Documente Academic
Documente Profesional
Documente Cultură
the Griinfeld
Contents
Preface
Seville Variation
11
41
52
4 Russian System
108
4 f4 System
135
Fianchetto System
1 52
Index of Variations
1 87
Preface
I have to admit that the title of this book has an air of sales-talk
about it, as similar titles generally do. Anyone trying to think up
a recipe for beating this or that opening will scarcely succeed, and
the Griinfeld Defence is no exception. But seriously: to chessplayers
whose repertoire includes the Griinfeld, the author imagines that
the present work will be of considerable interest and use. Hence
it is indeed quite possible that the book will help many readers
to score wins - with Black if they are adherents of the Griinfeld,
and with White if they are looking for a way to combat this
defence.
The book consists of a collection of games (or fragments of
games) which reflect the contemporary state of Griinfeld theory.
It is constructed around twenty-five paradigms of play by con
noisseurs of this opening (like many authors I have a weakness
for round figures), but the total number of examples is about ten
times higher! The notes to each of the twenty-five principal
games constitute a thorough discussion of the currently popular
variations. As a result, all the most fashionable systems occurring
in grandmaster practice in the last few years have found their way
into the book. The selected games are not arranged chronologically
but grouped according to themes. For this reason, the freshest
examples - those from the recently concluded duel for the world
chess crown, which incidentally are the most fully annotated - do
not form the culmination of the book, but are placed in the middle
of it . . . .
This book is written in the same format as the four-volume
work probably already familiar to the reader : The Open Game
(Semi-Open Game I Closed Openings I Semi-Closed Openings) in
Action. That is to say, the scores of all the principal games are
given in full, and in analysing the opening the reader will mostly
be able to study its relation to the middlegame or even the
8 Preface
Preface 9
many sources (Informator, books and magazines), and this is
reflected in the sheer quantity of references to games by masters
and grandmasters. In the majority of cases where a variation or
individual move is of major significance, its originator is mentioned.
In conclusion, I must thank Soviet Master Evgeny Gik, my co
author in many books, for his help in preparing the manuscript.
Anatoly Karpov
1 Seville Variation
Game No. 1
Kuzmin-Henkin
Moscow 1989
I
2
3
4
5
6
7
8
9
10
d4
c4
.!Llc3
cd
e4
be
c4
.!Llel
e3
0-0
.!Llf6
g6
d5
.!Llxd5
.!Llxc3
g7
c5
.!Llc6
0-0
g4
1
w
f3
.!Lla5 (1)
12 Seville Variation
But now Black can answer 1 2 .td3 with 1 2
cd 1 3 cd .te6,
and the exchange sacrifice (14 dS) is less dangerous to him, since
in several lines he has an important queen check on b6 (the details
can be found in any reference work on the Griinfeld). However,
in this situation White's exchange sacrifice is by no means forced.
We shall later acquaint ourselves with a recent game with 14 .l:l. c l
instead (Game No. 6 , Yusupov-Kasparov).
. . .
12
.txf7+
fg
..txfl
.l:l. xf7
.l:l.xfl +
Seville Variation 13
2
w
11V b6, and after 16 g1 11Ve6 17 11Vd3! I returned the pawn while
keeping all my positional trumps. The offshoots arising from 1 4
. . . c d 1 S c d will be examined i n d e tail i n Game No. S (apart from
1 S . . . "it"b6, the moves 1 S . . . "it"d7 and 1 S . . . eS have been seen).
The Seville match was not yet over when the new move 14
"it"d7 occurred in a game Chernin-Gavrikov, Lvov 1 987. That
game proceeded : 15 de J: f8 + ( 1 S . .. "it"xg4 16 f4!) 16 g 1 "it" xg4
1 7 f4 ( 1 7 "it"d3 c6 1 8 h3 is not bad either) 1 7 . . . "it" xd 1 + (but
not 1 7 . . . J: xf4? 1 8 .txf4 "it" xf4, on account of 19 "it"d8 +) 1 8 : xd 1
.txc3 1 9 d S (White gains nothing from 1 9 e6 : c8 2 0 J: d7
f7 2 1 gS+ e8 22 J: d3 .tb4, or in this line 2 1 .tgS? .tf6!
22 .txf6 xe6) 1 9 . . . .tf6 20 .th6 (the advantage is with White,
but Black manages to hold on) 20 . . . J: e8 (but not 20 . . . : f7 2 1
xf6 + : xf6 22 e S :rs 2 3 g4, and White wins) 2 1 xf6 + ef 22
J: d7 : xe4 23 J: g7 + h8 24 : c7 g8 2S J: g7+ (2S .td2 c6
26 : xb7 J: e7) 2S . . . h8 26 : c7 g8 27 J: g7 + h8 28 l: c7,
draw. Gutman suggests 20 g4! : e8 21 J: b 1 a6 22 f2, followed
by e2 and .td2, with the initiative.
In answer to 14 . . . "it"d7, a more logical move seems to be 15
g5 (3), as played in Karpov-Gavrikov, European Speed Chess
Championship, Spain 1 988.
Despite the 'non-serious' nature of the contest, the game is of
considerable interest : 15
"it"e6 16 eS! "it"c4 ! 17 g1 J: d8 18 "it" e 1
c6 1 9 .tf2 a 6 (wouldn't 1 9 . . . b S ! ? have been better?) 2 0 a4
aS 2 1 h4 .tf8 22 de! b3 23 .l:!.b l ! (more precise than 23 J: d 1
: xd 1 24 "it"xd 1 lbxcS) 2 3 . . . e6 (but now 23 . . . xeS loses t o 24
: b4 "it"dS 2S J: d4) 24 c6! be 2S d4 lbxd4 26 cd "it" xa4 27 "it"c3
(White has an obvious endgame advantage) 27 . . . J: c8 28 hS! gh
29 "it"h3 J: e8 30 "it"xhS J: e7 (30 . . 'lf c2 3 1 J: b7) 3 1 g6 11V c2 32
ooo
14 Seville Variation
3
B
Seville Variation 15
15
e5
16 Seville Variation
'ife2! : f8 1 8 'ifxc4+ 'Oxc4 1 9 e2, with the better ending for
White. One other possibility, 16 g1 was tried in Hansen
Fereec, Aosta 1 989. Black restored the material balance, but after
1 6 . . . 'ifxg4 1 7 'iVd3 'ife6 1 8 -*.g5! : f8 1 9 h3 cd 20 cd -*.xeS 2 1
de 'ifxe5 22 -*.c l ! 'ifxa1 2 3 "ird5 + e6 24 'ifxa5 : c8 2 5 'iVd2 b5
(doubtless a more accurate line was 25 . . . 'iVb 1 26 h2 'ifc2 27
1We3 "irxa2 28 'Oc3 "irb3 29 .i.d2, with a minimal plus for White)
26 a3, his position was fairly difficult.
,
4
w
16
.i.f2
In this case 16 'Of4 is weak, since the queen gets to e4. A game
Lichak-Asrian, Leningrad 1 990, went 16 . . . : f8 1 7 g1 "ire4 1 8
"irf3 "irxf3 1 9 gf .i.h6 2 0 10d5 -*.xe3 + 2 1 'Oxe3 cd 22 cd 'Oc6 23
: b 1 b6 24 g2 'Oxd4, and Black had an endgame advantage.
16
:rs
Seville Variation 17
With 17 g5, White achieves his principal aim of shutting off the
opposing bishop. Admittedly his king is dangerously placed
opposite the black rook, but this is just a temporary problem.
5
B
17
Wf7
"ife1
h6
In reply to 18
'if f5 Henkin gives these variations :
(a) 19 h4, and now :
(a1 ) 19
"ifg4 20 g1 h6 (or 20 . . . ltlc4 2 1 tt:Jg3! "if xh4 22 tt:Je4,
with a distinct plus) 2 1 gh .i.xh6 22 ltlg3! cd 23 cd "if xh4 24 tt:Jf5!
"ifg5 25 tt:Jxh6+ "if xh6 26 "if xa5, and wins.
(a2) 19
h6 20 gh .i. xh6 21 ltlg3, with a clear advantage.
(a3) 19
"if e4 20 g1 cd 2 1 cd ltlc4 22 J:t c l ! (but not 22 g3?
.i.xe5 23 de tt:Jxe5) 22 . . . ltle3 23 ltlf4!, and if 23 . . . l:t xf4 24 .i.xe3
,J xh4 25 J:tc8 + , White has a considerable initiative.
(b) Another playable line is 19 ltlg3 "if xg5 20 tt:Je4 'iff4 21 g1
cd 22 cd tt:Jc6 23 .1: b 1 ! with the advantage.
In a game Kuzmin-Malishauskas, USSR 1 989, Black played a
new move, 1 8
ltlc4 ( 1 8 . . . ltlc6 is also possible). There followed
19 tt:Jg3 tt:Jb2 20 tt:Je4 tt:Jd3 2 1 "ife3 (2 1 "ife2 tt:Jf4 !) 2 1 . . . tt:Jxf2 22
...
18 Seville Variation
gh
J.xh6
g1
lt:lc4
An inadequate alternative is 20 . . . J.d2 2 1 "it'xd2 "it'xf2+ 22
h 1 g7 23 "it'd3 lt:lc6 24 lldl .
21
lt:l g3!
J.dl
23
If 23
cd, then 24 e6! is immediately decisive. On 23
lt:ld2?,
White has 24 .z:td 1 cd (24 . . . "it'xa2 25 lt:le4!, or 24 . . . "it'f4 25 de!)
25 llxd2 J.xd2 26 "it'xd2 d3 27 -*.xa7.
24
25
de!
lle1
..txeS
"it'dS
lld1
ll xf2
30
llxd5
lt:lxe2
c6!
: xeS!
1 -0
llxel
f7
b6
Seville Variation 19
Game No. 2
Karpov-Kasparov
Belfort 1988
1 d4 ll:lf6 2 c4 g6 3 ll:lc3 d5 4 cd ll:lxd5 5 e4 ll:lxc3 6 be .tg7 7
.tc4 c5 8 ll:le2 ll:l c6 9 .te3 0-0 10 0-0 i.g4 1 1 f3 ll:la5 12
.txf7 + .:t xf7 13 fg .:t xfl+ 14 xfl 'it'd6 15 eS 'it'dS
16
i.f2
.:t d 8
Mter the 7th game in Seville, in which Kasparov played this
move for the first time, the rook manoeuvre to d8 became standard
practice in this position. On d1 the queen was quite conveniently
placed, but now the threat of 1 7 . f .txeS compels it to leave its
post. White gains nothing from 17 ltlf4 'ifc4+ ( 1 7 . . . 'iff7 1 8 ll:lh3!)
18 11Fd3 'ifxd3+ 19 ll:lxd3 cd 20 cd ll:lc6 21 ltlcS ll:lxd4 22 ll:lxb7
.:tdS 23 .:t e l : xeS 24 : xeS .txeS 2S ll:ld8 a6, with equality.
There are three ways for the queen to vacate the d-file : along
the d l -a4 diagonal (to a4 or c2), or with 17 'ifel. We examine 17
'ifa4 in the present game, and 17 'ifcl in Game No. 3. But first,
let us recall how the Seville game proceeded : 17 'ife1 'ife4 1 8 gS
'iffS! 19 h4 ltlc4 (he should have put pressure on the centre with
19 . . . ll:lc6 20 gl 'ife4, leading to sharp play) 20 g1 'ifg4 (20
. . . bS was more precise, as after White's next move, the black
knight will feel uncomfortable) 21 a4 h6 22 : a2! hg 23 'it'bl !
(White thus succeeds in exploiting the weakness of the a2-g8
diagonal) 23 . . . gh 24 'it'b3 'ife6 2S ll:lf4 'iff7 26 ll:lxg6 'ifxg6 (I
shall refrain from further comment until we reach one particular
critical position; this book is not the place for exhaustive notes
on the Seville games, they are no doubt well known to the reader
already) 27 'ifxc4+ h8 28 .:t b2! cd 29 cd 'ifg4 30 'iff7! .J:I. xd4
3 1 .txd4 'ifxd4+ 32 .:t f2 'ifxe5 33 : fs 'ife l + 34 :n 'ifeS 3S
hl? (although we are quite a long way out of the opening, it is
appropriate to take stock of the situation. With 3S 'iff4!, White
should come out on top : 35 . . . 'ifxf4 36 .:txf4 i.f6 37 : c4 g7
38 .:t c7 b6 39 fl . with a won ending) 3S . . . b6 36 'iff4 'ifhS (a
withdrawal that was impossible a move earlier. If now 37 'it'b8 +
h7 38 'ifxa7, the white queen is far from the scene of action,
and after 38 . . . 'ife2 Black has enough initiative to draw. But
White has another way to exploit the awkward position of the
black queen on the rook's file) 37 'iffS? (at this point, 37 .:t f3! i.f6
38 1Wb8 + g7 39 'ifxa7 1Wc5 40 1Wa6 was decisive) 37 . . . 1We2 38
.:t e l ? (here was White's third successive opportunity to win this
.
20 Seville Variation
game. He should have checked first - 38 "it'c8 + h7 - and only
then played 39 :t e l , threatening 'iltc2 + . If 39 . . . 'iit h 5, then 40
'iltc2+ h8 41 'iitd l ! etc.). The game lasted another 40 moves,
but all my efforts to overcome my opponent were in vain. Although
Black was on the brink of defeat more than once in this game, we
have seen that the opening was not to blame. After 1 7 'ilt e l , Black
had the means to obtain a perfectly reasonable position.
17
'ilta4
(6)
6
B
b6
Seville Variation 21
After 17
l:f8 1 8 <i>g1 'llf7 1 9 .*.h4! ll:lc4 20 'ilb3, the bishop
is again unable to break out and free itself: 20 . . . .*.h6 21 g5!
.i.xg5 22 .*.xg5 'ilf2 + 23 '>Ph 1 'ilxe2 24 h3, with a substantial
advantage to White.
18 'ilcl
There is nothing more for the queen to do on a4.
18
l:f8
7
w
22 Seville Variation
19
g1
1Wc4
The bishop cannot now break out onto the open board: 1 9 . . .
-th6 20 h4 1Wf7 2 1 .ltlg3, or 1 9 . . . .!t!c4 20 h4.
20
1Wd2!
1We6
h3
ll'gS! (8)
.!tlc4
b6
Seville Variation 23
8
B
24
.tg3
g5
. .
. . .
26
27
.tf2
ll:lg3
b5
.1:1. 7
24 Seville Variation
28
29
1Wg6
b4
f8
o!Lle4
30
xf2
'tiffS+
1Wc8+
1W xc5
<i>g1
o!Llg3
o!Llf5
.l:l. c 1
1-0
be
g8
<i>h7
1Wf7+
c2
.i.f8
g8
Rotterdam 1989
1 d4 o!Llf6 2 c4 g6 3 o!Llc3 d5 4 cd o!Llxd5 5 e4 o!Llxc3 6 be J.. g7 7
J.. c4 c5 8 o!Lle2 o!Llc6 9 J.. e3 0-0 10 o-o J.. g4 1 1 f3 o!Lla5 1 2
J.. xf7 + .l:l.xf7 13 fg l:Xfl+ 14 xfl 1Wd6 1 5 e 5 1Wd5 16 J.. f2
.l:l. d8
In the fifth game in Seville, Kasparov chose 1 6 . . . .l:l. f8 (see
Game No. 1 ). As we know, the rook move to d8 was first employed
in the 7th game of the match.
17
1Wc2
In the first game with this line, I chose 17 'tlfe l (see Game No.
Seville Variation 25
2); in Belfort 1 988, I preferred 1 7 'ilt'a4 (Game No. 2).
17
'ilt'c4
In Griinberg-Ilincic, Prague 1 989, an equal game resulted from
1 7 . . . .l:t c8 1 8 illf4 'ilt'f7 1 9 'ilt'e4 J.h6 20 g3 .l:t f8 2 1 >g2 J.xf4
22 gf 'ilt'xf4 23 'ilt'xf4 .l:t xf4 24 >g3 .l:t f8.
18
'it' b2
19
9
B
h4 (9)
26 Seville Variation
19
20
.1:1.8
g5!
'ilbl!
Seville Variation 27
A mistake would be 2 1 gh? lbc4 22 1i' c l 1Wf5 23 it'e l lbe3 + 24
g l lbc2.
21
22
23
it'el
gl
1i' e3
R.g7
23
10
w
24
lLig3!
White returns the pawn but seizes all the key squares with his
pieces. A familiar precept!
24
25
1t'xh4
lbe4
25
lLixf2
l:tdl !
cd
d3
28 eville Variation
29
30
31
32
33
If 33
. .
tl:lxd3
'lrf3!
e6
tl:lf4
tl:ld5!
'lra4
'lraS
tl:ld8
.i.eS
'IreS+
h1
1-0
Moscow 1989
1 d4 tl:lf6 2 c4 g6 3 tl:lc3 d5 4 cd tl:lxdS 5 e4 tl:lxc3 6 be .i.g7 7
.i.c4 c5 8 tl:le2 tl:lc6 9 .i.e3 0-0 10 0-0 .i.g4 11 f3 tl:laS 12
.t.xf7+ .:.x7 13 fg J: xfl + 1 4 xfl 'ir d6
15
g1
Seville Variation 29
The game proceeded : 1 6 'ifxa5 l:l. f8 + 1 7 e1 (but not 1 7 J.. f2?
on account of 1 7 . . . 'if h4! 1 8 g3 'ii' h 1 + 1 9 g1 'ifh2, and wins)
1 7 . . . 'ii' h 1 + 1 8 g 1 ! (on 1 8 .!.g1 ?!, Black has the very strong 1 8
. . . ..th6! 1 9 'ifxc5 'if xg2, threatening 20 . . . 'iffl mate) 1 8 . . . 'ifxg2
(but here, 1 8 . . . J.. h 6 comes too late : 1 9 d2 l:l. f2 + 20 d3
c4+ 21 .;;.xc4 J.. x e3 22 'ii' d 5+ .;;.g 7 23 l:l. e 1 , with the better
endgame chances for White; if he prefers, he can give perpetual
check by 23 f3 'ifxa 1 24 'ife4+) 19 'ii' b5 ..th6 (after 19 . . . cd 20
cd 'if xe4 2 1 'ife2 and 22 l:l. d 1 , White consolidates; but now, a
highly unusual distribution of forces comes into being) 20 'ife2
'ifg3+ 21 d2 l:l. f2 22 J.. x h6 l:l. xe2 + 23 xe2 'ifh3! 24 J.. e 3!
'ifxg4 25 l:l.fl ! h6! (a very strong move according to Henkin, who
gives the following variations : 25 . . . cd 26 J.. h 6! de+ 27 c1
'ifc8 28 d4, threatening e6 and l:l.f8 +; 25 . . . h5 26 f4,
followed by l:l. g 1 ; or 25 . . . g7 26 f4 g5 27 l:l. g 1 'ifh4 28
l:l. xg5 + ! . Now the game heads towards a draw) 26 J.. x h6 'ifxe4
27 l:l. f8 + h7 28 .!.g5 'ife6 29 a4 cd 30 cd 'ifa2+ 3 1 .;;.el 'ifxa4
32 lH7 + .;;.g 8 33 .J:I. xe7 'ii' b 4+ 34 f2 a5 35 J.. f6 'ii' b 6! 36 J.. e 5
a4 37 f4 'ii' b 2+ (Black loses after 37 . . . f8 38 J.. g 7+ !, or 37
. . . a3 38 e6!) 38 g1 ! 'ifc1 + 39 h2 'ii' b 2+ 40 g1 'ii' b 1 + 41
.;;.h 2 'ii' b2 + t-t.
So the extravagant 1 5 'ifa4 is not dangerous to Black. It must
be acknowledged that 1 5 e5, which we have already examined, is
the most precise; White can move his king if the need arises.
Nonetheless, 1 5 .;;.g 1 should also be studied.
15
'ife6
In Polajzer-Anka, Dortmund 1 988, Black played 15
l:l.d8.
It isn't clear that the rook on d8 is useful to Black after 16 'ii' d 3
'ife6 1 7 g5!. However, White replied 16 'if a4 'ifa6 1 7 .J:I.e 1 'ii' d 3 1 8
J.. f 2 c4 1 9 'ii' b5 l:l.f8 20 'ii' b l 'ii' d2 2 1 'ii' b3 b 5 2 2 'ii' x b5 e3 23
h3, and now, in Gutman's view, 23 . . . J.. h 6! would have given
Black a powerful initiative.
16 'ii' d3
'ifc4
So far, the play has followed the 1 1 th game in Seville, and the
text move is better than the capture on g4, which occurred (with
a transposition of moves and the insertion of the pawn exchange
on d4) in the 9th match game. In Chernin-Malishauskas, Lvov
1987 (played while the Seville match was still in progress), Black
didn't hurry to exchange queens, preferring 16
cd 17 cd l:l. d8.
After 18 g5 c4 1 9 J.. f2 b5 20 a4 (20 h3 was worth considering)
.
30 Seville Variation
Seville Variation 31
27 h3 :la4 28 a3. White has an extra pawn, which he eventually
exploited to win (though it took him 30 moves!).
Seirawan-Hort, Lugano 1 988, went 18
cd 19 cd eS 20 :lc1
(20 d5 h6 2 1 .i.cl d6 22 g3 : c8 23 .i.e3 :lc3 24 <i>f2 .i.f6!
2S a4 .i.gS led to equality in Dlugy-Nikoloff, Toronto 1 989;
Black can also play the immediate 20 . . . tLld6 21 g3 : c8 22
.i.e3, with adequate counterplay) 20 . . . bS 2 1 de .i.xeS 22 :ld 1
: c8 (22 . . . : e8 is more accurate and gives equality} 23 .i.f4 !
.i. g 7 2 4 :ldS a 6 2S f2 :le8 2 6 <i>f3 <i>f7 27 h 4 e S + 2 8 .i.xeS
: xeS 29 : d3 b4 30 f4. Here again, White has an extra pawn,
but this time Black managed to hold out.
Interestingly, his loss to Seirawan made such a strong impression
on Lputian that he took the first opportunity to play this variation
with White. In Lputian-Hansen, Dortmund 1 988, Black played
(from diagram 1 2} the immediate 18
e5. There followed 19 d5
b5 ( 1 9 . . . h6 20 .i.cl is sounder for Black} 20 : b1 :b8 21 <i>f2
aS 22 thc 1 h6 (22 . . . a3 23 :lb3 b4 24 cb cb 25 d6) 23 .i.e3
xe3 (in Lputian's view, 23 . . . tLld6 24 <i>f3 :lf8 + 25 <i>e2 c4 26
<i>e1 xe4 27 :lxbS xc3 28 :lxaS :ld8 would have given
equality; but not 23 . . . :lf8 + 24 <i>e2 xe3 2S xe3 :lfl 26
:la 1 ! and 27 b3!} 24 <i>xe3 c4 2S e2 .i.f8 26 tZ'lg1 .i.cS + 27
<i>e2 .i.xg 1 28 :lxg 1 <i>f7 29 a3 rj;e7 (29 . . . b4 would have
retained some saving chances} 30 :lb 1 , and White won the rook
ending. After the game, Lputian explained that at move 2 1 , the
correct course was 21 a4! b4 22 cb cb (or 22 . . . :lxb4 23 tZ'lc3
d6 24 :le 1 , preparing 2S .i.e7) 23 c1 .i.f8 24 b3 d6 2S
d2 b3 26 .i.e3 a6 27 <i>fl! :lb4 (28 . . . b2 29 <i>e2) 28 .i.c5
:lxa4 29 .i.xd6 .i.xd6 30 :lxb3, and White has the better chances.
.
cd
18
cd
e5
d5 (13)
d6
32 Seville Variation
13
B
Seville Variation 33
14
w
33 J:.f2+
A critical moment in the game. After the obvious 33 e2, Black
would scarcely be able to save himself: 33 . . . i.. b 6 34 a4 i.. x c5
35 ab J:. b8 36 d3! lhb5 (36 . . . a6 37 xc4 ab + 38 b3 i.. d6
39 J:. c6 e7 40 e2 J:. a8 41 c3 J:. a3 + 42 b2, or 40 . . b4
4 1 c1, with a won position) 37 xc4 J:.b7 38 lL!fl i..d 6 (38 . . .
a5 3 9 d2, o r 38 . . . i.. b4 3 9 e3) 39 d2, etc.
33 . . . gl 34 J:.f6 i..b6 35 J:.c6??
By retreating his bishop to f2, White would preserve the better
chances. Instead of this, I commit a fatal blunder. The sad fact is
that the possibility of the black knight jumping to the edge of the
board has escaped my attention.
35 . . . a5 36 i..xb6 xc6 37 i..cl J:.f8+ 38 e2 J:.fl 39
i..d6 J:.dl 40 i..c5 a5 41 !1 .:tel 42 i..d6 J:.c2+ 43 d3
J:.xa2 44 e3 fl 45 g4 c4 46 xe5+ xe5 47 i..xe5 b4
48 i..f6 b3 49 e5 J:.xg2 50 e6 + f8! 0-1.
Now at last we return to the game Naumkin-Neverov, which,
as it happens, was decided very quickly.
i..h 6
21
g3
.
22
i..cS
h4
i..f 2
e2
J:.dl
J:. c8!
i..f4
i.. d2
aS!
34 Seville Variation
27 gS bS 28 ltlg3 b4 29 lll fl J.c3 30 lll h 2 a4 31 d6 b3 32 ab ab
33 ltlg4 b2 34 d7 .IUS 35 :n <t;g7 36 ltlf6 lll d2 37 ltle8 + <t;f7
o-1
Manila 1990
1 d4 ltlf6 2 c4 g6 3 ltlc3 dS 4 cd lll xdS 5 e4 lll xc3 6 be J.g7 7
J.c4 cS 8 lll e2 lll c6 9 J.e3 o-o 10 0-0 J.g4 1 1 f3 ltlaS 12
J.xf7 + ll xf7 13 fg ll xfl+
14
xfl
cd
15
cd (15)
15
B
Seville Variation 35
15 . . . flb6 16 g 1
With the queen on b6, 1 6 h3 can be strongly answered by 1 6
. . . c4 1 7 ..t.f2 flb2! 1 8 gl l:t f8 1 9 l:t b l flxa2 20 l:t xb7 .th6!
with the initiative.
16 ... fle6
With the white king on g 1 , the line just mentioned no longer
gives Black anything : 16 . . . c4 1 7 .tf2 flb2 1 8 f4! l:t f8 1 9
d3.
17 fld3!
White returns the pawn while retaining all the positional trumps.
The position had first been seen (it arose by transposition) in
Alfeyevsky-Werner, corr. 1 984. That game went 17 g3 l:t d8 1 8
l:t c l ( 1 8 l:t b l .txd4 1 9 .txd4 c6 20 e2 xd4 2 1 xd4 flc4)
1 8 . . . flb6 19 e2 c6 20 l:t b l xd4! 2 1 l:t xb6 f3 + 22 f2
l:t xd l 23 l:t xb7 xh2 24 g4 g4+ 25 f3 h2 + , with a peaceful
conclusion by perpetual check.
1 7 . . . flxg4
After this, I succeed in obtaining a substantial plus. After 1 7 . . .
l:t d8 1 8 g5 c4 1 9 .tf2 b5, the play would transpose into
Chernin-Malishauskas which we have examined before (see page
3 1 ). (I should point out that the ideas and variations in Games
1 -5 have much in common, while sometimes being distinguished
by small nuances; I hope the reader will not get tangled up in the
'undergrowth'.) Instead of the capture on g4, a better idea is 1 7
. . . flc4, aiming to exchange o ff the strongest pieces. This idea was
employed in game 1 1 of the Seville match, with the difference that
the exchange on d4 took place later (see notes to Game No. 4).
White may avoid the queen exchange by retreating with 1 8 fld2
(exploiting the absence of the pawn from c3 and thus gaining a
tempo). After 1 8 . . . fla6 19 flc2 c4 ( 1 9 . . . flc4 is well answered
by 20 :t e l ! - emphasising once again that the removal of the c
pawns is favourable to White) 20 .tf2 .tf8 (20 . . . .th6 21 h4)
21 g5, and Black's position is devoid of counterplay.
18 l:t/1 l:tc8
The alternative 1 8 . . . l:t f8 is worse : 1 9 :t xf8 + .txf8 20 d5 b6
2 1 d4!
19 h3! fld7 20 d5 c4 21 ..t.d4
More precise than 2 1 d4 xe3 22 flxe3 l:t c4 23 e6 flxe6
24 de .td4 25 flxd4 l:t xd4 26 l:t f7 l:t xe4 27 l:t xe7 b5, with
equality; but 21 .txa7 b6 22 ..t.b8!? was worth considering.
36 Seville Variation
21 . e5
If 2 1 . . . lbe5, then 22 ..txe5 .txe5 23 lbd4 .txd4+ 24 -.xd4
b6 25 e5 -.c7 26 -.g4! is not bad for White.
22 de -.xe6 23 .t xg7 r;J;xg7 24 lb/4 -.d6 25 -.c3+ (16)
.
16
B
25 ... h6
Seville Variation 3 7
27 . . . g7 28 !i:Jf6 11d6 29 11 c3 11e5 3 0 11d3 11d6 31 11c3 11e5
32 1ib3 l:l. c7 33 11 d3 l:l./7 34 11 xc4 l:l. x/6 35 l:l. dl b5
The position is drawn, but Black is playing somewhat recklessly.
A simpler method is 35 . . . l:l. e6 36 l:t d7 + l:l. e7 37 l:l. xe7 + 11xe7
38 11d4+ f7 39 11xa7 'lrxe4.
36 l:l. d7+ h6 37 11e2 11 c5 + 38 h2 11e5+ 39 g3 11 c3 40
g2 11 c4 41 11e3 + g5 42 l:l. d2 11/1 + 43 h2 11/3
The sealed move. Once again, just as happened in game 7 of
the match, I persistently sought winning chances after resumption,
and reached an endgame a pawn up - but alas, a drawn result
was unavoidable. We will follow this interesting game to the end.
44 11d4 l:l. e6 45 e5 11/5 46 l:l. e2 a5 47 11d5 b4 48 11xa5 11d3 49
l:l.g2 11d4 50 11a8 11xe5 51 11/8+ g6 52 11xb4 h5 53 h4 gh 54
11xh4 l:l. d6 55 11 c4 l:l. d4 56 11 c6+ g7 5 7 11b7+ h6 58 11 c6+
g7 59 l:l. c2 l:l. h4+ 60 g2 11e4+ 61 11xe4 l:l. xe4 62 l:l. c7+
g6 63 l:l. a7 l:l. e3 64 h3 l:tc3 65 l:ta8 l:l. c4 66 a4 g5 67 a5
l:l. a4 68 a6 <i;h6 69 g2 l:l. a3 70 /2 g7 t- t .
15
eS!?
dS
11e1
38 Seville Variation
19
R.h6
Threatening . . . R.d2.
W'a6
20
lllg3
20 . . . W'xf2+ would be over-hasty : 2 1 W' xf2 R.e3 22 W' xe3,
followed by r:t cl.
21 <lo>h1
The threat was . . . r:t xf2, for example : 21 W'e2 r:t xf2! 22 <lo>xf2
R.e3 +; while if 2 1 lllfl , then 2 1 . . . lllb 2! is decisive.
21
W'a4
2 1 . . . W'a3 is also worth trying.
22
R.gl! (1 7)
17
B
The white queen's problems can be solved later. In Georgievlvanchuk, White played 22 "ife2, and there followed : 22
b6 23
h4 R.f4 24 lllfl (24 g5 llld 6 25 r:t e 1 was more precise) 24 . . . llld 6
25 l:l e 1 l:l c8 26 g3 r:t c2 27 W'f3 W'xa2 28 <lo>g1 R.h6 29 g5 R.g7
30 llle 3 r:t c7 3 1 lll g4 r:t f7 32 W'e3 W'c2 33 h5 lllc4 34 W'c1 W'xc1
35 r:t xcl gh 36 r:t xc4 hg 37 r:t c8 + R.f8 38 R.e l ! (lvanchuk gives
38 <lo>fl ! b5 39 <lo>e2 b4 40 R.c5 a5 41 r:t a8 b3 42 <lo>d2, with a
great deal of play left; or 38 . . . <lo>g7 39 <lo>e2 b5 40 r:t a8) 38 . . .
<lo>g7 39 R.c3 R.d6! 40 r:t c6 R.c5+ 41 <lo>g2 r:t f2 + 42 <lo> h 1 R.d4
43 R.b4 r:t f7! 44 r:t e6 r:t b7! 45 r:t c6 a5 46 d6? (lvanchuk considers
this the decisive mistake, although even after the more stubborn
46 R.a3 b5 47 d6 r:t f7! 48 r:t c7 R.b6 49 .D. c6 R.d8 50 R.b2 r:t d7
5 1 R.xe5 + <lo>f7, Black has a clear plus; not however 47 . . . R.b6??
48 r:t xb6!) 46 . . . ab! 47 r:t c7+ <lo>f8 48 r:t xb7 b3 49 r:t b8 + <lo>f7
50 d7 b2 5 1 r:t f8 + <lo>e6 52 d8 lll + <lo>e7 53 r:t fl xd8, and White
soon resigned.
. . .
Seville Variation 39
One other game is worth mentioning. In Naumkin-Miralles,
Voskresensk 1 990, Black answered 22 11re2 with 22
.i.g5, and
only after 23 lOft did he play 23 . . . b6. (He could have brought
about simplifications with the bold 23 . . . l0d6! 24 .i.c5 l0xe4! 25
.i.xf8 11rd4 26 .i.c5 ! 11rxa 1 27 11r xe4 11r xfl + 30 .i.g1 .i.d8,
threatening . . . .i.b6.) Naumkin now seized the initiative: 24 h4
.i.e7 25 g5 .i.c5 (25 . . . l0d6 26 l0d2) 26 .i.g3 l0d6 27 l0d2 .i.d4
28 J: c 1 ! 1ixa2 29 J: c6! J: d8? (it was better to take the central
pawn : 29 . . . l0xe4 30 11rxe4 11r xd2 3 1 J: xg6+ hg 32 11r xg6 + <i>h8
33 11rh6+, with a draw; sharp play would result from 3 1 <i>h2
11re3 32 11rg4!?) 30 <i>h2! aS (Naumkin gives these variations : 30
. . . 11r a 1 3 1 l0f3, or 30 . . . l0xe4 3 1 11rxe4 11rxd2 32 J: xg6+ hg 33
11rxg6+ ci>f8 34 11rf6 + ci>e8 35 11re6+ ci>f8 36 g6 .i.g 1 + 37
xg1 11rxd5 38 11rf6 + , with a substantial advantage to White) 3 1
h 5 11r a 1 32 l0f3 gh (nor can he save himself with 32 . . . 11rb2 33
l0xd4 11rxd4 34 11rg4 11rxe4 35 11re6+ .!On 36 J: c8!) 33 l0xd4 11rxd4
34 11rxh5 l0e8 35 g6 1 -0.
22
b6
25
d6!
l0xd6
40 Seville Variation
'irxb1
27 'irxd6
28 'irxd2
aS
28 . 'irb5 29 h3 'ird7 30 'irc3 'ire? 3 1 'irb3 is more tenacious,
although White's chances are still better.
29 'ird8 +
g7
h6
30 'irg5
White threatened the deadly 3 1 lDf5 + .
31
'irxe5 +
h7
32 h3!
J: d7
White was intending to bring his bishop into play after 33 h2.
But now the denouement comes instantly.
33
lDh5!
.
1-0
cd
.:t el (18)
cd
.i.e6
14
15
.i.xa2
"tlra4
. . .
19
B
17
e6
. .
. . .
18
.C. fd1
20
w
J.xbS!
White would lose after the hasty 24 c7? 'irxd3 25 l: xd3 etlxd3
26 cS('ir) : axeS 27 : xeS : xeS and 2S . . . l: c l + .
24
25
c7
etlxdl
'irdS!
J.xe8
etlxf2
c8('ir)
: xc8
: xc8
et:Jh3 + ! (21)
21
w
The knight has sped right the way across the board to get its
king out of trouble.
29
gh
29
30
g2!
1i'e2+
gl
31
i-t
In the present game we are still dealing with the main system
(Exchange Variation) of the Griinfeld, but turn our attention to a
line in which White postpones castling by one move, resulting in
play of a wholly different character.
In passing, we should also mention the innovation 10 .:t b 1 !?,
employed in Korchnoi-Kasparov, Reykjavik 1 988. There fol
lowed : 10 . . . aS 1 1 .td3 cd 1 2 cd b6 1 3 0-0 (immediate kingside
activity with 1 3 h4 could come up against counterplay with
obscure consequences : 1 3 . . . eS 14 dS fS 1 S hS f4 etc.) 1 3 . . . e6
14 'it' a4 ( 1 4 'it'd2 .tb7 1 S h4 .:t c8 1 6 .:t fc 1 1i'd7 17 hS : xc 1 + 1 8
.J:l xc 1 .J:l c8 leads to equality) 1 4 . .tb7 1 S .:t fd 1 .l:t c8 1 6 .td2
c6 17 .tc3 1i'h4! 18 .te l .D. fd8 19 f3 1i'e7 20 .tbS?! (in
Kasparov's view 20 .tf2 is better, leading to equality after 20 . . .
aS, but then Black also has 20 . . . .th6!?) 20 . . . a6! 2 1 .txa6
.txa6 22 1i'xa6 xd4 (the initiative is with Black, but after 23
xd4 .txd4+ 24 h 1 , or 24 .:t xd4 1i'cS 2S .tf2, White has
adequate defensive resources; instead, he commits the decisive
error) 23 .tf2? : a8! 24 1i'd3 (24 1i'c4 bS) 24 . . . .:t a3, winning the
queen and with it the game (2S xd4 .:t xd3 etc.).
. .
10
cd
11
cd
1ra5+
J.d7
12
n
An inferior choice is 12
: d8 13 h4 h5 14 1rb3 e6 1 5 d5 .!DeS
. . .
23
B
l:l. ac8
14
h5
e5
Without this central advance Black can hardly hope for counter
play. Let us look at the alternatives :
(a) 14
e6 1 5 hg hg 1 6 e5! (but not immediately 1 6 'ird3 b5 1 7
J.b3 b4 1 8 1W d 2 l:l. xc l + 1 9 xc 1 c6 2 0 'ird3 b4 2 1 e2 l:l. d8,
and Black has very strong counterplay in the centre. Lputian
Dvoiris, Simferopol 1 988, continued 22 J.h6? J.xd4! 23 'irh3
J.f6 24 J.e3 J.c8 25 tt:lf4 !i:Je7, and White got nowhere. 22 g1
was better) 16 . . . e7 17 'ird3 l:l. fe8 18 J.d2 'ira4 19 J.b3 l:l. xcl +
20 J.xcl 'irb4 2 1 'irh3 f8 22 J.h6 g8 23 J.xg7 + xg7 24
'irh8 + f8 25 l:l. h7 e7 26 'lrg7 d8 27 'irxf7, and it is all
over; Griinberg-Gauglitz, E. Germany 1 989.
(b) 14 . . . b5 15 J.b3 e5 (in the present circumstances, this counter
stroke in the centre is not good) 16 hg hg 1 7 de! (Black was vainly
hoping for 17 d5? d4 1 8 l:l. xc8 l:l. xc8 19 xd4 ed 20 J.xd4
J.xd4 2 1 1Wxd4 l:l. c 1 + 22 J.d1 f6 23 g3 'ira4! 24 'irxa4 ba 25
g2 l:l. a l 26 J.f3 l:l. xa2 27 l:l. c 1 a6, when the black bishop settles
on b5 and the pawn on a4 acquires formidable strength) 17 . . .
xeS 1 8 .l: xc8 J.xc8 ( 1 8 . . . l:l. xc8 1 9 'f4 g4 20 J.xf7 + !) 1 9 f4
J.b7? (loses at once, but then Black also has a hard time in other
. . .
l:l. xc4
17
xd4
eel 18 ..txd4 l:l. xc4 1 9 l:l. xc4 11f a6 ( 1 9 . . . J.. b 5 20 J.. x g7,
24
B
. . .
. . .
18
.l:l xc4
'lra6
25
w
'lrh3!
g1
'lrh7+
.l:l h6!
J.xc4+
f6
f7
1-0
Tbilisi 1989
lDf6
g6
dS
d4
c4
lDc3
lDf3
cd
e4
be
: bt
l
3
4
5
6
7
s
9
g7
lDxdS
lDxc3
c5
o-o
el
It is interesting that the 1 976 edition of ECO gave this set-up
no more than a cursory mention. Yet in the last few years, the
system with 8 : b 1 has virtually become the most popular choice
against the Griinfeld Defence, perhaps even surpassing the old
variation in which the bishop is brought out to c4.
Another method of developing the white pieces, behind that
strong pawn centre which characterises the Exchange Variation,
involves an early c1-e3. In this case, the light-squared bishop
is brought to e2 (later than usual), and the queen's rook may
occupy either c1 or d1 (after 1Fd1-d2). This variation was an
object of dispute in my last match with Kasparov - see Games
No. 14 and 1 5.
I would add that the popularity of the variation with : b 1 and
.te2 is above all associated with the names of the young
grandmasters "Boris Gelfand and Alexander Halifman.
9
1FaS
0-0
11fe6
11
.i.gS
An interesting alternative is 1 1
t2Jd7 12 .i.xe7 J:l. e8 13 .i.d6
cd 14 l'hxd4 l'hf6 15 f3 l'hd5!? 1 6 J:l. a 1 11f b2 1 7 11fa4 .i.f5 1 8 ed
.i.xd4+ 19 lrxd4 .D. xe2 20 g4 .D. d2 2 1 lrb4 .i.d3 22 lr xb2 .D. xb2
23 .D. fd 1 , with equality; Vaiser-Dvoiris, Bamaul 1 984. Vaiser
suggests 15 .J:l a1 lrb2 16 .lObS!? .ll:l xe4 17 .i.a3 ird2 1 8 .ll:l c 7.
12
ird3!?
b6
After 12
lrd6 1 3 lre3, or 12
.ll:l d7 1 3 lre3 .ll:l f6 14 .ll:l e S!,
the initiative is with White. On the other hand, 12
.D. d8!?
deserves to be tried.
. .
. .
13
d5
14
e5!
The point of White's play. He now obtains the two bishops and
et:lxeS
1W d 2
-*.xeS
'itxeS
1We3!
: e8
Wf6
: b8
d6
: bd1 (28)
eS?
21
. .
23
24
..th6
..tc6!
Wf6
:Z. e6?
..tg5
f4
W g7
h6
..te7
W xf4
Wa4
..td5
W xa7
..txb7
1-0
ef
g5
.!Oe5
..t b7
.!Lld7
Leningrad 1989
1 d4 ltlf6 2 c4 g6 3 ltlc3 d5 4 ltlf3 J.g7 5 cd llJxd5 6 e4 ltlxc3
7 be c5 8 l:l. b1 0-0
9
J.e2
b6
10
0-0
Another familiar plan is 10 J.e3 J.b7 1 1 e5 llJc6 1 2 h4,
10
11
J.b7
1rd3 (29)
29
B
11
J.a6
'lre3
. .
12
cd
e6 is too timid; after 13 de .*.xe2 14 'lrxe2 be 1 5 .*.f4
. . .
14
.*.a3
14 .*.xa6 lbxa6 1 5 'lra3 lbc7! gives White nothing; Halifman
14
Another possibility is 14
l:t e8 IS d5 .*.xe2 16 'lrxe2 'lra4 1 7
'lre3 lbd7 l S l:t fc 1 : ac8 1 9 h 3 l:t xc l + 2 0 l:t xc l llJf6 2 1 lbd2
.*.h6!, as in Gelfand-Malishauskas, Vilnius 1 988. However, Pribyl
gives IS : fd! as more accurate.
'ir xe2
.l:l. e8
.l:l. fcl
.l:l. c3
e6
.l:l. c8
.I:!. be l
a6 20 'ir xb6.
.l:l. xc8+
1 8 .l:l. bc 1 .l:l. xc3 1 9 .l:l. xc3 c!Llc6 20 "irbS .l:l. c8 2 1 .l:l. c4 'irb7 leads
to equality.
18
'ir xc8
19
.l:l. c l
'irb7
An alternative is 19 . . . 'ira6 20 "ire3 c!Lld7 2 1 .l:l. c7 c!Llf6 (2 1 . . .
'irc4
c!Lla6 (30)
30
w
This position had first occurred only a few days earlier, in the
game S. Ivanov-Pribyl from the same international tournament
(Leningrad 1989). After 2 1 dS ed 22 ed .l:l. d8 23 d6 c!LlcS! 24 xeS
be 2S 'ir xcS 'ira6 26 h4 .l:l. xd6 27 c!LlgS h6, the players agreed a
draw. The main game we are examining was to end in victory for
Black, but that was hardly the rightful outcome. For the moment
Halifman doesn't want to settle for a draw, and adds fuel to the
flames.
h4
e5!?
h5
.l:l. c3
'irc6
.l:l. xc6
.l:l. c7
..txf8
29
d5!
.l:l. d8
h6
gh
b5!
'irxc6
lt! b8
.U8
xf8 (31)
31
w
lt!d4
f4
.l:l. c5
.l:l. xd5
ed
.l:l. e8!
.l:l. e7
lt!a6
In the event of 33 .l:l. c6 lt!c7 34 .l:l. xh6, Black can save himself
with 34 . . . g7! (but not 34 . . . lt!e6? 35 lt!f5 .l:l. c7 36 .l:l. h8 mate)
35 .l:l. xh5 lt!e6 36 lt!f5 + g6 37 lt!xe7 + xh5 38 lt!xd5 g4.
33
34
35
36
37
.l:l. d6
lt! c6
f5
f6?
lt!c7
lt!e6
.l:l. c7
lt!g5
Game No. 1 0
Halifman-Henkin
Leningrad 1989
1 d4 lll f6 2 c4 g6 3 lll f3 J.g7 4 lll c3 d5 5 cd lll xd5 6 e4 lll xc3
7 be c5 8 .:t b1 0-0
9 J.e2
lll c6
10
d5
lll e5
'if d6!?
32
B
33
B
34
B
16 . . . ed
Other pawn captures are no better: 1 6 . . . ef 1 7 e6! followed by
1 8 0-0 and 19 g4; 1 6 . . . fe 1 7 fg ( 1 7 fe 'it' h4+ 1 8 d 1 doesn't
look bad either) 17 . . . hg 18 0-0 ed 19 cd 'it'd6 20 :t b3 .tg7 2 1
:t g3 etc.; 1 6 . . . gf 1 7 :t b3 :t e7, and now 1 8 .tb2, 1 8 d 6 and 1 8
'it'h6 are all highly unpleasant for Black.
1 7 fg
Practice has also seen 1 7 e6 d4 1 8 g4 b6 1 9 .tf3 : b8 ( 1 9 . . .
.ta6!?) 20 'it' g2 'flc7 2 1 0-0 gS 22 .tdS 'it'e7 23 h4, with the
initiative; Vaiser-Pribyl, Sochi 1984.
17 . . . hg, and now White has two lines :
(a) 18 0--0 : xeS 1 9 .tf3 d4 20 :t e l 'it'e8 2 1 .ta3 .tfS 22 : xeS
'it' xeS 23 .tdS + 'it' xdS 24 cd J.. x b l 2S J.. x cS, and Black is in a
bad way; Cebalo-Raicevic, Citorle 1 984.
(b) 18 cd : xeS 19 0-0 b6 20 d6 J..e 6 21 .tf3 :t c8 22 J.. b2 fS 23
'it'h6! and White has a won position; Agzamov-Pribyl, Sochi 1984.
ll
b6
. . .
e6
14
c4
e5 looks logical, but the following two examples from
15
i.b2
15
16
17
l% xb2
e5
i.xb2
l% e8
0-0
i. f3
l% e 1
ed
f6
fe (35)
21
fe!
A significant refinement.
21
22
23
24
25
.*.xd5+
'it'xd5+
e6
11Ff3
11Fc7
.*.xd5
g7
.l:l. ad8
.l:l. e7
.1:1. 2
11Ff6 +
11Fg5!
11Fd6
g8
'it'f6 +
11Fg5
h4!
11Fd5
g8
g7
.l:l. de8
.l:l. d8
11Fg5
Of course not now 33 .l:l. d2? 11Fg3!, and it is Black who wins.
:t dl
:t de8
11fc6?
36
w
36
:t d7!
0 0 0
36
b5
h5 doesn't help : 37 :t e3, followed by 38 :t g3. If 36
0 0 0
40
41
42
43
cb
hS
:t xe7
h6
h1
:t b1 !
11f f4!
11f xb5
11fb6
:t xe7
c4+
11fd8
:t c7
After 43 11fd5 :t c8!, or 43 11fd4 'tiffS!, Black could still hold on.
g5
43
11fxg5 + !
1-0
Budapest 1989
l d4 lt!f6 2 c4 g6 3 lt!c3 dS 4 lt!f3 ,j_g7 S cd lt!xdS 6 e4 lt! xc3
7 be cS 8 l:t bl 0-0
9
10
,j_el
cd
cd
'lfaS+
,j_dl
13
"Wet!
"We6
.tc4!
:e1
.tb4
'W xe4
'Wb7
.te6
J: xe6!
A typical exchange sacrifice giving White a dangerous initiative.
17
18
fe
g5 (39)
39
B
18
c6
lt!xe6
.tc3
w h8
But not 20 .td5? .l:l. fc8 2 1 lt!g5 h6, and Black has everything
in order.
.t6
20
'ir h6!
.l:l. g8
.l:l. el
.l:l. g7
g4!
lt!a5
.td3
.tal
'irc6
.1:1. 7
Things also go badly for Black in the endgame that results from
25 . . . 'irf3 26 g5 'irh5 27 'irxh5 gh 28 f4 .l:l. gg8 29 wf2 .tg7 30
lt!xg7 xg7 3 1 .l:l. xe7 + f8 32 l:t xh7.
.t g7
26
g5
27
28
29
d5!
.txg7 +
.txg6
1-0
'irxd5
g8
Game No. 1 2
Polovodin-Maslov
Leningrad 1990
1 d4 lt!f6 2 c4 g6 3 lt!c3 d5 4 lt!f3 .tg7 5 cd lt!xd5 6 e4 lt!xc3
7 be c5 8 l:t b1 0-0 9 .tel cd 10 cd 'ira5+ tt .td2 'ir x a 2
lt!d7
12
0-0
Over the last three years, a good many games and analyses
relevant to this move have accumulated. I shall here quote the
most important of them.
40
w
14
f6
. . .
15
.!Lic4!
Now after 23 . . . J:. abS 24 l:l. a7 a4, the game was eventually
drawn. Lagunov points out an interesting trap that Black could
have set : 23 . . . .:t eeS! 24 .ta3 (also after 24 l:l a7 l:l. xa7 25 .txa7
a4, the advantage is with Black) 24 . . . .tf5! 25 'ifb5 .td3! and
wins.
:at
xc4
bl!
t6
11rb2 loses the queen to 17 .txc4+ h8 18 .l: b l !. On
the other hand, t6 . . e3 transposes into the game after 1 7 : xa2
xd l .
. . .
t7
t8
t9
.l: xal
xd t
h8
.tc4+
: xdt (44)
44
B
.td7
J.d6
. .
22
.i.f7
a5
f4!
.l:l a6
J.d5!
26
27
J.xf8
l:l xf8
e6
J.c8
.tc6
l:l e7
Game No. 1 3
Gelfand-Ivanchuk
Tilburg 1990
I d4 .!Zif6 2 c4 g6 3 .!Zic3 d5 4 cd ll:lxd5 5 e4 .!Zixc3 6 be .i.g7 7
.!Zif3 c5 8 : bt 0-0 9 .i.e2 cd 10 cd 1W a5 + I I .i.d2 'lf xa2
12
0-0
. . .
45
B
46
B
IS
dS
lOaS
.tc5
e5!
.tf6
h3
.txf3
.txe5
.t xf3
lUeS
d6!
Now it turns out that Black cannot take the d-pawn, so he has
to part with the exchange.
20
11Vc4
The endgame which now arises is difficult for Black, but the
alternatives are thoroughly bad: 20 . . . ed 21 .td5 ll\c4 22 .l:t xb7,
or 20 . . . 11Ve6 21 d7 .l:t ed8 22 11Va4 lll c4 23 .txb7 .l:t ab8 24 .txa7.
21
22
d7
de(11V)+
11Vxc5
.l:t xe8
For the moment Black is quite well off from the material point
of view, with three pawns - more than the equivalent - for the
exchange. His pieces, however, are badly co-ordinated, and this
leads to more losses.
23
11Va4
lll c6
.l:t xb7
.l:t c8
-*.xc6
: xc6
.J: xe7
41
11r xb3 42 : xb3 f8 43 f3 e7 44
d6 46 : dl + we6 47 4 6 48 : c2
: bl
e6 45 e4
47
w
61
J: a8!
.l:l. xaS
1-0
d4
c4
l0c3
l0f6
g6
d5
cd
4
5
6
e4
be
lDxdS
lDxc3
j_g7
. . .
j_e3
c5
'ii' d 2 (48)
This is the position that formed the starting point for the
theoretical contest in the games (and analysis!) of the 1 990 match.
0-0
In the ninth game, where I tried this system for the first time,
Kasparov chose a more straightforward plan, exchanging pawns
in the centre. That game is also a very important one, so we will
now embark on a long digression to look at it in detail :
8 . . . cd 9 cd lll c6 10 r:J. dl
ECO only considers 1 0 lll f 3 .t.g4 1 1 r:J. d 1 (nor can White count
on anything after 1 1 .i.b5 0-0 12 .t.xc6 be 1 3 : c 1 .t.xf3 14 gf
1r d7) 1 1 . 0-0 1 2 .t.e2 : cs 1 3 0-0 b6 ( 1 3 . . . .t.xf3 14 .t.xf3
e5 1 5 d5 lll d4 doesn't look bad either) 14 d5 .t.xf3 15 gf lll e 5 1 6
r:J. c l 1Wd7 1 7 f4 lll c4!. After 1 S .t.xc4 1rg4+, Black is guaranteed
a draw; Haik-Granda, Dubai 01 1 9S6.
10 . 1Va5 (49)
The game Kozul-Dorfman, Marseille 1 9S9, in which the inno
vation 10 r:J. d1 was tried out, went 10
e6 1 1 lll f3 0-0 1 2 .i.b5!?
(the tempting 13 .i.h6 is parried by 1 3 . . . 1Wa5 1 4 1rxa5 lll x a5 1 5
.t.xg7 xg7 1 6 .t.d3 .t.d7 1 7 d2 : res, o r by the quite good
alternative 1 3 . . . f5!? 14 .i.xg7 xg7 1 5 e5 1rd5 16 .t.e2 b6
1 7 0-0 ib7, retaining counter-attacking possibilities in the
middlegame) 12 . . . id7 1 3 0-0 (after 13 ih6 ixh6 14 1V xh6,
the reply 14 . . . 1Wa5 + is rather risky on account of 1 5 : d2.
Similarly, 1 3 . . . lll e 5 14 lll x e5 ixb5 1 5 ixg7 xg7 is inadequate
in view of 1 6 h4. But 1 3 . . . f5 14 .t.xg7 xg7, as in the previous
note, is fully viable, since 15 e5 is met by 15 . . . lll x e5 16 de ixb5
17 1Wb4 1V b6 - the simple 17 . . . ireS is also possible - 1S lll g 5
: res 1 9 l:l. d6 ic6 20 lll x e6 + gS, with complications; or 1 S
lll d 4 r:J. ad8, with the idea of answering 1 9 1Vxb5 with 1 9 . . . : xd4!,
or 1 9 1re7+ with 19 . . . gS! 19 a4 id7, or 19 lll x b5 with 1 9
. . . r:J. xd 1 + 20 xdl a 6 2 1 1We7 + r:J. f7) 1 3 . . . a 6 ( 1 3 . . . lll e 5 is
unconvincing; although 14 de ixb5 1 5 1V b4 ireS 16 : fe l b6 is
satisfactory for Black, the unassuming 14 ie2 gives White the
better prospects) 14 ie2 1ra5 1 5 d5 1rxd2 (of course not 1 5 . . .
ic3? 1 6 de! ixd2 1 7 ixd2, with a won position) 1 6 : xd2 ed
1 7 ed lll e 7, and the ending clearly favours White. For example,
he can play to restrict the knight's mobility with 1S d6 lll f5 ( 1 S
. .
. .
. .
49
w
1 1 1i'xa5 xa5 12 !3
This endgame had arisen once before, in Yusupov-Gulko,
Linares 1 989. On that occasion White preferred 12 ..td3 0-0 1 3
e2 ..td7 1 4 l:t c 1 l:t fc8 1 S d2 e6 1 6 : xeS + : xeS 1 7 l:t c 1
l:txc l 1 8 xcl fS, and the minor piece ending was quickly
drawn after 19 e2 a6 20 ..tf4 c6 2 1 ..td6+ e8 22 c3 ..tf6
23 ..tc7 ..te7 24 f3.
12 . . . 0-0
Here 12 . . . ..tg4 is premature on account of 1 3 ..tb5 + .
1 3 ..te2 ..td7 (50)
14 ..td2!
Typical of Grtinfeld endings. If the knight retreats, then after
1 4 . . . c6 1 S dS eS (in the event of 1 S . . . d4 1 6 xd4 ..txd4
1 7 ..tb4 ..tf6 1 8 f4 aS 19 ..ta3 ..tc3 + 20 f2 l:t fe8 2 1 lt b 1 , or
20 . . . ..tb4 2 1 ..txb4 ab 22 l% d2, Black has many problems) 1 6
xes ..txe5 1 7 f4 ..tc7 1 8 : c t : res ( 1 8 . . . ..tb6 1 9 ..tb4 : res
20 d2) 19 ..te3 e6 20 ..tc4, White goes ahead with his policy
of constriction. 19 f2 is also playable.
14 . . . b6 15 0-0
After pondering it for a long time, I refrained from doubling
Black's pawns with 1 5 .t.xa5 ba, for he would then acquire an
adequate initiative. However, as Zaitsev has shown, 1 5 .t.a6!? was
more promising. The analysis of this endgame points to the
conclusion that the entire variation, beginning with 8 . . . cd, fails
to secure equality for Black.
15 . 'D.fd8?!
This appears to be the root cause of his subsequent troubles. A
more natural choice seems to be 1 5 . . . .l:fc8 1 6 : c 1 (if 1 6 .t.a6,
then 1 6 . . . 'D. c2 1 7 .t.xa5 : xa2 1 8 .t.b7 'D. b8. Therefore White
should first play 1 6 .t.xa5 ba, and then 1 7 .t.a6 'D. c2 18 'D. c 1
: xa2 1 9 : c7 'D. d8, but even so, Black i s risking nothing after 20
h3 'D. a3! 2 1 : xa7 'D. xf3 22 gf .t.xd4 and 23 . . . .t.xh3) 1 6 . . . .t.g4
( 1 6 . . . lDc6 is problematic; Black aims to answer 1 7 .t.a6 with 1 7
. . . lDxd4 1 8 .t.xc8 lDxf3 + 1 9 gf .t.xc8 20 g2 .t.e6, o r if 20 : c7,
then 20 . . . .t.h3. In either case Black has definite compensation for
the exchange, but is it sufficient for equality? Thus, in the last
variation, after 21 'D. fc l e5 22 'D. b7!, the rook cannot be dislodged
from the seventh rank : 22 . . . .t.c8 23 : xa 7!) 1 7 .t.a6 (on 1 7 d5,
Black has 1 7 . . . .t.b2, when neither 1 8 'D. b1 .t.xf3 19 gf 'D. c2 20
'D. fd 1 .t.c3 nor 1 8 : xc8 'D. xc8 1 9 'D. b 1 .t.xf3 20 gf 'D. c2 21 'D. d l
.t. c 1 presents him with any danger) 1 7 . . . : xc 1 1 8 'A xe l .t.xf3
19 gf .t.xd4 20 'D. c7. White regains the sacrificed pawn, but cannot
achieve more than that.
16 : ci .t.g4 (51)
Again a straightforward decision. Perhaps it was worth prepar
ing an exchanging operation with 1 6 . . . a6!? 1 7 'D. b l ( 1 7 : c7
.t.b5 1 8 .t.xb5 ab 1 9 d5 lDc4 20 .t.b4 .t.f6 2 1 .t.xe7 .t.xe7 22
.
1 7 d5 10b7
Devices such as 1 7 . . . e6 1 8 h3 .txf3 ( 1 8 . . . ed 1 9 hg de 20
.txaS ef 2 1 .txf3 ba 22 .txa8, and Black has a lost position) 1 9
.txf3 e d 2 0 e d fail t o help Black - h e i s still faced with the same
inferior ending.
18 h3
White would be wasting time with 1 8 .tb4 e6, while 1 8 : c7?!
is a mistake owing to 1 8 . . . tOeS! ( 1 8 . . . .l:t d7 19 : fc 1 , with the
initiative) 19 : xe7?! .tf6 20 : c7 10xe4.
18 .
.txf3
Nor is 1 8 . . . .td7 1 9 .tb4 e6 20 .ta6 any better for Black.
19 .tx/3 10c5 20 .te3 : ac8 21 .tg4! :t b8
Or 2 1 . . . : c7 22 .tf4 :t b7 23 :t fd 1 , and White has a large
plus, since 23 . . . 10xe4? loses at once to 24 : cS : xeS 2S .txc8.
White also retains a substantial advantage with 22 f4 (with a view
to 22 . . . hS 23 .tf3, and the pawns creep forward) 22 . . . fS!? 23
ef gf (23 . . . .l:t xdS 24 fg hg 2S .l:t fd 1 .l:t xd 1 + is also a grim outlook
for Black) 24 .txfS : xdS 25 : xeS! : dxcS (if 2S . . . : exeS, then
26 ..te6+ f8 27 .txcS : xeS 28 :t d 1 !, and Black is in a difficult
. .
so, by deploying his forces in the same way as in the game g4-g5, ..;. g l -g2, and the transfer of the rooks to the h-file - White
would clearly have the better chances.
28 TJ. a4
For good measure, Black's knight is now driven onto a bad
square, since 28 . . . TJ. a8 fails to 29 d6.
28 . . li:Ja5 29 g5!?
It is also worth considering 29 TJ. e4!?
29 . . . TJ. bc8 30 J:.e2 J:.d6
If 30 . . . TJ. c2, a good reply is 3 1 J:.d3 TJ. c3 (3 1 . . . TJ. b2? 32 J:.c1,
and the rook is trapped) 32 J:.a6, after which White again carries
out the plan beginning with g 1 g2
31 g2!?
More accurate than 3 1 J:.a6 TJ. c2 32 ..;.g2, which could be met
by 32 . . J:.c5 33 J:.d3 TJ. c3; whereas now, 3 1
TJ. c2 32 J:.d3
would immediately put Black in a critical position.
3 1 . J:.c5 (53)
.
. . .
. .
53
w
. . .
. . .
. .
. .
11
..i.h6!
A new, and quite effective, idea of Zaitsev's. By exchanging
bishops, White attempts to exploit the weakening of the dark
squares in his opponent's camp. I proceeded less convincingly in
the 1 3th game of the match, in which the diagrammed position
arose by a different move-order (9 .l: c l 1Wa5 10 ltlf3 e6). Now is
the best time to pause and examine that game.
l l d5
Practice has also seen 1 1 ..i.e2 cd 1 2 cd 1Wxd2+ 1 3 ltl xd2 ltlc6
1 4 d5?! ed 1 5 ed ltld4, with equality; or 1 1 d3!? .l:t d8 1 2 ..t.g5
f6 13 ..t.e3 cd 14 cd 1W xd2 + 15 xd2 ltlc6 16 .l:t c4 ..t.d7 17 : b l
..i.f8 1 8 e2 b6 1 9 ..i.d2 : ac8, with unclear play; Blees
Mikhalchishin, Budapest 1 990.
11 . . . ed 12 ed .l: e8!?
A game Lputian-Tukmakov, USSR 1 989, went 1 2 . . . ..t.f5 ( 1 2
. . . ltld7 1 3 ..i.d3 b5! 1 4 c4 1Wxd2+ 1 5 ltlxd2 be 1 6 ..t.xc4 should
have led to equality in Ftacnik-Hartston, Skara 1980 : 16 . . . ..t.b7
___
-*. xg7
cd
x g7
cd
xd2
e3
'iFxd2 +
:t d8
-*.d7 (58)
58
w
:t b1
:t abS
.i.d3
19
h5
hg
J: h2 (59)
f6
hg
59
8
Rather slow. White could have set his opponent more problems
with 21 g4!?. The threat to break up Black's kingside structure is
highly unpleasant. Replies which fail to help are 21
g5 (2 1 . . .
.i.c6 22 .i.c4!) 22 e5! t'i:Jd5 + 23 wd2 l:t h8 24 J:[ xh8! wxh8 25
.l:t h 1 + wg8 26 J: h6, and 21
.ll h8 22 .l:txh8 wxh8 23 g5 f5 24
l'i:Je5 .i.e8 25 ..ta6 b6 26 J: c 1 or 26 f3.
It seems Black would have to play 2 1
e5 22 de ..txg4 23
ef+ wxf6. White would then carry out a similar idea to one
which will arise in the actual game (with the difference that the
black pawn is on b6 and the white rook on h 1 ) : 24 t'i:Jd4 J: d7 25
: h4 : bd8 26 e5 + (26 : xg4 : xd4 27 e5 + wxe5 28 : xd4 J:[ xd4
29 f4+ J:[ xf4 30 J: b5 + t'i:Jd5+ 31 n xd5 + wxd5 32 wxf4 draws)
26 . . . wxe5 27 J: b5 + l'i:Jd5+ 28 l:t xd5+ J:[ xd5, but this plan loses
its force because the c6 square has not been weakened. There is
likewise no danger to Black in 24 e5 +!? wg7 (24 . . . we6 25
. .
. . .
. .
21
g4
de
e5!?
.t xg4 (60)
60
w
24
ef+
24 : bh 1 has been recommended as stronger. But let us see
whether it is. 24 . . . : xd3 + ! (inadequate alternatives are 24 . . .
.txf3 25 : h7 + ri>f8 26 : h8 + llJg8 27 .tc4! .txh 1 28 ef 'i>e8
29 .te6, and 24 . . . fe 25 llJxe5 .te6 26 : h7 + ri>f6 27 f4 : bc8
28 ttJf3! .tg8 29 ttJg5! .txh7 30 : xh7 : xd3 + 3 1 'i>xd3 : f8 32
e3 llJc6 33 : c7 llJe7 34 llJ h7 + ) 25 'i>xd3 .txf3 26 : h7 + f8
(61) .
27 : 1 h3 (forced, since 27 ef llJg8! 28 : l h3 .txe4 + ! draws; nor
does 28 : 1 h4 help, in view of 28 . . . g5!, and when the rook moves,
ll'ld4
62
w
26
f3
63
w
..tel
J:l ac4
..td7
J:l e8
J:l b2
.!LidS+
t-t
Black now has a tiny pull, but neither side can seriously count
on winning.
Game No. 1 5
Karpov-Kasparov
64
8
cd
cd
ll:lc6
h3
..td7
ll:lf3
..td3
ll:la5
..te6
J:[ fd1
b5?! (65)
a6
.Z: bc 1
..txd3
"it" xc8
.l:t xc8
'ilfxd3
.l: e8?!
"it"b7 22 "it"a3 .!Dc4 23 "it"xe7 1Wxe7 24
A passive move. 21
..txe7 J:l e8 was more flexible. It is possible that I would then have
had to revert to thy plan of .l:l d 1 -b 1 and a2-a4. After 22 .l: b1,
the reply 22 . . . h6 (l s bad on account of 23 ..td2!. The alternative
22 .z:t cl is less convincing, as after 22 . . . h6 23 ..tf4 tll c4, neither
of White's options promises him anything to speak of: 24 tll d 2
.l:t c8! 25 d5 .!Db2, or 24 d5 J:l c8 25 .!Dd4 tll b2 26 l:t xc8 + 1Wxc8 27
W'd2 .!Dc4 28 .!Dc6 tll x d2 29 tll xe7 + f8 30 .!Dxc8 .!Dxe4.
Another interesting variation is 21 . . tll c4 22 J.. xe7 .l: e8 (22
. . . .!Db2 23 W'b3 tll xd 1 24 J.. xf8 1Wc1 25 .ta3 'ilf a 1 26 "it" c2!) 23
. .
d5
tt.Jd2
tt.Jc4
24
66
w
26
l:l. c6!
. .
26
27
..te5
..t c3
..tb8
'ii' d4
..ta5
f6
Even now White has to play carefully; not 29 .1:1. b6?? .l:!. xc3! 30
.l:!. xb7 .l:!. c 1 + and mates.
29
30
31
'ii' c3
a3
..td6
.l:!. e8
g3
'ii' c5
..tc7
..tf4
.l:!. c7
d6!
g7
..te5
h5
.tal
'ii' d7
'ii' d8
d7
..td2
.l:!. b7
l-0
g5
.l:t8
..te5
Russian System
Game No. 1 6
Ehlvest-Ernst
Tallinn 1989
1
2
3
4
5
6
7
d4
ll'lf3
c4
ll'lc3
iib3
"ifxc4
e4
ltlf6
g6
J.. g7
d5
de
0-0
J.. g4
. .
J.. e3
1Wb3
Russian System 1 1 1
rt;g7 3 0 J:l. xc6 J:l. d2 3 1 g4 lDc2 3 2 fl lDd4 3 3 J:l. a6 lDf3 !- } .
1 9 'IJ.fe1 a4 20 'IJ. e4 J.h6 2 1 J.e5 a3 22 b3 lDa7 23 J:l. d7 J.cl
24 J:t xc7 J.b2 25 lDa4!
The material balance is restored, but White's threats are
mounting up.
25 . . . lDb5 26 J:l. xc6 J:l.fd8 27 'IJ. b6! J:l. d5
Cunningly thought out; if 28 lD xb2, then 28 . . . J:l. xe5! 29 J:l. xe5
ab 30 J:l. e 1 lDc3, and the b-pawn unexpectedly brings victory to
Black. But avoiding the trap is quite simple.
28 J.g3 lDc3 29 lDxc3 J.xc3 30 c6 J.d4 31 J:l. b7 1-0
Black resigned in view of the inevitable c6-c7.
lD b6
9
Alternatives are 9 . . . c5 and 9 . . . J. xf3
lDc6
10 J:l. d1
Practice has also seen 1 0 . . . e6, and 1 0 . . . e5 1 1 de lD8d7 with
sharp play. With 10 . . . lDc6, Black provokes the advance d4-d5.
.
11
dS
lDeS
12 J.el
lDxf3+
13 gf
J.hS
As practice has shown, 1 3 . . . J.h3 is less convincing in view of
14 J:l. g 1 ! and Black's bishop will soon have to retreat.
,
14
a4
69
w
1 12 Russian System
15
J:. gl !
Black could also g o after a different pawn - the one o n a 4 but after 1 5 . . . .txc3 + 1 6 be 'it'xa4 1 7 'it'xa4 lt:lxa4 1 8 d2,
White has quite enough initiative for it.
16
17
f4
d2!
'it' xh2
.txe2
lt:lxe2
c6
a5
lt:lg3
lt:ld7
cd
J:!. h1
'it' g2
..te2!
de
J:t xd7
f5!
e5!
'iff3+
26
J:t g l !
Once again the rook occupies the g-file, this time with even
greater effect.
26
f4
If 26 . . . J:t ad8 (26 . . . J:t ac8 27 "ili d l !), then 27 Wd5 ll xd7 28
"if xd7 f4 29 lZ:lf5 fe 30 fe is decisive.
27
28
29
30
31
lZ:lf5
'ifxe3!
"if h6
ll xg6
'if h4!
fe
'if xf5
'ifg6
hg
a6!
fe
b3
b6
e3
e4
b5
1 14 Russian System
34 . . . -*.c3 + 35 e2 -*.b4 doesn't help : 36 1Wxe4 -*.c5 37
1Wd5!
35 ""el -*.c3 36 1We7 -*.a5 37 .l:l. xa7 .l:l. ae8 38 1Wa3 -*.c3 39
1Wc5 b4 40 .l:l. e7 .l:l. c8 41 1Wd5 1 -0
Game No. 1 7
M. Gurevich-Kasparov
Moscow 1988
1
2
3
4
5
6
7
d4
c4
lt.:lc3
lt.:lf3
1Wb3
1Wxc4
e4
lt.:lf6
g6
d5
-*.g7
de
0-0
lt.:la6 (71)
71
w
Russian System 1 15
have also been seen! The moves 8 'ifa4, 8 b4, 8 ..te3, 8 h3, 8 e5,
8 l: b 1 and 8 ..td3 have completely gone out of use. 8 'it'b3, 8
..tg5 and 8 ..tf4 are also employed only rarely, and in any case
they often amount merely to a transposition - the bishop is
brought to e2 a little later. I shall now given one example of each
of these three alternatives that I have said to be the most common.
G. Georgadze-Tukmakov, Odessa 1 989, went 8 'it'b3 c5 9 d5
'it'b6! (9 . . . e6 is the usual continuation. The queen exchange gives
Black a comfortable game) 10 'ifxb6 (perhaps White shouldn't be
in a hurry to exchange; 10 ..tc4 or 10 e5 should be tested) 10 . . .
ab 1 1 ..tc4 e6 1 2 de ..txe6 1 3 ..txe6 fe 1 4 0-0 lt:lb4 1 5 h3 lt:ld7
1 6 l: d l lt:lc2 17 l: b l lt:ld4 18 lt:le1 (better 18 lt:lb5 e5 19 a3, with
unclear play. Now Black seizes the initiative and quickly achieves
a won position) 1 8 . . . lt:le5 19 a4 : ad8 20 ..tg5 l: d7 21 ..th4 h6
22 b3 g5 23 ..tg3 lt:lec6! 24 f1 l: fd8 25 lt:le2 lt:lxe2 26 xe2
lt:ld4+ 27 fl b5! 28 ab lt:lxb5 29 : xd7 : xd7 30 e5 lt:lc3 3 1 : b2
b5! 32 : c2 b4 33 : c l l: d2 34 : a t c4 35 lt:lf3 l: d8 0- 1 .
Flear-Ftacnik, Belgrade 1 989, went 8 ..tgS h6 9 ..th4 c 5 1 0
d5 b 5 ! 1 1 lt:lxb5 (ECO gives 1 1 'ifxb5 : b 8 1 2 'ife2 l:l xb2! 1 3 11Vxb2
lt:lxe4 14 : c 1 'ifa5, with advantage; if now 15 ..txe7 : e8 16 d6
lt:lxd6 17 'it'd2 : xe7 + 1 8 ..te2, Black has the very strong 1 8 . . .
..txc3 19 'ifxc3 'it'b5! 20 lt:l g 1 lt:lb4 2 1 fl : xe2! 2 1 lt:lxe2 ..ta6,
and although White is two exchanges up, Black's attack is
irresistible) 1 1 . . . 'if a 5 + 12 lt:ld2 l: b8 (theory recommends 12 . . .
lt:lxe4, but the rook move is more precise) 1 3 0-0-0 (after the
familiar 13 ..tg3 lt:lxe4 14 'ifxe4 : xb5 1 5 ..txb5 'ifxb5 16 'ife2
'ifxb2 1 7 0-0 lt:lb4, Black has excellent compensation for the
minimal material deficit. But in castling queenside, the English
grandmaster comes under a strong attack) 1 3 . . . lt:lg4! 14 1Wb3
(14 ..tg3 ..td7!) 1 4 . . . c4 15 ..txc4 ..td7 1 6 a4 : res 17 b 1
lt:l c 5 1 8 11V a 3 1Wxa4 ! 1 9 b 3 ..txb5! 0-1 . After 2 0 b a ..txc4 + 2 1
c 1 ..tb2+ 2 2 1W xb2 lt:ld3 +, White suffers big material losses.
Eingorn-H. Olafsson, Reykjavik 1 990, saw 8 ..tf4 c5 9 de 11Va5
(9 . . . ..te6 is also frequently played) 10 e5 lt:ld7 1 1 a3 1Wxc5 (72) .
This position was recently the object of a minor theoretical debate,
associated with three games.
12 1We4 (in Eingorn-Gavrikov, Tallinn 1 989, White played 1 2
lt:ld5, and after the inferior reply 1 2 . . . : e8 he quickly reached a
won position : 1 3 l: d 1 h6 1 4 h3 g5 1 5 1W xc5! lt:ldxc5 1 6 ..te3
etc. Soon afterwards, in Piket-Ivanchuk, Tilburg 1 989, Black
1 16 Russian System
72
w
d5
Russian System 1 1 7
73
B
. . .
10
11
ed
eel
: es (74)
The move 1 1 . . . ..tf5 might well transpose after 1 2 ..tf4 J:. e8;
alternative ideas after 1 1 . . . ..tf5 are discussed in Games No. 1 8
and 1 9.
After 1 1
b5 1 2 '1Vxb5! .:t b8 1 3 '1Va4 .:t b4 14 '1Vd 1 ! Black has
no compensation for the pawn.
1 1 . b6 is interesting -- the knight on a6 is at once freed
from its unwanted duty of defending the c-pawn. Tukmakov
Chiburdanidze, Biel 1 988, went 12 .:t d 1 ( 1 2 ..tg5 h6!) 1 2 . . . ltlb4
1 3 a3 ( 1 3 '1Vb3 ..tf5!) 1 3 . . . ..ta6 1 4 '1V b3 ..txe2 1 5 ltlxe2 ltlbxd5
16 ltlf4 c4 17 11t'xc4 '1Vc7 1 8 '1Vxc7 ltlxc7, and the position was
. . .
12 .t.f4
The fashionable development of the bishop on e3 (though with
Black having played 1 1 . . . .t.f5 instead of 1 1 . . . .l:l. e8) is covered
in Game No. 1 9.
After 12 l:tdl (to prepare a quick advance of the d-pawn) 12
.t.5, we reach a position that occurred twice in the Seville match
(75) . This is the appropriate moment to recall those games.
. .
75
w
Russian System 1 1 9
The 1 5th game o f the 1 987 match continued :
13 d6 h6
Not 1 3 . . . llJe4? 14 d7!
14 h3
In the 2 1 st match game, I played 14 .i.f4 (see below). A
suggestion of Gutman's is also interesting : 14 a3! llJd7 ( 1 4 . . . llJe4
1 5 .i.e3!) 1 5 1W a2 ! llJb6 1 6 .i.e3! .i.e6 1 7 1W b l . The queen has
escaped from pursuit, and the d-pawn continues to fetter the
opponent's pieces.
14 . . . llJb4 15 .i./4 llJ d7 16 : d2 a6 1 7 1Wb3 b5 18 1W d1 c4 19
a4! llJc5 20 ab llJbd3 21 .i.xd3 llJxd3 22 : xd3! cd
The bishop on f5 has to be left where it is, to restrain the
d-pawn; a weaker choice is 22 . . . .i.xd3 23 : xa6 : xa6 24 ba
1Wa5 25 1Wa4! 1Wxa4 26 llJxa4.
23 llJd5! ab! 24 llJe7+ <i;h7 25 : xa8 1Wxa8 26 llJxf5 gf 27
1Wxd3 'ile4 28 1Wxb5 : a8 29 .i.d2 : dB 30 1W c5 1W e6 31
.i./4 .i.xb2 32 llJh4 .i.f6 33 1W xf5 + 1W xf5 34 llJxf5 h5! 35 g4
hg 36 hg g6 3 7 g2 .i.b2 38 llJe7+ <i;f6 39 llJ c6 : d7 40 llJb8
: dB 41 d7 <i;e6 42 <i;f3 .i.a3 :f--J.
From diagram 75, the 2 1 st game of the Seville match went :
13 d6 h6 14 .i./4 llJ d7!
More accurate than 1 4 . . . llJhS 1 5 .i.e3, when the white bishop
has enticed the enemy knight onto the edge of the board and itself
settled in a good position.
15 : d2
Or 1 5 1W b3 llJb4 1 6 : d2, transposing; alternatively 1 6 .i.c4
1Wf6! 1 7 : d2.
15 . . llJb4 16 1W b3 .i.e6
Better than 1 6 . . . a6 1 7 a3 llJc6 1 8 llJdS!.
17 .i.c4 llJb6 18 .i.xe6 : xe6 (76)
.
76
w
77
8
13
.:. ad 1
li:Je4
..td3
15 . . . b4
. . .
. .
be
1 hb5
1ha6
..txc3
b5
ttJxc3
..txd3
'it'xd3
.l:l. fel
: xe1 +
lt)xe1
White has also a difficult position after 22 l:t xe1 lt)xd5 23 .l: d 1
lt)b6 2 4 9xf7 + xf7 2 5 .l:l. xd6 e7, but now Black succeeds in
increasing his advantage decisively.
22
23
.l: b8!
a3
Game No. 1 8
Belyavsky-Tukmakov
Odessa 1989
1 d4 lt)f6 2 c4 g6 3 lt)cJ d5 4 lt)f3 J.. g7 5 9 b3 de 6 9 xc4 0-0
7 e4 lt)a6 8 J.. e2 c5 9 d5 e6 10 0-0 ed
11
ed
J.. f5
12
J.. f4 (81)
. . .
82
w
Alternatives are 13
.l:t fe8 and 13
e8, but taking the b
pawn is risky; after 13
'if x b2 14 e4 'if b6 15 d6, or IS i.xf6
.i.xf6 1 6 .l:t ab l , Black is in great danger. On the other hand, in
Ivanchuk-Lputian, Irkutsk 1 986, White played the mistaken I S
. . .
d6
13
14
1i'b3
.!l:lb4
15
.J:l acl
Perhaps the queen should have returned home with 15 1i'dl.
15
.i.e6
Black has many possibilities - 15 . . . a5, 15 . . . .!l:ld3, 15 . . .
.i.d3 - but the move played is not bad either.
16
'ifa3
Again it was safer for the queen to return to its own camp : 1 6
W d 1 .!l:l 6d5 ( 1 6 . . . .i.xc3 1 7 be .!l:lxa2 1 8 .J:l c2 .!l:la4 1 9 Wd2) 1 7
.!l:lxd5 .!l:lxd5 1 8 .i.g3 .i.xb2 1 9 .J:l xc5 .!l:lc3 20 'ifc2 .!l:lxe2 + 2 1
'ifxe2 .i.a3 22 .J:l c3 .i.b4 2 3 .J:l c2, and White's chances are a little
better (Tukmakov).
16
.!l:lc4
.i.xc4
17
.i.xc4
18
.J:l fd1
b6
tL!e4
Here again, White had to play 1 9 'ifa4 a6 20 .J:l d2, and then
'ifa4-d l . It is becoming clear that the white queen is out of the
action.
19
20
.i.d5!
.!l:lfg5
.i.d2
Against the piece sacrifice 2 1 .!l:lxc5 be 22 .J:l xc5, Black has the
decisive 22 . . . 'ifb7!, since 23 .J:l cxd5 tL!xd5 24 Wf3 fails to 24 . . .
1i'b4 !
21
.!l:lxa2
22
23
24
25
26
ll a 1
ltlc3
..txc3
..txg7
xg2
h6
ltlxc3
..txg2
xg7
hg
1tg3
llfe8
1t xg5
h3
ll a3
.l:!. e4
.l:l. ae8
.l:t 8e5
1tg3
ll f3
ll d5
.1:1. 4
.l:t e8
a5
ll 4e6
30 . . . a5 is simpler.
31
32
33
34
1tb3
1Wf3
g1
0- 1
ll d8!
1tc6
.l:!. d7
.l:t exd6
Game No. 1 9
Bareyev-Lputian
Lvov 1990
1 d4 ltlf6 2 c4 g6 3 ltlc3 d5 4 ltlf3 ..tg7 5 1t b3 de 6 1t xc4 0-0
7 e4 ltla6 8 ..te2 c5 9 d5 e6 10 0-0 ed 1 1 ed ..tf5
.i.e3
85
w
13
l:t acl
The alternative is 13 l:t ad1 ; let us look at a few interesting
examples.
A game Ree-de Boer, played some time ago (Amsterdam 1 983),
went 1 3
.!D e4 14 .!Dxe4 l:t xe4 1 5 11fc1 .!Db4 16 a3 .!Dc2 17 .i.g5
f6 18 11fxc2 fg 19 .i.d3 lH4 20 11fxc5 g4 21 11fe3, and White gained
a clear advantage. An improvement is 1 5 . . . 11rb6 1 6 b3 l:t ee8 1 7
.!Dd2 .!Db4 1 8 .!Dc4, with a minimal plus for White.
Farago-Kozul, Montpellier 1 989, went 13
1i' b6 14 b3 l:t xe3 !?
( 1 4 . . . 11t'b4 leads to unclear play. In Ree-Chandler, Helsinki 1 984,
Black ended up in a difficult position after 1 4 . . . .!Dg4 1 5 -*.d2
l:t ad8 1 6 l:t fe 1 .!Db4 1 7 .!Da4 11rd6 1 8 .i.f4 ! 1Wf8 19 .i.c7 l:t d7 20
d6. Black also stands worse after 1 4 . . . .!De4 1 5 .!Dxe4 l:t xe4 1 6
11t'b5) 1 5 fe .!Dg4 1 6 .!Da4 11rd6 ( 1 6 . . . 11fa5 1 7 e4 .!De3 1 8 11rd3 .!Dxd 1
1 9 ef .!Dc3 20 .!Dxc3 is in White's favour) 1 7 11rf4 ! (better than 1 7
e4? -*.d7 1 8 11f c 1 b 5 1 9 .!Db2 .!Db4, with sufficient compensation
for the exchange, for example : 20 .!Dd3 .i.d4+ 21 <ot h 1 .!Dxa2,
with advantage to Black) 17 . . . 11fxf4 1 8 ef .!Db4 ! (but not 18 . . .
.!De3, in view of 1 9 .i.xa6 ba 20 .!Dxc5) 1 9 l:t d2 .!De3 20 l:t c 1 .!Dbxd5
. . .
. .
14
1Vb3
J: ed8
lL!h4 (86)
19
-*.a4
'lff6!
l0f3
-*.xf3
gf
-*.d2
b6
22
de
"ife3!
"ife4
ef+
l:l. cdt
'lfg4!
.I:Xd2
l:l. d6
l:l. b8
xf7
'lfg5+
t-t.
5 4 ..t f4
System
Game No. 20
Lukacs-Ftai!nik
d4
c4
ll:lc3
J. f4
ll:lf6
g6
d5
e3
c5
This system was used in the first game of the 1 986 match, but
with a different move-order (e2-e3 was postponed for three
moves) : 4 ll:lf3 .ig7 5 .if4 c5 6 de "ifa5 7 l:l c 1 (previously, 7 cd
had been played; Kasparov responds to the novelty very precisely)
7 . . . de! 8 e3 "ifxc5 9 'ira4+ ll:lc6 10 .ixc4 0-0 1 1 0-0 .id7 1 2
'irb5 'irxb5 1 3 J.xb5 l:l ac8 1 4 l:l fd 1 l:l fd8 1 5 h 3 h 6 1 6 fl a6
1 7 J.e2 J.e6 1 8 l:l xd8 + l:l xd8 19 ll:le5 ll:lxe5 20 J.xe5 l:l d2
!- ! .
de
Winning a pawn with 6 J.xb8 l:l xb8 7 'ira4+ .id7 8 'ir xa7 is
too risky: 8 . . . cd 9 'ir xd4 0-0 1 0 cd 'ira5 1 1 'ird2 b5! 1 2 .id3 b4
1 3 ll:lce2 'irxd5.
6
7
'ira5
l:l d (87)
88
B
J.xc4
de
0-0
Wxc5
tll f3
.t.b3
10 tll b 5 will be mentioned in the notes to Game No. 2 1 .
10
tll c6
. . .
. . .
13
1We2
The ninth game of the 1 986 match went 13 ll) d4 J.d7 14 Wel
ll)xd4 1 5 ed e6 16 J.dl "ii' b6, resulting in equal chances. Another
four moves - 1 7 l:l. fd 1 J.c6 1 8 J.e3 iraS 19 ..td2 Wb6 20 J.e3
WaS - and we agreed a draw.
An attempt at strengthening White's play was made in Peturs
son-Ivanchuk, Reggio Emilia 1 989/90 :
16 J.e5 J.c6 1 7 l:l.fd1 l:l.fd8 18 1We3 l:l. d7 19 1Wg5
It is worth considering the sharp 19 d5 J.xd5 20 J.xf6 J.xf6
2 1 ll)xdS ed 22 Wf3 ..txb2 23 l:l. xdS! Ivanchuk examines two
alternatives for Black: 19
ed 20 Wd4 ll)e8 21 ..txg7 .!L\xg7 22
J.xdS l:l. ad8 23 WeS J.xdS 24 l:l. xdS l:l. xdS 2S .!LlxdS, and 19
.!Llxd5 20 ..txdS J.xdS 21 J.xg7 xg7 22 .!Lle4! l:l. e8 23 Wf4
Wd8 24 .!LieS l:l. de7 (but not 24 . . . eS? 2S 1Wd2 l:l. d6 26 .!L\xb7 J.xb7 27
1Wxd6 irgS 28 f3! J.xf3 29 1W d2, with advantage to White) 2S
ll)e4 l:l. d7 - with equality in both cases.
. . .
. . .
liJe4
eS (90)
liJd5
In Rizhkov-Epishin, Leningrad 1 986, Black played 1 4 . . . liJc5
at once, which led to equality after 15 j_c4 e6 16 b4 1i'a3 1 7
liJe7 + liJxe7 1 8 be j_e4.
90
w
. .
. .
. . .
142 4 f4 System
a draw : 2S . . . fe+ 26 cot>xh3 'iPg8 27 f4 e2 28 fe ef('if 29 .l: xg7+
cot>xg7 30 'ifg6+ etc.
25 . . . .l: xf7! 26 'ifxj7 fiJg5!
After the game I discovered that the computer had indicated
26 . . . 'if bS here, but there would follow 27 fiJg6 + .,Ph7 28 fiJe7
'ife8 29 11t'xe6 fOgS 30 11t'fS + cot>h6 31 .l: h 1 ! and in spite of his
extra peice Black has no defence.
27 fiJg6+ cot>h7 28 fiJxe5 liJxf7 29 liJxf7 'it>g6 30 fiJd6 fe
The position has become simplified, the tension has abated, and
Black has even emerged with a slight edge, though its significance
is purely symbolic.
31 fiJc4 ef 32 .l: xf2 b5 33 fiJe3 a5 34 .,Pg3 a4 35 .l: c2 .1:/8 36
'it>g4 d4 37 .l: e2 xe3 38 l:t xe3 .1:/2 39 b3 .l: xg2+ 40 .,Pf3
.l: xa2 41 ba J-J.
15 .i.h2
fOeS
15
fiJf6 is very strongly answered by 16 .l: xc6! fiJxdS 17 .l: d6
fiJe7 1 8 e4, with a clear advantage; Pinter-Rogers, Szirak 1 986.
But a more popular continuation is 15
e6 16 .l: fd1 (92) . Let
us look at some interesting games with this line from recent years.
92
B
e4!?
. . .
93
w
17
"it'e3
ab
J:t c5
"it'xc5
lll xb3
J.. xe4
"it' xc5
20 lll f6+ J.. xf6 2 1 "ifxc5 .l:l. d5 22 "ife3 was more exact.
'lfe4
gf
J.g3
'lfb5
'lfxb7
: xd5
.i.xf3
: fd8
: dl
l0d4
l0e2 +
gl
xg3
'lfxa7
l0xg3
J.b6
: xb2
'lf b6
The position has levelled out, and the game gradually heads
towards a draw.
29
30
: at
: d3
g7!
: a6
'lfe7
gl
J.g5
J.f4+
: dl
'lf b6
l-t
Game No. 2 1
Belyavsky-Gavrikov
Moscow 1988
1 d4 l0f6 2 e4 g6 3 l0e3 d5 4 J.f4 J.g7 5 e3 e5 6 de 'lfa5 7
: ct de
94
B
.lt.:lf3
. . .
96
w
1 1 ll:\c7 (after this, Black seizes the initiative. Unclear play results
from 1 1 .t.xe6 11t'xb5 1 2 .i.c4 11t'xb2 1 3 0-0; if 1 2 .i.b3, then 1 2
. . . ll:\e4) 1 1 . . . .i.xc4 1 2 b3 (in Popchev-Lalic, Bosna 1 988, White
opted for 1 2 ll:\d2, but after 1 2 . . . b5 1 3 b3 ll:\d5 14 ll:\xa8 ll:\c3 1 5
.l:l. xc3 i.xc3 1 6 be 11t'xc4 ! 1 7 11t'e2 .i.xd2 + 1 8 11t'xd2 ll:\c6 1 9 11t'e2
11t'b4+ his position was hopeless) 1 2 . . . 11t'a5 + 1 3 11t'd2 11t' xd2+ 1 4
ll:\xd2 i.d3 1 5 ll:\xa8 ll:\d5 1 6 ll:\c7 .l:l. c8 1 7 ll:\f3 ll:\xf4 1 8 ef i.b2
19 d2 i.xcl + 20 .l:l. xc 1 i.e4 21 ll:\b5 .l:l. xc l 22 xc1 i.xf3 23
gf ll:\c6, and Black realised his advantage in the ending; Inkiov
Lputian, St John 1 988.
10
11
0-0
11t'aS
ll:\a6!?
ll:\eS
I3
i.. xf7 + !
ltJxf7
b4!
J: xf7
xf7
lll dS
1t' xdS+
1t'xb4
ltJxdS
ltJe6
Better than 17 . . . i.. e 6 18 1t'xc5 1t' xc5 19 J: xc5 i.. xa2 20 i.. e 5!
b6 2 1 J: c2 i.. b 3 22 J: c3 i.. x e5 23 .l:t xb3 aS 24 f4 i.. g7 25 J: xb6
a4 26 J: f2 a3 27 J: a2 i.. b 2, with advantage to White (Gavrikov).
IS
J: c4
Wb2 (98)
19
J:t fcl?
19
The point is that after 20 'it'xd7 J:t d8 21 'it'a4 b5!, Black wins. If
White had played 19 ..tg3, he could have answered 19 . . . ..td7
with 20 'it'xd7 J:t d8 21 J:t f4 + . However, in that case Black has 1 9
. . . ..tf6, and i f 2 0 J:t fc l , only then 20 . . . ..td7!
20
h3
White loses not only after 20 'it'xd7, but also after 20 J:t 4c2
..tc6!
ll d8
20
..tc7
J:t c8
.l:t 4c:Z
'irxb5
ib6
.l:t xc6
'ir b5
ixb5
.l:t c6
ixc6
Fianchetto System
Game No. 22
Karpov-Timman
d4
c4
.!0(3
g3
.!0(6
g6
.ig7
c6
cd
ltlc3
ltle5
cd
o-o
e6 (99)
99
w
. . .
ltlc6
13
:ct
. .
. . .
'1Vb4!
0-0
b3
c5
J:l. fd 1
..txd4
cd
..txd4
'lf xd4
J:l. xd4
J:l. dd1
'lf xd4
ll:lb6
..tb7
e4!
de
ll:lxe4
J:l. c5
g7
J:l. fd8
: xd8
lDc3
xg2
: xd8
J.. x g2
: c8?
White's positional advantage lies in his queenside pawn majority,
and Black has to play very accurately to avoid reaching a
dangerous ending. The correct move here was 26
ll d2!, and if
27 : c7 : cl 28 lDb5? : xc7 29 lDxc7 f6, Black is not at all
worse. A more precise line for White is 28 a4! a5 29 lDb5 : xc7
30 lDxc7 f6 31 lDa6 lDd5 32 lDc5 lDb4! 33 f4 (33 f3 e5!)
33 . . . lDc6 34 f3 e7 35 e4 d6 36 lDb7+ c7 37 lDc5
(37 lDxa5 f5 + !) 37 . . . d6, with a draw.
27
: xeS
lDxc8 (101)
. .
101
w
28
f4?
fS!
ltla4
>f6
.!ticS
ltld3
ltlb6
ltld7
ltlb4
I should have brought the king forward : 32 >f3 e5 33 fe +
ltlxe5 + 34 ltlxe5 >xe5 35 h4! with a draw.
32
33
ltlc6
34
35
ltld4
fe +
eS
a6
ltl xe5
e7
n
36 f2 is answered by 36 . . . etlg4 + .
36
37
38
39
40
d6
e2
etlcl
dS
e4
a4
b4
.!Of3
etld4+
lt!xd4
bS
xd4
t- t
After 42 . . . ab 43 ab c5 44 e3 xb5 45 f4 c4 46
g5 d3 47 h6 e3 48 h4! f3 49 xh7 xg3 50 xg6,
both sides' resources are completely exhausted.
Game No. 23
Karpov-Timman
c4
lt!f3
g3
J.g2
cd
.!Oc3
tOeS
g6
J.g7
c6
dS
cd
o-o
e6
o-o
103
8
11
f6
b6
12 d3
'fle7
13
b3
13
..td7 would transpose into the Nikolic-Nunn game that
we have already seen. So it may be said that we are only now
breaking new ground.
. . .
14
15
16
17
18
19
a4
.tel
e3
..ta3
l:t c l
..txf8
..td7
l:t fd8
..te8
'flf7
..tf8
'fi xf8 (105)
105
w
f5
gS
g3
26
Wet
An inaccuracy, committed after much thought. The sheer variety
of possibilities put me in rather a quandary, and I decided to
decline the pawn sacrifice although Black's position would almost
have become critical if I had calmly accepted it. After 26 xd5+
(26 .l:l. xd5 ed 27 ed .i.f7! 28 .l:l. xd8 + Wxd8 gives Black counterplay
for the pawn) 26 . . . h8, White has a wide choice :
(a) 27 xc6 J.xc6 28 1W c1 ed 29 e4 !, or
(b) 27 .l:l. fcl ed 28 e4! . This strategic operation involving the return
of the pawn and the inevitable break with e4-e5 secures White a
considerable plus.
(c) 27 de xeS 28 .l:l. xc7 1Wxc7 29 .l:l. c 1 'fle7 30 xeS 1W xeS 3 1 1W d4!
is also good.
(d) 27 1Wcl .l:l. xd5 28 .l:l. xdS b6 (28 . . . b4 29 l:t cS!) 29 de b4 30
xb4 .l:l. xcl 3 1 .l:l. xc l , or 27 . . . b6 28 .l:l. xc6 J.xc6 29 J.xc6 ed
(29 . . . .l:l. d6 30 dS .l:l. xdS 3 1 .i.xdS) 30 ed .l:l. xd4 (30 . . . .l:l. d6 31 dS
.l:l. xdS 32 .l:l. e 1 !) 3 1 .l:l. e 1 1Wd6 (3 1 . . . .l:l. xd3 32 .l:l. xe7 xe7 33 Wet
xc6 34 1We8 + g7 3 S hS+ h6 36 1Wf8 + and mates) 32
.l:l. e8 + g7 33 l:t e6 - with a very strong initiative for White in
either case.
(e) 27 .i.g2 (in some of the foregoing variations the white king's
: c2
lilf2
1i'd2
l:tfcl
e4
lt:ld6
.l: dc8
aS
lt:lb4
: c3
lild1
: xc6
: xc6
lilc3
f2
40
el
e1
d1
-.d7
: c6
: xc6
-.xc6
f8
e7
f8
-.cs
g7
lll a 2
11rxa2
lll x a2
11rc7
f2
1i' b2
.tel
el
f8
e7
d8
c8
d2
ct
b7
11re7
46 . . bS!? is sharper.
.
47
48
Ill eS
f2
11Vhl ! (107)
107
B
h4!
hg
11rxh7
11re7
11V b4
11V xb3
fg
xa4
11rc6
1hg5
'fll e7
'fll xd6
.t. dl
a4
'fll d6
ltlxd6
.tb5?!
It was hardly a good idea to block the path of his pawn. After
59 . . . b5, it is most likely that the game would soon have been
drawn; at any rate Black would not be risking anything. Here are
some sample variations : 59 . . . b5 60 f6 (60 ltle2? b4 61 ltlc1 c7)
60 . . . .txg4 61 .txg4 ! a3 62 .te6 c6 63 ltle2 (not 63 f7 ltlxf7
64 ltlxe4 ltlg5! 65 it:lxg5 a2, and Black wins) 63 . . . a2 (63 . . . b4 64 ltlcl)
64 .t.xd5 + xd5 65 ltlc3 + c4 66 ltlxa2 b4 (nor does he gain
anything from 66 . . . b3 67 ltlc1 + b2 68 ltle2 b4 69 d5 b3
70 g3 c2 71 ltld4+ c3 72 ltlxb3 xb3 73 f4 c4 74
e5) 67 ltlxb4 xb4 68 g3 c4 69 f4 d3 (69 . . . d5
70 g5 e6 7 1 g6) 70 e5 ltlf7 + 71 e6 ltlg5 +, with a
draw.
60
61
ltle2
a3
ltlcl (108)
108
8
c7
. .
1 70 Fianchetto System
67 g7 f7 68 f6 .tc4 69 .tg4 g8 70 .te6 + t;)f7 7 1 f5 .tb5
72 g6 .te8 73 .txd5 b5 74 .ta2 b4 75 .tb3.
63 .te2
But not 63 f4? t;)b2! threatening . . . t;)d3 + .
.te8!
63
The strongest defence. 63 . . . li)xe3 is inadequate : 64 .txb5 t;)c2
65 g5 t;)xd4 66 f6 d6 67 g6 e6 68 g7 f7 69 .te8 + g8
70 f4, and wins.
64
f4
Podgayets revealed a striking variation culminating in a
problem-like mate : 64 g5 t;)xe3 65 f4 t;)c2 66 e5 t;)b4 67
f6 d6 68 g6 t;)c6 69 ..tb5 e3 70 g7 t;)e7 7 1 ..txe8! e2 72
li)xe2 a2 73 t;)c3! t;)g8 + 74 f7 t;)h6+ 75 f8 a l (V) 76 t;)b5
mate (109) .
109
B
Fianchetto System 1 71
66 gS e7 (66 . . . lilc4 is a mistake in view of 67 ..txc4 de 68
f6 ..ta4 69 g7 ..tb3 70 f6 c3 7 1 lilxc3 a2 72 lilxa2 ..txa2 73
gS) 67 h6 f6 68 lilb4 ..tf7 (inadequate alternatives are 68 . . .
ll:ld3 69 lilxdS + f7 70 gS a2 7 1 g6+ f8 72 g7 + f7 73 ..thS+
g8 74 lll e 7 mate, and 68 . . . ..tc6 69 ..tfl ! lildl 70 gS + f7 70 . . . xf5 7 1 ..th3 mate, is pretty - 7 1 g6+ etc.) 69 ..tfl ! lild l
(it is too late for counterplay with 69 . . . lild3 70 gS + e7 7 1
lilc2 a 2 7 2 g7) 7 0 g5 + e7 7 1 g7 lilxe3 7 2 f6 + e6 73
..th3 + lilfS + 74 ..txfS + xf5 75 xf7 e3 76 g7! e2 77 ll:lc2
a2 78 f7 e 1 (1i') 79 lilxe l a 1 (1i') 80 f8(1i')+. However, after the
correct 65 . . . b5!, nothing can be found for White.
lild3
65
ll:lal
69
-*.b3
bS
..tdl
b4 (IJO)
Bringing the king a little closer to the pawns does not work :
68 . . . d6? 69 ..tb3 lilcl 70 h6 b4 7 1 f6 e6 72 g7 lilxb3
73 lilxb3 ..ta4 74 lilcS + ! d6 75 f7 a 1 (1i') 76 f8(1i')+ c6 77
1i'c8 + d6 78 'ifd8+ c6 79 1i'd7 + b6 80 1i'b7 + aS 8 1
1i' a6 mate.
1 10
w
lilcl
..tc4
..tbS
1 72 Fianchetto System
d5 74 a1 c4 75 .tf6 c3 76 e7 b2 77 xeS xa1
78 f6 b2 79 f7 a 1 (W) 80 f8(W) Wa4+ 81 f7 b3 + . We had
reached this position in our analysis, and Zaitsev suggested 82
.tf6 W xe3 83 e5 .tc2 84 Wf4 .td3 85 d5 with chances of
success. But in the post-mortem, Timman pointed out the more
effective 82 d5! Wxe3 (82 . . . W xd5+ 83 'Oftg6 with a technically
won ending) 83 W b4 + ..tt c 2 84 g5 Wf3 + 85 g7 e3 86 Wc4+
b2 87 W b5 + c1 88 d6 e2 89 Wc5+ b2 90 Wb6+ c1 91
d7 e l (W ) 92 d8(W) Wec3 + 93 Wdf6 etc. According to computer
analyses, this ending is a win for White.
e7
72
..tg8
73
h6
gS
g6
b3
1-0
Haninge 1990
I have played the White side of the Griinfeld so often, especially
the g2-g3 system, that when the symmetrical variation arose in
Fianchetto System 1 73
the present game, it was with some interest that I handled it for
the other colour.
lt:l f6
1
lt:lf3
2
3
4
5
c4
g3
.i.gl
d4
g6
.i.g7
0-0
c6
8
9
10
11
lt:lc3
cd
lt:l e5
0-0
f4
.i. e3
d5
cd
e6
lt:lfd7
lt:lc6
lt:lb6 (1 1 1)
111
w
1 74 Fianchetto System
Another possibilty is 14 a4! de 15 aS .!Dbd5 16 .!Dxe4 .:t b8 17
'iVb3 ..te8 18 .l:tfcl .!Dc6 1 9 'iVa3 .!Dcb4. Nikolic-Hulak, Zagreb
1 987, continued 20 .:t c4? .!Da6 2 1 .!Dd6 .!Dac7 22 .:t ac l .!DbS 23
.!DxbS ..txbS 24 : cS ..te8 2S b4 b6, and White achieved nothing;
but Gutman's suggestion 20 .!Dc3! gives him the advantage. At
move 1 7, an alternative is 1 7
..tc6 18 .!DeS .!Dc7 19 ..txc6! with
the initiative; this is stronger than 18 .:t fc l a6 ( 1 8 . . . .!Dc7! at once
is more precise) 19 .:t c4 ( 1 9 .!DeS! is correct) 19 . . . .!Dc7 20 .!Dc3
..txg2 21 xg2 .!bedS, with equality; Andersson-Hulak, Wijk
aan Zee 1 987.
14 . . . 'iVxd7 15 e5 .:l.fc8 16 :t el ..t/8 17 ..tj3 .:l. c7
An accurate move which equalises. I would have answered the
incautious 1 7 . . . a6 with 1 8 ..te2 .:t c7 1 9 g4 ! ..th6 20 'iVd2, and
if 20 . . . .!Dc4, then 2 1 ..txc4 : xc4 22 .!De4 ! . But now this idea
doesn't work : 18 ..te2 : ac8 19 g4 ..th6 20 'iVd2 .!Dc4 21 ..txc4
.:t xc4, and both black rooks are on the c-file.
18 b3 : ac8 19 'iVd2 lbc6 20 'iVb2 a6 21 ..te2 'iVe7! 22 l1Jb1 liJb4
23 lbc3 lbc6 24 l:lJb1 lbb4 25 .:l. c5 t:tJd7 26 : xc7 : xc7 27 lbc3
l:lJ c6 28 l:lJb1 t:tJb4 29 l:lJc3 lbc6 30 lbb1 i-J.
In the third game, Kasparov varied with 1 2 . . . CfJe7! (a refinement on
12 . . . ..td7 of the first game) 13 a4 aS 14 'iVb3 (Black would have
had more problems after 14 e4 de l S ..txe4 .!DbdS 16 'iVb3.
Greenfeld-Birnboim, Tel-Aviv 1 988, continued 16 . . . f6 17 .!Dc4
h8 1 8 .:t fe l CfJb4 1 9 .:t ad 1 .!bedS 20 .!De3 : a6 2 1 ..txdS ed 22
.!DexdS ..tg4 23 .:t d2 .:t e6 24 .!Dxb4 .:t xe l + 2S ..txe l ab 26 'iVxb4
: e8 27 ..tf2, and White gained the advantage) 14 . . . ..td7 1 S
.:t fc 1 ..tc6 1 6 .!DbS .!Dbc8! 1 7 e 3 .!Dd6 1 8 .!Dxd6 'iVxd6 1 9 .tel
.:t fb8 20 ..tfl f6 2 1 t:tJf3 'iVd7 22 'iVc2 .!DfS 23 ..td2 .!Dd6 24 b3
.:t c8 2S 'iV d l h6 26 .te l gS 27 : a2 'iVe8 28 : ac2 .U8 29 ..td3
g4 !-!.
. . .
12
b3
..td7
f6
'iV d2
In Portisch-Korchnoi, Black first played 1 3 . . . .:t e8 1 4 .:t fc l ,
and only then 1 4 . . . f6. After l S .!D d 3 .:t e7 1 6 h 1 ..te8, the
players agreed a draw. Evidently on that day they were just not
in the mood for a fight; the result has nothing to do with the
Fianchetto System 1 75
position. In the present game, I carry out a regrouping in the
centre as Black, without wasting time on rook manoeuvres.
14
15
16
IL!d3
.l:t acl
ILleS
IL!d6
-*.f2 (1 12)
JJ2
B
IL!e5
h1
g4!?
f5
"W/e7
.l:t fc8
IL!xe4
IL!xd7
IL!e4
fe
1 76 Fianchetto System
The knight cannot hold out on eS for long, but perhaps it was
worth exchanging it for the black one : 2 1 .l:. c3 xeS 22 de .l:. xc3
23 'ii' x c3 .l:. c8 24 'ii' d 2 b6 2S .tg3, intending f4-fS.
2 1 .tg3!? at once is also interesting : 2 1 . . . xeS 22 de .l:. xc l
23 .l:. xc l .l:. c8 24 .l:. xc8+ .txc8 2 S 'ii' c 3 (White's aim i s t o carry
out f4-fS, but the situation does not yet call for it : 2S fS? ef 26 gf
gf 27 'ii' xdS + ..te6, or 2S 'ii' d 4 b6 26 fS ef 27 gf gf 28 'ii' xdS +
J. e 6 2 9 'ii' a8 + f7, with a good game for Black i n either case)
2S . . . 'ii' d 7 (but not 2S . . . 'ii' d 8 26 .th4!, or 2S . . . .td7 26 'ii' c 7)
26 e3 b6 27 a4 .tb7, and again the chances are about equal.
21
22
'ii' xd7
a3 (1 13)
l /3
B
.th3
aS!
'ii' e7
.l:. al
d8
fS
gS
.tg3
c6
e3
fe
.:. rs
1We2
.:t d8
1Wxe6
h6
.:. n
: act
.:t xc6
.:t bl
lll e7
.:t dc8
.:t c6
1Wxc6
. .
34
35
36
37
38
..tn
1W b2
.tel
.tel
..tf8
lll c8
1Wd7
..td6
.:. cl
lll e7
: f8!
1 78 Fianchetto System
40
g2
b4
.-c2
a4
.-f7!
.-dl
43
44
45
46
gh
J.xh4
hS!
tt:Jh4 + !
Fianchetto System 1 79
46
47
48
et
gh
h3+ !
hl
x2
: cl
hl
2
: xfl
: xhl+
: gl+
: fl
hl
hl
hl
fl
:ci
..tg3
h7
h6
: hl +
: 2
..td6
: gl +
: g8
: g3
Game No. 25
Vaganian-Kasparov
Barcelona 1989
1
2
3
4
5
6
103
c4
g3
..tg2
10c3
cd
106
g6
..tg7
0-0
d5
10xd5
1 15
w
Here White has the choice between the slow e2-e3, as in the
actual game, and the more committal d4-d5 and e2-e4. Let us
first consider the latter.
9 d5
Another possibility, seen more rarely, is 9 ..tf4.
9 . li)a5 10 e4
The modest 10 'ii' c 2 quickly leads to equality; here are two
examples :
Tukmakov-Halifman, Simferopol 1 988, went 10 . . . c6 1 1 de
<'i) xc6 12 .l:l. d1 ..tf5 1 3 e4 ..td7 14 -*.f4 .l:l. c8 1 5 'ii' e 2 'ii' e 8 1 6 h3
-*.e6 17 <'i)d5 ..txd5 18 ed <'i)b4 19 <'i)e1 'ii' d 7 20 'ii' d2 <'i)a6 !-! .
Plachetka-Smejkal, Trnava 1 989, varied with 1 2 . . . 'ii' c7 1 3
<'i)b5 'ii' b8 1 4 ..tf4 e 5 1 5 ..te3 ..tf5 1 6 'ii' c5 .l:l. c8 1 7 <'i)xa7 <'i)a4 1 8
<'i)xc6 be 1 9 'ii' e 7 t-t.
In Nikolic-Kasparov, Skelleftea 1 989, White played instead 1 0
..tf4, and there followed 1 0 . . . c6 1 1 de 'ii' xd 1 1 2 .l:l. axd 1 <'i)xc6
1 3 <'i)b5 ..tg4! 14 b3 e5! 1 5 ..te3 e4! 1 6 <'i)fd4 lDxd4 1 7 <'i)xd4 <'i)d5
1 8 h3 <'i)xe3 19 fe ..td7. Black's position is the more pleasant,
though the game was eventually drawn.
.
i. g5
11 . . . h6
Black failed to j ustify 1 1 . . . i.g4 in Hansen-Kasparov, Thessa
loniki 01 1 98 8 : 1 2 h 3 ..txf3 l 3 1W xf3 cd 14 tL!xd5 tL!xdS 1 5 .J:I. ad 1
1W c7 1 6 ed ..txb2 ( 1 6 . . . tL!c4 1 7 1We2 .J:I. fe8 1 8 .J:I. c 1 b 5 1 9 b3 "ii e 5
20 "ii xe5 tL!xe5 21 d6, and Black is in a bad way) 17 .J:I. fe 1 .J:I. fe8
18 l: xe7! .J:I. xe7 1 9 d6 .J:I. e 1 + 20 .J:I. xe 1 "ii xd6 21 .J:I. e7 .J:I. f8 22 "ii d5,
and although Black did manage to save himself, his position does
not inspire much confidence.
12
i./4
cd 13 ed
tL! ac 4 ! . In Pastircak
forward
13 . . . tL!ac4 14 1We2 g5
If 1 4 . . . tL!xb2, then 1 5 tL!e5 ! .
15
i. c l
e6! 16 de
Better 1 6 .1:1. d 1 , although 1 6 . . . ed 1 'l tL! xd5 tL!xd5 1 8 "ii xc4 tLJ b6!
1 9 "ii b3 W f6 gives Black slightly the better game (Popovic).
.J:I. d8
22
tL! c6 be
9
9
. . .
e3
e5
10 d 5 tL!a5 1 1
tL!d4 i.d7. I n Saloniki, Portisch played 1 2 "ii e 2, which led to
e4
c6
.i.gS
14
.te3
ed
f6
cd
.tg4
b3
l U7
fS
.tb4 (11 7)
117
B
be
e4
J: c l
1i'd7?!
19
19
or
.!Oxc4 20 .!O xe4 fe 2 1 J: xc4 .txf3 22 1i' d 2 11' d7, with equality;
19 . . . ef 20 .txf3 .txf3 2 1 11' xf3 2 .!O xc4. The latter, however,
20
21
22
23
24
.!Ob1!
.txf3
11'xf3
g4
1i'b3
ef
.txf3
f4
11'a4
24
25
f3
h5 ( 1 18)
26
27
h3
1i'd3
J: e8
/ 18
w
Index of Variations
1
d4 lL!f6 2 c4 g6
Seville Variation
3
lL!c3
4
cd
5
e4
6
be
7
.tc4
8
lL!e2
9
.te3
10 0-0
lO . . . ed 1 1
f3
II
12
.txf7 +
13
fg
14
xfl
1 4 . . . ed
d5
lL!xd5
lL!xc3
.tg7
c5
lL!c6
o-o
.tg4
lL!a5
.l:txf7
l:t xfl +
'ii' d6
ed 'ife8 15
. . 'ii' d 7 34
1 5 . . 'ii' b6 35
1 5 . . . e5 1 6 de 3 7
1 6 .:t e l 37
16 d 5 37
1 4 . . . 'Wd7 1 5 de 13
1 5 g5 'ii' e 6 13
1 5 . . . .:t d8 1 4
15
15
1 5 h3 1 4
14
. . .
1Ve8 15
15
15
e5
28
g 1 .:t d8 29
1 5 'if a4
'it'd5
1 5 . . . 'it'e6 15
16
..tf2
1 6 ltlf4 16
J:l. 8
16
1 6 . . . J:l. d8 1 7 ltlf4 19
17 1i' e 1 19
17 'it' a4 ltlc6 20
1 7 . . . ltlc4 1 8 ltlf4 20
1 8 g5 20
1 1 . . . : r8 21
17 . . . b6 18 "ifc2 J:l. c8 21
18 . . . 'it' c4 22
1 8 . . . : rs 1 9 ..tg1 22
1 7 'it'c2 J:l. c8 25
1 7 . . . 'it'c4 1 8 g5 25
1 8 1i' e4 25
1 8 'it'b2 25
17
g5
1 7 g1 16
'it'f7
17
1 7 . . . 'it'e4 1 7
18
'it'el
h6
1 8 . . . 'it'f5 1 7
1 8 . . . ltlc4 1 7
19
gh
18
47
47
1 2 . . ..td7 1 3 h4 h 5 48
.
1 3 . . . e5 48
1 3 . . . : ac8 48
13 . . .
11
12
13
14
15
16
17
f3
..td3
cd
:cl
'if a4
d5
aS
cd
..te6
..txa2
..te6
..td7
: res
'ii' b4
1 7 'if a 3 b5 42
e6
17
1 7 . . . b5 42
1 7 . . . b6 42
18
: fdl
1 8 de 43
1 8 d 6 43
1 8 c3 43
1 8 d4 44
18
ed
44
7 ..te3 c5
f3
8 'ii' d 2 cd 84
8
. .
. 'if a 5 90
8 . . . 0-0 9 d5 9 1
9 : c 1 91
9 f3 c d 93
9 . 'if aS 1 0
. .
9 ...
l:t c 1 93
..t g4 1 0 g5 104
50
8
9
c5
0-0
cd
: bt
..te 2
9 . . . ..t g4
53
'ilf a5 1 0 l:t b5 53
1 0 0-0 'ilfxc3 53
10 . . . 'ilf xa2 1 1 ..tg5 53
9 . . . b6
1 0 ..te3 57
1 0 0-0 ..t b7 1 1 W' d 3 5 7
9 . . . tt:J c6 10 ..te3 61
1 0 d 5 61
10
'ilfa5+
cd
11
..t d2
1 1 'i!f d 2 67
11
W'xa2
12
0-0
..tg 4
1 2 . . . lZJ c6 68
1 2 . . . b6 1 3 'ilf c l ..tb7 68
9 ...
1 3 . . . 'ilf e6 1 4 ..tc4
69
. tt:Jd7 1 3 n a t 71
12
..
12
. . . W' e 6 1 3 W' c2
1 3 ..t b4
13
..tg5
71
77
79
Russian System
tt:Jc3
3
d5
4
tt:Jf3
..tg7
5
'ilfb3
de
6
'ilf xc4
0-0
7
e4
lZJa6
7 . . . ..tg4 8 tt:Jg5 108
8 tt:Je5 108
8 ..te3 tt:Jfd7 9
9
8
8 W' b 3
8 ..tg5
8 ..tf4
8
9
..te 2
1 15
1 15
1 15
c5
d5
: d l 109
W' b3 lZJb6
111
de
1 16
9
10
10
.tg5
10
11
e6
0-0
116
ed
.t f5
ed
1 1 . . . b5 1 1 7
1 1 . . . b6 1 1 7
11
..
1 2 .l:t d 1 1 18
1 2 .tf4 b6 121
12 . . . .tf5 - see
.l:t e8
12
below
.tf4
1 2 l: d 1 l: e8 1 18
.tg5
1 2 .te3
12
132
132
: es
12
1 2 . . . 'i!V b6
1 2 . . . tt:Jd7
13
13
13
126
129
l:t adl
tt:Je4
. . . 1V b6 121
. . . tt:Jd7 121
14
.td3
14 tt:Jb5 121
14
15
.txc3
be
125
4 .t f4 System
3
4
tt:Jc3
.t f4
e3
de
d5
6
de
.tg7
c5
6 .txb8 135
6
7
'i!Va5
: ct
7 cd 136
7 'i!V a4 136
135
. . .
Fiancetto System
.tg7
tt:lf3
c6
g3
4
4 . . . d5 5 cd tt:lxd5 6 tb c3 0-0 7 0-0 tt:lc6 8 d4 tt:lb6 9 d5 180
9 e3 ll e8 181
e5 182
9
d5
.tg2
5
. . .
cd
cd
ltl c3
0-0
tt:le5
e6
8 . . . .tf5 153
0-0
9
9 .tg5 154
9
9 . . . c6 160
10
tt:lfd7
f4
1 0 tt:lf3 160
10
10 . f6 160
10 . tt:lxe5 161
tt:lc6
. .
. .
11
.te3
1 1 . . f6 164
.
tt:lb6 1 73
'Si nce his surprise choice of the GrOnfeld Defence in the 1 986
World Championship match, Gary Kasparov has regularly
adopted the GrOnfeld in tournaments and matches around
the world. Against Karpov, however, this opening has proved
an unfortunate choice a,nd Karpov has won more World
Championship games against Kasparov in this opening than
in any other. Highlights of this book i nclude:
a me i n Action
:>en Game in
London Wl H OAH
ISBN 0 -7 1
-'=-PUBLISHER'sJ
\
net
prlca
,l
111