS (D) bl) 27 . . . e6 turns out to be inac
curate, as the bishop should be
placed on e6, not the pawn; after 28
g6 fxg6 29 �d4 'ifxe4 30 'ii'g 8
it'xd4 31 'iff7 + 'it>d8 32 'ifxf8+
'it>c7 33 il'xa8 the white king easily
steals away from the 'perpetual'
check: 33 ...'ii'dl+ 34 'it>b2 'ii'd4+
35 c3 'ii'f2+ 36 .i.c2 "ii'b6+ 37 �al
'ifgl+ 38 .i.bl .
b2) 27 ....i.e6 ! 28 g6 (28 �d4 is
not dangerous for Black either, as
after 28 . . .'ii'xe4 29 �xe6 fxe6 30
g6 .::tc 8 31 g7 il'e l + 32 �b2 White
is forced to submit to the perpetual
check; or 28 .i.xe6 fxe6 29 g6
'ii'xe2 and then . . . "ii'el+) 28 . . . fxg6
(28 . . . "ii'xe2 29 g7) 29 .i.xe6 "Yi'xe2
30 .:n 'ii'el+ 31 �b2 'ifb4+ 32
.i.b3 'i!fxb3+! (32 . . .�d7 33 .::t xf8
.::t xf8 34 'fi'xf8 'ifxe4 with three
12 Moscow University Ch 1969
pawns for the piece) 33 axb3 �xf7 c l ) 25 . . ..i.e6 26 lLld4 (but not
with very unclear play. Moreover, the continuation 26 .i.xe6 fxe6 27
maybe Black does not have to give lbd4 'ii'f7 !) 26 ... 'ii'xe4 27 lLlxe6+
up his queen. fxe6 28 :Cl+ 'iti>e8 29 'ii'g8+.
It is amusing that old and, as we c2) 25 ...e5 26 lLlg3 ! 'ifte8 27 .:n
shall see, not particularly precise 'itxg3 28 'i!t'g8+ lLlf8 29 'ii'xf7+
variations for some time caused �d8 30 'ilixf8+ .i.e8 31 'ii'xd6+
experts to change their attitude to and Black is forced to lay down his
wards 16 ...'ikxc3, and in some arms.
theoretical publications this move c3) 25 . . . e6 26 ltJd4 'iixe4 (al
has been considered quite play ternatively, 26 .. .'iff4 27 lLlf5) 27
able. But all the same, microscopic .:n .i.e8 28 lLlxe6+ 'ifte7 29 lLlc7.
analysis allowed me to return, and I 24 g6!
think definitively, to the old assess The courageous pawn sacri
ment: taking on c3 loses for Black. fices itself, guaranteeing a breach
If after 23 ...lLlg6 24 'i!t'xh7+ �f8, in the enemy defences. The hasty
instead of 25 'ii'h6+ or 25 .:h6, 24 ifxh7+ would have left the
White should make the quiet rook black king at large: 24 . . . �f8 and
move, 25 l:hl ! ! , then Black would now not 25 ii'h8+ (nor 25 lLld4 ow
be completely helpless. ing to 25 . . .'ilidl + 26 'iti'b2 'ii'xd4+)
Having thought of this rook ma 25 .. .ri;e7 26 'ii'xa8 due to 26...'ii'xh5.
noeuvre, it was easy to find sup 24 .•• lLlxg6
porting variations: Not 24 ... fxg6 (24 . . . hxg6 is even
c) 25 .:hi ! ! (D) and then: worse: 25 'ilih8#) 25 'ii'xh7+ 'ifilf8
26 'ii'h 8+ �e7 27 .:h7+ lLlf7 28
'iVxa8.
However, if, as in the game,
Black takes on g6 with the knight,
it seems that everything is in order
for him: White's attack has been
repulsed, and furthermore he is
two pawns down.
25 'i!t'xh7+ 'itf8
26 .:rs ! ! (DJ
This unexpected rook move was
like a thunderclap in a clear sky for
B Black ! The elegant geometrical
Karpov - Hort 13
.I �
-
� �
� -�� - - �
�
�' -
• �
� �
� ... .�
w�' �
-
��\Wr
w
. . -�·
• • •••
.
- -
�
� � .
�� � -
·�· . .
-�· . .
.<it?. • •
B B
idea immediately decides the fate 31 'ii'xg6+ ri;e7
of the game. Two lines - the a2-g8 32 'ilg5+!
diagonal and the f-file - intersect After 32 exf5 .:fs Black could
on one critical square - f7. The still resist.
threat is 27 'ifxf7#, and moreover, 32 ri;e8
the rook is supporting the queen 33 exfS .:c8
along the file, while the bishop - in 34 'ii'g8+ q;e7
the event of 26 . . .exf5 - is support 35 'ii'g7+ 1-0
ing her on the diagonal. Because of
this Black is forced to part with his Game 2
own queen. Karpov - Hort
26 'ilxb3+ Moscow Alekhine mem 1971
27 axb3 exfS Sicilian, Keres Attack
28 lbf4! (D)
Another elegant manoeuvre. 1 e4 cS ,
The aS-rook is not defended, and 2 lbf3 d6
thanks to this White can com 3 d4 cxd4
pletely destroys his opponent's de 4 lbxd4 lbf6
fences. 5 lbc3 e6
28 ... .:d8 6 g4 (D)
29 'ii'h6+ The Keres Attack - my favourite
A last finesse - the g6-pawn will weapon against the Scheveningen
drop with check. in the 1970s and 1980s. This vic
29 �e8 tory is perhaps one of the clearest I
30 lbxg6 fxg6 gained in this variation, although
14 Moscow Alekhine mem 1971
8 f4
8 .ie3 a6 (two other popular
lines are 8 ...4Jb6 and 8 ....ie7) 9 h4
(theory has also minutely studied 9
a4, 9'tlfd2 and 9 :g1) is an impor
tant alternative ; 8 4Jdb5 4Jb6 9
.if44Je5 10 'ilih5 has been played
in its time; now 10 .. 4Jg6 ! (after
.
B B
10 . . . lDc5 1 1 lDxc6 bxc6 1 2 �xc5 constant defence by Black. But
gives White a clear advantage. then my e4-pawn would also in
1 1 'ilt'xd4 e5 some cases need watching, which
12 'ilt'd2 exf4 might restrict my light-squared
13 �xf4 lDe5 bishop. Now this bishop is free for
1 3 . . . 'ilt'b6 14 llg3 'ilt'xb2 1 5 llb1 its own action, all the more so as its
Wa3 1 6 lDd5 is obviously bad for black counterpart is no longer on
Black. the board.
14 �e2 �e6 16••• lbg6
Black could not generate any ac Black's position is appreciably
tivity with 1 4 . . . 'ilt'a5 as the end worse, so Hort seeks a tactical so
game which follows is clearly in lution to his problems. Castling on
White's favour: 1 5 lDd5 'ilt'xd2+ 16 either side ( 16 0-0 or 16 ...'ilt'c7
...
1 9 �d1 (D)
22 l:.g4!
B The h2-pawn is disturbing no
The white king is not troubled one (for the time being !). Now the
by the loss of his castling rights. At main thing is not allowing the en
the same time, the black king emy queen an active position on
maintains that possibility right to the kingside, where my pieces are
the end of the game, but he does rather precariously arranged.
not manage to make use of it. 22 ... 'iVf6
19 ... gxh6 22 ...�xh2 would have been a
20 �xh6 �f6 relatively better choice, restoring
20 ...Wf6 would probably not the material balance. But I can un
have worked out for Black in view derstand my opponent; he wanted
of the fact that the h4-bishop has to develop.
perceptibly lost its mobility. Hort 23 h4!
was intending idealistically to re Now not 23 ... li)xh4? because of
build his forces, but in chess there 24 �g7. In general it will be diffi
are two sides, and I in my turn took cult for Black to win back this until
some necessary countermeasures. recently so helpless pawn, which is
21 c3 �eS (D) now gradually finding its strength.
Black seems to have achieved 23 . ..0-0-0 was also impossible in
his aim - he is threatening to play view of 24 �g5.
22 ...Wh4+, while after 22 �g5 "ilfb6 23 ... 'ii'fS
23 �e3 Wc7 my achievements will Defending against the threat
not be that great. But White finds a ened 24 h5, Black prepares to cas
very strong retort. tle long.
Karpov - Hon 17
24 l:tb4! (D)
w
nothing after 27 ......xh5 28 l:txf6
B 'iVh l + 29 .tfl (the rook defends
A beautiful place for the rook ! both bishops) 29 ... t[jg8 30 •e l+ ,
-�· ••• . . . ••
�-
� �.....
- · - .
· �·····
- · -
·�···- . �-···- .
• • • • • • • •
·- �·
- - .
- . --��-�
� .
-
- . . - . . . .
�"�
U �-�U "
��,� U ·
�"� "�rf.�
� U " ,
�
- -
�
��-
� � :�
� �
- -�-
� � :�
�
w w
12 tbxc8 l:l.fxc8 game after 24 'ifxf7+. As you can
Twenty-odd years after this game 'I
see, if the rook were on e8, Black
I decided to test another capture- would have had to resign immed.i-
1 2 ... 'ifxc8 (Salov-Karpov, Buenos ately.
Aires, Sicilian theme tournament 14 i.d3
1 994) 1 3 e5 ltJd5 14 i.c 1 i.c5 and 14 i.g5 has apparently not been
now White, instead of the theoreti- tested in practice, for example,
cal recommendations 1 5 i.d3 or 14 .. .'ii'e5 1 5 i.h4 'ii'xe4 1 6 .i.g3
1 5 c4, decided on an aggressive l:l.a8 17 i.d3 leads to a complex po
policy - bringing his queen over to sition.
the kingside with the standard Si- 14 ... i.d6
cilian manoeuvre 'ii'd 1-d3-g3: 1 5 15 �h1
'ii'd 3 0-0 16 'it'g3l:l.e8!? 17l:l.dl aS 15 f4 has also been seen more
1 8 b3 a4 1 9 i.h6 i.f8 20 h4 'f!lc7. than once. If 1 5 ... e5 16 f5l:l.xb2 17
Thus, pressure on the g7-point is g4, then 1 7 . ..'ifa5 18 �h 1 i.c5 19
parried, and the ending arising is i.c 1 l:l.xa2 20 l:l.xa2 "ii'xa2 21 g5
highly favourable for Black. ltJe8 22 'ifh5 d5 ! 23 g6 lLlf6 is
13 i.xa6 l:l.d8! (D) good for Black.
At the time when this game was 15 .i.e5
played, the rook move to d8 was 16 c3 l:l.xb2
presented as a novelty, and more- 17 'it'cl ltJg4! (D)
over a very important one. It looks 18 f4
as though the rook is no worse on Of course, 1 8 'W'xb2 loses after
e8, but run ahead for a second and 1 8 ...i.xc3 ! .
have a look at the situation i n the 18••• ltJxe3
20 Leningrad IZ 1973
21 ..• eS
Later it was established that here
the best move is 2 l . .. g5 !? with
roughly the following variation: 22
g3 'ifd6 ! (an important intermedi
ary manoeuvre) 23 i.e2 i.e5 24
'ifxf7+ Wh8 25 l:td l 'it'c7 26 'ife7
(26 .tg4 d5 ! 27 'ifxe6 i.xc3 28
exd5 cxd5 29 i.f3 We5 30 Wxe5+
i.xe5 with a draw) 26 ... i.xc3 27
i.g4 W'c8 ! 28 i.xe6 l:te8 29 Wc5
w l:txe6 30 Wxc3+ �g8 3 1 'ii'c4 'ii'b 8
19 'li'xbl 32 a4 'ife5 Razuvaev-Matulovic,
1 9 fxe5 is no good in view of Tbilisi 1 973.
1 9 'ifb6 with a clear advantage.
... 22 g3 'ifd6!
19 ... i.xf4 23 i.e2 i.gS
But not 19 ... lLlxfl ? 20 fxe5 l2Je3 24 'ilfxf7+ �h8
2 1 W'e2 'it'b6 22 :tel . 25 a4 (D)
20 W'f2 Incidentally, 25 i.g4 is also a
He has to return the exchange, promising coninuation for White,
because after 20 l:tf3 lLlg4 Black's for example, 25 ...i.f6 26 �g2
chances are better, as 2 1 h3 is im l:tf8 27 'ii'xd7 'ii'c5 28 �h3 Wc4 29
possible due to 2 1 ...i.c 1 ! !. l:tf2 preserving an advantage for
20 .•. lLlxfl White; Adorjan-Matulovic, Novi
21 :xn (DJ Sad 1 973.
B
Smejkal - Karpov 21
Strangely enough, I had this po Wh4 'ii'f6+ 35 'ii'xf6 gxf6 immedi
sition on the board when I was pre ately loses a pawn to 36 Jld7 and
paring for the game. True, it turns the white king will easily steal his
out to be weak for Black. I had to way through the holes in the pawn
apply more than a little force to ex barrier) 34 'i!fe7 ! Jle3 ! (the strongest
tricate myself from it, and then I continuation because co-operative
managed to swindle my opponent. variations such as 34 ... h6 35 i.e6
25 Jle7 \i'fl + 36 �h4 lead to victory for
26 aS l:.f8 White), and now Black, by defend
27 'i'c4 l:.xfi+ ing against the direct threat, creates
28 Jlxfi 'ili'f6! counterplay:
The only way for Black to or a) 35 Jle6 'ii'f l + 36 �g4 'i'e2+
ganize his position. The bishop be (but on no account 36 ... h5+?? be
longs on c5, and so needs support cause of 37 �h4) 37 �h3 'ii'f l + 38
from the queen on f8. Wh4 g5+ 39 �g4 (after 39 'itth 5?
29 �gl 'ili'f8 Black forces mate: 39...'ii'e2+ 40
30 Jle2 Jlc5 Jlg4 'fi'xh2+ 4 1 Jlh3 'ii'xh3#)
31 Jlg4 'iif2+ 39 ...'ii'e2+ 4(1 'itth 3 'iWfl +, etc., is a
32 �h3 d6 (D) perpetual check.
b) White cannot even advance
his passed pawn with 35 a6 since
after 35 . ..h6 threatening ...'iWfl +
and ...Jlg5+, he would be forced to
give perpetual check himself.
c) 35 Jle2! (not 35 ...'ii'xe2?? 36
ir'f8#).
d) 35 'i!fxd6 also leads to per
petual check after 35 ... 'iWfl+ 36
�h4 g5+ 37 �h5 'iWf7+ 38 'itth6
ii'g7+ 39 �h5 'i'f7+.
Smejkal was counting on win
w ning the c6-pawn and quietly going
The critical moment. about attaining an advantage. The
33 Jld7 move played is perhaps the correct
33 'ii'e6 g6 (Black cannot move practical decision.
into an ending with opposite-col 33 g6
oured bishops, since 33 ... 'iif l+ 34 34 Jlxc6 'ittg7
22 Leningrad IZ 1973
• • • •
-� �
�-� -� �-�
:� -� .,.
-
�- . . • •
• .i. . • • B
. ·�· . 42 Wh2+
•
. �tlli
-
• �
. �� 43 �f3 .i.d4
•
• �
-
�..wrll.� "
�,@-� u 44 .i.dS .i.cS
45 .i.c6 .i.d4
.
. �-�
� �
� -
� - � 46 .i.b7? (D)
w White has fallen into the trap I
Precise play by Black has al set for him and made my problems
most equalized the position, but the easier. 46 .i.d5 ! was correct. I do
presence of queens on the board not at all want to say that after this
makes the situation dangerous for move everything is in order for
both sides. White, because that is really not
Smejkal - Karpov 23
the case. In the adjourned position Black to begin his pawn movement
Black can try various ways of with his bishop on c5 or d4, while it
achieving victory, but White has is better for White to keep his own
defensive resources for every one bishop on c6 or dS. As you can see,
of them. a system of corresponding squares
has arisen. Smejkal feared this and
accordingly spent more time ana
• • • • lysing a different continuation, so
�· -·
.i.. �• ��
he did not find this area of corre
- - �··
- - spondence, and took his bishop too
far away. I will show the difference
. .
- - �
- .
-
•
• - A �-� •
-L:l· •
in later notes .
46 gS!
• • •�o 47 �g4 hS+
�
-
� .
� • .
• mu - 48 �rs
The only move. After 48 �xg5
. . . .. . 'iVxg3+ 49 �xhS (49 �fS 'ii'g4#)
B .
49 . . ..i.f2! White has no defence
The position which has arisen against mate.
on the board is a kind of zugzwang 48 'iVxg3
for White. The king cannot go to 49 �e6 'iif2 ! (D)
the only free square, g4, in view of
the mate. The white queen cannot
move to either g2 or e2, as in that
case he will be forced to advance
the g3-pawn by means of ...'iVhS+,
and definitively compromise the
position of the king. In fact, only
the bishop can move. In his turn,
Black has to find a plan to
strengthen his position. As we saw
earlier, simple checks here achieve
nothing. I chose a plan linked with
advancing my kingside pawns.
Analysis revealed that this con
tinuation is not the only one, but it
is effective enough. It is better for
24 Leningrad IZ 1973
52 .tc8 "fle7
•
• • .0 • • .
53 .trs �h6 • • • .i.
• • mu •
-
54 "ii'n "ilc7
. . .
SS 'ii'el 'ii'cS+
56 �e6 (D) B'ii'B • B
B
White has occupied the last line
of defence, but this blockade is
only temporary.
63 .•• �h4
64 .tgl 'iVgl
Black forces an exchange of
queens, after which the road to vic
tory becomes quite short.
65 'ii'xgl .i.xgl
66 �d6
Black would have achieved an
B effective finale after 66 �f6 .td4
56 �gS 67 .tfl g2! 68 .txg2 �g3 fol
57 wn 'ii'a3 lowed by ...h4.
58 'ii'el .tcs 66 .i.d4
59 'ii'dl+ 'ii'e3 67 a7 .txa7
60 'ii'aS .tb6 68 �xeS �g4
Karpov - Polugaevsky 25
69�dS h4
70eS h3 •• • •••
71.txh3+ �xh3 --
� ·�
� - -�.�. -
72e6 .tcS! •• ·-*-- •
0-1
This encounter won the regular
. . - .
competition in lnformator for the � • • �a •
most important theoretical game. -
�!@�� -
��� � � �
Game@
" . i. -
u �· �� � "
� u
Karpov - Polugaevsky a ••• •<att
Moscow Ct (6) 1974 w
Sicilian, Najdorf .:td 1 .:te8 17 tbf5 .td8 1 8 tbd4
tbg6 1 9 .:tffl tbe5, Black had a
1 e4 cS lovely game. Later on both 14 'ii'd2
2 tbf3 d6 and 14 tbd5 were tested several
3 d4 cxd4 times. The second of these moves
4 tbxd4 tbf6 deserves special attention, for ex
5 tbc3 a6 ample in the encounter Hulak-Por
6 .te2 eS tisch, Indonesia 1 983, where after
7 lbb3 .te7 14 tbd5 .txd5 1 5 exd5 tbfd7 1 6
8 0-0 .te6 .:tb4 .:tfe8 1 7 a5 .tf6 1 8 .t g1 .tg5
9 f4 'fic7 19 tbd2 tbf6 20 tbfl g6 White
10 a4 lbbd7 could have gained an advantage by
1 1 �h1 0-0 attacking on the queenside: 2 1 c3
12 .te3 exf4 .:te7 22 ifb3 h5 23 h3 lbh7 (pre
13 .:txf4 tbeS (D) paring ... f5) 24 .:tb6 followed by 25
14 aS .:ta4.
Fourteen moves have gone and I 14 ... tbfd7
have not yet made a single com It was scarcely worth moving
ment. This is not surprising - this this knight away from the centre.
position is one of the standard pat The back rank is free of pieces, and
terns in the Najdorf Variation, and an initial manoeuvre by one of
it has been seen in practice many Black's rooks suggests itself -
times. In the fourth game of the ....:tfe8, ....:ae8, or ....:ac8. For ex
match I played the weaker 14 tbd4, ample, after 14 . . . .:tac8 1 5 tbd4
and after 14 ....:tad8 1 5 'ii'g 1 .:td7 1 6 tbfd7 16 'ii'd2 .:tfe8 17 tbf5 .tf8 1 8
26 Moscow Ct (6) 1974
w
Polugaevsky spent more than an
hour wondering whether to take or
not. He might have been thinking
that if I was prepared to sacrifice, it
meant that I had analysed every
thing thoroughly at home, but on
the other hand, he could not find the
forced loss at all, so he kept search
ing through it again and again. If
Black declines the sacrifice he
Karpov - Polugaevsky 27
26 'ii'a7
l:.ab8
Wc6
• • a.:•� The decisive mistake. 26 ...t:hc6
w is no good either: 27 'i!i'c7t:hb4 28
24 l:.ed1 t:hb4 'ti'xc2 lhxc2 29 J.b6 with advan
This is the essential difference. tage to White. But 26 ...'ii'e2 would
If on my 2 1 st move I had played have given Black definite counter
l:.fel, the white rooks would now play, although here too after 27
be on bl and d l. Consequently, J.b6 White has an advantage.
Black would only have two reason 27 J.f4
able possibilities: 24. . . lhf2+ 25 Later I discovered that in the
J.xf2 'i!i'xf2 26 e5 J.e7 27 l:.fl , and press centre Furman was at this
28 Moscow Ct (6) 1974
B B
34 ti'b5 The so-called Modern Schev
35 lbh6+ Wh8 eningen is seen these days only
36 lbxf7+ l:.xf7 rarely, and this match with Spassky
Or 36 ...�g8 37 lbh6+ �h8 38 did not play the final role in this. In
.:.Xf8+ and mate. my duels with Kasparov in 1 984
37 l:.xf7 ..tf6 and 1 985, where the Sicilian De
38 ti'f2 Wg8 fence arose regularly, my opponent
39 l:.xf6 gxf6 preferred the Classical Scheven
40 ti'xf6 1-0 ingen with a more flexible move
order - first ... a7-a6 and ...ti'd8-c7,
Game 5 then ... lbb8-c6 and ...l:.f8-e8, and
Karpov Spassky
- only then .....tc8-d7. Premature de
Leningrad Ct (9) 1974 velopment of this bishop is playing
Sicilian, Scheveningen into White's hands. The alterna
tives are 9 ...ti'c7 or 9 ...e5.
1 e4 c5 10 lbb3
2 lbf3 e6 Black was preparing to ex
3 d4 cxd4 change knights in the centre and
4 lbxd4 lDf6 advance the bishop with .....tc6.
5 lbc3 d6 Thus this knight retreat is quite ac
6 ..te2 ..te7 ceptable and more precise than the
7 0-0 0-0 other well-known moves 10 �h1
8 f4 lbc6 and 10 'fi'el .
9 ..te3 (D) With the bishop on d7, Black's
9 ..td7 counterplay on the queenside falls
30 Leningrad Ct (9) 1974
8
21 ••• .th4
An unpleasant endgame awaits
Black after 2l ...'i'e6 22 1fxe6 fxe6
23 l:taf l .
22 l:td2 'i'e7
23 l:tO!
8 I was not tempted by the chance
18 ••• lbd7 of winning the exchange; after 23
In the event of 1 8 . . . 'Mi'e6, White .i.c5 'ii'g 5 24 l:td7 lLlxc2 25 i..xf8
could have developed the initiative l:txf8 Black has active counterplay.
32 Leningrad Ct (9) 1974
The d-file does not matter; the The time has come to invite the
d8-square is securely covered. black knight to move away from its
Therefore I have to generate ac familiar place.
tivity in other directions. 26 llJa6
23 l:l.fd8 (D) 27 l:l.e2! (D)
w
24 llJbl! White does not intend to ex
The craftiness of this move lies change rooks yet; the major pieces
in the fact that I have managed to are needed for an attack on the f
choose the most suitable moment file. This also frees a square for the
to move the knight into a more ac knight transfer llJd2-f3-g5, and
tive position (this can happen par moreover threatens 28 g3 �f6 29
ticularly quickly if Black swaps l:l.ef2 l:l.d6 30 �g5.
rooks). It is amusing that conse 27 .:r.rs
quently this unexpected knight 28 llJd2 �d8
retreat has become an almost char 29 llJf3 f6
acteristic illustration of my work. In defending the e5-pawn, Black
24 ••. 'iiib7 is simultaneously trying to cover
25 'itth 2! the f-file. But White's attack is al
It is a rare occurrence when, in ready unstoppable.
the middlegame, the mobility of 30 l:l.d2!
your king limits the activity of an The white rook's 'hesitation'
enemy bishop. might seem illogical. First he occu
25 �g7 pied the d-file, then abandoned it,
26 c3 and now returns in decisive fashion
Karpov - Korchnoi 33
w
10 h4 .:cs
In Game 1 examined above, the w
then fashionable line .. .'i!i'a5 and A very topical position in those
....:tfc8 was played. Transferring days. Here White has many possi
the f8- rook to c8 was later totally bilities: 1 6 .:tdg1, 16 lhd5, 16 .th6,
displaced by putting the queen's or 16 e5. However, I chose the
rook there. Besides, in recent years modest knight retreat to e2, as I had
Black has tended to meet h4 with in mind one new idea which I had
Karpov - Korchnoi 35
w
lt:Jd5, mating. After the alternative
23 . . . ..te6, I had prepared 24 lt:Jxe6
fxe6 25 lt:Jxf6+ exf6 26 ifxh7+
r.fi>f8 27 ifxb7 'fi'g5+ 28 �bl lle7
29 'ii'b8+ lle8 30 'ii'xa7 (but not 30
llh8+?? <l;g7 and Black even wins,
due to the threat of 3 1 . . .'ii'g l#)
30 . . .lle7 3 1 'ili'b8+ lle8 32 'ili'xd6+
with a form of 'windmill' .
w 24 eS!
21 lidS! Cutting off everything on the
Not, of course, 21 lt:Jd5 llxd5 and fifth rank. I was almost dazzled by
the main guardian of Black's for the wealth of apparently effective
tress, the f6-knight, remains alive. possibilities, but only this continu
21 ... llxdS ation appears to be decisive. The
22 lt:JxdS lle8 straightforward 24 lt:Jxf6+ exf6 25
Now the queen can no longer lt:Jh5 'ili'g5+ (the point !) 26 'ifxg5
manage to return to her camp: fxg5 27 lt:Jf6+ �g7 28 lt:Jxe8+
22 . . .'ifd8 23 lt:Jef4 'ii'f8 24 lt:Jxf6+ ..txe8 would not have won.
exf6 25 'fi'xh7#; if 22 ... lbh5, then 24 ... ..txd5
White wins by 23 lbxe7+ �h8 24 After 24 ... dxe5 25 lt:Jxf6+ exf6
lt:Jxc8. 26 lt:Jh5 mate is inevitable.
23 lt:Jef4 ..tc6 (D) 25 exf6 exf6
It is necessary to control the d5- The main thing is to be strong to
square, as otherwise lt:Jxf6+ and the finish ! It is still not too late to
Karpov - Vaganian 37
lose: if 26 lLlh5 (to meet 26 . . . gxh5 support, . . . b7-b5, has already been
by 27 l:tg1 + and 28 "ilg7#) then the prepared. But this rook's pawn
sobering 26 ...l:te1+. move has one essential defect - it
26 "ii'xh7+ <iitf8 does little to assist piece develop
27 "ilh8+ ment.
After 27 .. .rt;e7, 28 lLlxd5+ ii'xd5 6 dxcS
29 l:te1+ would be decisive. Apparently the simplest solu
1-0 tion to the problem. However, 6
In the second half of 1 974 this .te2 also gives White reasonable
game won the best game prize in prospects.
lnformator, with 89 marks out of 6 •.. .txcS
90 ! In the almost thirty-year his 7 lLlb3 .tb6 (D)
tory of this contest, never have the The bishop has many possible
grandmaster jury been so unani retreats - besides this, it could also
mous! go to e7, d6 and a7. It is difficult to
say which one of these is the safest.
Game 7
Karpov - Vaganian
Skopje 1976
French, Tarrasch
1 e4 e6
2 d4 dS
3 lLld2 cS
4 exdS exdS
5 iLlgf3 a6
In those days I almost always
chose the variation with 3 lLld2 in
this opening. And my opponent, w
for whom the French Defence was 8 .td3
then and is to this day a fundamen 8 .tg5 gives White nothing:
tal weapon for Black in reply to 1 e4, 8 ... lLle7 9 'ii'd2 lLlbc6 10 .te2 0-0
had prepared a quite rare continu 1 1 0-0 h6 12 .te3 l:te8 1 3 l:tad1
ation. The aim of it is to prevent the lLlf5 ! 14 .txb6 'ii'xb6 15 l:tfe 1
bishop appearing on b5. Besides .te6.
that, when the occasion arises he is 8 lLle7
prepared to play . . . c5-c4, and its 9 0-0 lLlbc6
38 Skopje 1976
w
Vaganian is trying to become ac
tive on the queenside by forcing
the white knight from b3. He prob
ably underestimated the ensuing
pawn sacrifice, after which the
B game takes on a forcing character.
11 h6 17 J.b1
12 h3 J.hS It may seem superfluous to re
13 J.e3 0-0 peat that White's main idea is to
Karpov - Vaganian 39
control d4, but he has to watch it c3-pawn, and keeping the major
constantly. Therefore the move black pieces out of the game, but is
'it'e3, which enters into White's also creating the threat of a bishop
plans, would in this specific case sacrifice on g6, with a mating at
be unsuccessful: 1 7 'it'e3 ti'xe3 1 8 tack).
l:.xe3 i.xf3 1 9 l:.xf3 a4, and Black 22 ... l:.d7
has everything under control. After 22...l:.e8 I was planning to
17 i.xf3 continue 23 i.xg6 fxg6 24 ti'xg6+
18 'it'xf3 a4 'it>f8 25 ti'xh6+ 'it>g8 26 l:.e6.
19 lbd4 'it'x b2 23 i.fS! (D)
Not 19 ...lbxd4 20 l:.xd4 'it'xb2, The point!
as after 2 1 l:.b4 Black loses a piece
(the rook not only defends the
bishop, but also blocks the a3-f8
diagonal).
20 lbxc6
Black's final piece is diverted
away from defending its king.
20 lbxc6 (D)
B
23..• l:.e7
After 23 . . .l:.c7 White would
continue with the simple 24 l:.xd5;
23 ... gxf5 loses due to 24 l:.d3 (it
must be precisely this rook, cover
ing the b1 -h7 diagonal, as the other
white rook has to control the e-file)
w 24. . .f4 (24 . . . lbe7 25 l:.xe7) 25
21 'ii'fS g6 'ii'xf4 ti'c2 (White was threatening
22 'it'f6 26 'ilfg4+, but 25 . . .f6 is probably
The white pieces are very har more stubborn) 26 l:.g3+ 'it>h7 27
moniously placed (in particular, 'ilff6 l:.g8 28 l:.xg8.
the queen is not only defending the 24 l:.xe7
40 Skopje 1976
27 .:tel!
The black rook must not be al w
lowed onto his third rank, a s a fter 32 .:te7!
....:ta6-g6 he could gain reasonable A final finesse. The black rook
counterchances linked with the is now diverted from the a-file.
dangerous passed a-pawn, by giv 32 .:tf8
ing up some pawns. 33 'ifg4+ 'i1i>h7
27 ... 'iixa2 34 .:teS 'li'h6
Tatai - Karpov 41
Tatai Karpov
- in view of 10 ltJxdS exd5 1 1 ltJc3.
Las Palmas 1 977
English Opening
1 ltJf3 c5
2 c4 ltJf6
3 ltJc3 d5
4 cxd5 ltJxd5
5 g3 g6
6 ..tg2 ..tg7
The so-called Modem Variation
of the English Opening. The stand
ard continuation here is 7 0-0, but. ..
7 ii'a4+ w
An entertaining idea. White 10 'ii'b5 c4
strives to make immediate use of 1 1 ltJa4
his small advantage in develop A logical continuation of the
ment and the lack of harmony plan which was begun on move 7 .
among the black pieces. 1 1 'iWcS is repulsed by means of
7 ••• ltJc6 1 l . . . ..tf8, while 1 1 ltJcS 0-0 1 2
After 7 .....td7 8 Wc4 ltJb4 9 0-0 ..txc6 bxc6 1 3 'ii'xc6 e5 would give
ltJ8c6 10 'ii'xc5 it is not easy for Black a dangerous initiative.
Black to prove that he has compen 11 0-0
sation for the pawn. 12 ltJxb6 axb6
8 ltJg5 13 ii'xc4 (D)
In the event of 8 Wc4 ltJdb4 ! 9 The critical position. White has
0-0 'ilfaS ! 1 0 ltJe4 ii'a6 ! 1 1 'ilfxc5 got his way, having won the c
b6 1 2 'ile3 0-0 1 3 ltJe1 ..te6 14 pawn. Black meanwhile has man
ltJc3 l:tac8 B lack has more than aged to remove his king from the
42 Las Palmas 1 977
w
22 a3
22 0-0 allows Black a pleasant
choice between the two possibili
tie s 22 ... -tg4 and 22 . . . ii'd2. The
latter is more promising, as White
w is obliged to go for the variation 23
17 d3 -txe4 -th3 24 -tg2 -txg2 25 <iPxg2
1 7 e3 gives White a hopele ssly .:.xe2 26 'ilfc1 'ii'd5+ 27 'ittg 1 -td4
weak position after either 17 ... ltlc6 with good attacking prospects for
followed by 1 8 . . . ltle5 or 1 8 ... ltlb4, Black.
or 17 ... ltlf3+ 1 8 -txf3 exf3. 22 ••• -tg4
17 ••• bS Black guesses his opponent's
18 -te3 crafty little idea: 22 ... bxa3 23 0-0,
. Again the only move. 18 e3 ltlf3+ making use of the undefended po
1 9 -txf3 exf3 20 ltlxb5 �a5+ 2 1 sition of the aS-rook.
ltlc3 b5 serves no purpose at all. 23 �c2 (D)
18 b4 Ruling out the threat 23 . . .-tf3,
19 ltld 1 .:.es because of 24 -txf3 exf3 25 �b3+
20 dxe4 fxe4 and 26 �xf3 .
21 -txd4 23 ••• 'it'd3!
But not 21 -txe4? in view of the It is rare that one can place the
reply 2 1 . . . .:.xe4 22 'it'xe4 -tf5 and queen under attack from a pawn .
... ltlc2+. 24 exd3
44 Las Palmas 1977
- - -·
• •• • • •
- - • ••
B B B B
B �
� .
� . ...
. ... .
.
Allowing the attack to develop �
�f . �g� .- ;,- -� ��
�
��-� --- D�.i.r,;� ;;;_
,, �
Garne 9
Karpov - Korchnoi
Baguio City Wch (14) 1978
Spanish, Open Variation
1 e4 e5
2 .!iJf3 .!lJc6
3 .i.b5 a6
4 .i.a4 .!iJf6
5 0-0 .!lJxe4
The Open Variation of the Span
ish, one of the most popular open w
ings in both my matches with I thought of an interesting idea
Korchnoi for the world chess crown. linked with transferring into the
One could almost write a whole sort of endgame where opposite
book about the games we played coloured bishops do not diminish,
with this opening. In this encoun but on the contrary even increase,
ter, for the first time in our theoreti White's advantage.
cal dispute I managed to use an 15 .i.xe4 dxe4
extremely valuable novelty in this 16 .!lJxc5 exf3
variation. 17 .i.f4! (D)
6 d4 b5
7 .i.b3 d5
8 dxeS .i.e6
9 c3 .i.cS
10 .!iJbd2 0-0
1 1 .i.c2 .i.f5
12 .!iJb3 .i.g4
1 2 . .i.g6 has the same value.
. .
13 h3 .i.h5
14 g4! ? .i.g6 (D)
In opening manuals at the time
this position was said to be totally
in Black's favour. In fact, if you are B
referring to the rniddlegame, the If White exchanges queens im
open position of the white king is mediately by 17 Wxd8, then he
the basis for this assessment. But loses a tempo and Black has time
46 Baguio City Wch (14) 1978
29 �e6 Game 1 1
30 c4 lbb4 Timman - Karpov
31 'ifxcS 'ifbs Montreal l979
32 �n l:r.c8 English, Four Knights
for White, he has absolutely no lnformator it also won the prize for
choice. the best game ! In the almost thirty
18 fxg3 'ii'xd6 year history of these prizes, this
19 � 'ii'h6 double has only happened twice.
20 ..td4 'ii'h2+
21 'it>e1 'ii'xg3+ Game 1 2
22 'it>d2 'ii'g2 Korchnoi - Karpov
23 ltlb2 ..ta6 Merano Wch (9) 1981
24 ltld3 ..txd3 QGD, Orthodox
25 'it>xd3 l:.bd8
26 ..tn 'ii'e4+ 1 c4 e6
27 'it>c3 cS! (D) 2 ltlc3 dS
3 d4 ..te7
4 ltlf3 ltlf6
5 ..tgS h6
6 ..th4 0-0
7 :et dxc4! (D)
w
Clearing the final approach to
the white king.
28 ..txcS 'ii'c6
29 'it>b3 l:.b8+
30 'it>a3 :es w
31 ..tb4 'ii'b6 Amazingly, this simple move
0-1 had never been seen in grandmas
The merit of the novelty used in ter practice before. Chess really
this duel was recognized - it won is inexhaustible - even in such a fa
the lnformator prize for the most miliar opening as the Queen's
important theoretical game. Re Gambit it is still possible to think
markably, in the very same issue of up something new on move seven.
56 Merano Wch (9) 1981
B w
A typical position with an iso An important link in Black's
lated d-pawn has arisen. But there plan; one knight temporarily finds
is one important nuance which is itself at the edge of the board,
favourable for Black. In situations while the second has managed to
like this, White's dark-squared take control of d5 . Here White
bishop does not usually hurry to could rid himself of the weak
get into the game, but waits for the pawn: 1 3 d5 exd5 1 4 lL!xd5 lL!xd5
best moment. The position of the 1 5 i.xd5 lL!f4 16 i.e4 'ii'xd 1 1 7
Korchnoi - Karpov 57
:cxd1 i.e6 with equality. Korch an isolated pawn. In that case his
noi did not want to give up his ad knight would be an able and reli
vantage as White so easily, and as a able defender of his pawn, while
result he fell into a difficult posi- also attacking the enemy d4-pawn.
tion. At the same time the function of
13 i.b3 lbf6 the white bishop is limited.
14 lbe5 i.d7 18 lbxc6 :xc6!
15 'ilfe2 :cs Very precisely seizing the c-file,
16 lbe4 in case White wants to move his rook
White cannot find a clear-cut to the kingside. After 1 8 ...lbxc6 1 9
plan, and fails to take account of d5 exd5 20 i.xd5 the initiative
the fact that every exchange of mi would have passed into Korchnoi's
nor pieces reduces the dynamic po hands.
tential of the d4-pawn. 19 :c3 'Mid6
1 6 :fe l looks expedient, inci 20 g3
dentally preventing 16 . . .i.c6 be It was not worth hurrying to
cause of 1 7 lbxf7. In that case I open an escape hatch. Now White's
would have preferred to reply with prospects on the kingside have dis
1 6...:c7 or 1 6...i.e8. appeared.
16 lbxe4 20 :ds
17 'ii'xe4 i.c6! (D) 21 :d1 (D)
58 Merano Wch (9) 1981
w
The unpleasant threat of . . .lLlb4
has arisen. White stubbornly does
B not wish to play a2-a3, as he is
Korchnoi - Karpov 59
more he had very little time left, ter 32 .:.b3 Black should continue
so he exchanged off minor pieces, 32 ....:.cs and 33 ....:.c4.
depriving himself of any tactical
chances whatsoever and condemn
ing himself to passive defence.
27 .i.xd5 .:.xd5
28 .:.b3
Beginning an unsuccessful op
eration, as a result of which the
white rook turns out to be 'offside' .
If Korchnoi had planned to play f2-
f4, he should have done it straight
away, without moving the rook
from the d-file.
28 B
29 '6'c3 32 ... bxa4
30 f4 (D) Not for the sake of winning a
pawn, but to distract the white
pieces from the kingside.
33 '6'a3 aS
34 .:.xa4
Or 34 .:.c4 'ii'b5 35 'iWxa4 'iWxb2.
34 ... '6'b5!
Chaining the queen to the de
fence of the unfortunately placed
white rook, and incidentally threat
ening to invade on e2. I thought
that White now had to choose 35
b3, in order to free the queen. This
B would probably have led to a win
The only way of preventing ning rook ending after 35 ....:.bs, al
... e6-e5, but at the price of weaken though I will not hide the fact that I
ing his own king. wanted to finish the struggle in the
30 ..• b6! middlegame. I was helped in this
60 Merano Wch (18) 1981
13 a4!
The fact that Korchnoi spent 54
minutes thinking testifies to the
strength of this novelty. The his
tory of game 14 was being re
peated.
13 ... .i.e7 (D)
The rook move 1 3 . . . l:b8 loses a
tempo, and after 14 axb5 axb5 1 5
tbe4 .i.e7 1 6 .i.e3 White has a seri
ous initiative; however, 1 3 . . . .i.c5
w 14 tbe4 .i.b6 1 5 tbxd4 .i.xd4 1 6
The Open Spanish underwent tbg5 0-0 1 7 axb5 tbxg5 1 8 .i.xg5
fundamental tests in Baguio and .i.xf2+ 1 9 l:xf2 'ii'xg5 20 bxa6
Merano. In the final stage of the �xe5 deserves attention, as in the
second of these matches a telling game Tischbierek-Chekhov, Pots
blow was inflicted on B lack. Now dam 1 985, when after 2 1 �h 1?,
after 1 3 tbxd4 1Wxd4 14 1Wf3 l:d8 21 ... �b5 ! would have led to an
1 5 a4 White achieves only an even even game, although it is true that
position, as in the old game Capa after 2 1 'iWfl ! ? White would have
blanca-Lasker, St Petersburg 1914. preserved slightly better chances.
In game 14 of this match I had al
ready used a novelty here, taking
Korchnoi unawares: 13 tbe4 ! . Hav
ing thought for a record amount
of time, 78 minutes, he answered
1 3 ....i.e7 14 .i.e3 tbxf3+ ( 1 4...tbf5
was the more accurate move my
opponent played in game 1 6, when
he came out of the opening with no
losses), and after 1 5 it'xf3 0-0 1 6
l:fd 1 �e8 17 tbf6+ ! .i.xf6 1 8 exf6
it'c8 1 9 fxg7 l:d8 20 h4 ! he found
himself in a critical position. I then w
gained my fifth victory. For this de 14 tbxd4 (D)
cisive encounter I had prepared yet The manoeuvre 14 tbe4 has no
another dangerous surprise. independent significance, as in this
62 Merano Wch (18) 1981
17 ••• 'iVxd1
18 lUxd1 l:[tbS
19 l:[d7 (D)
w
24 . . . lbc4 25 ..tc5 ..txc5+ 26
lbxc5 gxf6 27 lbd7 .z:r.cs 28 lbxf6+
is no good.
B 25 .:.act .z:r.xc7
19 ... ..tfS 26 .z:r.xc7 l:[dS (D)
It was necessary to continue
1 9... ..td8, although after 20 a5 lDf8
2 1 l:[d3 Black's position is not the
most pleasant.
20 fS lbd8
21 aS!
Emphasizing the passive nature
of the black pieces; 2 1 .z:r.xc7 bxa4
22 ..td4 l:[b4 23 .z:r.d 1 .z:r.ab8 is not so
clear.
21 ... lbc6 (D)
22 e6! fxe6
23 f6! lbeS w
Defending against the unpleas 27 h3!
ant threat of 24 f7+ and 25 tbg5. There is no hurry, and this quiet
Simplifying by 23 ....z:r.d8 24 .z:r.xc7 move deprives Black of the chance
.z:r.ac8 would have led to a hopeless of 27 ....z:r.d1 + 28 � lbg4+.
endgame for Black after 25 f7+ 27 ... h6
�h8 26 .z:r.xc8 .z:r.xc8 27 .z:r.c t . He must not allow the knight to
24 .z:r.xc7 .z:r.cs invade g5.
64 M1rano Wch (18) 1981
28 :a7! 32 b4 :ds
The most clear-cut way of real 33 :xa6 �f7
izing the advantage. 28 et:'lcS gxf6 34 :a7+
29 et:'lxe6 :d 1 + 30 'iltf2 i.b4 would Avoiding a trap: after 34 :xe6
have activated the black pieces. et:'lxa5 35 bxa5 :cs the rook ending
28 ... et:'lc4 may turn out not to be winning.
The rook is tied to the back rank: 34 ... �g6
28 ...:d1 + 29 'iPf2 :b1 30 i.d4 lLlc6 35 :d7
3 1 fl+ 'iPh7 32 :as et:'lxd4 33 :xf8 Since exchanging rooks leads to
�g6 34 et:'ld6, and the f-pawn will the loss of a piece - 35 ...:xd7 36
become a queen. et:'lxd7 et:'ld6 37 lbb6 - Black is
29 i.b6 :bs forced to give up the open file and
In the event of29 ...et:'lxb6 30 fl+ resign himself to the loss of his last
�h7 3 1 axb6 Black cannot cope queenside pawn.
with two passed pawns. 35 ... :es
30 i.c5 i.xcS+ 36 a6 :as
31 et:'lxc5 gxf6 (D) 37 :b7 'iPfS
38 :xbS 'iPeS
39 :b7 'iPdS
• • ••• 40 :r7!
a . • • - Precisely here; after 40 :e7 eS
··�
W& �
� -
�·-
� � - � 41 :n f5 ! Black is still holding on.
"'� .
u - - ·�
�
40 ...
41 :r6 (DJ
fS
••• • •
. . . ·� 1. . • • •
" -
u -� ·
� W& � �
� -
�
� � • • • •
W& •
•
� W& -
� ·
�
�• ••a •
w
At this point Black even has an
. �·· · ·
extra pawn. However, White's pos �··
u• -
W& •
� �
W& •
•
�
session of the seventh rank, the dif . . . ·�
ference in possibilities of the
knights, and the abundance of weak
. . -�·
nesses in the black camp leave him • • • =
no hopes of saving himself. B
Karpov - Yusupov 65
Game 14
Karpov - Yusupov
USSR Ch (Moscow) 1983
Spanish, Open Variation
Garne 1 5
Karpov - Kasparov
Moscow Wch (9) 198415
QGD, Tarrasch
B
Kasparov's opponents in the
Candidates matches had played a
variety of moves here - 12 1t'a4, 1 2
'j/fc2 and 12 a3, and Black achieved
a good game every time. I instead
used a rarer continuation, which I
had specially prepared for the en-
counter. B
12 lbaS 15 .tcs
13 'ili'cl .tg4 16 .txcS .:txcS
14 lbfS 17 lbe3!
GM Lajos Portisch thought of Compensation for the isolated
this move. It is better to dispatch d-pawn usually takes the form of
70 Moscow Wch (9) 198415
active piece play, but in this case Now the white rook and knight
the black pieces are quite passive. are very unusually positioned, in a
Thus, the opening dispute has con letter T (it is as if the d3-rook is
cluded in White's favour. suspended between epaulettes on
17 ••• .ie6 c3 and e3), and his pieces are
The fork 17 ...d4 is not danger moreover directed at the most vul
ous because of 1 8 :ad 1 . nerable spot in Black's fortress. All
18 :ad1 Wc8 the same, the d-pawn is withstand
White was already threatening ing the pressure.
1 9 lbexd5 ! �xd5 20 e4. 21 ... lLlc4
19 Wa4 22 lLlxc4
1 9 Wb1 :d8 20 :d3 deserved 22 lLlexd5 �xd5 23 lLlxd5 .txd5
attention, when 20 ... d4 is bad due 24 .ixd5 gives nothing in view of
to 21 :Cd l �6 22 .ixc6. 24 . . .:dxd5 ! 25 lbd5 :xd5 26
19 :ds lbd5 �b6 27 'it'd4 lLlxd5 28 'it'xd5
20 :d3 a6 'it'c 1 + 29 �g2 'it'xb2 with an level
21 :rd1 (DJ queen ending.
After 2 1 Wd l Black would have 22 ... :xc4 (D)
to reply 2 1 . . .11t'c6, as 2 1 . . .lLlc4 22
�exd5 lLlxd5 23 �xd5 .ixd5 24
.txd5 �xb2 25 .ixf7+ �xf7 26
:xd8 lLlxd I 27 :xc8 :xc8 28
lbdl would lead to a rook ending
with an extra pawn for White.
w
23 WaS
The natural move 23 'ifb3 would
probably have been stronger. After
23 . . . d4, not 24 'ii'xb7 'it'xb7 25
.txb7 :b8 26 .txa6 dxc3 27 .txc4
c2 28 :d8+ lbeS ! , but the quiet 24
Karpov - Kasparov 71
w
33 .:txc4 .:txc4
34 .:td4 wrs
35 .i.e2 .:txd4
36 exd4 We7 (D)
Insufficiently energetic play by
White has almost allowed Black to
stabilize the position; here 36 .. lbe4
.
37 lt::!a2 .tc8
38 li)b4 'itd6
39 f3 lOgS
40 h4 lt::!h6
41 'itf2 lt::!f5
42 lt::Jc2 (D)
• .i.. • •
• • •••
·�
-� -
� � �·�
�
-
� ·
�
w
· · · · ·�· · have been so easy. However, Black
�
-� �·
� u � � �
-
� u has taken the h4-pawn. Exchang
�- �
�
� -
��8� u ing pawns (if it happened) would
BlLl · ii.= • make achieving a draw simpler. In
fact, the break g3-g4 is harmless,
• • • • and White only has the f4-square
B left through which he can penetrate
Here the game was adjourned, the enemy camp; his pieces, the
and Black sealed the move 42 ... f6. king and knight, cannot both pass
After 42 ...lt::!g7 43 g4 f6 44 .td3 g5 through it at the same time. How
45 .tg6 ! hxg4 46 h5 White has a ever, White has at his disposal a
dangerous passed h-pawn, but study-like route.
42 ... .td7 would probably have 47 lt::!g2! !
been more reliable. Psychologically i t was impossi
42 f6 ble for Black to foresee such a
43 .td3 g5 move. It is all too easy to assume
44 .txf5 .txf5 that White will automatically re
45 li)e3 .tbl capture the pawn on h4. The unex
46 b4 gxh4 (D) pected knight manoeuvre is based
Black's bad bishop (his a6-, b5- on a pawn sacrifice. By giving up
and d5-pawns are placed on the one of his own pawns White clears
same coloured squares) determines a square which is needed for his
White's stable advantage, but is it king. The material balance is soon
decisive? After 46....tg6, piercing restored, after which the white
a hole in the fortress would not pieces invade.
Karpov - Kasparov 73
47 ... hxg3+
48 �xg3 �e6
49 lDf4+ �s
50 lbxh5 �e6
51 lbf4+ �d6
52 �g4 .tc2
53 �hS .tdl
54 �g6 �e7 (DJ
After 54... .txf3 55 �xf6 Black
loses the d5-pawn after all.
w
. - - . 66 lbe7 .*.b7
.
- . - .
- - 66...�b3 67 d5 �xa3 68 d6
• • - ·� · i.h3 69 lDd5 will not save Black
17 ... .i.b7
After the game this move pro
voked unanimous criticism from
commentators. In fact, 1 7 ....i.d7,
controlling the b5-square, is more
reliable, for example 1 8 �fl (there
is another possibility linked with
moving the bishop to a6 and the
knight to e5) 1 8 .. .l:tfd8 1 9 l:.b3
l:.ac8 20 l:.a3 l:.c7 2 1 c4 .i.a4 22
l:.bl .i.e8 23 l:.a5 and White had a
minimal advantage in Novikov B
Sturua, Lvov 1 985. 19 ..• l2Jd7
But in the game itself White's In the final analysis this loses a
advantage from the start was also pawn; 1 9...l:.ac8 20 l:.a5 l:.c7 21 c4
scarcely noticeable. Therefore, .taB is more resilient.
glancing at this almost symmetri 20 l:.aS l:.tb8
cal position, it is difficult to under 21 c4 .i.c6
stand immediately why the bishop 22 lbel!
is better placed on d7 than on b7. Paradoxically, before going on
1s �n .i.ds the attack, the white pieces first of
At the cost of a tempo it would all retreat.
have been possible at this point to 22 l:.b4
prevent the rook from landing on 23 .i.d1 ! (D)
b5 : 1 8 ...�c6, but after 1 9 lDe5
.i.a4 20 .i.b5 (20 .i.a6 l:.fd8 2 1
�e2 threatening 22 l2Jd3 i s also un
pleasant for him) 20 ... �xb5+ 2 1
l:.xb5 l:.fc8 22 l2Jd3 c4 2 3 l2Jb2
Black has not escaped difficulties.
19 l:.bS (D)
The white rook is headed for an
ideal position - the a5-square. The
a2-pawn cannot be taken in view of
the reply 20 c4 and the bishop is
trapped - the cl -rook is already in
fluencing the course of events.
76 Moscow Wch (27) 198415
. - -
If 23 'Lld3 immediately, then
23 ...l:.a4, and the pawn is safe.
Now, however, the march 'Llel ••• - -
d3xc5 cannot be prevented. The full R.t.li .t. R R
extent of White's resources is now
revealed, and the hidden strength
a • ••••
of the quiet move 1 7 l:.fc I becomes -�- - -
clear. .ii.. B � R
23 l:.b7
�·
.� -�-
B W B � �'7}.,
- - - -
24 f3 l:.d8
25 'Lld3 g5
26 ..tb3 w
The final preparations; after the 36 l:.xgS ..trs
hasty 26 'Llxc5 'Llxc5 27 l:.xc5 l:.b2 37 �e3
28 l:.xc6 l:.dd2 all White's gains The black h-pawn will in the
would have dissolved. long run cause White distinct
. 26 �f8 trouble, and it would have been
27 'Llxc5 'Llxc5 worth fixing it with 37 h4. For ex
28 l:.xc5 ample, the following variation is
White still faces a large and la possible: 37 . ..l:.f7 38 �e3 ..tg4 39
borious task, but the preliminary c5+ �c6 40 ..ta4+ �c7 41 ..te8
results of the engagement are not l:.h7 42 ..tg6 l:.h6 43 ..te4 with an
difficult to sum up: a pawn is a easy win.
pawn. 37 h4
28 l:.d6 38 �d4 e5+
29 �e2 �e7 39 �c3 ..tb1
30 l:.d1 l:.xd1 40 a3 l:.e7
31 �xd1 �d6 41 l:.g4 (D)
32 :as f5 White's final move was sealed.
33 �e2 h5 (D) In the process of home analysis it
Kasparov is trying with all his became clear that the adjourned
might to gain some counterplay. game was exceptionally sharp, and
With his last move he has weak that Black could gain counterplay.
ened the g5-pawn, and I immedi Therefore, in order to find a clear
ately make use of this. cut route to victory, I had to solve
34 e4 fxe4 various problems straight away. It
35 fxe4 ..txe4 is no accident that this ending has
Karpov - Kasparov 77
B
54 l:.h8
ss l:.dl .i.a2
56 l:.el+ �f4 w
57 l:.e4+ �g3 A curious manoeuvre, which al-
58 l:.xeS �xg4 lows Black to make use of the
59 l:.e2 1-0 white queen's position on c2 in or-
der to simplify - 9 cxd5 lbb4 ! 10
Game 17 'ii'b3 tbxd5.
Karpov - Kasparov 9 l:.dl cS
Moscow Wch (4) 1985 10 dxcS 'iiaS
QGD, Orthodox 1 1 cxdS tbxcS
12 'ii'd2
1 d4 dS After 1 2 d6 .i.d7 followed by
2 c4 e6 .. .l:.ac8, or 1 2 dxe6 .i.xe6, Black
3 tbc3 .i.e7 has sufficient compensation for the
4 tbf3 l£lf6 pawn.
Karpov - Kasparov 79
ltlxc3
.i.e6
llac8 w
22 .i.g4
80 Moscow Wch (4) 1985
w
33 l:lc2 .:.Xc2
After either 33 ...'it'f7 34 l:le2 or
w 3 3 ...'it'c6 34 'ife2 l:lc8 35 l:lxc3
Karpov - Kasparov 81
w
Here the game was adjourned,
and painstaking analysis showed
that White has huge attacking re
sources.
41 'ii'e6+ �h8
If 41 ...�f8. then 42 .i.g6 'ili'f4 43
B l:tel with the deadly threat 'ife8+.
36 'ii'c3 42 'ii'g6 Wg8
37 exdS exdS 43 'ii'e6+ �h8
38 .i.bl! 'ii'dl 44 .i.fS 1Wc3
39 'ii'e5 45 'ii'g6 �g8
39 'ii'e6 was an alternative, but 46 .i.e6+ �h8
then a white piece has to go onto a 47 .i.fS �g8
dark square at some point! Obvi 48 g3 Wf8
ously there is no reason to pursue 49 �gl 'iff6
the pawn - 39 'ii'xd2 .i.xd2 40 l:tdl so 'ifh7 'ili'f7
.i.g5 4 1 :Xd5 l:td8 and now the op 5 1 f4 was threatened, winning a
posite-coloured bishops work in piece, and if 50...g6, then 5 1 .i.xg6
B lack's favour. 'ii'g7 52 f4 .i.f6 53 l:tdl .
12 Mo�eow Wcla (4) 1985
w
w If 55 ...i.e5, then 56 i.h3 l:tf6 57
54 l:tf3! l:te3 ! l:txf2+ 58 <j;gl is the end.
One imprecision - 54 l:te3 - 56 'ifc8 l:tf6
would be enough to allow Black to 57 'ifc5+ <j;eS
free himself by 54 ...g5 ! . 58 l:tf4 'iib7+
54 ••• <j;e7 59 l:te4+ rj}f7
The king is forced to abandon It looks as though Black would
his refuge. If the f-file is recovered, escape without a scratch after
the appearance of the rook on e3 59 ...l:te6, because of 60 i.xe6
will quickly be decisive: 54 ...i.f6 'ii'xe4+, but again a problem ma
55 l:te3 g5 56 'iWxh6+, and the noeuvre is decisive: 60 'ifc4! l:txe4
rook's action is covered; 54 .. .l:tf6 6 1 'ii'g 8+ <j;e7 62 'ifxg7+ and 63
55 l:te3 g5 (55 ...l:txf5 56 'ii'h 8+ 1Wxb7.
Kasparov - Karpov 83
Game 1 8
Kasparov - Karpov
Moscow Wch (5) 1985
Spanish, Zaitsev
1 e4 e5
2 lLlf3 lLlc6 w
3 �b5 a6 13 axb5 axb5
4 �a4 lLlf6 14 :Xa8 i.xa8
5 0-0 �e7 After 14 ... :xa8 1 5 lLlg5 the
6 :et b5 black knight is forced to retreat to
7 �b3 d6 the back rank allowing White to
8 c3 0-0 seize the centre: 1 5 ... lLld8 16 lLldf3
9 h3 �b7 exd4 ( 1 6 . . . h6 1 7 lLlxf7 ! lLlxt7 1 8
10 d4 :e8 dxe5; 1 6 ...c 5 17 dxe5 dxe5 1 8
This rook move was brought 'it'xd7 lLlxd7 1 9 lLlxf7 ! c4 20 lLlxd8
into practice by my long-time sec :xd8 2 1 i.a2 :e8 22 �e3 and
ond Igor Zaitsev, and the variation White had a clear advantage in
is therefore named after him. Black lvanchuk-Portisch, Linares 1990)
fortifies the centre, and does not 17 e5 !.
spend time on the prophylactic 15 d5
... h7 -h6. True, there is a danger that Now, as he has taken on aS with
White will repeat moves with 1 1 the bishop, after 1 5 lLlg5 Black can
lLlg5 :f8 1 2 lLlf3 :e8 1 3 lLlg5. reply 15 ... :e7, and then ...h7-h6.
Thus, if he needs a victory, Black 15 ... lLla5
should choose something else. In game 46 of our first match I
11 lLlbd2 i.f8 retreated with 15 . . .lLld8, and after
B4 Mo1cow Wch (5) 1985
B
21 tt:'Jc7
22 .:Z.f2 tt:'Jd7
23 g5 'ike7
B 24 h4 'iie3
14 g5 25 .:Z.d1 tt:'Jb5
15 ..tg3 ..td6 26 'iixe3 .:Z.xe3
16 'iid3 tt:'Ja6 27 �h2! tt:'Jb6
17 b3 'iie7 28 tt:'Jg3 tt:'Jc8
18 ..txd6 29 tt:'Jn .:Z.e7
The immediate 1 8 f4 would have 30 l:td3 tt:'Jcd6
led to a sharp struggle: 1 8 ....:Z.ae8 31 tt:'Jg3
BB Moscow tt 19B6
B w
Both 12 .i.b3 and 1 2 li)e4 used 1 S l:le1 l:ldS 19 'iif4 li)d5 20 li)xd5
to be played, but in both cases con cxd5 2 1 li)e5 .i.xe5 22 l:lxe5 .i.e6
vincing routes to equality have 23 'it'e3 White gained an advan
been found for Black. This modest, tage.
but poisonous peripheral pawn 15 l:le1 aS
move was first used by Kasparov in In the event of 15 ... l:leS 1 6
the 23rd game of our second l:lxeS+ 'fixeS 17 'iid2 the game fol
match. The idea behind it is that the lows the variation examined above.
light-squared bishop for the time 16 a3 'iid7
being stays where it is, and so after In game 22 of the return match
1 2 ... l:leS he still has the possibility with Kasparov I replied 1 6 . . .l:le8,
of 1 3 'iib 3. Besides this, the g4- and after 17 l:lxeS+ ..xe8 1 8 'iid2 I
square is taken under control, and tried a new idea - 1 S ... li)d7. How
after opening the centre the cS ever after 1 9 1i'f4 .i.g6 20 h4 'ifdS
bishop will have problems devel 2 1 li)a4 I ran into difficulties. An
oping. Thus Black finds himself in important novelty was used in
a peculiar zugzwang. Gurevich-Van der Sterren, Balcu
12 exd4 1 9S6: 1S ...'ifd7 19 l:le 1 l:leS ! 20
13 exd4 li)b6 l:lxeS+ 'fixeS 2 1 ..f4 .i.e6 ! , and
14 .i.b3 .i.f5 (D) Black equalized.
In the above-mentioned game 17 li)eS .i.xe5
against Kasparov I continued with 18 l:lxe5 l:lfe8
14 ... l:leS and brought the bishop 19 1i'e2 l:lad8
out on the next move. After 1 5 l:le1 Black has decided that there is
.i.f5 16 l:lxeS+ 'fixeS 17 'iid2 'iid7 no hurry to push his rook's pawn.
90 Moscow tt 1 986
AxeS
a4 (D)
w
When I played 27 d5, I envis
aged 29 ll:!e7+ here, but while I
w was checking the variation I turned
It looks as though Black's delay my attention to the possibility of
in pushing his a-pawn makes no 29 .. .llxe7 ! 30 'it'xe7 lbxb2 3 1 'it'e8+
difference, and White must remove �h7 32 'it'xb5 ll:ldl 33 'it'c4 f5 with
the bishop to a2, when the variation a study-like draw: the knight leaps
22 ll:!xa4 ll:!xa4 23 .i.xa4 'ibd4 24 between b2 and d l , and Black
'it'xf5 'it'xa4 cannot worry him. saves himself. In the event of 34 a4
However, White finds an attractive lDb2 35 'it'c6 ll:!dl the queen has to
intermediary manoeuvre. return to c4, as after 36 aS .i..e 4!
22 'it'cS axb3 and then 37 . . . b2 Black actually
23 :e7 'it'd6 takes the upper hand.
Yet another zwischenzug. Before 29 'it'c3 :ds
he gives back the piece, the white 30 lDf4 Adl+
rook penetrates the seventh rank. 31 �b2 ll:!d2
Here 23 . . .'it'xd4 loses because of After 3 1 ...Abl 32 a4 ll:!xb2 33
24 Ae8+! �h7 25 'it'xf5+, while if axb5 the pawn cannot be stopped.
23 ...'it'c8, then 24 'it'xb6 with an 32 b4
obvious advantage for White. Preventing perpetual check by
24 Ae8+ lhe8 the knight.
25 'it'xd6 ll:!c4 32 .i.f5
26 'it'b4 b5 33 f3 g5
27 dS! cxd5 34 bxg5 bxg5
28 ll:!xd5 .i.c2 (D) 35 ll:!e2 (D)
Karpov - Kasparov 91
B
This variation came into prac
tice after the famous game Petro
sian-Fischer, Buenos Aires Ct (2)
1 97 1 , which continued 10 ...'1i'a5
1 1 .i.c4 t'Dd7 12 t'De2 �5 1 3 .i.a2
.i.f5 14 .i.xe5 ! .i.xe5 1 5 .!Dd4
tWxc5 1 6 .!Dxf5 gxf5 17 0-0 with a
dangerous attack for White. Later
various lines were proposed for
Game 2 1 Black, the most effective being
Karpov - Kasparov 1 2 ... .!Dxc5 1 3 0-0 0-0 14 f3 e5 ! 1 5
London/Leningrad Wch (5) 1986 .i.g3 b 5 1 6 .i.a2 'li'b6! 17 Wh 1 a5
Griinfeld Defence with sufficient counterplay.
10 ... 'ilt'xd2+
1 d4 t'Df6 Kasparov prefers to exchange
2 c4 g6 queens. Until now this move had
3 lDc3 d5 resided somewhat in oblivion, and
4 .i.f4 .i.g7 only recently have new ideas been
5 e3 c5 found for Black.
6 dxc5 'it'aS 1 1 Wxd2 lbd7
7 .:.et t'De4 12 .i.b5
t2 LDndol'lllAnlngrad Wch (5) 1986
z • • •••
-··
- -� i.�
-i�
-i
• • ••• w
17 c6!
• o�• • He does not keep the extra
-
�
� -
� � ·
� -
� - pawn, but in returning it, White
-
• �
u "
�� -
��
• will elicit maximum profit by cre
• mi � -� A � -� ating a passed pawn and limiting
- � � � L.l � �
�
�
- �
- . � .:
. the movement of Black's rook and
light-squared bishop.
w 17 ••• bxc6
14 e4 f5 18 d6 c5
15 e5 e6! ? Now the light-squared bishop
A n important moment. Why finds some freedom, but the dark
did Kasparov reject the previously squared one begins to suffocate.
tested 1 5 ...l:l.ac8 1 6 c6 bxc6 17 d6 Perhaps the fate of the bishops
exd6 1 8 exd6 l:l.f6 with an advan should have been dealt with differ
tage to Black (Schmidt-Gross, Nal ently - 1 8 .. . g5 1 9 -txg5 -txeS 20
eczow 1 984)? The answer is that cS l:l.cb8, and Black maintains
1 5 . . . l:l.ac8 would be met by 16 c4! hopes of counterplay.
l:l.xcS 17 -te3, as in Seirawan 19 h4! h6
Adorjan, New York 1 987. 20 le!h3! (D)
16 c4 l:l.fc8 (D) After the more natural move 20
After 1 6 ... gS 1 7 -txgS -txe5 1 8 lZ1f3, the reply 20 . . .-tc6 ! would
lbf3 -tg7 1 9 l:l.b 1 and l:l.he 1 White have led to a complex game with
Karpov - Kasparov 93
B B
chances for both sides. However, I 23 lbf2 a2
managed to find a most logical 24 lbd3 lla3
solution to the position. The white 25 llal g5
knight is concentrating on the only 26 hxg5 hxg5
true trajectory towards its optimal 27 .i.xg5 (D)
square, d3 ! As soon as it has
reached its goal, the g7-bishop will
be caught in a trap constructed by
the white e5- and d6-pawns and the
f4-bishop. Incidentally, White pre
vents the freeing advance ... g6-g5.
20 ••• aS
The passed a-pawn is not dan
gerous, so Black should have im
mediately occupied another file
with 20...llcb8.
21 f3 a4
22 llhel ! (D) w
Overprotecting the e5-pawn. 27 ••• rM7
After the hasty 22 t:Df2 Black can After 27 . . . llb8 the duel might
muddy the waters with 22 ... g5 ! 23 conclude with a study-like draw:
hxg5 hxg5 24 .i.h2 f4 25 lbd3 .i.e8 28 .i.f4 llbb3 29 lbxc5 llb2+ 30
26 .i.g1 .i.g6 27 lbxc5 .i.xe5 28 <itc1 llxg2 3 1 .i.d2 .i.h6 ! 32 .i.xh6
lbxe6 .i.xd6 29 lbxg5. llc3+ 33 �d 1 lld3+! 34 lbxd3
22 ••• a3 .i.a4+ 35 �c 1 llc2+ and Black
� I.IMtlll£lrtlrtii'IMI Wch (17) 1986
Game 22 w
Karpov - Kasparov The critical position in the
London/Leningrad Wch (1 7) 1986 Smyslov Variation. Almost every
Grilnfeld Defence thing has been tried here, includ
ing 12 'it'xd6, 1 2 lLlb5, 1 2 d5, 1 2 h3
This miniature is a classic example and 12 0-0. In the key game Bot
of a battle which was decided en vinnik-Fischer, Vama OL 1962, af
tirely in the process of home prepa ter 12 h3 .i.xf3 1 3 gxf3 .:tfd8 14 d5
ration. White's novelty cannot be lLle5 1 5 lLlb5 'it'f6 1 6 f4 lLled7 17
called unusual - it consists of a e5, Fischer gave Botvinnik an un
simple one-square advance by the pleasant surprise with 17 ...'ifxf4 ! .
h-pawn, but careful analysis has However, after mistakes b y both
shown that after this move Black's sides this fascinating game eventu
defence becomes noticeably more ally ended in a draw. However, it
difficult. turns out that White has at his dis
1 d4 lLlf6 posal one other interesting move,
2 c4 g6 which has not been tested in prac
3 lLlc3 dS tice until now.
4 lt:lf3 .i.g7 12 eS!
5 'iWb3 dxc4 At first glance this looks para
6 'iWxc4 0-0 doxical. White not only hopelessly
7 e4 .i.g4 weakens his e5-pawn, but also al
8 .i.e3 lLlfd7 lows an exchange of queens. How
9 .:tdl lLlc6 ever, things are not so simple as
10 .i.el lLlb6 that.
11 'iVcS 'iWd6 (D) 12 ... 'iWxcS
Karpov - Kasparov 95
14 h3!
In game 1 5 , where the novelty
1 2 e5 was used for the first time, I
played 14 �b5, and after 14 ... :b8 ! w
1 5 �xc7 e6 ! Black managed to 16 .txc6! bxc6
equalize completely. In fact, the 17 .td4 .tf4 (D)
threat to surround the knight by After an exchange of bishops
means of 1 6...a6 forces White to with 17 ....txd4 1 8 :xd4 :bs 1 9
lose a tempo by playing 1 6 �b5, b 3 a5 the white rook can penetrate
and the black c8-knight quickly the seventh rank, but after 20 :d7
transfers to a comfortable position. (20 :a4 is stronger: 20 ...:as 2 1
The result of that encounter obvi �e2 :ds 2 2 :d 1 :xd 1 23 �xd 1
ously satisfied my opponent, who followed by the transfer of the
thought that White had no way of knight to c4) 20... �a7 21 :xc7
96 London/Leningrad Wch (17) 1986
lllb� 22 ltlxb5 .:.xb5 23 .:.xc6 a4! away the bishop, whereupon .i.e5
Black has counterplay. and .:.d7 will make inroads into
Black's position. In view of the
atrocious position of the knight on
c8, Black has great difficulty op
posing this plan.
19 ... a4?!
The pawn continues its march,
but this is fiddling while Rome
bums. 1 9 ...e6 would not help in
view of 20 lbe4 when White's
pieces co-operate superbly. It is
hard to see what Black should actu
ally play here - maybe 1 9 .. .f5 to
w deny White's pieces access to e4,
18 0-0 aS?! but to commit oneself to such an
It was necessary to move the e extreme measure would be diffi
pawn. My preparation showed that cult at the board.
after 18 ...e5 19 .i.e3 (19 lbe2 lbe7!) What is not in doubt, is that the
19 . . ..i.xe3 ( 1 9 ... lbe7 is also possi pawn advance chosen by Kasparov
ble) 20 fxe3 White has a definite carries no real threat, and consti
advantage. However, soon after the tutes a fatal waste of time.
match this position was seen in 20 .:.e4!
Karpov-Timman, Tilburg 1 986, Originally I had intended 20 a3,
and the Dutch grandmaster proved to fix the pawn on a4 so as to gang
that Black has chances to hold the up on it. However, it occurred to
position. The game continued thus: me that I could ignore the pawn;
20 ... lbe7 21 .:.d7 lbf5 22 .:.xc7 especially since if (following 20
:res ! 23 .:.d7 :adS 24 .:.fd 1 .:.xd7 a3) the white knight captures on
25 .:.xd7 lbxe3 26 .:.c7 .:.bs 27 b3 a4, his black counterpart could
:dB with a swift draw. True, the spring out to b5 via a7, with some
feeling remains that somewhere chances of real resistance. Clearly
White could have played some there would be no point complicat
thing stronger. ing matters in this way, when sim
19 :rei ply centralising my pieces is so
White plans to dominate the strong.
centre. The move g2-g3 will chase 20 ... .i.h6
Karpov - Kasparov 97
Game 23
Karpov - Kasparov
London/Leningrad Wch (19) 1986
Griinfeld Defence
18 ... 'ii'xb2
Black's misfortune is that nei
ther 1 8...lllxf2 nor 1 8 ...lllg3 works w
because of 1 9 'ii'b 5, attacking the 20 lbd2!
undefended rook on e8. Playing for the pin - the main
19 l:.de1 motif of White's idea.
The straightforward 19 d6 (inci 20 ••. 'ii'a4
dentally threatening to play 20 g4) 21 'ilfc4
19 ...lld8 ! 20 We3 h5 2 1 llb 1 'ilfxa2 Forcing an exchange of queens;
22 l:.xb7 'it'd5 23 l:.xa7 lbxd6 24 21 lbxe4 is not so clear: 2 1 . . .l:.xe4
.i.xd6 'ilfxd6 25 lbg5 is unpleasant 22 l:.xe4 .i.xe4 23 'i!fd2 c4 with
for Black, and it is therefore prob counterplay.
ably worth playing. However, I de 21 ••• 'ilfxc4
cided to pin the knight, limiting the 22 lbxc4 .i.c3
enemy pieces' activity. Of course 22 ...b5 23 lbd2 lbf6 24 l:.xe8+
19 l:.fe1 does not work in view of lLlxe8 25 l:e1 , etc., loses immedi
1 9 . . .'ilfxf2+ 20 �h1 'ilfxel + ! 21 ately.
l:.xel lbf2+ 22 �g1 l:.xe1+ and 23 liJd2 .i.xd2
23 ... lbxd3. 24 .i.xd2 .i.d7 (D)
19
..• 'i!fb4? (D) During the game many people
This is the decisive mistake. Af thought that Black had got himself
ter 1 9 . . . lbf6 20 l:.xe8+ lbxe8 2 1 out of his mess: the d2-bishop is
Karpov - Kasparov 101
- - �. -·
· · --*- - · - ·
- - -·-
- ·8- -
- -·- -
-
A �� -%m·;; -"'B
. "" �-
A � �
LJ � %
�
0+
�
LJ
, y B��
/�
'""'
!.;.;
% -""'
%- , !/;::"
37 �e2
w
�f7
.
. . � .: �
. � � 38 ..td3 �e6
w 39 �c4 lLle5+ (D)
25 .i.f4! .i.b5
26 f3! g5
After 26 . . . .i.xfl 27 'itxfl lLlf6 - - - -
28 l:lxe8+ lLlxe8 29 .i.e5 ! the d -
- . -
- . -••
pawn finally shows what it is capa � •.•
- - - •
�
ble of: 29 . . . f6 30 d6 ! and he has to
give back the piece.
.
. .
. - �
-
23 .i.b4!
There is no sense in rushing, as
Black is totally helpless.
23 ..• 'ifb8
Again, if 23 ...e6, then 24 l£lxf6+
.i.xf6 25 1i'g6+ is decisive; after
Karpov - A. Sokolov 105
25 f5 !
The phalanx of white pawns fi-
nally begins its forward march.
25 lbxd4
26 lbxd4 lbfl+
27 �xfl .:.et
28 g4 �8 w
29 �3 .:.n+ 34 a4 .:.bl
30 �g3 .:.et 35 lbd4 i..e7
Checking from behind achieves 36 h5
nothing: 30....:.g1 + 3 1 �f4 .:.n + 32 I spent a lot of time on this
lbf3 , while Black already has to move. Of course, the continuation
bear in mind the direct threat of 3 1 36 g5 hxg5+ 37 hxg5 would have
lbc6. given me chances for victory, but it
3t �f4 is more useful to fix and 'seal up'
After 3 1 g5, in the event of Black's kingside and his pawns. I
3 I .. ..:.g1+ (or 3 l ...�e8 32 lbf3) 32 am not exaggerating when I con
�f4 .:.n+ 33 lbf3 things are not firm that at this point I had already
looking good for Black, but after calculated a beautiful combination
exchanging rooks with 3 I . . . .:.c3 32 a dozen moves long. But at the
.:.xc3 .i.xc3 33 lbc6 aS 34 �f4 same time I will not hide the fact
�e8 his resistance could have been that I foresaw the possibility of the
prolonged. pawn break.
3t h6 36 a6
32 h4 �e8 37 � i..cS
/OB LlnaNs Ct ( 10) 1987
B
The sealed move, which seemed
the strongest. The rook lies in an
ambush, anticipating the black
king wishing to cross the f-file.
42 ••• .Ubl
Preparing for 43 �d5, in which
case after 43 ....Ugl 44 l£Jd3 .Uxg4
45 lbxc5 bxc5 46 �xc5 fxe6 47
fxe6 �e7 Black would have held
the position. However, White's
knight goes off in quite a different B
direction. This is White's quintessential
43 l£Jd5! .Ugl ! idea. He is threatening 46 l£lc7+
Karpov - A. Sokolov 109
Game 26
Kasparov - Karpov
Seville Wch (2) 1987
English, Four Knights
w
1 c4
I must confess that I was not in It was precisely this move, sug
itially expecting Kasparov's con gested in its time by Igor Zaitsev,
version to the English Opening in which plunged my opponent deep
this match. However, this surprise into thought (until then Black had
at the start strangely turned into a always taken automatically on f3).
gift for me straight away - at least, 10 d3!
Kasparov spent almost an hour and The correct reply, so White has
half on his first ten moves ! It is in not wasted his thinking time. Let
teresting that the novelty I used us look at two other possibilities:
was prepared at the beginning of a) 10 dxe3 b6 ( 10 ... h6 1 1 lbh3
the 1 980s for the Merano match. d5 12 lbf4! dxc4 1 3 e4 is worse) 1 1
However, instead of 7 lbg5, Korch e4 and then:
noi retreated the knight to e 1 , and a l ) 1 l .. . .ta6 immediately is
so it remained unused. dangerous : 12 f4 h6 13 lbxf7 !
1 eS cl;xf7 14 e5 .txc4 (14 ... lbg8? 1 5
2 lbc3 lbf6 .td5+ '1tf8 1 6 .ta3+ lbge7 17 'iid3
3 lbf3 lbc6 d6 18 'iih7 ! winning) 15 exf6 gives
4 g3 .tb4 White an advantage.
5 .tg2 0-0 a2) 1 1 ... h6 12 lt:Jh3 .ta6. The
6 0-0 e4 c4-pawn is doomed, and Black can
7 lbgS .txc3 be quite satisfied at the results of
8 bxc3 :es the opening.
Kasparov - Karpov Ill
b) 10 d4 iLia5 1 1 'ifd3 b6 12 c5
.i.b7 ! and Black has the initiative, · ·i.� · -··
as in H.Olafsson-Naumkin, Bel �··
• �a • ••
d .•
grade 1988. .
. .
. -
- .
.
10..• dS
The only move, because other
�
- ··�
. -
·�
� �
�
- - .
wise the e3-pawn will quickly be
• �"' '/�!iJ-
Ao .
% oA B
"�' '�"'
surrounded. /,- !'?
' !1,::: �
1 1 'ii'b 3! 8 Bl · 8 -�B
Yet another precise move; after
1 1 cxd5 iLixd5 two lines are possi
�
� �
- •
.:*
�
ble: w
a) 1 2 iLie4 f5 1 3 c4 fxe4 14 d4, by directing the knight to e5,
cxd5 exf3 15 l:txf3 ilJd4 1 6 l:txe3 White is better.
.tg4 17 .te4 'iff6 1 8 .tb2 'ifb6 ! 19 16 .tb2 .tg4
.txd4 'ii'xd4 20 'iic l .txe2 21 1itg2 17 c4 (D)
.txd3 with a winning position for Of course, it is very tempting to
Black. threaten the enemy king, but this
b) 1 2 'ii' b3 ilJa5 ( 1 2 ...'iixg5 1 3 gives Black the initiative. 17 iLif3
f4 iLixf4 1 4 l:txf4 leads to a sharp 'ifd7 1 8 .tal .l:te7 ! ? 1 9 l:tfc l .i.h3
game) 1 3 'iia3 'iixg5 14 'iixa5 'ii'e5 is not dangerous either - Black has
( l4 . . . b6 15 f4) 1 5 d4 ( 1 5 f4? iLixf4 sufficient counterplay. However,
16 'iixe5 iLixe2+ 17 1ith l l:txe5 1 8 by continuing with the quiet 1 7
.i.b2 .:r.b5) 1 5 ...'ifd6, and the game l:tfe l White would have retained
is roughly even: 1 6 .ta3 'ifc6 17 f4 good chances.
'iixc3 1 8 'ifa4 iLif6. 17 dxc4
11 ilJaS 18 .txf6 gxf6
12 'ii'a3 c6 19 ilJe4 litg7
13 cxdS cxdS (D) 1 9....:Xe4 ! ? 20 .i.xe4 f5 ! 2 1 .tf3
14 f4 ilJc6 ilJd4 22 dxc4 i.xf3 23 exf3 e2 24
15 l:tb1 'ii'c7 l:tfel 'ifxc4 is also interesting.
After 15 ....i.g4 16 .l:tel (the most 20 dxc4
precise; after 16 l:txb7 .txe2 17 Now the advantage fully trans
lbxf7 'ifa5 ! 1 8 'ifd6 .i.xfl 1 9 fers to Black. The continuation 20
iLih6+ lith8 20 ilJf7+ White has to iLixf6 ! ? litxf6 21 .:r.b5 l:tad8 22
settle for perpetual check) 1 6... h6 'ifc3+ lite? 23 'ifxc4 would have
( 1 6 . . .'iid7 1 7 h3) 1 7 iLif3 'ifd7 1 8 led to uncertain complications.
1 12 S1vlll� Wch (2) 1987
20 ... l:.ad8
More reliable than 20....txe2 2 1
lbxf6! .
2 1 l:.b3 (D)
Apparently the decisive error.
After the continuation 2 1 lbc3
li:)d4 22 li:)d5 there is still a stub
born fight to face. Its direction de
pends on the queen's choice of
whether to go to d6 or c4:
a) 22... 'ii'd6 23 'irxd6 l:.xd6 24
B l:.xb7 .txe2 25 l:.e1 .i.xc4 26 lbxe3
Here many commentators have ex l:.de6 27 l:.b4 lbe2+ 28 �f2 .i.d3
amined the more natural 20 'ii'c 3, 29 lbd5 lbxf4 30 lbxf4 :Xe1 3 1
and in this case after the correct re lbxd3 l:.8e2+ 32 �f3 l:.e3+ 3 3 �f2
ply 20 . . .'ird8! (20 .. .'1Ve7 2 1 l:.xb7 ! l:.1e2+ 34 �1 :Xg2 35 �xg2 lhd3
'ii'xb7 22 lbxf6 �f8 23 lbxg4 ! and 36 l:.a4, and as a result chances are
White has the upper hand) 2 1 dxc4 equal.
'ii'd4 22 :Xb7 l:.xe4 23 'ii'xd4 l:.xd4 b) 22 ...'ii'xc4 23 li:)xe3 'ii'xe2 24
24 .txc6 the chances are roughly li:)xg4 'ii'xg4, with a tense situation
even. on the board.
Another dangerous manoeuvre
is 20 lbd6, and Black has to play
carefully: 20 . . .l:.e6 2 1 lbxc4 l:.d8
22 f5 l:.ee8 23 l:.b2 lbd4 24 l:.xb7
lbxe2+ 25 �h1 lbxg3+! 26 �g1
lbe2+ (neither 26 ...'ii'xb7 27 .txb7
lbxfl 28 �xfl .txf5 29 lbd6, nor
26 . . . lbxfl 27 l:.xc7 e2 28 'ii'xa7
e1 'iV 29 l:.xf7+ �h6 30 l:.xf6+ !
�g5 3 1 'ii'g7+ �f4 32 'ii'h6# will
do) 27 �h l 'ii'g3 with perpetual
check. However, in the last vari
ation Black could take a risk:
27 ...'ii'c 8 ! ? 28 lbd6 l:.xd6 29 'ii'xd6
.txf5 with enough compensation
for the exchange.
Karpov - Farago 113
B
13 ••• ti::lfS
14 l:tb1!
This move has been given an ex
clamation mark not because it is so
cool or particularly strong, but due
to other considerations: this game
is unique in the number of short
moves made by the white rooks -
out of the 27 moves remaining in
the game, 1 5 are made by the 16 .b7
rooks! 17 .d3 ti::lce7
14 ••• l:tc8 18 l:tg1 l:tc4! (D)
Karpov - Farago 1 15
• • •• • •
....... . . .
. -·- -
. -
- ··� ��· ·
-- ��
�
-
�-
- --- � �}, . -
"
u �
-
� -·
�\Wr �
rff// � -�
� �- "'
· � mltj•
- - . � u
•
,Wit : �
-
·� �
� i.. "
�
w B
Black's rook is ideally placed 24 h3 ii'd8
on c4: it is working both along 25 'ii'g3!
White's fourth rank and the c-file, Following Black's example of
and furthermore it is covering the rearranging his pieces on the c-file,
fl -a6 diagonal from the possibility I also decided to swap round my
of transferring the queen to a6 or queen and rook.
the bishop to a4 (or b5). 25 'ilfe7
19 g4 ll'lh4 26 .:o ii'cS
20 l:.g3 ,.c8 27 'itd1
21 l:.b2 ll'lhg6 White is planning to begin ma
22 l:f3! (D) noeuvring the knight, therefore it
Overprotection of the f4-pawn. is best to remove the king from the
The white rook on g3 was beauti possibility of a check from c4 be
fully fulfilling its defensive func forehand, simultaneously defend
tion on the third rank and the g-file, ing the pawn on c2. Incidentally,
but now it is time to get down to ac transferring the king to dl now is
tive operations, for which it is nec better grounded than it was on
essary to regroup. move 10.
22 ••• ll'lc6 27 ••• l:.a4 (D)
Black cannot repeat the position In a complex position Black is
with 22 ... ll'lh4 because of 23 l:.f2 ! , losing the thread of the game. It
and then 2 3. . .l:.xg4 i s impossible in was necessary to play 27 ... d4 28
view of 24 'ilfh3 ! , when White is ll'lc l l:.c3 29 .i.xc3 dxc3 30 l:.b5
better. 'ilfd4+ 3 1 l:d3 'ilfxf4 32 'ilfxf4 ll'lxf4
23 m .:hs 33 l:.xc3 ll'ld5, and Black holds on,
1 16 Brussels World Cup 1988
w
19 l:tfxd1 i.cS
19 . . . i.d6 is more stubborn, pre
venting the white rook from pene
trating the seventh rank.
20 i.xcS 'iWxcS
21 l:td7 fS w
22 l:tad 1 lDc6 25 .:Xc6 'iWxaS
White was threatening l:td8+, 26 l:txe6 �a7
and in the event of 22 . . . c6 23 h4 ! 27 g3 (D)
(but not 23 l:txg7 lDd3 ! 24 l:txd3 27 h4 ! is even stronger, com
'iWe5 ! ) White has an obvious posi pletely depriving Black of counter
tional advantage. play on the kingside.
23 lDa4
23 lDe2 !? is also good: 23 ...'ir'b5
24 l:tcl lDe5 (the most obstinate
continuation is 24 ... lDa7 ! , though
here as well Black experiences se
rious problems) 25 l:tdxc7 lDxf3+
26 gxf3 Wxa5 (Black cannot con
tinue 26 . . . 'iWxe2 due to 27 l:tc8+
�xb7 28 l:t l c7#) 27 l:tc8+ �xb7
28 l:txh8 'ir'd2 29 �fl is insuffi-
cient for Black.
23
24 l:tcl
120 Btlfort World Cup 1988
27 g5! 39 .i.g2
28 ltxh7! ltb8 Co-ordination between the rooks
29 hJ g4 and the bishop creates decisive
30 hxg4 fxg4 threats.
31 .i.g2 'it'a1+ 39 'it'd7
32 ..th2 'it'xb2 40 lth8 c4
33 lthh6 'it'a2 41 .i.e4 1-0
34 ltef6 c5
34...ltxb7 would have lost: 35 Game 29
.i.d5 ! 'it'a5 (35 ...'it'xd5 36 ltxa6+ Karpov - Kasparov
..tb8 37 lth8+) 36 .i.xb7 ..txb7 37 Belfort World Cup 1 988
ltxa6 'ifxa6 38 ltxa6 ..txa6 39 f4!. Griinfeld Defence
35 ltf4 (D)
1 d4 l£lf6
2 c4 g6
3 lbcJ d5
4 cxd5 lbxd5
5 e4 lbxc3
6 bxcJ .i.g7
7 .i.c4
The plan of 7 lbf3 c5 8 ltb1 0-0
9 .i.e2 has enjoyed great popularity
in recent years. Another plan of
development for White involves
supporting his mighty pawn centre
B with an early .i.e3, followed by
35••• 'it'd2? ! bringing the light-squared bishop
35 . . .c4 was necessary, although out to e2.
it is true that after 36 .i.d5 White 7 c5
has taken three black pawns for his 8 lbe2 lbc6
own on b7, and finally an endgame 9 .i.e3 0-0
would arise with a rook, bishop 10 0-0 .i.g4
and pawns on f2 and g3, against the It is well known that in the event
queen, which is good for White. of 10 . .. cxd4 1 1 cxd4 lba5 1 2 .i.d3
36 .i.n ! Lb7 .i.e6 1 3 d5 White can sacrifice the
37 ltxa6+ ..tb8 exchange, gaining a dangerous at
38 ltf8+ ..tc7 tack.
Karpov - Kasparov 121
w
12 i.xf7+
I specially prepared this cap
ture on f7 for the Seville match.
Immediately after it had ended
Kasparov remarked that the vari
ation is hopeless for White. But if
you judge from what happened in
the five games with it in Seville,
you cannot say that it is all that
easy for Black to solve the prob
lems of the opening. Moreover the w
proposed post-match duel in gen The threat of a capture on e5
eral ended miserably for him. Only forces the white queen to vacate
later was it established that in the d 1 . In this game I moved her to a4,
event of i.xf7+ and the exchange while later against Timman I
122 B1(fo11 World Cup 1988
played 17 Wc2, a game which also provoke such a useful pawn move
proved to be very engaging. as . . . b6? The fact is that after 1 7
17 . . . Wc4 1 8 'ifb2 .i.h6 19 h4, and 'ifa4 b6 1 8 'ifc2 in some variations,
now: for example in reply to 1 8 ...'ifc4,
a) 19 ...'iff7 20 �g1 (20 g5 ! ltX4 White switches the queen to e4
2 1 e6 ! is correct; thanks to this and, by attacking the knight, wins
zwischenzug, White diverts the an important tempo. Thus, moving
queen, and does not allow Black to the queen out to a4 has its advan
double on the f-file: 2 1 . . .'iff5 22 tages. But I will not undertake to
�g3 Wxe6 23 Wxb7 llf8 24 lle 1 ! decide definitively where the
�e3+ 25 �g1 l:txf2 26 gxh6 cxd4 queen is more comfortable, al
27 cxd4 llf8 28 'ifb2 �4 29 llxe6 though I have tried both possibili
1 -0 Vyzhmanavin-Emst, Stock ties in practice.
holm 1 99 1 ) 20 ...l:tf8 2 1 �g3 �c4
22 We2 Wxf2+ ! 23 'ifxf2 .i.e3 24
'ifxe3 �xe3 and a roughly even • • •••
endgame arose in Karpov-Kaspar � -
-·�
� - -·
� -
ov, Amsterdam 1988. • • •••
- �- N �
b) 19 ... llf8 20 g5 ! (a universal
move which guarantees White an
- - u �
-
�
advantage in this case) 20...'ifd3 2 1 •• D ·�·
'ii'b 1 ! 'ife3 2 2 1We 1 .i.g7 2 3 �g1 -
�
�
� �
u �
-
� �
� -
�
�
We4 24 �g3 ! (White returns the
pawn but seizes all the critical
� · ·tb· � �p�
squares) 24...'irxh4 25 �4 llxf2 a • •�•
(a desperate sacrifice, similar to B
what happened in the Belfort game 17 ... b6
I am commentating on) 26 �xf2 White gets a better game after
cxd4 27 lld 1 d3 28 'ire3 ! �c6 29 17 ...�6 1 8 'irb3 c4 19 'ifxb7.
�xd3 'ira4 30 'irf3 'ifa5 3 1 e6 �d8 If Black prefers 17 . . .�c4, then
32 �4 .i.e5 33 llJd5 'ifc5+ 34 �h1 after 1 8 �f4 1Wf7 1 9 g3 �d2+ 20
1 -0 Karpov-Timman, Rotterdam �g2, White has the advantage in
1 989. the event of 20. . .lbe4 (Gutman) 2 1
17 'ifa4 (D) 'ifc2 �xf2 22 'irxf2, but Black has
Now after . . .b6, the queen will the strong reply 20 ...g5 ! as demon
(as we see) retreat to c2. The ques strated by Gutman. He further sug
tion is, why does White choose to gested that White should play 1 8
Karpov - Kasparov 123
B
22 h6
22 . . ...tf6 does not work due to
w 23 exf6 exf6 24 l'Df4.
124 Be/fort World Cup 1988
43 ... fxe6
44 l:xe6 �h8
45 .:eel ..d7 (D)
w
33 'ibc4 •hs
34 •e4 .i.fl
35 g4 ..h4
36 w ..d8 (D) w
46 aS! .i.xa5
46 ... .i.c7 also loses; 47 .i.xb4
cxb4 48 ..xc7 !.
4 7 •xc5! lbd3
48 •xas lbxel +
49 l:xel .:ed8 (D)
B
37 .:n .i.b6
38 l:del l:e8
39 l:e2 •c7
40 .:et �7
41 •c4 c5
42 �g2 lbb4 w
43 e6! 50 �g3! �7
White makes the decisive break. 51 ..a3 l:d7
128 USSR Ch (Moscow) 1988
19 ••• �e7
The path to the queenside is
fraught with danger, e.g. 1 9 ... �c8
20 l:.c l + �b8 2 1 'ii'c7+ �as 22
li)d4 'ii'f6 23 .txa6 ! l:.b8 24 'it'a5
'it'd8 25 l:.c7 winning.
20 'ii'b4+ �6
If 20 ...�d8, then 2 1 li)d4 'iVf6
22 .txa6 bxa6 looks reasonable.
Then after 23 l:.c l ! (23 li)c6+ �c7
24 l:.c 1 l:.a8 25 lLla5+ �d8 26 'ii'b7
�e7 27 tLlc6+ �d6 28 'ir'b4+ �c7 B
only leads to a repetition of the 24 tLlgS
position) 23 ...l:.eg8 24 l:.c6 there is Threatening 25 g4 ..Wh6 26 l:.xe6
no defence against the threat of 25 and 27 lLlxf7+.
l:.xa6 and 26 l:.a8+: 24 l:.hf8
a) 24...'it'e7 25 'iia5+. 25 .te2! ..Wh6
b) 24 ... 'ii'e5 25 l:.xe6! . 26 .tf3 l:.e7
2 1 'ii'd4+ Retreating the king does not
A pure geometric pattern. The help: 26 ...�c7 27 ..Wf4+ and 28
white queen is moving like a pen tLlxe6; 26 ... �c8 27 l:.c 1 + �d8 28
dulum. 2 1 .txd7 does not work in .txd5 .txd5 29 ..Wxd5.
view of 21 ...'ir'd3+. 27 'ii'b4 lLlf6
21 .•• �e7 28 'ii'd6+
22 .td3! 'ii' bS 28 'ii'f4! would have been imme
22 ...'ir'f6 cuts off the path of the diately decisive, as it attacks the
king's retreat after 23 'iVb4+ [edi knight and simultaneously threat
tor's note: for example 23 .. . �d8 ens 29 tLlxe6+ and 29 Wfb8+.
24 'ii'xb7 'ifxf3 !? 25 'ir'a8+ (25 28 ••• l:.d7
gxf3?? .th3+ 26 �g1 l:.xe 1 + 27 29 'ii'f4 tLlg8 (D)
.tfl l:.xfl#) 25 ... �e7 26 'ifxe8+ If 29 ... 'ir'g7, then 30 l:.xe6!.
l:.xe8 27 gxf3]. 3 0 .tg4! �c8
23 h4! (D) Taking on g4 leads to an elegant
It is almost as if White is con mate: 30 ....txg4 3 1 'ifb8#.
structing a study on the theme of 31 .txe6 fxe6
domination. 32 l:.cl+ �d8
23 •.. �d8 33 tLlxe6+ �e7
Karpov - Malaniuk 131
3 i.g2 g6
4 c4 i.g7
5 .!Df3 d6
6 0-0 0-0
7 .!Dc3 'ife8 (D)
w
34 'ii'xf8+ 'ii'xf8
35 .!Dxf8 �xf8
36 .Uh3 .!De7
37 h5 �g7
38 h6+ �6
39 llf3+ �e6 w
40 .Ue1+ �d6 One of the most topical posi
41 .Uf6+ �c7 tions in the Leningrad Variation of
42 g4 .!Dc6 the Dutch Defence. Black is in
43 .Ue8 d4 tending ...e5, creating a flexible
1-0 pawn centre. e8 is the most suitable
square for the queen, as from there
Game 32 she is influencing the centre, and
Karpov - Malaniuk might come in useful on the king
USSR Ch (Moscow) 1988 side.
Dutch Defence 8 b3!
The appearance of the bishop on
Malaniuk is one of the outstanding a3 could essentially hinder Black's
experts in this opening, and there plans. Other lines are less danger
fore I prepared particularly thor ous for him.
oughly for our game in this event. 8 ... .!Da6
As a result I managed to come up An immediate march by the e
with an interesting novelty. pawn, 8 ... e5, has not yet been fully
1 d4 f5 investigated, but it is evidently not
2 g3 .!Df6 without danger for Black.
132 USSR Ch (Moscow) 1988
29 l:l.xe5 dxe5
30 'iVxeS rM7
31 d6 j_f5
32 c5
Black is in total zugzwang.
32 h5
33 g4 hxg4
34 hxg4 j_d3
If 34 ...j_xg4, then 35 'iVf6+ 'it>e8
36 'ii'xg6+ and 37 'ifxg4 is deci-
sive.
35 j_d5+! 1-0
An effective final blow ended
the game; taking the bishop is im
possible in view of 35 . . . cxd5 36
'iVxd5+ 'it>e8 37 'it'e6+ mating.
Game 33
Karpov - Hjartarson
Tilburg 1988
QGD, Slav Defence
Game 34
Hjartarson - Karpov
Seattle Ct (3) 1989
Spanish, Zaitsev
29 h6!
By stealing along the edge of the 1 e4 e5
board, this pawn undermines the 2 lLlf3 llJc6
black kingside. 3 �b5 a6
29 ••• lbg2 4 �a4 lLlf6
Attempting to preserve a dy 5 0-0 �e7
namic pawn chain by means of 6 :et b5
29 . . . g6 is unsuccessful on account 7 �b3 d6
of the weakness on h7: 30 .:thd 1 ! 8 c3 0-0
.Ugd4 (30 ....:txd 1 3 1 .Uxd 1 .:th4 32 9 h3 lieS
�b5 .Uxh6 33 l:td7+ �b8 34 �a6 10 d4 �b7
with decisive threats) 3 1 lixd4 1 1 lLlbd2 �f8
lixd4 32 �g8 ! lid? 33 �c2 and 12 a3
Black will be put in zugzwang. As well as 1 2 a4, this restrained
30 hxg7 l:txg7 pawn advance is also seen fre-
31 lien .Ud6 quently. White guards b4 from in-
32 lih6 vasion by the black knight, and
Black's fortress has been de then prepares to develop the initia-
stroyed, and this sums up the result tive on the queenside.
of the opening. 12 ... h6
Hjartarson - Karpov 137
13 ..tc2 liJb8
14 b4 liJbd7
15 ..tb2 aS! ? (D)
A move I had prepared specifi
cally for this match; 1 5 ...g6 is a
more widely used continuation.
- ... -
. jilli
... • ... d
-
·
�
� - -
... �· ·
�
� .t. �
.
. � - � - �
� ·d
- · � - -� . 20 bxc5
B
�-/� �
� fj -
�- · d
�
• 20 lbxb5 immediately is much
� ·� � u . 1":\ . fj
.t.z.J. more precise: 20 ...cxb4 21 axb4
� ... /�
'. . �
� --'<% � -�� 0A
�
- : � l:lxal 22 ..txal d5 23 exd5 (23 e5
••
�
R �. m
·· · 7• • 1'1
..txb4 leads to an approximately
� . � � equal position) 23 ...l:lxel+ 24 1Wxel
w liJxd5 25 'lie4 liJ7f6 26 ..txf6
16 ..td3 c6 liJxf6 27 "ilxb7 "ilxd3 with a sharp
17 liJb3 axb4 game, in which Black has enough
18 cxb4 compensation for the pawn.
1 8 axb4 liJb6 1 9 liJa5 "ilc7 20 dxc5
would have led to equality. 21 liJxb5 ltJxe4!
18 ••. exd4 22 'lic2?!
19 liJfxd4 (D) This is a serious error. White had
Now Black seizes the initiative. to play 22 "ilf3, when the following
Things are also fine for him after variations are possible:
1 9 ..txd4 c5 20 ..txf6 liJxf6 2 1 a) 22 ...liJd6? fails to 23 l:lxe8
..txb5 l:be4. The correct continu 'tWxe8 24 liJxd6 ..txd6 25 "ilxb7
ation is 1 9 liJbxd4 c5 20 bxc5 dxc5 l:lb8 26 '1id5.
2 1 liJxb5 ltJxe4 22 liJe5 liJxe5 23 b) 22 ...'lib6 23 ..txe4 ..txe4 24
..txe4 with a roughly even game. l:lxe4 'lixb5 25 a4 is unclear.
19 ••• c5! c) 22...l:lb8 should be answered
A very promising pawn sacri by 23 ..tc4, with a complicated po
fice, with which Black takes aim at sition. Instead after 23 ..txe4?!
the enemy centre. ..txe4 24 l:lxe4 l:lxe4 25 'lixe4
138 Slcelleftea World Cup 1989
Game 35
Karpov P. Nikolic
-
s g3 0-0 17 'ird1
6 .tg2 i.xd2+ First and foremost in order to
7 Wxd2 d6 control the h5- and g4-squares.
8 ltlc3 e5 After the immediate 17 f4 Black
9 0-0 l:te8 could have chosen the exchange
10 e4 .tg4 1 7 . ..exf4 in order to gain counter
11 dS .txf3 play by means of ...ltlg4 (or ..ltlhs)
.
• •.t.• -··
m B mu · -
- · ··- m
••• • •
. .
- - -� - .
. � . .
-
d£• R�
� � U"'l.J.
8� • B 8_rJ
•
. .\lllr
- �� . �
.. �
-�
B w
17 ltleS! lDdS 2 1 ... f5 is no good because of 22
18 l:.g3 '6'f4 ! (or 22 'ii' h5 g4 23 ltlxg4); in
White does not hide his inten the event of 21 ... ltlf6 White has the
tions. strong 22 'itf3, or 22 '6'f4! ltlh5 23
18 ••• .i.fS 'ii'f3 ltlxg3 24 ltlxf7 'iixd4 25 l:.e7,
This move is a novelty. Theory etc.
gives the continuation 1 8 ....i.e6 19 22 hxgS!
'iid2 �h8 20 l:.e4 and 21 l:.h4, 22 Wh5 is insufficient: 22... fxe5
which is somewhat preferable for 23 hxg5 .i.g6 24 gxh6 �h7 and the
White. black king is safely covered.
19 'tWhS .i.h7 22 ••• hxgS
The paradoxical move 19 ......e6!?, In the event of 22 ... fxg5 the re
suggested by Taimanov, is interest ply 23 f4 simply becomes more
·
ing. valid.
20 'tWg4! gS (D) 23 f4
After 20 . . . g6 there is the very After the continuation 23 l:.h3
strong reply 21 h4. 20 . ..'it'f6 does fxe5 24 l:.xh7, White would have
not work because to 21 ltld7. After won beautifully in the event of
20 ....i.g6, the following variation 24...�xh7: 25 .i.c2+ �g7 (25 ...�h6
is possible: 2 1 ltlxg6 'tWxg6 22 26 'iWh3+ �g7 27 'iWh7+ �f6 28
'iixg6 fxg6 23 l:.xg6 l:.ae8 24 l:.xe8 dxe5+ Wxe5 29 'iWh6+) 26 'fi'xg5+
l:.xe8 25 �1 l:.e4 26 l:.d6 l:.xd4 27 cltf7 27 dxe5 We6 (27 ...'6'b4 28
l:.d8+ with a big advantage to e6+ �e8 29 .i.g6+; 27 .. .'ile7 28
White. 'iWg6#; 27 ....:tae8 28 'iWh5+) 28
21 h4 (6 .i.f5 . However, Black can hold on
144 London Ct (8) 1989
.
- - ... �
. - .
.
• • • • ••
. . •:•
. · · �'� .
• • • •
• • • •
80 • .8•
.
. . �
�� �
. �
w
After 30....:d8 White has the de-
cisive 3 1 .:d 1 .
31 .:e3 l:.c2
32 .:d7+ �g6
33 .:xb7 .:e8
The rook ending is hopeless for
Black, but resigning the game was
tantamount to resigning the match
too, and so Yusupov played on,
hoping for a miracle. If 33 ...1:.ff2,
then 34 .:g3+ �h6 35 .:bs ..th7 36 Game 37
.:b4 is decisive. Karpov - Timman
34 a3 d4 Kuala Lumpur Ct (4) 1990
35 .:d3 .:xe5 Griinfeld Defence
36 .:xd4 .:g5 (D)
If 36 ....:ee2, then 37 .:g4+ �f5 1 d4 lLlf6
38 .:gg7. 2 c4 g6
37 .:d6+ �h5 3 lLlf3 j.g7
38 .:h7+ �g4 4 g3 c6
39 .:d4+ �5 s j.gl d5
40 .:d5+ �g6 6 cxd5 cxd5
41 .:g7+ �xg7 7 lLlc3 0-0
42 .:Xg5+ �f6 8 lLleS e6
43 .:b5 a6 9 0-0 lLlfd7
44 .:b6+ �e7 10 f4 lLlc6
146 Kuala Lumpur Ct (4) 1990
A �
Q � " �� -
�- U �
�
� w
�
" V
�- �
�-
r.... � rt�
��� -
�"l.JU
- � i. " 26 Wet
�
"� V
- -� m
' �\lllr � m �
Obviously, the continuation 26
�
� �- � � �-: � .i..xd5+ �h8 27 l:tfcl exd4 28 e4 !
is stronger, returning the pawn;
w the inevitable break e4-e5 would
20 g4 'ile7 have guaranteed White a serious
21 'ild2 .:tac8 advantage. Killing two birds with
22 ltle2 l:tc7 one stone (attempting to control
23 l:tc5 ltlc8 the c- and f-files simultaneously)
24 f5 g5 does not work, and Black, by liqui
White is very slightly better af dating the danger in the centre,
ter the continuation 24.. . ltld6 25 achieves a playable game, which
lLldf4 gxf5 26 gxf5 ltlxf5 27 ltlxd5 for a long time will be typified by
exd5 28 l:txf5. manoeuvring. However, thanks to
25 ltlg3 such tardy action the game quickly
25 fxe6 'i!he6 26 ltlc3 ltl8e7 27 moved into an ending filled with
e4 is not so clear. fascinating events.
Karpov - Timman 147
26 b6 33 liJdl l:.c6
27 l:.c2 e4 34 l:.xc6 l:.xc6
28 liJf2 liJd6 35 l:.xc6 'ifxc6
29 'ii'd2 l:.dc8 36 l2Jc3 �f8
30 l:Ucl aS 37 �f2 �e7
31 �n 38 �el �f8
If it did not deprive the white 39 �dl 'ii'c8
queen of mobility, this move would 40 �el �g7
have been very perilous for Black. 41 l2Ja2 l2Jxa2
Attempts to rebuild with 3 1 'ii'c3 42 'ifxa2 'ii'c7
are repulsed by means of 3 1 . .. �d7 43 �f2 �f8
and 32 ...l2Je8. 44 'W'b2 �e7
31••. li)b4 (D) 45 �e2 �d8
3 l . . .li)xd4 !? suggested itself, 46 �el �c8
but after 32 exd4 e3 33 'ii'd3 exf2+ 47 �d2 �b7
34 �xf2 l:.xc2+ 35 l:.xc2 l2Je4+ 36 48 'ii'c l 'W'e7
�f3 a draw is not far away (swap 49 �el �d7
ping the good c6-knight for the bad so � l2Je8 (D)
one on f2 is not favourable for
Black). Timman's position in the
match obliged him to maintain the
• • •• •
pressure on the board in the hope ···i.- • .
of seizing the initiative from me. -� . . � - .�
�
- ·
-·· -8--
8.�� Ri-
u ·8-·
.8.
d d � u �
�
•
d •.li.m
d � � u
.
. �
- .
. .
.
w
51 'W'hl !
I did not want to agree to a draw,
so I decided to bring about some
confusing complications, by leav
ing the queenside to its fate, while
actively occupying the kingside.
148 Kuala Lumpur Ct (4) 1990
� i. •
..ti>f4 winning.
�
-
� " •
� . 64 ..ti>f4 (D)
B Here is a striking variation after
Here Timman sealed a move he 64 g5, ending in a problem-like
had been thinking about for almost mate: 64 ...lbxe3 65 <itf4 lbc2 66
half an hour. Black's activity on the <j;e5 lbb4 67 <ifi>f6 <iti>d6 68 g6 lbc6
queenside puzzled many commen 69 i.b5 e3 70 g7 l1Je7 7 1 i.xe8 ! e2
tators, and the following day there 72 lbxe2 a2 73 lbc3 ! lbg8+ 74 <j;f7
were even headlines in the local lbh6+ 75 <ifi>f8 a1 'iW 76 lbb5#.
Karpov - Timman 149
. .... . .
�
-
�� - � ·
� -
� .
• • • •
�
-
/'·�
� - ·
�·-
� ���
� �
�
- � � ... -
��-
�
" -� ·
�
�
• • •o •
B
.... . . .
65 �g5 � .JL. •
After 65 ltJa2, the long variation w
65 . . . �d6 66 �g5 �e7 67 �h6 69 .i.b3 tiJcl
�f6 68 tiJb4 .i.f7 69 .i.fl ! tiJd I 70 After 69 ...'it>d6 70 cJ;f6 Black's
g5+ �e7 7 1 �g7 ltJxe3 72 f6+ king is kept away and (he white g
�e6 73 .i.h3+ tiJf5+ 74 .i.xf5+ pawn is irrepressible.
�xf5 75 rJifxf7 e3 76 'iii>g7 ! e2 77 70 .i.xd5 'iii>d6
ltJc2 a2 78 f7 e l 'ii' 79 liJxel al 'i!l 71 .i.c4 .i.b5
80 f8._,+ �xg5 8 1 'iWf2 promises There is also the interesting pos
White good prospects, but after the sibility of 7 l ...b3 72 .i.xb3 liJxb3
correct 65 ... b5 ! nothing appropri 73 liJxb3 'it>d5 74 tiJal 'it>c4 75 cJ;f6
ate can be found. 'it>c3 76 cJ;e7 cJ;b2 77 'it>xe8 'it>xal
65••• tiJd3 78 f6 cJ;b2 79 f7 al 'iW 80 f8'iW
It is too late for 65 ...ltJc4 66 �f6 'ifa4+ 8 1 'it>f7 'i!i'b3+. This position
ltJxe3 67 cJ;e7. arose in our analysis, when we
66 tiJb3!? a2 considered that 82 'it>f6 Wxe3 83
150 Brussels Ct (4) 1991
would have given chances for Only now does Timman make
victory, but in our post-mortem, the decisive error. 73 ...b3! 74 .i.xb3
Timman demonstrated the more lbxb3 75 lbxb3 .i.c4 76 lbal �f6 !
striking continuation 82 d5 ! 'ii'xe3 would have led to a draw.
83 'ifb4+ �c2 84 g5 'iff3+ 85 �g7 74 .i.e6 i.d7
e3 86 'ifc4+ �b2 87 'ifb5+ �c 1 88 75 g5 b3
d6 e2 89 'ifc5+ �b2 90 'ffb6+ �c 1 76 g6 1-0
9 1 d7 el'if 92 d8'ff 'ffec3+ 93 If 76 ...i.xe6, then 77 fxe6 b2 78
'ifdf6, etc. g7+ �g8 (78 ...�e7 79 g811f bxal'ff
72 i.g8 �e7 80 'it'f7+ �d6 8 1 'ifd7#) 79 e7 �
73 �h6 (D) 80 g8'it'+ �xe7 8 1 'it'g5+ �f7 82
In the event of 73 f6+ �f8 74 'ifd5+ and after two more checks
.i.e6 b3 (74 ... .i.a4 leads to the White takes the b2-pawn.
same position as in the game, but
74 . . . .i.d7 ! will save Black, and if Game 38
75 i.xd7? then 75 ...b3 76 �f5 b2 Karpov - Anand
77 g5 lbe2! 78 g6 lbg3+ 79 �g5 Brussels Ct (4) 1991
lbh5 ! 80 �xh5 bxal'ff 8 1 g7+ � QGD, Semi-Slav
82 .i.e8+ �xf6 and Black unex
pectedly takes the initiative) 75 1 d4 d5
.i.xb3 lbxb3 76 lbxb3 i.c4 77 lba1 2 c4 c6
�f7 78 �f5 White has the upper 3 lbf3 lbf6
hand. 4 lbc3 e6
5 e3 lbbd7
6 'ifc2
After 6 .i.d3 dxc4 7 i.xc4 b5 8
i.d3 the classical variation of the
Meran arises. Moving the queen
forward to c2 is a standard Anti
Meran possibility.
6 000 .i.d6
7 .i.e2
Besides this move, White has a
choice between 7 i.d3, 7 g4 and 7
b3.
7 ••• 0-0
Karpov - Anand 151
• -
d � .: d
·.i.�
�
Anand had probably only reck
oned on 23 bxc3, and then 23 ...b2!
8 24 i.xb2 .:l.xb2 (24....!Dxe4 25
I had this position on the board 1i'xe4 .:l.xb2 26 'ii'x e5) 25 'ii'xb2
during my preparation for the .!Dxe4 26 .:l.d2 .!Dxd2 27 'ii'xd2 e4
game. The attack by the black 28 lbel e3 ! would have led to a po
queenside pawns, which is inci sition with chances for both sides.
dentally forced, seems dangerous, After this bold knight move the en
but in fact these foot soldiers will emy pawn reaches White's second
soon be stopped. rank, but fortunately its career
17 ••• b4 comes to an end there.
18 .!Da4 23 ••• c2
1 8 lL!b5 •d8 ! 1 9 .!Dxd6 'ii'xa5 24 .:l.d3 (D)
20 .!Dxc4 'ii'c7 !? is not so clear. I chose this continuation, al
18 ... 'ii'd8 though I also examined 24 .:l.el
19 .:l.a7 b3!? .:l.e8 25 .!Dc6 'ii'b6 26 'ii'f3 which
After 1 9 . . . .!Db6 20 i.e3 b3 2 1 has roughly the same value.
'ii'b l it i s more difficult for Black to 24 ••• 1t'e8
create counterplay, and he does not Now 24 . .. .:1.e8 loses: 25 .!Dc6
have full compensation in the event 'ii'b6 26 e5 .!Dxd5 27 .:l.xd5 'ii'xc6
of 1 9 ... c3 20 bxc3 b3 2 1 'ii'e 2 lL!c5 28 \i'c4.
Karpov - Short 153
35 lbd5 li:)xcJ
-.t.- -·- 36 l:td3 li:)a2
� -
- � -�··· 37 lbb3 .i.c8
- - - -�
' !:'-'
38 .i.c4! g6
- - 8 - �· •
40
41
l:txcl
l:tcJ
li:)xcl
l:te8
- ·-:- - 8 42 lbcl l:te7
B • -'iVD
� 8.
'�
!:'-' .- !i!
43 .i.n -
10
a m -i.=
� + if? ' !i!
Game 39
B -
Karpov Short
25 li:)c6 l:tb6 Linares Ct (7) 1992
26 .i.e3 li:)xe4 QGD, Tartakower
27 .i.xcS li:)xc5
28 l:te3 'iVd7 1 d4 d5
29 'iVc4 l:txc6 2 c4 e6
29 ...'iVd6 30 l:tc3 li:)d7 3 1 l:txb3 3 ll)cJ li:)f6
is no better. 4 .i.g5 .i.e7
30 dxc6 'iVd1 5 e3 0-0
31 l:te1 'iVd6 6 ll)f3 h6
32 'iVcJ! 7 .i.h4 b6
White has a clear-cut plan to win The Tartakower Variation of the
a pawn: the queen makes way for Queen's Gambit, also known to
the bishop on c4, after which the Russian players as the Bondarev
threat of l:ta5 arises. sky-Makogonov Variation, is very
32 ••. 'iVd5 widely used in modem practice,
32 ... .tf5 33 'it'e5! 'iVxe5 34l:txe5, and moreover many grandmasters
etc., does not help. willingly play it as both Black and
33 c7! White. When the h-pawn attacks
33 l:ta5 would have been too the bishop, you always have some
hasty in view of3 3 ... li)e4! 34 l:txd5 hesitation: should you exchange on
li:)xc3 35 l:tc5 li:)a4 ! 36 l:tb5 li:)xb2 f6, or wait until ... b6 has been
37 l:txb3 li:)dl and Black unexpect played, and temporarily retreat the
edly wins. bishop to h4? In my encounters
33 .i.b7 with Kasparov, this bishop di
34 l:taS li:)e4 lemma arose 24 times (equivalent
154 Linares Ct (7) 1992
B
b i ) 1 3 ...'ifd6 I 4 a4 lLld7 1 5 a5
l:tad8 I 6 axb6 axb6 17 l:ta7 'iib 8
IS .:r.a2 b5 I9 lLle1 .i.e7 20 lLld3
.i.d6 2I g3 lLlb6 22 .i.f3 .i.c8 23
l:tfal .i.f5 24 lLlc5 lLlc4 25 lLle2
gave White a secure plus in P.Nik
olic-Short, Manila IZ 1 990.
b2) Black tried to improve with
1 3 . . . a5 in Karpov-Short, Amster
dam 1 99 1 . The game continued I4
w a3 lLld7 1 5 b5 c5 1 6 lLlxd5 .i.xd4
1 1 b4 cS ( 1 6 ... .i.xd5 I7 'ifxd5 lLle5 1 8 'iib3
By playing I l ...c6 Black could ltlxf3+ I9 .i.xf3 .:r.cs 20 .i.c6 is no
have avoided hanging pawns, but good). After 17 .:r.ad 1 lLle5 ! 1 8
Karpov - Short 155
20 �n (DJ
A modest novelty, which I had
prepared specially for the match. It
is more precise than 20 �f3, as in
the game Yusupov-Beliavsky, Lin
ares 1 988, where after 20 ... .:tab8
2 1 a4 .:tfd8 22 'iic 3 �c8 23 .:tbc 1
a6 24 'iixc6 'iixc6 25 .:txc6 axb5
26 axb5 the players agreed a draw.
w
26 .:te3! lbe6
26 .. ..:tc2? loses directly: 27 .:te8+
'iii>h7 28 'iixc2 1fxc2 29 �d3+.
27 .:te5 'ii'g4
28 �e2! 'ii'g6
29 g3 .:tbc8 (D)
. �. . . • •
B • • •••
20 .:tabS • ·41\·�-
:res
21 a4
22 .:tb3! �xb5
·�•·a •
23 axb5 lbd8 (D) • � -� IJ
. ·
� -
� ·
�
�
• • • B
� �
24 'ifa2!
Now the white rook on the third
�� �-
'ii' IJ
· � ��
/� u u �
rank prepares to move to one of
three important squares, a3, e3 or -
·
� ·
� -
� .:. ·
� -
� �
� �
f3. w
24 ••• .:tc7 30 �h5!
If 24 ....:tb7, then 25 .:te3, threat The d5-pawn is going nowhere,
ening 26 'ifxd5 and 26 .:te8+. The and so there is no point hurrying to
weak d5-pawn is causing Black a take it: 30 .:txd5? .:tc2; 30 jfxd5
lot of trouble. .:td8 3 1 'ji'e4 'ji'xe4 32 .:txe4 .:tcd7.
25 .:ta3! 'ji'fS 30 ••• 1i'f6
Karpov - Kamsky 157
Game 40
Karpov - Kamsky
Moscow Alekhine mem 1992
w Grtinfeld Defence
34 'ii'e4 l:tcc3
35 �g2 �g7 This game, played at the Moscow
36 d5 lbc5 super-tournament celebrating the
37 'ilt'd4 l:tc2 l OOth anniversary of Alekhine's
38 l:te8 (D) birth, went, as they say, wonder
fully well. A bitter struggle was
conducted over the whole board,
• . •:• • first on one flank, then the other,
� -
- �
�� ·.·�
- and in the end the black pieces be
•
mu -
. ."" -
�
� -
- ... came totally disorientated.
· � - � - •:n
-
1 d4 lbf6
2 c4 g6
-
�� - � . � -� � 3 lbfJ .tg7
- �
.
- �
-
�
- � � u 4 g3 c6
B .I Bi.B�· First White (by playing 4 g3 in
B
Supporting the knight with the
pawn is more logical than retreat B
ing it to f3. 15 .•• l:.ac8
10 •.• ltlc6 1 5 ...l:.fc8 was Kasparov's pref
1 0 ...f6 is also played quite fre erence in Karpov-Kasparov, Se
quently, but practice has shown ville Weh ( 1 ) 1987. This game went
that it has more disadvantages than 1 6 l:.cl .i.f8, when White could
advantages. have sacrificed a pawn with 17 g4
1 1 ..t.e3 ltlb6 l:.c4 18 f5 !?, achieving a serious in
Here as well 1 1 . . .f6 is possible itiative. Instead the game contin
(see game 37). ued 17 .i.f3 l:.c7 1 8 b3 l:.ac8 1 9
12 .i.f2 .i.d7 'ilfd2 ltlc6 20 'ilfb2 a6 2 1 .i.e2 'ilfe7
In Karpov-Kasparov, Seville 22 ltlb1 l£Jb4 23 ltlc3 ltlc6, ending
Wch (3) 1 987, Kasparov preferred soon in a draw by repetition.
Karpov - Kamsky 159
Now the black pieces are getting when White simply has an extra
under each other's feet. The knight pawn.
is blockading his own bishop, a 24 lbbl!
fact of which I instantly made use. Now the bishop has to desert the
However, 22 ...lbxd4 23 W'xd4 is important a3-f8 diagonal.
no better, since the b6-knight is 24•.. .tb2
hanging (23 ....tc5 24 Wf4 .txf2+ 25 'ii'd 2!
25 1fxf2). The white queen's shuttling is
23 'ii'h6 (D) causing Black more than a little
Stronger than 23 lbbl lbc2! 24 bother.
W'h6 ..tf8. 25•.. lbc2 (D)
The startling 25 ... a5 ?! can be re
futed by means of 26 a3 (the trap is
26 ti'xb2 l:r.c2 27 'ii'a3 l:r.xa2)
26 . ..l:r.c2 27 W'e l ii'bS (27 ...lbc6
28 l:r.d2 !) 28 axb4 .:.e2 29 it'xe2
"it'xe2 30 bxa5 (or 30 .:.d2 'ii'xg4 3 1
.:.xb2) 30...lbd7 3 1 .:.d2 .
........,
• • - • • a"i
·- ····
23 'iWe8
• ··8�·
�
�-��
u -
·�J � �
� -
�
•••
-�· . -
In the new situation 23 . ..lbc2
does not guarantee equality, as af
ter 24 lbe2 the black rooks have no
� ··· fli.B
entry square for their invasion.
23 .. .'ii'e7 is no better: 24 lbbl .tb2 attJa : a : =
25 ..th4 'ii'f8 26 'ii'd2 .:.c2 27 1Wel , w
and the black pieces are soundly Black has been demonstrating
tied up on the queenside, despite the maximum activity for the last
their superficial activity. 23 ...lbd3?! five moves. However, the threaten- ,
does not work either, in view of 24 ing-looking tandem of the bishop
lbxd5 ! lbxd5 25 .:.xd3 .tf8 26 on b2 and knight on c2 has to be
'ii'g 5 h6 27 fxe6 ! 'ii'xe6 28 .txd5, defended, and only reduces the
Karpov - Kamsky 161
· · - •••
•l •
� ···• • · mit
-·
• •••••
�
-·· �
�z . �-� � -
�
�� �u
�
- - u
�
� �tffi � "'
" �
o � • m •.t
-
--� �- �
� r• �
-
�
.l[). • .:. •�
B
38 b4 w
39 axb4 axb4 49 f6+! i.xf6
40 llxc2 llxc2 50 i.h6+! �h6
41 :n ftxfl 51 'ili'xf6 'ifcl
42 'ili'xfl .i.a3 52 g5+ �h5
Alas, after 42 ....i.c3 there is the 53 �gJ!
immediately decisive 43 f6 'iff8 44 Strangely, 53 �h3? would have
'ili'c2 i.el 45 'ifc7 'ili'e8 46 i.fl let victory slip from my grasp:
Kamsky - Karpov 163
for B lack to take on b2, altho1:1gh 20 f4 (in order to close the c l-h6
the position of the c-pawn is hardly diagonal from checks) 20...gxf4 2 1
better for White than in the game i.d4 seems dangerous, for exam
itself. ple (D):
.. . . . .
�
- •.t. . - ··� �
-
�
-
- � �·- -
-
� � -
� -
� - - � - mu ·
�
�
-
�
� - �
� - � -
�
Wl1
- u � i.�
·
� �
�
� � -� -
, �
� �
�,\&��-�"u
. � �:- � �- �
�
w
. B
14 'ii'xe5+ a) 21 ...ifd5 loses on account of
15 i.e3 b6 22 l:.e5 Wxg2 23 l:.xc5! bxc5 24
16 0-0-0 g5 ifxc5+ �e8 25 ..txf6.
17 'ii'a4 cS b) 2 l ...'it'c7 is also in White's
18 libel ..td7 favour: 22 i.eS ifc6 23 i.xf6+ (a
19 'ii'a3 l:.hd8 (D) rare combination arises after 23
..txf4 ! ? 'ilt'a4?? 24 ..td6+! �e8 25
l:.xe6+! fxe6 26 i.g6#) 23 ...�xf6
24 i.e4 Wa4.
c) 2 1 . . .'it'g5 ! 22 l:.e5 ifh4 ! res
cues Black, and 23 g3 does not
lead White to his goal in view of
23 . .. fxg3 24 hxg3 ifxg3 25 l:.xcS
bxcS 26 ifxcS+ �e8 27 i.xf6
iff4+.
20 ••• 'ii'c7
The black queen abandons the
centre of the board, and restrains
all approaches to the king.
20 g3 21 ..td4
166 Dortmund 1993
have the two bishops, I want to After 1 8 ... �d3, White should
open up the position ! Generally, play the simple 19 .txd3 cxd3 20
the idea of sacrificing a pawn with .:r.ad l , and Black loses both pawns
e4-e5 is not new, but I am not in the centre. 1 8 . . b4 is no better:
.
w
18 .te3!
Quietly finishing his develop
ment, as the e6-pawn is going no
where.
18 ... 0-0
1 70 Linares 1994
w
Now 3 l . . ..i.d7 does not work:
32 l:te7 .i.xh3 33 ll)hS, etc.
32 ll:)fS l:txd6
33 ll)xd6 .i.g6
34 .i.xg6 ll)xg6
35 ll)xc4 l:td8 (D) B
35 ...l:tf8 36 l:te4 ll)f4 37 h4 �h7 39 ••• ll)f8?
38 ll)eS is suicide. Of course, the defence 39 ... �f8
40 h5 ll)e7 4 1 l:te5 �e8 would
etc.
20 e3 1-0
This victory in the final round
was my ninth at Linares; it was
even more remarkable that I won
six straight games in the first six
w rounds ! With four draws (I man
18 b4! lt:Jxa4 aged to avoid being beaten) I
Beliavsky is confused, and this reached an inconceivable score,
is another consequence of White's 1 11 1 3 . And, incidentally, the sec
1 4th move. He should have played ond and third placed players, Kas
1 8 . . . lt:Jce4 1 9 .i.xf6 lt:Jxf6 (but not parov and Shirov, were 2112 points
1 9 . . ..i.xf6? 20 .i.xe4 .i.xd4 2 1 e3 behind me.
.i.xal 22 l:lxd6 l:lxd6 23 .i.xb7 Without false modesty I can say
with a big advantage) 20 .i.xb7 that in the whole of chess history,
l:lab8 2 1 'i'xa6 'ilt'xb4 22 lt:Jb5 . tournaments where on the one
White would have taken the upper hand, all the stars of the chess
hand, but the game would have world are gathered together, and on
continued. the other, the winner has demon
19 ...b3! (D) strated such notable superiority
Now it's all over: Black loses a over the remaining contestants, can
piece. be counted on your fingers.
Karpov - Beliavsky 1 75