Professional Documents
Culture Documents
Burgess New Ideas in The Alekhine
Burgess New Ideas in The Alekhine
Graham Burgess
(I)
An Owl Book
Henry Holt and Company
New York
Henry Holt and Company, Inc.
Publishers since 1866
115 West 1 8th Street
New York, New York 100 1 1
10 9 8 7 6 5 4 3 2 1
Introduction 4
Symbols 5
1 White does not play 2 e5 6
2 2 e5 Miscellaneous 15
3 Chase Variation 21
4 Unusual Fourth Moves for White 27
5 Exchange Variation 33
6 Four Pawns Attack without 6...lbc6 42
7 Four Pawns Attack with 6...lbc6 46
8 4lbf3: Introduction and 4...dxe5 5lbxe5lbd7 54
9 Alburt Variation: 4...g6 61
10 4....tg4: Old Main Line 78
11 Flohr-Agzamov Variation: 5...c6 85
12 Kengis Variation: 4... dxe5 5lbxe5 g6 104
13 The Main Line: 6 .tc4 112
Index of Variations 128
Introduction
2) Lines to which I can add some I would like to thank the follow
thing original. After all, that's what ing people: Robert Timmer, Peter
an author's for! Whayman, Ian Rogers and Jonathan
3 ) Lines in which a lack of de Tait for generously supplying mater
tailed knowledge is disastrous. This ial; John Henderson, for sending me
is self-explanatory ! As a pleasant a copy of his large Alekhine data
side-effect, this means that I am fo base, although I only really referred
cusing on the more exciting lines. to it in the few cases where I was
4) Linesfor which systematic cov short of good material from my own
erage is possible. I shall be referring sources; John and Petra Nunn for ef
back to TCA quite frequently, but ficient typesetting; and finally my
this book can be used on its own as a mother for proofreading on Christ
guide to the main new systems in the mas Day !
Alekhine Defence - for these lines I
have provided the 'essential knowl Graham Burgess
edge' . London, February 1996
Symbols
avoid this transposition : 4 'ii'f3 lDc6 the position after 6 bxc3 work for
allows White to secure a pleasant ad Black. For example, 6...lbc6? 7 d4
vantage with 5 lbxe4 lbd4 (the reply lDxe5 8 f3 lbc5 9 lDg3 leaves the
5 . . . lbxe5 ?? loses a piece, for exam black knights well and truly forked.
ple to 6 'ii'g 3 +-) 6 'ii'c 3 dxe4 7 lDe2 Then Bagirov played the blindingly
;!; (Adams). simple 6 lbc5 7 d4 lbca6 8 lDf4 c6
...
5
••. lDc6 ought to be good for White, but
It has been known for a long time should be compared with the note to
that the reply 5 dxc3 is not a blun
..• Black's 5th move. The only practical
der, as once was claimed: 6 1lt'a4+ example hardly tested Black's idea:
(apparently Black resigned at this 7 . . . .tg4 8 d4 e6 9 h3 .th5 1 0 'ii'b3
point in a game Grondechevsky l:tb8 1 1 g4 i... g 6 1 2 lDf4? lbxc3 0-1
Okhlin, Moscow 1 973, which may (a distinctly premature resignation)
have done much to perpetuate the P.Maarten-J.Petrov, Halle U-20 Wch
myth) 6 . . . lbd7 7 'iWxe4 lbc5 is at 1995.
least OK for Black. However, until a2) 6 .tg4 and now:
...
recently this was regarded as merely a21 ) 7 cxd4? ! .txf3 8 gxf3 l/!g5
a curiosity, since no one could make 9 .tg2 lbe6 10 f4 lbcxd4 II 1i'a4!
/0 White does not play 2 e5
�d7 1 6ll'lf4 i.e7 (Black has organ 'it>f2 ll'lc8 26 ll'lxc8 �xeS 27 llc4
ised his position and has a firm grip 'it>c7 28 l:tg3 b5 29 l:tc3 g6 gave
on d5 ; he will now look for play on Black a pleasant ending in J.S 0ren
board sides of the board) 1 7 �e2 a5 sen-Bagirov, Berlin 1 992.
1 8 a4 g5 19 ll'lh3 h6 20 f3 ll'ld5 2 1 b22 1 2) 14 0-0 0-0-0 15 i.e3 i.d5
ll'lf2 ll'lb6 2 2 axb5 cxb5 2 3 ll'le4 16 f4 (the fact that Black cannot
l:thb8 24 ll'lc5+ �e8 25 g4 i.g6 26 meet this with .. .f5 means that some
h4 gxh4 27 f4 a4 28 i.a2 b4 29 l:tcfl thing has already gone wrong for
a3 30 i.b3 axb2 3 1 l:tb1 h5 32 d5 ? Black) 1 6 . . . e6 1 7 f5 l:td7 1 8 ll'le4
(this is disastrous, underestimating i.e7 19 fxe6 fxe6 20 l:tf3 ± J.S0ren
the power of Black's passed pawns ; sen-Kindl, Berlin 1 992.
instead the continuation 32 g5 l:ta1 b222) ll bS (I prefer this move)
•..
lt'Jxd4 17 i.xd7+ 'it'xd7 1 8 'ti'xd7+ greedy, and may offer White danger
�xd7 1 9 llf7+ �c6 20 i.e3 looks ous compensation, but it's entirely
risky for Black. conceivable that Black's defensive
c) 12...a6! 1 3 e6 axb5 14 exd7 + resources are sufficient (read: Fritz
�xd7 1 5 'Wxb5 �c8 looks fine for found an answer to everything I tried
Black. for White here ! )
12 ... i.xe6 a2) 1 4 ..a 6 1 5 i.xc6+ bxc6 1 6
.
12 ... fxe6 1 3 'it'e4 was assessed as 'it'xc6+ (of course not 1 6 lt'Jxc6??
± by Stoica, but I am not convinced i.d7) 16 ... 'iixc6 17 lL'lxc6 i.d5 gives
that 1 3 . . .lt'Ja5 ( 1 3 . . . g6? 14lt'Jg5 e5 1 5 Black risk-free winning chances ow
1\Vd5 1Wc8 1 6 fxe5 e 6 1 7 1li'f3 i.e7, ing to his bishop-pair.
W.Watson-Neil, British Ch (Nor b) 14 lt'Jxc6 i.d7 miraculously
wich) 1 994, and now 1 8 'ii'f7+ �d8 recovers the piece:
1 91Wg7 l:tf8 20 lL'lxh7 is simple and b l ) 15 lt'Jd4 c6 and here are a
strong- Pein) 14 i.xd7+ 'it'xd7 1 5 couple of sample variations: 16 i. e2
lt'Jg5 e5 16 fxe5 lL'lc6 i s i n White's fa- b5 17 lL'lxb5 (White cannot hold on
vour. to the knight) 17 ... cxb5 gives Black a
13 lt'Je5 i.d7 pleasant game; 16 i. c4 'ir'xg2 ! 1 7
Here Fritz came up with the star llfl b 5 1 8 lL'lxb5 'i!Ve4+ 1 9 � f2 (only
tling resource 13.. .'ii'd 5 ! (D). move) 19 . . . 'ihf4+ 20 'ite 1 1We4+ is
This is the sort of move computers good for Black, who could obvi
find far more easily than humans - ously take a draw if he wished.
14 White does not play 2 e5
that keeping up to date with a main 7 c4li::Jx b3 8 axb3 gave rise to an un
line just in case you reach it via a very usual variety of Four Pawns Attack
rare transposition isn't your idea of in Xheladini-A.G.Panchenko, B ern
fun, it makes sense to play some 1 994. Then 8 . . . dxeS 9 fxeS cS 10 dS
thing else. The following lines are e6 1 1 li::Jc 3 exdS 12 cxdS 'iih4+ 1 3
my suggestions: g 3 'iid4 1 4 li::Jf3 'iix d 1 + l S �xd 1
c 1 ) 3 c6 makes a good deal of
••• .tg4 gave B lack a very satisfactory
sense if you meet 3 d4 d6 4 .llc4 with position.
4 . . .c6. d) 3 g3 has been popularised by
c2) 3 li::J b6 4 .ll b 3 (D) and now
••• Rozentalis, who also introduced 1
Black can play: e4 cS 2 c3 li::Jf6 3 eS li::JdS 4 g3 into
grandmaster practice. After 3 d6 4 •••
B
li::Jf3 'ii'e6+ 8 'ii'e2 'i!Vxe2+ 9 �xe2
.tfS = 10 c4? ! li::Jb4 l lli::Je 1 ? ! li::Jxa2 !
1 2 .td2 ltla6?? 1 3 l:t.xa2 and White
wins, Ulfarsson-Halldorsson, Haf
narfjordur 1 99S.
d2) 4 cxd6 S .tg2 lLlf6? ! (Sha
•••
avoid, since 6 exd6? c4 would be very li::Jf3 'iie 7+ 7 'ii'e2 c6 8 li::Jc 3 ltla6
embarrassing, while 6 c3? ! c4 forces gave Black full equality in the game
the game into obscure territory: 7 Disconzi da Silva-Lima, Brazilian
.txc 4 li::J x c4 8 'ii'a4+ li::Jc 6 9 'ii'xc4 Ch (Brasilia) 1 99S) 6 . . . dS 7 .!Df3
dxeS or 7 .t c2 dxeS 8 li::JxeS 'ii'dS. .te7 8 0-0 .tfS (8 . . . 0-0 9lDeS.!Llbd7
2 e5 0.d5 Miscellaneous 17
22 ... b3+ 23 'ifi>e2 'ii' b 5+ 24 'ifi>e 1 .l:tc4 bxc6 9 h3 .i.h5 10 g4 .i.g6 1 1 'ii'e 2
0-1 S.Madsen-Granberg, Danishcorr. e6 12 0-0-0 d5 (this position ought to
Ch 1 9 8 1 (comments based on notes be quite reasonable for White) 13 h4
hy Niels Granberg). h5 14 .:tdg1 'ii' b 8 15 ltJd2 hxg4 1 6
4 dxc3 'Wxg4 'ii'b 5 17 .i.g5 .l:tb8 1 8 b3? 'ii'a5
4 bxc3, is also normally answered 1 9 ltJc4 'Wxc3 0- 1 Rahls-Bagirov,
nowadays by the solid 4 ... d5: Kusadasi 1 990.
a) 5 d4 .i.f5 (5 . . . c5 is preferable) c) 6 .'ii'd7 !? (intending . . . 'ii' f5) 7
.•
1 8 0.a5+ 'iti>a7 ! 1 9 0.c6+ 'iti>b6 -+) the knight eyes the c5-square, cer
14 ... bxc4 15 0.xf7 l:txd5 16 'Wih5 gxf4 tainly has its points.
1 7 'iixf5+ l:txf5 1 8 0.xh8 .tg7 and 13 0.h4 .th7
Black won shortly in a computer After the game Agdestein sug
game Fidelity-Mephisto. gested the move 13 .te4 as a possi
...
'Wa4+ �f8 22 axb3 'ii'xb3+ 23 'ii'xb3 tion: 20 'Wc6+ (20 hxg6!?) io...�f8
:xb3+ 24 �a2 l:.a3+ 25 �b1 .txc2+ 21 hxg6 Wbs 22 'i'f3(22 'ii'xb5 .:Xb5
20 2 e5 lbd5 Miscellaneous
This is rather a quirky system. White tactics are the rule, rather than the
d1arges his c-pawn up the board to exception. The conclusion at the mo
L·), an d while this costs no tempi ment is that B lack has nothing to fear
(since meanwhile Black is playing in the critical lines, but strong nerves
.. li1d5 -b6-d5), it is hard to see how are definitely required, with a sense
1 his can improve his position. There of the bizarre a desirable optional ex
arc glaring holes on d3, d4 and d5, tra!
and t he pawn will become exposed
on c5 . In fact, the first time I read Game 3
about Alekhine's Defence in some Sveshnikov - Khmelnitsky
general book, 4 c5 was given a ques- Sibenik 1 990
1 ion mark, with the explanation that
Black's knight could return to the 1 e4 tllf6 2 e5 tlld5
centre .. . 3 c4 lllb6
So what is White's idea? The in 4 c5 llld5 (D)
iiiat point is that White will argue
I hat the d5-knight is a target, and so
I he c5-pawn's real role is in cutting
w
oil th e knight's retreat. Eventually
Black will have to resolve the ques-
1 ion of the d5-knight - if by ex
d�anging knights on c3, then this
will open lines for White, quicken
lllg h is development.
The pawn on c5 itself may also
prove a tasteless morsel if Black
grahs it by . . . e6 and . . . �xc5 , if White
··an gain time on the bishop and at 5 .tc4
lack the dark squares it has neglected 5 lllc3 is the main alternative:
• •n th e king side. If Black plays . . . d6 a) 5...e6 6 d4 gives B lack the
or ... b6, then an exchange of pawns problem of avoiding a line of the c3
will open lines. Sicilian (which could easily arise
Black must not underestimate after 6 . . .d6 7 cxd6 cxd6). One idea is
1 he Chase Variation. Some precise 6 . . . lllxc3 7 bxc3 b6 8 cxb6 (8 'ii'g4? !
knowledge is essential, as intricate bxc5 9 .tg5 ? ! .te7 10 .txe7 'ii'xc7
22 Chase Variation
1 1 'il/xg7 l::t f8 is good for Black due (this insipid move abandons the plan
to his central control) 8 . . . axb6 9 li'lf3 of a kingside pawn storm) 15 . . . e6 1 6
(9 'ifg4?!) 9 . . . iLb7 10 .i.d3 d6 1 1 0-0 0-0 ( 1 6 tt'lg5 i.g6 with . . .h 6 to fol
.i.e7 1 2 l::te 1 tt'ld7 1 3 .i.c2 (it is hard low) 1 6 . . . i.e7 1 7 a4 l::t c 8 1 8 'ii'c 3 c5
to believe that Sveshnikov's sugges 19 i.e3 'it.>b8 left White a pawn down
tion 1 3 .i.e4 i.xe4 14 l::t xe4 dxe5 1 5 for nothing much in Vaassen-Etmans
dxe5 tt'lc5 1 6 l::td4 'il/c8 1 7 i.g5 is Dutch corr. Ch 1 990- 1 . Instead 15
promising for White) 13 . . . l::t a5 14 i.f3 e6 16 h4 ( 1 6 li'le2 looks the best
tt'ld2 dxe5 15 tt'lc4 l::t a7 (here Svesh try, though I'm sure White is fighting
nikov gave 1 5 . . . l::td5 16 i.e4 exd4 17 for equality) is more spirited, but
.i.xd5 exd5 as good for Black, but he does not give B lack problems :
was missing the very neat tactical se 16 ... tt'lb6 1 7 g4 i.e4 1 8 i.xe4 dxe4
quence 1 8 .i.a3 tt'lc5 1 9 tt'la5 i.a8 20 1 9 'ii'xe4 and White has won back
cxd4, which gives White a substan the pawn, but Black's position is ob
tial advantage) 16 tt'lxe5 tt'lxe5 1 7 viously pleasant, while 16 .. .i.e7 1 7
l::t x e5 .i.f6 18 l::te 1 'ifa8 1 9 f 3 0-0 20 g 4 i.e4 1 8 i.xe4 dxe4 1 9 g 5 f5 20
i.f4 g6 21 i.b3 c5 22 dxc5! t Svesh exf6 gxf6 21 'fi'xe4 looks attractive
nikov-Morozevich, Alushta 1994 . for Black too.
b) 5 ... c6, as in the next note, is a b22) 9.. .'fi'xc7 (this must be one
fighting move. Then 6 .i. c4 (6 tt'lxd5 of the most natural moves to get the
cxd5 7 d4 d6 8 cxd6 exd6 is level) is 'N' symbol in lnformator) 10 i.xf7+
critical: 'it>d8 11 'ft'e3 (best) and now (D):
b 1 ) 6 ... e6 (a speciality of Kengis)
7 'ii'g 4 (for other moves, see TCA)
7 . . . f5 8 'il/g3 b6 9 cxb6 axb6 1 0
B
tt'lge2 .i.a6 1 1 d 3 'ife7 1 2 0-0 'ilif7 1 3
.i.xd5 cxd5 14 tt'la4 tt'lc6 1 5 i.d2
(Blatny considered 15 tt'lxb6 l::tb 8 1 6
tt'la4 tt'lb4 necessary) 15 . . .b 5 1 6
tt'lac3 b 4 1 7 li'ld 1 .i.e7 1 8 b 3 0-0 1 9
li'lb2 .i.b5 2 0 l::t fc 1 l::t a 3 gave B lack
excellent play in Sveshnikov-Gluz
man, Bern 1 992.
b2) 6 ... d6 7 'ii' b 3 li'ld7 8 tt'lxd5
tt'lxc5 9 tt'lc7+ gives Black two good b22 1 ) ll ... d5 1 2 d4 tt'le6 1 3 li'lf3
options: g6 14 h4 ! tt'lg7 1 5 e6? (Sveshnikov
b21 ) 9 ... 'it>d7 10 'il/e3 'it>xc7 1 1 d4 suggests 15 h5 ! tt'lxh5 16 l%xh5 gxh5
i.e6 ( l l . . .d5 was discussed in TCA) 17 e6 with compensation) 15 . . . 1i'd6
1 2 b3 d5 1 3 i.e2 tt'ld7 14 f4 (this has 16 tt'lg5 h6 1 7 li'lf3 tt'lxe6 ! 18 'ii'e5
been reckoned to give White com i.g7 (not 18 . . .'fi'b4+? 19 i.d2 'ii'xb2
pensation) 14 . . . i.f5 and now 15 li'lf3 20 'fi'xh8 'fi'xa1 + 2 1 'it.>e2 'ii'x h 1 22
Chase Variation 23
hilities:
b222 1 ) Sveshnikov suggested 12
d4 'ifxfl 1 3 dxc5 , but it is easy to be
sceptical about White's compensa
tion after 1 3 . . . dxc5 .
b2222) 12 Jth5! ? 'with compen l::t xfl 14 'ii'xc4 would have been
sation' - Sveshnikov. Play may con roughly level.
tinue 12 dxe5 13 �f3 and now:
.•. b) 6 �f3 d6 7 cxd6 exd6 8 0-0
b2222 1) 13 g6 14 d4 ( 1 4 .tg4)
•.. dxe5 9 �xe5 i.e7 10 d4 0-0 1 1 �c3
1 4 .. . gxh5 ( 1 4 . . . �d7) 15 dxc5 'ii'a 5+ i.e6 12 'ii'f3 �xc3 1 3 i.xe6 fxe6 14
( 1 5 .. J:tg8 16 0-0) 16 Jtd2 'ifxc5 1 7 'ifxc3 'ii'd5 1 5 i.e3 �a6 16 b3 l::tad8
�g5+ followed b y 'ii'xh5 causes 1 7 .U.ad 1 i.d6 1 8 �c4 i.b8 19 i.c 1
problems for Black. 'ii' h5 20 h3 �c7 gave Black a com
b22222) 13. i.d6 14 d4 (14 �xe5
.. fortable position in Yurtaev-Bagirov,
.llxe5 15 'ii'xc5 =) 14 . . . exd4 15 'ii'xd4 Frunze 1 979.
with good compensation. c) 6 'i!Ve2 b6 7 �c3 e6 (grabbing
b22223) 13 ..e4 14 �g5 �d3+ 15
. the pawn is too risky) 8 cxb6 axb6
<J.>n Jlc5 16 'ii'xe4 �xf2 17 'ii'h4 9 d4 Jla6 1 0 �f3 ( 1 0 i.xa6 �xa6
�e7 (avoiding 17 . . . �d7 1 8 d4 �xh 1 1 1 �xd5 exd5 12 �h3 'ii'h4 +)
19 dxc5 'ii'e5 20 �f3 'ii'xc5 2 1 Jtf4 10 . . . i.e7 1 1 0-0 0-0 12 �e4 f5 1 3
±) 18 �fl+ �d7 when White should exf6 i.xf6 14 l::te 1 i.e7 1 5 �e5 �f4
opt for a draw by repetition with 19 1 6 'ii'g4 i.xc4 17 i.xf4 i.d5 1 8 l::te 3
lile5+ , since 19 'ii'xe7+? �xe7 20 i.xe4 19 l::t xe4 l::t f5 20 �d3 'ii'f8 2 1
l/:'lxh8 �xh 1 2 1 d4 i.xd4 22 Jlg5+ a 3 c 5 2 2 dxc5 bxc5 gave Black a
'>t>f8 23 l::t d 1 is insufficient after good pawn centre in Scotto-Shaba
either 23 . . . e5 or 23 . . . c5 . lov, Saint Martin 1 993 ; White was
5 ... e6 pushed off the board as follows: 23
5 ...c6 (D) is an altogether more l::tc 1 �c6 24 i.e3 l::tb 8 25 f4 'ii'f6 26
imbalancing move: l::t a4 .U.d5 27 'ii'e2 'ii'f5 28 �f2 i. f6
a) 6 d4 d6 7 cxd6 exd6 8 �f3 29 g4 'iVg6 30 .l:r.c2 i.d4 3 1 l:f.d2 h5
uxe5 9 0-0 !? Jle7 (9 . . . e4 10 �g5 32 h3 l::tb 3 3 3 l::t a8+ �h7 34 .llxd4
.lle7 1 1 'ii'h5 i.xg5 12 i.xg5 'ii'd6 �xd4 0- 1 .
13 �c3) 10 dxe5 0-0 1 1 'ii' b 3 (Skrip d ) 6 �c3 i s discussed under 5
chenko-Burgess, Biel 1992) and now �c3 c6 6 i.c4.
ll . . .�b6 1 2 e6 ! ? �xc4 1 3 exfl+ 6 �c3
24 Chase Variation
ll...f6, 1 1 . . .0-0 and l l . . .f5 (which bxc3 b6 14 ltlf3 gives White com
looks soundest). pensation.
c) 9 .tf8 ( ' ! ' from Khmelnitsky
••• c3) 11 ... lLlds!? (in order to deny
- it is certainly playable, but allows White the possibility of strengthen
White compensation) 10 i.g5 (Svesh ing his pawn centre with bxc3) I 2
n ikov suggested the move 10 bxc3 !?, i.xd5 i.e7 1 3 .txe7 "il/xe7 looks quite
which gives White a certain amount reasonable for Black.
of extra freedom as compensation) 9 .txc3+
I O h5 (best; 10 ....te7 1 1 .txe7 'fixe7
••• 10 ct;n (DJ
12 'flixg7 ;!; ; 1 0 . . . f6? loses to 1 1 exf6
gxf6 1 2 'ii'h 5+ <i;e7 1 3 i.xf6+) 1 1
1ff4 (D) and now:
II
10 ••• "il/e7!
This move is forced according to
Khmelnitsky, but this may not be the
cl) 11 ... f6?! (very interesting, but case: 10 llg8 1 1 i.g5 'ii'xg5 (though
•••
not best, and not a try for advantage) this is forced for Black, Khmelnitsky
12 exf6 gxf6 (and not 12 . . . i.d6?? 1 3 neglects to mention the possibility;
fxg7 +-) 1 3 i.xf6 i.b4 ( 1 3 . . .i.h6 14 certainly not 1 1 . . .f6? 1 2 .txf6 +-)
�f3 i.d2+ 1 5 <i;fl ! +-) and now 1 2 'ii'x g5 .txa1 gives B lack rook,
White can choose between 14 .t xd8 bishop and two pawns for the queen.
tiJds+ 15 axb4 lLlxf4 16 i.xc7 lLlxg2+ Also, while the reply 10 <i;f8 looks
•••
17 ct;n when Black may go a pawn very greedy, no knock-out blow for
down, but White's unsightly pawns White is apparent.
will provide compensation, and 14 11 llbl
axb4 llf8 15 .txd8 ( 1 5 bxc3 leaves After this move White is defi
White a little better due to the weak nitely worse. He should have tried II
ness of h5) 15 . . . llxf4 16 .txc7 ( 1 6 "il/xg7 'ii'f8 1 2 "illf6, which Khmelnit
bxc3 <i;xd8 ;!; ) 16 . . . llxd4 when the sky gave as unclear - an inadequate
black position hangs together. assessment without a great deal of
c2) ll ... .te7 (the only move given further evidence, since White is so
by Khmelnitsky) 1 2 i.xe7 "il/xe7 1 3 much material down. For example,
26 Chase Variation
Conclusion
The Chase Variation is a great exam
ple of 'chess from another planet' ,
but isn't Black just better in the main
lines?!
4 Unusual Fourth Moves for White
In this chapter we consider all lines b) 4 f4, on the other hand, has
in which after 1 e4 tDf6 2 e5 tDd5 3 been quite popular in the 1990s, with
d4 d6 White plays neither 4 c4 (Ex Kupreichik as the main protagonist:
change or Four Pawns) nor 4 tDf3 b 1) 4 i.f5 5 tDf3 e6 6 .id3 .ixd3
•••
4 ..tc4
There are a few other moves here
worthy of attention:
a) The unpopularity of 4 .ie2 is
astonishing, considering its excel
lent score in practice. I know of only
one recent example : 4 . . . dxe5 5 dxe5
i. f5 6 c3 e6 7 ti:)f3 tDd7 8 0-0 .ie7 9 b2 1 ) s ... .trs 6 lbf3 e6 7 i.dJ
a3 0-0 1 0 h3 a5 1 1 c4 tD5b6 12 tDc3 !l.. x d3 8 'ii'x d3 c5 9 0-0 i.c7 (tJ h 6 . . .
Jung Karl, Moscow 1 99 1 , Tal rec dxe5 d5 is a sharp line that has gener
ommends 18 lLlc4! ? �d6 1 9 lLle3 ;!;. ally been considered quite playable
b23) 5...c5 6 lLlf3 cxd4 (6 ... �f5 ? ! for Black, but after 9 lLlc3 .tg4 1 0
7 � b 5 + lLlc6 8 c4 ! lLlc7 9 �xc6+ ! lLlxd5 e 6 White tried a promising
bxc6 10 lLlc3 e6 1 1 'it'a4 'it'd? 12 �e3 new idea in Kalod-Hudecek, Czech
cxd4 1 3 lLlxd4 c5 14 lLldb5 ! lLlxb5 Ch (Luhacovice) 1 993 : 1 1 �g5 ! ?
15 lLlxb5 0-0-0 16 0-0 �e7 17 b4 ! .txf3 1 2 gxf3 'iixg5 1 3 lLlxc7+ 'it>f7
cxb4 1 8 c5 �c2 1 9 'it'a6+ 1 -0 Ku 14 .txe6+ 'it>g6 15 f4 'ii'g 2 16 'iid 3+
preichik-Karolyi, Dortmund 1 99 1 ; ( 1 6 f5+ ! ?) 16 . . . 'it>h6 17 0-0-0 l::.d 8
6 . . . .tg4 ! ? 7 �b5+ lLld7 8 0-0 e6 9 and now rather than 18 'iie 3 l::. x d 1 +
.tg5 .te7 10 �xe7 'ikxe7 1 1 'fid2 0-0 1 9 l::. xd 1 �e7 20 .th3 'iix h2, which
1 2 .txd7 'iixd7 1 3 dxc5 'fie? 1 4 was unclear, White could have tried
'iid4 .txf3 1 5 l::. xf3 l::. ad8 1 6 lLla3 18 'iig 3 l::. x d 1 + 1 9 l::. x d 1 'iix g3 20
lLle7 = Carr-Crocker, Halifax 1993) hxg3 when Black is in extreme peril.
7 'iixd4 lLlc6 8 'ii'e4 ! ? (Kupreichik's b) 5....tf5 is preferred by a num
latest idea) 8 . . . g6 (8 . . . e6 9 c4 ;!;) 9 ber of Alekhine specialists:
.tc4 lLlb6 (9 . . . lLldb4 10 �b3 .tf5 1 1 b 1 ) 6 e6 (not one of the better va
'iie 2 might be a shade better for rieties of e6 pawn sacrifice) 6 . . . .txe6
White) 10 .tb3 (10 e6?! f5 !) 10 ... .tg7 7 �xe6 fxe6 8 lLlf3 lLlc6 9 0-0 'it'd?
Unusual Fourth Moves for Wh itt' ."J
gxf3 l:tf8 1 2 tt:'!c3 'ii'h4 1 3 tt:'!e4 �e7 for example 5 dxe5 c6 6 tt:'!c3 �c6
1 4 �h6 1 -0 Djuric-Speck, Adelaide 7 tt:'!f3 tt:'!d7 8 tt:'!xd5 �xd5 9 �xd5
1 990 is a cautionary tale) 9 tt:'!xd4 'ii'a5+ 1 0 �d2 'fixd5 1 1 'fie2 c6 1 2
l!Vxd4 1 0 0-0 g6 1 1 'ii'f3 �g7 1 2 0-0 �e7 1 3 :fdl 'ii'b5 1 4 c4 'il'a6 1 5
�d2 0-0 1 3 �c3 'ii'h4 14 tt:'!d 2 a5 1 5 �g5 l:td8 1 6 �xe7 'it;xe7 1 7 'il'd2 ±
a4 tt:'!d7 1 6 l:tfe l l:ta7 1 7 �c4 b6 1 8 Mellado-Viot, Enghien- lcs- Bains
�b5 tt:'!c5 1 9 b3 l:td8 2 0 �c6 �d7 1 995.
30 Unusual Fourth Moves for White
26 ... lle6!
Now Black has a very big plus.
Black is substantially better here. 27 'ii'f2 l:g6 28 g3 1i'd7!? 29
The rest of the game is a superb dem lDxaS!? h5 30 lDb3?
onstration by Miles. He gradually re White's last chance for counter
groups and strengthens his grip on play was 30 f5 ! i. xf5 31 lDxb7 ! or
the position befor� advancing on the 30 lDxb7.
white king. 30 h4 31 l:c3 i.e7 32 f5 i.xf5
•••
that Black really should avoid it alto 14 .i.e3 f5 15 .i.c2 f4 16 .!Dxf4 .!bexc4
gether. Consider: 1 7 .!bxe6 'iixe6 1 8 .i.d4 ± West
a) 8 .!bc6 9 d5 .!be5 10 0-0
.•• Fuller, Sydney 1 992 .
.!Dxd3 1 1 'ii'x d3 .i.f5 1 2 'ii'd 1 0-0 1 3 b2) 9 e5! 1 0 .i.e3 gives Black a
•••
and 0-0-0 shows an independent .i.e2 ltJbd7 1 3 0-0 ltJc5 14 lbd4 i.xel
idea) 9 ltJge2 ltJc6 10 0-0 e5 1 1 d5 15 'ilixe2 l:te8? ( 1 5 . . . 'ii'a5 16 i.d2 1
lLlb4 1 2 b3 ltJxd3 1 3 'ii'x d3 f5 14 f4 Raetsky) 1 6 b4 ltJcd7 17 ltJb3 ! ±
ltJd7 1 5 l:tad 1 b6 ! ? 1 6 'ilid2 .i.b7 1 7 Raetsky-Dovzhik, Lipetsk 1 993.
fxe5 dxe5 1 8 i.g5 .i.f6 1 9 .ih6 i.g7 c) 9 e5 10 dxe5 and now:
...
and so is prepared to meet either of ltJxe6 .i.xe6 17 i.xe6 fxe6 and 1 4 h6 ...
This looks like rather a good nov 1 3 ltJf3 l:te8 14 0-0 e6 15 dxe6 i.xe6
elty, making use of the slowness of 16 'ilid2 ± Sergienko-Kopylov, Voro
White's build-up, in a line that had nezh 1989) 1 3 i.xg4 fxg4 14 ltJgc2 c5
been giving Black some problems: 1 5 dxe6 i.xe6 16 0-0 'ii'd7 ( l 6 d5 . . .
a) 9 .if5? ! 1 0 .ie2 d5 ? ! 1 1 c5
... 17 c5 ltJd7 1 8 ltJd4 ltJxc5 19 lbch5
ltJc8 12 .i.f3 ! .i.e6 13 ltJge2 ltJc6 14 'ilie7 20 l:Ie 1 ± Sergienko-Shash k i n ,
ltJf4 'ilia5 1 5 0-0 l:td8 1 6 a3 ! ± Kis Riazan 1 982; 1 6 . . . ltJc8 17 i.d4 4\c7
lov-Brachenko, Odessa 1 980. 1 8 .ixg7 �xg7 19 �h 1 lLlf5 20 li1d4
b) 9 ltJ6d7 10 ltJf3 (or 10 d5 ! ?)
... ± Kislov-Bragin, Voronezh I Y75 ) 17
1 0 . . . ltJf6 1 1 d5 ( 1 1 h3 ! ?) 1 l . . .i.g4 12 'ii'd2 l:.ad8 18 i.d4 .l:tfe8 I 9 l::t 1 CI lAX
40 The Exchange Variation
�f3 �c6 1 3 �g5 = ) 1 2 . . . .txd4 1 3 not take the queens off, catching the
'ifxd4 �c6 +. white king in the centre has more
10 �ge2 e5 1 1 dxe5 dxe5 is rea merit) 12 dxe5 ( 1 2 c5 exd4 1 3 cxb6
sonable for Black; note that 1 2 'ifxd8 dxc3 gives Black an extra pawn; 1 2
l:txd8 1 3 �xb6 axb6 14 a4 doesn't d 5 �f5 also looks rather good for
work because B lack is too active for Black) 1 2 . . . dxe5 1 3 c5 and now
White to have the opportunity to at 13 �6d7 14 �c4+ �h8 1 5 �e4
•••
Chekhov didn't fight a little harder than with the knight on e2, since here
to prevent White implementing it) the king's bishop is free to move, e.g.
1 2 . . . �e5 13 �xe5 ! �xe5 14 �e2 e6 to c4 following c5 .
15 dxe6 ( 15 �f3 !?) 15 . . . �xe6 1 6 �f3 11 f4
d5 17 cxd5 �xd5 18 �xd5 ( 1 8 0-0!?) 11 �4? g5 12 � f4 + 1 3 �xg7?
1 8 . . . �xd5 1 9 .tb4 .txf3 20 'ili'xf3 fxe3 -+; 1 1 dS g5 1 2 f4 ( 1 2 �d4? !
�b2 (20 . . . l:tf7 2 1 0-0 'iff6 22 l:tfd 1 f4) 12 ... g4 1 3 �f2 e5 14 fxe5 dxe5 ao.
;!; ) 2 1 0-0 ! �xc 1 2 2 �xf8 'ifxf8 23 11 .•. � e6
The Exchange Variation 41
Conclusion
As we have seen, the Exchange Vari
ation is far from the harmless nonen
tity its name would suggest: bold,
imaginative play is required from
both sides . The most important stra
tegic addition to White's armoury is
B oth sides have their problems : the idea of preparing to meet . . . lbc6
Black's position is somewhat in dis with d5 . Black is yet to establish
array, but the white knight on h3 is a completely reliable paths to equality
by-stander at present. versus this concept.
6 Four Pawns Attack without
6 . . . ttJc6
should play 10 . . . ..tb4 - but I feel that leaves Black material up.
the critical line is 1 0 d6. This move b) 6 l2Jc3 e6 7 l2Jf3 and now:
would constitute my main reason for b 1 ) 7 l2Ja6 8 l2Jf3 (8 exd6 cxd6
...
to advance his centre pawns, rather lLle5 18 'ii' xf4; 16 l:te8+ (consid
•••
than trying to restrain them as one ered best by Fritz3) 17 'iPf2 i.xd6 18
normally would. The idea, of course, i.xc4 i.c5+ 19 'iii>fl i.xf3(19 ...i.e3
is to destroy the pawns before they 20 'ii'c2) 20 gxf3 l:te3 21 i.e6+ fxe6
land- just like Space Invaders! 22 'ibc5. None of these look espe
7 d5 e6 cially convincing for Black. I actu
8 lLlc3 ally won the game quickly, but this
White must obviously avoid 8 was somewhat fortuitous.
d6? 'ii'h4+. Other tenth moves for White are
8 exdS less critical, e.g. 10 i.e3 i.b4 and
9 cxdS c4 (D) now:
10 lLlf3 a) 1 1 d6 (illogically bolting the
The queen sacrifice 10 d6! ? lbc6 cathedral door after the dark bishop
1 1 lbf3 (11 i.f4 g5 12 lbe4 gxf4, etc., has gone) 1l. . .lbc6 12 'ii'h5 0-0 ! '!
44 Four Pawns Attack without 6... tLlc6
This is the traditional main line of line with 9 . . . i.g4. In return for this,
the Four Pawns . Rather than try to White misses out on the chance to
smash White's centre to pieces im meet 9 �f3 i.e? with the sharp 1 0
mediately, Black first puts d4 and e5 d5, but i t seems o n the current evi
under pressure to restrain further ad dence that White should not go in for
vances by White. After further prepa the complications anyway.
ration Black envisages a freeing 9... 'ii'd7
break with . . . c5 or .. .f6. In the next game we discuss the
current main line, 9 . . . i.g4. Here are
Game 8 the alternatives :
Tasc R30 Mephisto Genius 2
- a) 9 �b4 was Alekhine's origi
...
1 7 . . . tt:'lxf3 1 8 'ii'xf3 .::!.b 8 19 .tf6 with 25 'ii'xe5 tt:'lf6 26 'ii'x b5+ �f8 27
an overwhelming attack. .::!. xh 1 .l:td8 28 �c 1 �g8 29 e7 .l:tc8
c) 9 ... .te7 used to be the standard 30 .txf6 gxf6 3 1 'ili'f5 1 -0 was Fer
move. There does not appear to be nandez Garcia-Leko, Debrecen Echt
anything particularly wrong with it, 1 992) 18 gxh4 'ii'f6 (D) is a very
but it is not terribly fashionable deeply analysed position, but since it
nowadays. The wildly sharp lines has hardly been played at all in the
following 10 d5 fail to trouble Black 1 990s, I shall just give a few exam
if he is supremely well-prepared, so ples:
White should try the quieter ap
proach:
c 1 ) 10 .te2 0-0 1 1 0-0 f6 12 exf6
.txf6 1 3 'ii'd 2 'ii'e7 14 .::!. ad 1 ( 1 4 c5
tt'ld5 15 j_f2 .::!. ad8 1 6 .::!. fe1 tt:'ldb4 1 7
.tc4 tt:'lc2 1 8 j,xe6+ j,xe6 1 9 'ii'xc2
tt:'lxd4 20 tt:'lxd4 j,xd4 2 1 j,xd4
.::!. xd4 22 tt:'lb5 .::!. df4 23 tt:'lxc7 'ifxc7
24 l:txe6 �f2 = Garma-Alburt, New
York Open 1 993) 14 . . . �ad8 15 'ii'c 1
h 6 1 6 �h 1 .::!. d7 ( 1 6 . . .�h8 is normal,
but is under some pressure in the line
17 h3 .th7 1 8 j,g1 .::!. fe8 1 9 .::!.fe 1 c21 1 ) 19 i.e2 'ii'e5 20 i.g5 c5 2 1
'ilif7 20 'ii'f4) 17 .::!.fe1 'fin 18 d5 tt:'le7 e7 cxd4 2 2 exf8'it'+ .l:txf8 2 3 'ii'xd4
( 1 8 . . . exd5 is more robust) 1 9 dxe6 'ii'xd4 24 .::!.xd4 tt:'l4d5 (a novelty; the
.txe6 20 tt:'le4 t Narciso-Mellado, older 24 . . . tt:'lc6 was reckoned to leave
Spanish Ch (Linares) 1 993. White a little better after 25 .l:tf4) 25
c2) 10 d5 and now: tt:'lxd5 i.xd5 26 a4 i.b3 27 a5 tt:'ld5
c2 1 ) 10 exd5 1 1 cxd5 tt:'lb4 12
••• 28 j,g4 tt:'lf6 29 j,h3 .:.e8 30 ..tg2
tt:'ld4 .td7 13 e6 fxe6 14 dxe6 .tc6 .::!.c 8+ 31 �b1 l:tc5 32 l:td2 l:.xa5 33
15 �g4 .th4+ 16 g3 .txh1 17 0-0-0 i.xb7 .l:tc5 34 .l:td8+ �f7 and Black
(after 17 .tb5+ c6 1 8 0-0-0 0-0 1 9 had his fair share of the play in
gxh4, Black has the option of trans Texier-Solozhenkin, Noumea 1 995.
posing to line 'c2 1 2' by means of c2 1 2) 19 i. b5 c6 ! ? ( 1 9 . . . c5 is
1 9 . . . 'fif6) 17 0-0 (playing 17 . . . 'ii'f6
••• well-documented, and entirely satis
often simply transposes, but reduces factory for Black) 20 i.g5 (20 l:txh 1
Black's options - after 1 8 .tb5+ c6 cxb5 2 1 tt:'ldxb5 tt:'l6d5 22 i.d4 'ili'f4+
1 9 gxh4, he should transposes to line 23 'ili'xf4 .l:txf4 24 l:tg1 g6 25 .:r.g5
'c2 1 2' with 19 . . . 0-0; instead 19 . . . h5 ? .l:td8 26 a3 tt:'lxc3 27 bxc3 tt:'lc6 28
20 'ili'g3 cxb5 { 20 . . . 0-0 2 1 .te2 .td5 .te5 .l:te4 29 tt:'ld6 .l:te2 30 i.g3 b6 3 1
22 a3 ± } 2 1 .tg5 'ii'g 6 22 'ili'd6 ! h5 .l:txe6 3 2 tt:'lc4 tt:'le7 3 3 hxg6 hxg6
tt:'l6d5 23 l:tfl ! tt:'ld3+ 24 �d2 tt:'le5 0- 1 B. Christensen-Granberg, Danish
48 Four Pawns Attack with 6. . . ll:k6
1 3 . . .lDxe3 1 4 fixe3 must be good for both very powerfully met by 12 d5)
Black, but in Ananse-Cheiron, Pad 12 c5 lDd5 13 lDxd5 'ifxd5:
erborn microcomputers Web 1 995 a) 14 lDg5 (the main line, but
the computer playing Black then maybe not such a problem for Black)
drifted; as a plan, queenside castling 14 . . . .i.xe2 1 5 'ii'xe2 lDxd4 1 6 .i.xd4
and a quick .. .f6 suggests itself. 'ii'xd4+ 17 �h 1 'ir'd2 18 'ii'xd2 l:txd2
10 :ct? ! .i.g4 1 1 c5 ( 1 1 .i.e2 .i.xf3 19 l:.xf7 .i.xc5 20 lDxe6 .i.b6 2 1
1 2 gxf3 gives Black a choice be lDxg7 .i.d4 (this position has gener
tween 1 2 . . . 0-0-0 1 3 c5 lbd5 1 4 ally been regarded as equal) 22 e6
lbxd5 fixd5 1 5 .i.c4 fid7 1 6 .i.b5 .i.xb2 23 e7 (new move) 23 . . . .i.xg7
fid5 = and 12 . . . l:.d8 ! ? 13 c5 lDc8 14 24 l:.e1 l:.e8 25 l:.xg7 (looks good for
.i.b5 a6 15 .i.a4 b5 16 cxb6 lDxb6 oo, Black) 25 . . . h6 26 'it>g1 :xa2 27 l:.g6
as analysed by Kupreichik) 1 1 . ..lbd5 'it>d7 28 l:r.xh6 and now, in the game
1 2 lDxd5 fixd5 1 3 .i.c4 'ir'e4 14 fid2 Dark Thought-Cheiron, Paderborn
( 1 4 'ii'e 2? lDxd4 ! is winning for microcomputers Web 1 995, B lack's
Black, but White should play 1 4 most sensible course was 28 ... :xe7
�f2 ! ? .i.xf3 1 5 'ii'xf3 'ir'xf3+ 1 6 29 :d1 + �e8 30 l:th8+ � 3 1 :h7+
gxf3 0-0-0 1 7 :hd l = - Kupreichik) �e8 32 l:th8+ with a repetition.
Four Pawns Attack with 6... lLic6 49
Other moves for White are no bet be very good for Black in all vari
ter: ations :
16 liJh6 lUxe3 17 'ii'xe3 'ii'e7 +. a) 19 'iixg5 'ii'd 3+ 20 'it>f2 .i.c5+
16 liJd6+? .i.xd6 17 exd6 liJxe3 21 'it>e1 'ii'd4 (2 1 . ..iLb4+ repeats, but
0- 1 Bullockus-Oakley, corres. 1 984. Black has better) 22 'i'i'f5+ 'it>b8 23
16 liJg3 .t b4+ 17 i.d2 ( 17 'it>f2 i.e2 'ii'xh4+ and Black regains the
'ike7 intending 1 8 . . . liJxe3 ) 1 7 . . .'ii' a4 piece with a huge advantage.
works well for Black. b) 19 e6 'ii' b 5+ (or 1 9 . . .'ii' d 2+ 20
16 ... liJxe3 'ii'xd2 l:txd2+ 2 1 'it>e3 gxh4 22 a3 iLa5
16 .. .'ii'a4 works nicely in case of 23 b4 ±) 20 'it>f2 (20 'it>f3 'iVd5+ -+)
17 i.g5 i.b4+ 1 8 'it>f2 i.c5+ 1 9 'it>g3 20 . . . l:td2+ 2 1 iLe2 (2 1 'ifxd2 'ii' b 6+
( 1 9 'it>f3 f6 20 b3 'ii'c 6 -+) l 9 . . . liJc3 22 'ii'e 3 iLc5 23 liJf5 fxe6 wins for
20 'ii'e 2 l:td3 -+, but after the simple B lack) 2 1 . . . gxh4 22 exf7 .:tf8 +.
17 liJf3 I don't see anything promis c) 19 liJf3 g4 is the main point.
ing for Black. Having seen what Fritz was up to, it
17 'ii'xe3 i.b4+ wasn't too difficult to coax the fol
The critical idea that B l ack must lowing variations out of it:
avoid is the exchange of queens in c l ) 20 liJg5 l:ihe8 -+.
such lines as 17 ...i.c5 18 'ii' h 3. c2) 20 liJh4 'ii'e7 21 liJf5 (2 1 g3
18 'it>f2?! l:td2+) 2 l ...l:f.d2+ 22 'ii'xd2 'ii'xe5+ 23
18 'it>e2!? (D) was Murray Chan 'ii'e3 'il'xb2+ 24 'it>d3 (24 'it>d 1 l:td8+
dler's proposed improvement: 25 i.d3 'ii'x a1+ -+) 24 . . . 'i'c3+ 25
'it>e4 (25 'it>e2 'iic2+ 26 'fi'd2 'ii'xd2#)
25 . . . !1e8+ -+.
c3) 20 lUe1 'ii'f5 (20 . . . l:the8 2 1
B
liJct3) 2 1 l:f.c l (2 1 liJct3 l:.xd3 22 'ii'xd3
'ifxe5+ -+) 2 1 . . ..:the8 22 liJd3 l:txd3
23 'ii'xd3 'ii'xe5+ 24 'it>f2 (or 24 �d1
l:td8) and now 24 ... 'iff6+ mates in
seven moves at most: 25 'iff3 g3+ 26
'it>xg3 iLd6+ 27 'iii>h 3 'iih 6+ 28 'iti>g4
l:tg8+ 29 �f5 'ii'g6#.
The rest of the game was a superb
attacking demonstration by Genius 2.
Then such lines as 18 ... 1\VbS+ 1 9 18 ... 'iie 7!
'it>f2, 1 8...'ii'd2+ 1 9 'ii'xd2 l:txd2+ and 19 'ii'h3 +
18 ...'iVg4+ 1 9 liJf3 'ii'c4+ 20 'it>f2 are 19 liJf3 is impossible as 19 . . . iLc5
no good at all for Black. picks off the white queen.
After more than seven hours ' de 19 �b8
liberation, Fritz came up with the 20 liJf3 iLcS+
astonishing 18 ... g5! , which seems to 21 �g3 !:.d2!
Four Pawns Attack with 6. . . lL!c6 51
22 .l:.cl 10 i.e2
The rook is immune as 22 lL!xd2 10 'ifd2 is the main alternative,
allows a forced mate: 22 . . . 'ifxe5+ 23 but it has never performed well in
�f3 'ife3+ 24 �g4 h5+ 25 �h4 practice, and allows Black good
i.e7#. counterplay:
There is also a forced mate after a) 10 'i¥d7 1 1 i.e2 ( 1 1 0-0-0 i.b4
...
Conclusion
9 lLlf3 i. g4 looks quite satisfactory
for B lack. Little wonder then that
White is choosing 9 i.e2 increas
ingly frequently. Then the line
9 . . . i.e7 10 lLlf3 may offer White a
small edge, while 9 . . . 'ii'd 7 1 0 i.e2
l:td8 could be a very important line,
although virtually no one has been
playing it !
8 4 liJf3 : Introduction and 4 ... dxe5
5 liJxe5 liJd7
Now we come on to the main line of if the sacrifice is not good for White,
the opening, by far the most popular then the whole line with 4 �f3 is far
in practice: 4 �f3 (D). less fearsome. The play following
the sacrifice is almost purely tactical,
so if you wish to play the line, please
study the material in this chapter
B
very carefully, and supplement it
with your own analysis. The current
verdict is inconclusive: there is no
clear win for White, nor can one
demonstrate that the sacrifice is un
sound. My hunch is that Black is
OK, but some of the messy positions
will be cleared up over the next few
years as computers become more
It is this apparently modest move powerful.
that more than any other puts players Among the other fourth moves for
off Alekhine's Defence. As we shall Black we touch upon in this chapter
see in this and the remaining chap are 4 ...c6 and 4 ... �6. Tony Miles has
ters of the book, Black has a variety enjoyed considerable success with
of interesting responses. White cer 4 . . . c6. This move largely aims to
tainly does not enjoy a simple advan give Black a playable game and to
tage that can be maintained by transfer the weight of the battle to the
'natural' moves. middlegame; it is no attempt to seize
Our main focus of interest in the equality from the word go, though
current chapter is the notorious line some of his games have shown the
4 . . . dxe5 5 �xe5 �d7, inviting White quick counterpunching resources at
to drag the black king all over the Black's disposal if White plays too
board with the knight sacrifice 6 ambitiously. On the other hand,
�xf7 �xf7 7 'it'h5+ �e6. Although 4 . . . �c6 challenges White directly.
rarely played, an enormous amount White cannot hope for much if he
of analysis has been devoted to this avoids the critical pawn sacrifice 5
variation, for the simple reason that c4 �b6 6 e6. This line, however, has
4 &iJf3: Introduction and 4. . . dxe5 5 llJxe5 &Dd7 55
Game 1 0
Szilagyi - Krantz
corr. 1992
for the exchange) and now 15 dxe5 e l l ) 5 li:Jc6? ! 6 e6 fxe6 (the fact
•••
�xd3 1 6 �xd3 li:Jc5 is an example that White has played li:Jc3 instead of
of Black's idea, while 15 g4! ? �xg4 c4 must constitute a substantial im
16 ii'xc6 ii'b6 is OK for Black. provement over the line 4 . . . li:Jc6 5 c4
b) 4 li:Jc6 has been extremely
••• li:Jb6 6 e6 fxe6) 7 �d3 g6 8 h4 �g7
unpopular recently. Here is the only 9 h5 li:Jxd4 10 li:Jxd4 �xd4 1 1
notable new game: 5 c4 li:Jb6 6 e6 .i.b5+ c6 1 2 ii'xd4 e5 1 3 ii'd l cxb5
fxe6 7 ll:Jg5 e5 8 d5 li:Jd4 9 �d3 .i.f5 14 hxg6 ± Berelovich-Hetey, Gron
1 0 �xf5 li:Jxf5 1 1 �e3 g6 1 2 li:Jc3 ingen 1995 .
( 1 2 g4 ll:Jg7 13 ii'f3 is the theoretical c l 2) 5 �g4 6 h3 �h5 (D) and
•••
Conclusion
At the risk of stating the obvious, a
great deal of detailed knowledge is
needed to venture S . . . l?:Jd7 with con
fidence. Black will spend a long pe
riod in an extremely precarious
situation, any slip being fatal . If this
22 l?:JbS+? suits your style, then, barring a refu
22 cxd7 .ixgS 23 l?:Jb S ++ �xd7 tation being discovered, I can recom
24 l?:Jxd4 .ixc 1 2S �xeS l:.h6 seems mend the line.
9 Alburt Variation: 4 ... g6
'iWh4 f5 ( 1 3 . . . 'it'b5 might be the best I:.e 1 , with a small plus for White in
try) 14 lt:Jc3 (D) and now: both cases.
b1) 14 lt:Jc6 15 i.g5 'ii'd4 1 6
... 8 0-0 0-0
'ii'xh7 'ii'e 5+ 1 7 'ittf l 'ii'g7 1 8 'ii'xg7 9 .:tel (D)
l:txg7 1 9 I:.e 1 i.d7 20 lt:Jb5 l:.c8 2 1 9 ... lt:Jc6
i.f4 e 5 2 2 i.xe5 lt:Jxe5 2 3 ltxe5+ 9 i.g4 is a very natural alterna
...
13 i.e3
White plans both 1 4 lL'lh4 and 1 4
d5 followed by 1 5 i.d4.
13 ... d5
Anand analysed instead 13 ... �h7
14 d5 lL'la5 1 5 i.d4 ! and 13 ... lL'la5 14
lL'lh4 ! i.d7 ( 1 4 lL'lxb3 15 axb3 a6 1 6
. . .
Now Black has serious problems. White intends 'iie 3 and f3, ex
18 ... tt:'lxb3 ploiting Black's unfortunate line-up
Blatny analysed the continuation on the e-file.
18...tt:'lxd2 1 9 .t xd5 at length, con 29 'il'a4
cluding that White was winning in 30 b3! cxb3
view of Black's unfortunately placed 31 axb3 'iVb5
pieces . Or 31...'ili'xb3 32 f3 .
19 tt:'lxc4 32 �e3 ! .td6
Afterwards Anand preferred 19 33 tt:'lh5 ! 1-0
'iWxb3! tt:'lxd2 20 .txd2 e6 2 1 f3 �xb3 Black's resignation was based
22 axb3 .tc2 23 .:txa7 .txb3 24 .:txb7, upon the line 33 . . .gxh5 34 'ii'xh6 .te5
with a sound extra pawn. 35 'it'g5+ .tg6 36 .:txe5 hxg4 37 'iif6
19 ... �a6! l:he5 38 .txe5 <Ji>f8 39 'ifd8+ 'ife8
Presumably Anand had missed that 40 .td6+.
this move, Black's only chance, was
possible. Fortunately, he is still win- Game 1 2
ning. Kveinys - Speelman
20 �xb3 dxc4 Moscow OL 1994
20 ... .:txc4 is no good since White
then has time for 2 1 f3 . 1 e4 tt:'lf6 2 e5 tt:'ld5 3 d4 d6 4 tt:'lf3 g6
21 �dl .tc6 5 .tc4 tt:'lb6 6 .tb3 .tg7
22 .:txe7 .:tfe8 7 'iie2
23 l::txe8+ .:txe8 This is the most popular move
24 .te3 nowadays. In the next game we con
Instead after 24 d5 .ta4 Black's sider all lines in which White plays
bishops may yet have their say. the move a4, whether now or on sub
24 ... �5? sequent moves.
Timman had to try 24... g5 25 tt:'lh5 7 ... 0-0
'ii b 5 26 'it'd2 'ii'd 5 and hope that 7...tt:'lc6 8 0-0 (8 h3 0-0 returns to
Anand would get bogged down in the main line) is best met by 8 ... dxe5
the technical difficulties. 9 dxe5 tt:'ld4 10 tt:'lxd4 1i'xd4 as here
25 d5! 1 1 .:tel .tg4 ! is irritating for White,
Now Black is getting pushed off and 11 e6 .txe6 12 .txe6 fxe6 1 3
the board. 'iixe6 'it'c4 1 4 'iixc4 tt:'lxc4 is fine for
25 ... i.d7 Black, e.g. 15 c3 0-0-0 1 6 a4 l::t hf8 1 7
25 ... l1d8 is met by 26 .td4 ! .txd4 f3 .te5 1 8 tt:'la3 tt:'lxa3 1 9 l::t x a3 .tf4
1/z- 1/z
27 1i'xd4 1i'xb2 28 .:tel - Anand. Dolmatov-Neckai', Bern 1 994.
26 'il'd2 .trs Instead 8 ... 0-0 9 c3 .tg4 1 0 .tf4
27 .td4 .tf5 gives White more chances of achiev
28 g4! .te4 ing an advantage, for example
29 .:tel 10 . . . a5 1 1 a4 dxe5 12 dxe5 .txf3 1 3
A/burt Variation: 4. . . g6 67
'ii'xf3 l2Jxe5 14 'ii'x b7 l:tb8 1 5 'ii'e4 'ii'f6+ �g8 25 'ii'e7 'ii'b 6 (25 . . . t2Jxb3
t2Jbd7 1 6 'i!Vc2 l2Jc5 17 l:td 1 'ii'e 8 1 8 26 axb3 'ii'c 6 leaves B lack totally
t2Jd2 'it'c6 1 9 ..t xe5 i.xe5 20 ..tc4 t; passive) 26 l2Jg4 �g7 27 ..tc2 i.e8
Mortensen-Polzin, Lyngby 1 99 1 . (B lack has no defence against the
8 h3 t2Jc6 threatened i.xg6, e.g. 27 . . . l:taf8 28
Black quite frequently plays in i.xg6 'it>xg6 29 'ii'f6+ �h5 30 l2Jxh6
stead S aS here, when 9 a4 is nor
... forces mate) 28 i.xg6 h5 29 i.xh5
mal - see the next game. .l:txh5 30 t2Jf6 l:tf5 3 1 l:te3 'ii'x b2 32
9 t2Jc3 (D) l:tg3+ 1 -0 P. Popovic-Damljanovic,
This move is a little unusual, but Belgrade 1 992. It is worth noting
has its points . Instead after the nor that the opening in this game was a
mal 9 0-0 Black may try: Modern Defence: I e4 g6 2 d4 i.g7
a) 9 t2Ja5 gives White a choice
... 3 t2Jf3 d6 4 i.c4 t2Jf6 5 1i'e2 0-0 6 e5
between 10 l2Jc3, reaching the posi dxe5 7 dxe5 l2Jd5 8 h3 t2Jb6 9 i.b3
tion after White's l Oth move and 10 l2Jc6 10 0-0. Could it be that Alek
..tgS, discussed in TCA. hine players are avoiding this line of
b) 9 a5 should probably be met
••• the Alekhine Defence?
by 10 a4, for which see the next
game. Instead, 10 c3 a4 1 1 ..1c2 ..1e6
1 2 l2Ja3 ..1d5 1 3 exd6 exd6 14 ..te3
8
l:te8 15 l2Jh2 'it'd7 left White strug
gling to maintain the balance in the
game J .Polgar- Khalifman, Amster
dam Donner mem 1 995: 16 'ii'd2 l2Ja5
1 7 ..1f4 l2Jac4 1 8 l2Jxc4 ..txc4 1 9
l:tfe l l:txe l + 2 0 'ii'xe l 'ii' b5 2 1 'ii'c l
h5 2 2 t2Jf3 ..1d3 2 3 ..1xd3 'ii'x d3 24
..1h6 ..tf6 1h-1h. In any case, White
should certainly avoid 10 a3? a4 1 1
..t a2 dxe5 1 2 dxe5 t2Jd4 1 3 t2Jxd4 9 ••• tlJaS
'ii'xd4 14 l:te l l:ta5 ! (Anand). 9 dxe5?! 1 0 dxe5 t2Jd4 ( 1 0 . . . i.e6
.••
.ib3 ! ? 0-0 9 h3 lt:)a5 10 0-0. What f4 l:.xg5+ 23 fxg5 'ii'd 7, with com
makes Speelman's notes to the game pensation.
interesting is that he was apparently 15 ••• b6
completely unaware that he had B lack has no way to capture on e5
transposed to a main line of Alek without giving himself an unpleas
hine's Defence, both during the ant pawn structure since 15 lt:)xe5
•••
game, and when annotating it. It is a 1 6 lt:)xe5 .ixe5 17 .ixh6 is good for
rare treat to a get top-class grand White due to Black's weak kingside.
master analysing a main line of an Nor does the exchange sacrifice
opening as if it were virgin territory, 15 l:.xf3 1 6 'ili'xf3 lt:)xe5 look par
.•.
unshackled by any notion of the ' ac ticularly good for Black.
cepted wisdom' on the position. 16 e6! ? lt:)f6
10 •.. h6! ? 17 lt:)xf6+ l:.xf6
This i s a new move, preventing 18 lt)d4 (D)
i. g5 . While it is hard to believe the
move should give Black full equal
ity, it certainly provides more op
B
portunities than the continuation
10 lbxb3 1 1 axb3 .if5 12 .if4 f6 1 3
.••
Speelman directly targets e5, but he feared the response 19 'it'e4!, but I
since after White's next move he is think he underestimated his pros
unable to destroy the pawn, he was pects here:
later critical of the idea. He also con a) 19 l:.b8 20 lZ)c6 .ib7 21 lha7
..•
attractive for Black) 2 1 gxf5 l:.xf5 22 l:.xe7 25 l:.a8 'i!Vxa8 26 'ii'x a8+ �h7
Alburt Variation: 4. . . g6 69
27 i.c3 and it's not so easy for the 23... a5 would be answered by 24
black rooks to co-ordinate. �a3 ! intending �b5-c3-d5 .
a2) 23 l:c7 24 l:a8 ! 'ifxa8 25
.•. 24 b4 'iVf5
�xe7+ is the same trick in reverse 25 'it;>h2
order. 25 bxc5 i.e4 ! is OK for Black
a3) 23 ... 'iVf8 24 i.c3 l:f5 25 �e5 since 26 g4? allows 26 .. .'ii'xc 5+.
±. 25 ... 'iVe4!
a4) 23...'it;>h7 24 i.c3 (24 �e5 Speelman considered 25 ... i.e4?
'iVc8 25 �d7 l:xe6 26 'iVxe6 'iVxd7 unplayable due to 26 g4 (26 . . . 'it'xe6
= ) 24 . . . l:f8 25 i.xg7 'it;>xg7 26 �e5 27 .libe l ), but Black is unfortunate
l:c7 looks rather unclear. that after 26 . . . i.d3 27 gxf5 i.xe2 28
b) 19 ... d5 forces White to play l:f2 i.xc4 29 fxg6 i.xe6, White has
precisely to avoid being seriously 30 l:e l , which ties Black up some
worse, viz. 20 �c6 dxe4 2 1 �xd8 what.
i.xe6 22 �c6 i.xc4 (22 . . . a5) 23 26 l:be1 l:txf4
�xe7+ 'it;>f7 24 bxc4 'it;>xe7 25 i.c3 27 'iVxe4 l:xe4 (D)
l:f7 26 i.xg7 l:xg7 27 l:fel 'it;>f6 28
l:xe4 l:e7 =.
19 i.c3 ! a6
w
Black must prevent �b5 .
20 f4! ?
Speelman suggested 2 0 �c2! .
20 . . . 'iVf8
21 �c2 c5!
21...l:xf4 22 l:xf4 'ifxf4 23 i.xg7
'it;>xg7 24 l:fl is 'terrific for White' . I
can believe that Black is in trouble
here, e.g. 24 . . . 'iVg5 25 l:f7+ 'it;>g8 26
�e3. A tense ending has been reached.
22 i.xf6!? I do not intend to present detailed
22 b4 cxb4 (the reply 22 . . . l:xf4 analysis - the interested reader
might be worth considering here, be should refer to Speelman's notes in
cause by comparison with the last the March 1 995 Chess Monthly.
note, the black queen has the e4- 28 bxc5
square at her disposal) 23 �xb4 28 l:xe4 i.xe4 29 �e3 cxb4
l:xf4 24 l:xf4 'iVxf4 25 l:fl i.d4+ works out OK for Black.
26 �h l 'iVe5 27 'it'g4 i.xc3 28 bxc3 28 ... l:txc4 29 �e3 l:xc5 30 b4!
(Speelman) gives White play for l:e5 31 �c4 l:tg5 32 l:e2 i.d5 33
the pawn. l:f4!?
22 'iVxf6 33 �xb6 l:e5 ! 34 l:c2 i.xe6 gives
23 l:ab1 l:f8 Black some winning chances.
70 Alburt Variation: 4. . . g6
33 b5 34 tbb6
.•• 7 a4 (D)
34 h4 ! ? was analysed to approxi
mate equality by Speelman.
34 ... .tb3 35 tbd7?
35 tbc8 ! is better, though 35 . . .l:r.e5
36 tbxe7+ �h7 37 l:r.fe4 .txe6 is
safe for Black.
35 ... l:r.f5 ! 36 l:r.xf5 gxf5 37 �g3
.tc4 38 l:r.e3 .td4 39 l:r.a3 .txe6 40
tbb8 .te5+ ! ? 41 �f2 �f7 42 tbc6
.tb2 43 l:r.xa6 .tdS 44 tba7!? .tc3 45
tbxb5 .txb4 46 l:r.al f4! 47 l:r.cl .tc5+
48 �fl e5 ! ? 49 tbc3 .tc6 50 l:r.al ? !
�e6 5 1 l:r.bl ! .ta7 52 l:r.al .td4 53 It was Paul Keres who first drew
l:r.a3 e4 54 tbe2 .tcs 55 l:r.b3 �e5 attention to the potential of this pawn
Speelman suggested instead get thrust. White hopes to force a weak
ting the f-pawn moving directly, so ening in Black's queenside; Black's
that it cannot become blockaded on task - by no means an easy one - is
f2. to demonstrate that the weakening of
56 l:r.b8 f3 57 gxf3 exf3 58 tbg3 White's own queenside is no less
.td7 59 h4? significant.
59 tbhl ! ! would have forced Black 7 ... aS
to find the winning line 59 . . . .txh3+ This is the normal reply to 7 a4 .
60 �e l .tg2 61 tbf2 �d4 62 l:r.h8 Alburt experimented with the moves
.tb4+ 63 �d l .tfl 64 l:r.xh6 .te2+ 7 . . . d5 and 7 . . . c6, but never really
65 �c l .tel ! . proved them viable, and so they have
5 9....th3+ 60 �el f2+ 6 1 'i!te2 �4 vanished from practice. The only
62 l:r.g8 ! ? .tg4+ 63 l:r.xg4+ �xg4 64 move to make White think twice
tbe4 �h3! 65 �fl h5 about playing 7 a4 is 7...dxe5!?. Then
This is a position of reciprocal after 8 aS (otherwise White is worse)
zugzwang. Black has a choice:
66 tbf6 �xh4 67 tbe4 �h3 68 a) 8... tbd5 is not an ambitious
tbf6 h4 69 tbe4 �h2 70 tbf6 �g3 71 move; White should be somewhat
tbe4+ �f4 72 tbxd6 h3 0-1 better after 9 tbxe5 0-0 10 0-0:
al) 10 ... .te6 1 1 l:r.el tbd7 1 2 tbf3
Game 1 3 tb7f6 (White is a little better, but
Nunn Howell
- now seems to get carried away) 1 3 c4
Isle of Man 1994 tbb4 14 .td2?! tbd3 ! 15 l:r.xe6? fxe6
1 6 'ii'e 2 e5 ! (rescuing the knight) 1 7
1 e4 tbf6 2 e5 tbd5 3 d4 d6 4 tbf3 g6 tbxe5 tbxe5 1 8 dxe5 tbd7 1 9 .tc3
5 .tc4 tbb6 6 .tb3 .tg7 tbc5 20 .tc2 tbe6 21 'ft'e4 'i!th8 22
A/burt Variation: 4. . .g6 71
John Nunn, in The Pircfor the Tour 14 i.xe6 fxe6 1 5 lLJd2, followed by
nament Player. (The move order un bringing the knight to f3 . For exam
der discussion was l c4 g6 2 d4 if.. g7 ple: 15 ... 1:tad8 1 6 lLJf3 'ii'c4 17 'ii'xc4
3 lLJf3 d6 4 i.c4 lLJf6 5 'jWe2 0-0 6 e5 tt:Jxc4 1 8 lLJg5 ± Timmerman-Weij
dxe5 7 dxe5 lLJd5 8 h3 lLJb6 9 i£.b3 erstrass, corr. 1 99 1 , or 15 ...1:tf5 1 6
tt:Jc6 10 0-0 lLJd4 1 1 lLJxd4 'jWxd4 12 tt:Jf3 'ii'c4 1 7 b 3 'i¥xe2 1 8 l:txe2 l:taf8
l:te 1 a5 13 a4.) 1 9 i£.d2 lLJd5 20 l:tae 1 and again the
b) 13...1:ta6 was successful in the e6-pawn is sick; B orge-Simonenko,
game B olzoni-Simonenko, Manila Manila OL 1 992.
A/burt Variation: 4. . . g6 75
'ii'd 7, Chetverik gave 15 liJf3 .ie6 try, and one that looks OK: 15 lbb5
16 .l:td 1 'ii'c 8 17 'ii' b5 .ixb3 18 'ii'x b3 i.xb5 16 'ii'x b5 c6 17 'ii'e 2 lbd5 1 8
as slightly in White's favour, while c 3 'ii' b 6 1 9 .ic4 .l:tad8 20 .ig5 h6
15 ltJe4 'ii'c 6 16 ltJg5 h6 17 liJf3 21 i.e 1 e6 22 h4 .l:td7 23 g3 .l:tfd8
.l:tfd8 1 8 .ie3 'ii'e4 ! gave him ade (B lack's position is solid; there now
quate play in I.Almasi-Chetverik, follows some tame j ockeying for
Budapest 1 994 . position) 24 'it>g2 'ii'c5 25 i.b3 'fib6
f) 13 .id7 and now:
•.• 26 .ic4 'ii'c5 27 i.b3 'ii' b 6 28 'ii'c4
fl) 14 ltJ c3 (D) (was praised by lbe7 29 .ie3 'ii'c7 30 'ii'e 2 ltJf5 3 1
Chetverik, but I find some, well actu i.f4 'ii' b6 32 .ic4 'ii'c5 3 3 .l:tab 1 .if8
ally most, of his supporting analysis 34 .ib3 112-112 Leko-Timman, Wijk
incomprehensible): aan Zee 1 996.
f2) 14 c3 'ikh4 ( 1 4 . . . 'ii'c 5 15 e6
.ixe6 16 i.xe6 fxe6 17 'ii'x e6+ .l:tf7
1 8 .ie3 'ii'c6 19 i.xb6 cxb6 20 liJd2
B
.l:td8 2 1 'ii'x c6 bxc6 22 .l:te2 was
slightly better for White, Wahls-Tim
mer, Bern 1 994) 1 5 liJd2 i.c6 1 6
ltJe4 i.xe5 1 7 .ig5 'ii' xe4 1 8 'it'xe4
.ixe4 19 .l:txe4 ltJd7 20 .l:td 1 .id6 2 1
.ixe7 lbc5 22 .ixf8 Wxf8 112-112 Ste
fansson-Egger, Moscow OL 1 994.
10 dxe5 lba6 (D)
10 ltJc6 transposes to the pre
.•.
Black the bishop pair, and White Instead Nunn analysed 17 c6 1 8•••
more space - not too bad a deal for lLlxg7 'iii> x g7 1 9 .l:td4 'ii'd7 20 .l:.h4
Black. 'iii> g 8 2 1 i.h6 .l:tfd8 22 .l:te l ;!;. The
1 1 ... ltJcS text move weakens Black' s king
12 .l:td1 position, but it is surprisingly resil
12 i.a2 (Fritz) is possible since sient.
12 . . . ltJcxa4 ( 1 2 ... lLlbxa4 13 i.e3 b6) 18 .l:txdS "ii'c6
1 3 i.e3 lLlxb2 ( 1 3 . . . ltJd7) 14 i.xb6 19 "ii'd3
cxb6 15 'ikb5 lLld l 16 c3 traps the White threatens to establish a flrm
knight. grip on the position by i.f4 and l:tc l .
12
••• 'ike8 Instead 1 9 .l:t bS would be met by
13 ltJc3 lLlxb3 1 9 ... b6 intending ... 'ii'e 6, and 19 .l:td3
13 .id7 14 ltJd4 lLlbxa4? (instead
..• by 19 . . ..l:tfd8 20 l:tg3 'ii'e6 (Nunn).
1 4 . . . lbxb3 is better, transposing to 19 e6
the note to Black's 14th) 1 5 lLlxa4 20 .l:td7 i.xeS
lLlxa4 ( 1 5 . . . i.xa4 16 i.xa4 lLlxa4 1 7 21 i.h6 (D)
lLlb 5 wins material) 1 6 e 6 (Nunn) is
a disaster for Black.
14 cxb3 i.e6
B
14. . i.d7 15 ltJd4 gives Black the
.
game and the post mortem, that b2) 22 'ii'e3 is Fritz's suggestion.
B lack was without a defence here, b3) 22 f4 'iii> h 8 23 .ixf8 l:txf8 24
and it was only when he came to �c3+ 'ii'xc3 25 bxc3 .ixf4 ;!; (Nunn)
write up his lnformator notes that should be a draw.
Dr. Nunn discovered that providing 22 l1bl
convincing variations to bury all of 22 .11 c l is less good in view of
Black's defensive tries was none too 22 . .'ii'e4 .
.
had always been considered suspect recaptures, but the choice here is
since B lack was reckoned to have largely a stylistic one) 9 . . . d5 1 0 'be2
no follow-up if White played simply it.e7 1 1 ctJf4 0-0 1 2 g3 ctJb8 1 3 b3
6 0-0 (for example 6 . . . e6? ! 7 c4 fol (an idea familiar from the Advance
lowed by 8 exd6 and 9 d5 is very Variation of the Caro- Kann; Black's
good for White). However, the new preparation of . . . c5 has been so tor
idea 6 . . . dxe5 7 lbxe5 lbxe5 may tuous that White has been able to
force a re-evaluation. Here is the re prepare c4 as a powerful response)
cent practical material: 13 . . a5 14 c4 c6 15 .l:tb1 ctJa6 16 'iie2 !
.
early d5 are not too convincing) 7 . . . St.xe2 8 't!Vxe2 'ir'd6 9 'bxc6 'ifxc6
7 .. .fxe6 8 it.e3 (8 ctJg5 i.xe2 9 'ir'xe2 10 c4 'ir'a6 ? ! 1 1 .l:te 1 ± Malishau
lbxd4 10 'ir'e4 is not good - the black skas-Fioramonti, Bern 1 992) 8 dxe5
king can walk to d7 and if necessary (8 St.xg4 lbxg4 9 ii'xg4 g6) 8 . . .it.f5 9
c7) 8 . . . it.xf3 (8 . . . g6 9 ctJg5 St.xe2 1 0 c4 'bb6 10 'bc3 e6 1 1 it.e3 it.e7 1 2
'ifxe2 'ii'd7 1 1 'iff3 lbct8 1 2 lbct2 h6 1 3 'ii' x d8+ it.xd8 (but not 1 2 . . . .l:txd8 ?
ctJh3 yields compensation) 9 it.xf3 1 3 ctJb5) 1 3 g4? ! ( 1 3 .l:tfd 1 c6 1 4 .l:td2
lbxc4 10 0-0 (Pogorelov 's sugges it.c7 is the critical line) 1 3 . . . .tg6 1 4
tion 10 it.g4 d5 1 1 St.xe6 'ii'd 6 1 2 f4 f5 1 5 exf6 it.xf6 1 6 ctJb5 0-0-0! +
'ir'h5+ g 6 1 3 'ir'xd5 i s interesting, but 1 7 .l:tad l ( 1 7 lbxa7+ <it'b8 1 8 'bb5
should be OK for Black) 1 0 . . . d5 1 1 it.xb2 + Blatny) 1 7 . . . it.xb2 1 8 f5 ?
lbc3 g6 1 2 it.g4 'ir'd7 1 3 b3 ! ? ctJd6 exf5 1 9 gxf5 .l:txd 1 20 .l:txd 1 it.xf5 2 1
(this returns a pawn; 1 3 . . . lbb6 1 4 c5 lba4 gave Black an enormous ad
lb b 5 puts more pressure on e 6 - but vantage in the game Korenev-Pan
still, B lack is two pawns up) 1 4 chenko, USSR 1 99 1 .
lbxd5 ctJf5 1 5 ctJf4 St. g 7 1 6 it.xf5 ! 6 0-0 it.e7
exf5 1 7 d5 ! gave White good com 7 c4
pensation in Gil.Garcia-Pogorelov, Before the main line was per
Albacete 1 995 . ceived to be so good for White, the
b) 6 0-0 and now: form of Exchange Variation 7 h3
b 1 ) 6 'bb6 (this move was tried
••• it.h5 8 c4 'bb 6 9 exd6 cxd6 10 lbc3
without success by Mikenas in the 0-0 (note that 10 . . . lbc6 is premature
1 960s) 7 h3 ! it.xf3 8 it.xf3 (the fact since 1 1 d5 secures an edge) was
that White has the option of chop quite popular. Nowadays it occurs
ping off the c6-knight severely limits only rarely. Here are a few examples:
B lack's plans; otherwise exchanging a) 11 b3 'bc6 12 it.e3 ( 1 2 d5 exd5
on e5 would be a good idea) 8 . . . e6 9 1 3 lbxd5 lbxd5 1 4 'ii'x d5 it.f6 ! ex
lbc3 (9 exd6 is thought to give White ploits the fact that White has played
a pleasant edge no matter how Black b3) 12 . . . d5 1 3 c5 ctJc8 ( 1 3 . . . lbd7 ! ?)
80 4. . . .i.g4: The Old Main Line
ter castling. However, White found a gen PCA 1 993) 15 t2Jd6 ! t2Jdxc5 1 6
good response: 14 'ii'b5 ! 'ii'c7 15 .i.f4 .l:i.xc5 .i.xd6 1 7 .l:i.b5 ! , with at least
a6 16 'ii'b 3 0-0 1 7 l:tfd 1 h6 1 8 .l:i.d2 ;!;. some advantage for White in each
9 .i.e3 (D) case.
White can also reach the analo b) 9 a6 seems the best hope for
•.•
This has been the most popular The main line is 12 l£Jc6 (D) :
.••
(25 . . . bxc5 26 dxc5 c6) 26 cxb6 cxb6 is unpleasant for Black - Sharnkovich.
27 bxa6 ± Herrera-Diaz Perez, Cu a3) 13 i.xc5?! 1 4 dxc5 d4 1 5
..•
ban Ch 'hf (Villa Clara) 1 995. 'ir'h5 (this move was criticised i n
4. . . i.g4 : The Old Main Line 83
lnformator 29, where it was claimed tt:\ec6 22 i.e3 l:tad8 23 l:td2 lLle6 left
that 15 i.xh7+ 'it>xh7 1 6 'ii'h 5+ 'it>g8 Black with no difficulties in the end
1 7 tt:\e4 intending lLlg5 was simply ing whatsoever.
overwhelming, but 17 . . . f6 ! 18 exf6 13 'it>h1 'it>h8
l:txf6 ! is unclear, as Shamkovich dem 14 l:tg1
onstrated) 1 5 . . . g6 ( 1 5 . . . h6? 1 6 l:tad 1 I think Blatny 's idea must be to
"fie7 17 tt:\e4 ! dxe3 1 8 lLlf6+ 'it>h8 1 9 meet 14 f5 gxf5 15 l:tg1, which led to
lLlg4 ! intending 2 0 lLlxh6 i s over a quick win for White in the game
whelming) 1 6 'fi'h6 dxc3 17 l:tad 1 Kiik-Janes, Tartu 1 986 (see TCA),
"file? ( 1 7 . . .'ilie8 1 8 f5 ! exf5 1 9 f4 ! in with 15 ... l:tg8, e.g. 16 l:txg8+ 'ilt'xg8
tending 20 l:tf3 and 2 1 l:th3 is very 17 i.h6 (D) when B lack can avert
strong) 1 8 bxc3 l:td8 ( 1 8 . . . f5 1 9 exf6 immediate disaster, but may still find
l:txf6 20 i.e4 ±) 1 9 i.e4 'fi'e8 20 f5 ! himself rather over-stretched :
exf5 (20 . . .'fi'f8 2 1 f6 ! ) 2 1 i.xc6 ! bxc6
22 l:txd8 "fixd8 23 i.g5 '§'f8 was the
game Shamkovich-Alburt. Shamk
8
ovich indicates that 24 l:td 1 ! would
then have been very good for White:
24 lLle7 25 l:td7 lLld5 26 c4 or
•••
haps overly sharp 14 f5 was played back to Kiik-Janes after 18 'iid 2 fol
in Hendriks-B osch, Dutch Ch 1hf lowed by l:tgl .
(Enschede) 1 995. Then 14 . . . exf5 1 5 c) 17 ...b6 1 8 "fid2 f6 1 9 i.h5 'ii'd8
i.f3 i.g5 1 6 lLlxd5 f4 17 tt:lxf4 i.xf4 20 l:tg1 i.f8 21 i.f7 lLle7 22 exf6 lLlg6
1 8 i.xf4 'ilt'xd4 1 9 'ikxd4 ( 1 9 i.g3 23 l:txg6 hxg6 24 'ilt'g5 and White
lLJ8e7 20 'fi'c2 was played in Tseshk forces mate.
ovsky-Alburt, Daugavpils 1 97 8 , but d) 1 7 lLld7 1 8 'iVb3 gives White
•••
current position, the reply 20 . . .'ii'd7 tt:\d7 20 i.h5 i.xh6 21 'ilkxh6 'fi'f8 22
would give White more problems) 'ifxf8+ lLlxf8 23 i.xf7 (23 lLlb5)
19 . . . lLlxd4 20 i. g4 lLle7 2 1 l:tad 1 23 . . . lLle7 24 lLlb5 l:tc8 25 lLlxa7 l:ta8
84 4. . . .tg4: The Old Main Line
26 tLlb5 lha2 27 tLlxc7 .l:.xb2 looks .l:.ag1 .l:.xg2 27 .l:.xg2 'ii'f 7 28 .l:.f2
like the best try, but I can't see Black tLlg6 29 .thS 'ilie7 30 tLlg2
surviving after 28 tLlxe6 tLlxe6 29 30 tLlxg6+ hxg6 3 1 i.g4 is a better
.ixe6 .l:.xf2 (probably bad) 30 .l:.b1 try, e.g. 3 l . . .fxe3 32 'ii'x e3 .l:.xf2 3 3
tbc6 3 1 .l:.xb7 tLlxd4 32 .txd5, while 'ii'xt2 rj;g7 3 4 .th3 "fkf7 3 5 'ii'e3 - it's
28 .txe6 .l:.xf2 29 .tf7 tLlfg6 30 not easy for Black to make progress.
tLlxd5 tLlxd5 3 1 .txd5 .l:.d2 32 .txb7 30 ...fxe3 31 'ii'xe3 .l:.xf2 32 'ii'xf2
.l:.xd4 also looks hopeless. 'ii'f8
So what is Blatny's idea? Now White's position is falling
14 .th4 apart.
15 .l:.g4 tLlc6 33 'ikg1 tLlf4 34 'ikf2 tLlg6 35 'ilig1
16 'ii'g 1 .l:.g8 'ii'rs 36 i.e2
17 'ii'd 1 f5 36 i.xg6 hxg6 leaves White with
18 .l:.g2 tLl8e7 virtually no moves.
19 a3 .l:.g7 36 tLlf4 37 .tf3 tLlxg2
.•.
Conclusion
Looking in the crystal ball, I see a
w
young, talented player studying at
the position after 5 . . . e6 and thinking
'this position can't be too bad for
Black - there must be ways to make
it work ! ' And he will work furiously
at developing new plans for Black,
new move-order tricks to avoid the
bad lines and reach the good ones.
I know neither who nor when, but
21 tLld2 g5 22 tLlf3 gxf4 23 tLlxh4 state this with certainty, for such is
fxe3 24 fxe3 ltf8 25 'ii'd2 f4 26 the way with chess openings.
1 1 Flohr-Agzamov Variation: 5 ... c6
The old books tended to recom d4-d5 at the right moment, it may
mend 6 l2Jg5 , with the idea that the j ust be that in those games, Black
position after 6 . . . i.xe2 7 'ii' x e2 is could have stopped there being a
very good for White - it is hard to 'right moment' . The main specific
imagine this view being challenged. reason for seeking an alternative to
Therefore Black must play 6 . . . i.f5 , 6 . . . l2Jb6 is the line 6 c4 l2Jb6 7 l2Jg5 ,
when White sharpest attempt is 7 e6, when the position after 7 . . . i.f5 is far
leading to very messy positions, as more pleasant for White than that
we shall see in Game 1 5 . In TCA, I after 6 l2Jg5 i.f5 - the black knight
presented some new analysis after 7 is further from the kingside, which
i.d3, arguing that this move was makes a crucial difference if White
harmless for Black. This analysis starts immediate action there. Thus
has, as far as I am aware, stood the Black is condemned to a dull de
test of time, and consequently there fence after 7 . . . i.xe2 8 'it'xe2. While
have been few recent outings for 7 this is considerably healthier for
i.d3. Indeed, in practice when White Black than the line 6 l2Jg5 i.xe2 7
loses time to exchange the light 'i*'xe2, it is still no fun.
squared bishops, he can easily get White has other methods after 6
into the same sort of mess defending c4 l2Jb6. One is a form of Exchange
the e5-pawn as discussed above. One Variation, popularised by Yudasin, 7
idea for B lack is to play . . . h6 and exd6. However, this should not
. . . g5, threatening to kick away the trouble B lack provided he keeps his
knight, which will then be on f3, and knights flexible and does not balk at
preparing . . . i.g7 to attack the hap playing . . . d5 when necessary. It is
less pawn once again. hard to see why in many games
Another major option for White is Black avoids this standard thrust, in
6 c4, the logic being that if White stead shuffling pieces aimlessly -
wishes to play this move, then he with the pawn structure d4 & c5 vs
may as well do so when the e7- d5 & c6, White can find play on the
square is unavailable to the black queenside, but Black has counter
knight. Then 6 . . . l2Jb6 is by far the play against d4 and often some
most popular move, presumably be cheap threats against c5 too. Con
cause it keeps White's centre under sider the following position, in
fire, but if 6 . . . l2Jc7 does not give which Black's threat of . . .l2Jc5 forces
Black too passive a position, then it White to waste some time (D):
deserves to be given serious atten White has a more critical line: 6
tion. Ian Rogers told me that he felt c4 l2Jb6 7 l2Jbd2 (as played by John
that the retreat to c7 was fully play Nunn in quite a few games over the
able. Although there have been sev years), a maj or point being that
eral games that have suggested that 7 . . . dxe5 8 l2Jxe5 gives White far too
White should be better if he plays much activity. Black has little choice
Flohr-Agzamov Variation: 5. . . c6 87
but to answer 7 . . . lLl8d7 . Then 8 exd6 Although B lack's main aim is to at
gives White a slightly improved ver tack and undermine e5 , to do so with
sion of Yudasin's approach, while 8 a move like . . . b5 is a rather desperate
lLlg5 is the interesting move. Then measure at the best of times.
8 . . . ..tf5 looks suspect in view of White may be able to find some
Klovans's forceful reply (see the advantage in lines where he retains
notes to Game 1 7 ) , so 8 . . . i.xe2 9 e6 the option of playing lLlc4, but first
(quite a shock if you haven' t seen the plays some useful moves, for in
idea before ! ) 9 . . . f6 is necessary. This stance 1 1 'ir'e2 (instead of 1 1 lLld2-c4
is OK for B lack, but he really must reaching the diagram we have just
know it in some detail - see TCA. seen). This is not terrifying for
Coming back to the move 6 0-0, Black, but counterplay is hard to
this has been quite popular in the last generate. For this reason, I have
few years, in part due to a much-pub mentioned in the notes to Game 1 8 ,
licised win by Gufeld, in which he a n old and forgotten game b y Keres
quickly brought the knight to c4. In in which he was successful with
fact the idea is not new; Amason had 6 . . . e6 after 6 0-0. I am aware of no
played the same sequence in 1 9 8 3 . other games with this move, so it de
However, h e had the misfortune to serves at least a second look.
try it against Agzamov, who replied
sensibly, equalised, and went on to Enough chat; on with the enter
win. Gufeld 's opponent instead al tainment!
lowed a very attractive sacrificial
breakthrough (D): Game 15
This is the position where Agza de Fi:rmian Burgess
-
i.b4+ 25 'it>fl l2Jc3 26 l2Jxc3 .ll x c3 29 . . . 'it>xg7 (29 . . . 'it>h8 30 .l:Ih7+ 'it>xh7
27 .l:Id l does not look sufficient for 3 1 �d7+ 'it>h8 32 'it'xc8+) 30 'Wd7+
Black. 'ifi>f6 3 1 i.g5+ 'it>xg5 32 .l:r.g l + 'it>f4
16 ..tn 33 �f7+ 'it>e4 34 'Wxe6+ 'it>xd4 35
17 l2Jxf8 .l:Ihxf8 .l:Ig4+ 'it>c5 36 �c4#.
18 gS liJdS 29 .l:Idg1 .l:Ic4 (D)
19 l2Jxd5 cxd5 (D) I decided I might as well allow a
20 'fibS+ pretty mate. 'That's the editor in you
20 'ikf3+ can be met by 20 . . . 'it>g6 ! , coming out ! ' - de Firmian.
a s pointed out b y d e Firmian ( I con 30 .l:Ixg7+ 'itt'S
fess that I had intended the inferior 31 .l:Ih7 eS
20 . . . 'it>g8 2 1 �xd5+, which gives 32 'i!fxbS 'ii'xbS
Black less compensation), when 33 .ll h6+ 'it>e8
Flohr-Agzamov Variation: 5. c6 91 ..
move, but really it has no advantages said for Black simply getting on with
over castling immediately) 10 lbc3 it since his pieces are already well
0-0 (I can ' t see anything better) 1 1 placed to start action against White's
.i.f4 d5 1 2 c5 lb6d7 (Black's ideal pawn centre. 11 c5 lb6d7 12 b4? !
piece deployment is bishop on f6 and (this seems rather pointless since
knights on e6 and f5 , so I was White is not going to get far with a
tempted to play 12 . . . lbc8, but felt I queenside blitzkrieg; 1 2 .i.f4 .i.f6 1 3
needed more control over e5 ) 1 3 'ii'd2 i s more flexible) 1 2 .i.f6 1 3
...
a3? ! lba6 (holding up the b3-b4 ad .i.f4 .l:r.e8 (it's never very clear what
vance, and envisaging the manoeu the rook can achieve on the e-file in
vre . . . lbc7-e6 gaining time on the such positions; sometimes it just
f4-bishop) 14 'ii'b l ? (neglecting d4) helps White to seize control of the e
14 . . . .i.f6 15 .l:r.d l ( 1 5 Wd3 is neces file, whereas if Black fails to oppose
sary, but Black is doing very well the e-file, a build-up of major pieces
Flohr-Agzamov Variation: 5 . . c6 93 .
'ifxd4 26 'ife7 tbf6 (abandoning so he should not delay this move un
the queenside to its fate; instead duly) 1 5 d5 tbbd7 1 6 h3 i.h5 1 7
26 . . . tbe4 27 'ifxd7 'ifxf2+ 28 �h 1 tbe4 i.g6 1 8 tbfd2 tbf6? ( 1 8 . . . f5
tbg3+ 29 �h2 'iff4 30 'ife6+ �g7 3 1 may not be too bad for Black, since
� g 1 i s not particularly clear) 27 White cannot play 1 9 c5? fxe4, viz.
'ifxb7 l:tf8 28 'i¥c7 (28 'ifxc6 tbge4 20 cxd6 i.xd6 21 dxc6+ i.. f7 or 20
29 l:tc2 tbxf2 30 l:txf2 tbe4 wins for dxc6+ i.f7 21 cxd7 i.xb3 22 dxe8'if
Black: 3 1 'ifxd5+ 'ii' xd5 32 i.. c4 'ifxe8 23 tbxb3) 1 9 i.b6 'ii'd 7 20
'ir'xc4 33 bxc4 tbxf2 34 �xf2 �f7 , tbxf6+ i.. xf6 21 i.xa5 gave White a
etc . , but 28 l:tc2 tbge4 2 9 'ifa6 i s clear extra pawn in Yudasin-Fernan
more robust) 28 . . . f4 (Agdestein de dez Garcia, Dos Hermanas 1 993.
cides to go for the throat) 29 'ir'xa5 f3 9 •.. 0-0
30 'ifc3 'ii'h4 3 1 l:tc2 tbh5 (3 1 . . . tbfe4 10 l:te1 i.. f6!
3 2 'ife3 tbxf2 ! also wins : 33 l:txf2 This novelty was a recommenda
'ifg3 or 33 'ii'xf2 fxg2 34 'ti'xg2 tbf3+ tion in The Complete Alekhine. The
35 �h 1 tbe1 ) 32 �h2 (32 'ir'e3 fxg2 idea is to keep the queen's knight
3 3 i.. x g2 tbf4) 32 . . . fxg2 33 i.. x g2 flexible and to put pressure on d4 .
l:txf2 (33 . . . tbf4) 34 l:txf2 'ir'xf2 35 10 tb8d7? ! 1 1 a4 a5 ? ! ( 1 l . . .d5) 1 2
...
is the logical way for B lack to con positions the decision to play . . . a5
tinue ( 1 l . . .i.xf3 12 i.. xf3 tbc4 might must never be taken lightly. This
be viable too). Instead 10 tb8d7?! ... move weakens the queenside, and if
1 1 a4 (now that Black cannot follow it is not linked to a specific plan to
. . . a5 with . . . tba6-b4) l l . ..a5 1 2 tbbd2 hold up White's queenside expan
l:te8 1 3 l:tfe1 tbf8 14 'ifb3 l:tb8 (once sion, then is better omitted.
B l ack starts having to make passive 12 l:ta3
defensive moves on the queenside, White continues much as in the
something has gone wrong; games aforementioned Yudasin game. This
like this one reinforce my view that is less effective here for two reasons:
Black has no plan better than . . . d6-d5, Black is able to play . . . tba6 (and if
Flohr-Agzamov Variation: 5... c6 95
supplying this game by e-mail with l:r.f7 l:r.e8 19 "ii'f3 paralyses Black) 15
brief comments. tLle4 'i'xe6 16 tll g 5 "ii'g 8 17 0-0
tllxc4 (allowing a cheapo, but there
was nothing good: 1 7 . . . lll 8 d7 1 8
tllf7 ; 1 7. . .h6 1 8 lllf7 l:r.h7 1 9 i.d3 g6
20 "ii'c 2 l:r.g7 21 lllx h6) 18 l:r.xf8+
<li>xf8 19 "ii'fl + won material in
Klovans-Zetterberg, Gausdal 1 995 .
8 exd6 exd6 gives White an im
proved version of the type of ex
change variation we saw in the
previous game. Nevertheless, I think
that the same principles as I outlined
there should allow Black a decent
8 h3 game: in particular not holding back
I am rather suspicious of this with the move . . . d5, and only consid-
move, since the best reply to 8 0-0 is ering playing . . a5 when White has
.
13 .id3 l?Je7
14 It.el c5!
14... l?Jg6?! 1 5 .ixg6 hxg6 1 6 .tg5
makes it hard for Black to organise
his position.
15 'iVf4
15 'ir'e4!? looks more logical, given
that . . . h6 and . . . g5 is part of Black's
plan.
15 ... l?Jc6 ·
15 ... h6 is safer, and should lead to
similar play to the game, but denying Black must make an important deci
White the possibility of 16 .ixh7 . sion:
16 .id2 a) 27...l?Jxe5?! 28 lia8+ �d7 29
16 .ixh7! ? l?Jdxe5 is messy. lixd8+ (29 l?Jxe5+? 'ir'xe5 30 l:.xb7+
16 ... h6 �e8 3 1 .:txd8+ �xd8 3 2 'ir'd 1 + lDd4
This is a standard method of ob wins for Black) 29 . . . l:.xd8 30 l:d 1 +
taining counterplay in the Flohr-Ag and White will pick u p one o f the
zamov line. loose knights.
17 .ic2 g5 b) 27...'ir'b8?! ( ' ! ! ' Fernandez, but
18 'ir'e4 0-0-0 this does not appear justified) and
19 .ic3 i.e7 now while 28 l:Ia2 l?Jxe5 29 l?Jxe5
20 a3 'ir'xe5 + 30 l:.xc2 'ir'h2+ 3 1 ..t>fl 'ir'h1 +
Fernandez considered 20 i.a4 32 ..t>e2 'ir'xb1 -+ works well, and 28
lDb6 2 1 .ixc6 'ii'xc6 22 'ir'xc6+ bxc6 l:.bxb7 "i!Vxb7 29 l:.xb7 l?Jf6 ! -+ is a
equal . In return for his weakened tremendous way to exploit White's
queenside, Black has control of the vulnerable king, 28 li aS! l:.h6 29
d-file, and prospects of opening the "i!Ve4 l:.dh8 30 'iii> f l lih l + 31 l?Jg1
kingside with . . . h5 and . . . g4. looks unclear.
20 h5 c) 27 ... 'iVc5 seems very strong:
21 b4! g4! 28 l:.a8+ l?Jb8 29 'ir'e4 (29 'ir'f4 'ir'b6 !)
22 hxg4 29 . . . l?Jb4 is just an extra piece, while
After 22 b5 gxf3 23 bxc6 "i!Vxc6 28 l:.axb7 l?Ja3 destroys the co-ordi
B lack has no problems. nation of White's rooks.
22 hxg4 26 l:Idg8
23 'i¥xg4 cxb4 27 'ir'f4 f5
24 axb4 i.xb4 28 .tbl?
25 .ixb4 l?Jxb4 (D) 28 .ixf5 exf5 29 l:.xa7 is better,
26 .te4 'but accurate defence should prevail '
26 l:.ebl l?Jxc2 (26 . . . l?Jxe5 ! ?) 27 - Fernandez.
llxa7! ? is the critical line, when 28 ... l?Jc5
Flohr-Agzamov Variation: 5. . c6 99.
B
zamov:
c l ) 12 .!Dxb6 'ii'xb6 1 3 'ii'e2 i.c5
14 a3 a5 15 c3 (Howell-Burgess,
Hastings Challengers 1 99011) 15 ...0-0
intending .. .f6 looks OK for Black.
c2) 12 i.xd5 cxd5 13 .!Dd6+ i.xd6
14 exd6 �xd6 1 5 'it'g4 (S.Pedersen
Burgess, Odense rpd 1993) 15 ...:g8 !?
1 6 i. d 2 .!Dc4 1 7 .tc3 ( 1 7 .tf4 'ii'c 6)
a) l l b5? ! 12 i.xd5 ! cxd5 (Gu
... 17 . . . g6 18 :ad 1 gives White com
feld analysed 12 . . . bxc4 1 3 i.xc4 pensation .
.!Dxe5 14 i.f4 ! i.d6 15 .txe5 i.xe5 16 c3) 12 tLld6+ i.xd6 1 3 exd6 'it'xd6
'ii' h 5 +-) 13 .!Dd6+ i.xd6 14 'i¥xd5 ! 14 b3 Vc7 ( 1 4 . . . '1\Ve7 ! ? appears very
0-0 ( 1 4 . . . exd5 1 5 exd6+ �d8 1 6 sensible) 15 Vd4 0-0 16 i.b2 .!Df6
dxc7+ �xc7 1 7 ne7 gives Black an 1 7 'ii'h4 'ii'e7 1 8 i.e4 h6 19 ne3
appalling ending; 14 . . . nc8 15 exd6 (White has compensation, but it's
l\Vc4 { else 1 6 nxe6+ } 16 'Wb7 Vc6 hard to believe Black should be in
1 7 'i¥xa7 +- Gufeld-Goh, Penang much trouble) 19 . . . .!Dfd5? 20 :g3 !
1 99 1 ) 15 'ii'x d6 ( 1 5 exd6 is good if e5 2 1 'fi'xh6 f5 22 i.xd5+ .!Dxd5 23
White can maintain the d6-pawn; :e1 +- Haba-Ratolistka, Czech Ch
this type of position can be quite hard (Luhacovice) 1 993.
to assess) 1 5 . . . �b7 (B lack intends c4) 12 'ii'e2 can be answered by
. . . l:tfc8, . . . .!Df8- g6, . . . nc4 with pres 12 . . ..!Dxc4, reaching line 'c5 ' , but
sure against c2, e5 and White's king) gives Black the useful extra options
1 6 .te3 nfc8 17 ned 1 .!Df8 1 8 c3 12 ... .te7 (as suggested to me by Kaj
.!Dg6 lfz- 1/z Deszczynski-Kadziolka, Rosell) and 1 2 . . . i.c5.
Warsaw 1 993. I don't really believe c5) 12 'ii'd4 .!Dxc4 (else White
B lack's compensation in this line. plays 1 3 .!Dd6+) 13 'fi'xc4 0-0-0 and
b) ll .te7!? 12 ti:ld6+ �f8 (of
... now:
course 12 . . . .txd6? 1 3 exd6 is wholly c5 1 ) 14 a4 i.e7 (14 . . .h6 ! ? intend
unacceptable for B lack) 1 3 c4? ! ( 1 3 ing . . . g5 looks better) 15 aS ? ! ( 1 5 c3
Flohr-Agzamov Variation: 5 .. c6 101
.
so decided to plough on, and find out move, just getting the knights tan
what the trick was. gled; still, 2 1 .td4 h5 22 .if3 .ie7
14 f4 ttJd7 23 .ixa7 .if6 boils down to approxi
15 fS (D) mate equality.
21 .id4
21 .txe6+ ttJxe6 22 'fi'xe6+ <Ji;b8
leaves White unable both to save his
B
bishop, and to prevent a winning
. . . .ixg3 sacrifice, e.g. 23 .ic3 .ixg3;
23 'fi'e2 .ixg3 24 hxg3 'it'xg3+ 25
<Ji;h l l:f.d6; 23 .td4 .txg3 24 .txa7+
<Ji;a8 25 'ii'e3 .ixh2+ 26 <Ji;h 1 .U.he8 is
a slaughter.
21 .ieS
22 .ixeS 'fixeS
23 'it'xeS ttJxeS
It looks as though Black is about 24 .ih3 <Ji;c7
to be torn limb from limb, but now 25 b4 ttJcd3
comes the anticipated surprise . . . 26 a3 ttJxc4
Flohr-Agzamov Variation: 5 . c6 103
. .
In view of its tremendous popularity, Black should play 6 . . . c6, rather than
this variation, characterised by the 6 . . . �e6, although both may be vi
moves 4 tL!f3 dxe5 5 tL!xe5 g6 (D), able. Moreover, his games are su
should be regarded as the current perb models of how Black should
main line of Alekhine's Defence. handle the resultant positions. Many
players readily move the knight from
d5 , perhaps to hit the c4-bishop.
Kengis tends to leave the knight on
d5, unless forced to remove it, since
the knight radiates such power from
d5 . Black's counterplay frequently
comes from a minority attack on the
queens ide. He also tends to keep his
queen flexible, often staying on d8
for quite a while before choosing be
tween c7 (most often), c8, d7 and b6.
Kengis's trainer has for many
That I have named this variation years been Alekhine's Defence ex
after the Latvian GM Edwin Kengis pert Vladimir B agirov. It is notable
should not prove controversial. He that in the 1 990s he has himself
has played the line for many years, taken up the Kengis line, after many
with extremely good results: on my years of playing the 4 . . . �g4 main
database he has a 65% score with the line - so it seems that the teacher has
variation against IM/GM strength learnt from the pupil in this case.
opposition, and only one loss ! I would This chapter deals with all of
reserve the name 'Larsen Variation' White's alternatives to 6 �c4. By far
for 4 . . . dxe5 as a whole, and perhaps the most dangerous of these is 6
5 tL!xe5 tL!d7, even though Larsen 'ii'f3 , despite it being hardly ever
was not the first to play that line. played. The notes here give some
Kengis's main theoretical contri ideas for Black after the moves
bution has been in demonstrating 6 . . . �e6 7 c4 tL!b4 8 'ii'x b7, but I'm
that after the main continuation, 6 afraid that John Henderson's idea,
�c4, the subject of the next chapter, 8 . . . 'ii'xd4? ! 9 'ii'xa8 tL!c2+ 10 �e2
Kengis Variation: 4. . . dxe5 5 lhxe5 g6 105
would you believe that White has a 14 f3 l2Je3+ 15 �e1 l2Jc2+ 16 �e2 !
plausible losing move? Well, here it (this is much clearer than 1 6 �d2)
is: l l .id2?? l2Jxe5 ! 12 dxe5 lLlxc3 16 . . . l2Jd4++ 17 �d 1 'ii'c 2+ 18 �e 1
0- 1 Chachalev-Ayupbergenov, Vol (almost all of Black's pieces are en
gograd 1 994 . The d2-bishop drops prise) 1 8 . . . .1g7 1 9 1i'xb8+ �d7 20
after either 1 3 .ixe6 1i'xd2 or 1 3 'i!Vb2 (simplest; greedy approaches
'ii'x c3 .ixc4 14 'i!Vxc4 'itxd2. may be viable too) 20 . . . 'itxb2 2 1
d2) 7 c4 l2Jb4 8 'itxb7 (D) is the .ixb2 l2Jxf3+ 22 gxf3 .ixb2 2 3 fxg4
critical attempt to refute Black's and White should win this ending.
play: d22) 8 l2Jc2+ 9 �d 1 l2Jxa1 1 0
...
mally, Black could not hope for 1 7 i.xc5 i.xf3 1 8 i.d4 when White
much by playing these moves, but won material in the game Ernst-Ba
White has set herself up for it here - girov, Helsinki 1 992, 16 0-0 17 l:b2
.•.
'ii'f7 1 5 �e4 i.f5 1 6 �g5 �d7 1 7 (Ernst), 19 ....l:r.xa2 20 l:r.xe7 i.d3 and
�xb6 �xd 1 + 1 8 .l:r.xd 1 axb6 1 9 c5 19... i.d3.
h6 20 i.c4+ 'iilh 8 21 h4 (2 1 �f7+? Back to Anand and Adams in their
'iil h 7 22 cxb6 �a5 23 i.d5 c6 24 PCA Candidates match.
i.c5 cxd5 25 i.. xf8 .l:r.xf8 +) 2 l . . .b5 6
..• i.g7
22 �f7+ 'iilh7 23 �g5+ 'iilh 8 24 6 �d7 has been tried by Stefan
•...
�f7+ 'iilh7 25 �g5+ lf2- lh J.Polgar Loffler, and seems fairly logical. It is
Agdestein, Isle of Lewis rpd 1 995. natural to compare it with the line
Kengis Variation: 4. . . dxe5 5 ltJxe5 g6 109
b2) 17 c4 i.xf3 ( 1 7 . . . lLJc3? is just Anand felt this was better than 19
too cheeky: 1 8 1i'xd8+ .l:f.xd8 19 i.g5 c4 c5 intending ... lLJc6 or 19 i.g5
lLJe2+ 20 'iii> f l .l:f.e8 21 lLJxe5 .l:f.xe5 22 lLJg6 intending 20 f4 h6.
i.e3) 1 8 1i'xf3 ( 1 8 i.xf3 lLJb6 high 19 lLJd5
lights the deficiencies of the white 20 c4 lLJe7
queenside) 1 8 . . . lLJb6 followed by 21 i.e3 .l:f.ad8
. . . 'Wd4 leaves B lack with a very ac 22 b4 .l:f.d7
tive position. 23 �f2 lLJc8!
b3) 17 �b3! i.xf3 18 i.xf3 lLJb6 23 b6 is met by 24 a4 ! intending
.•.
The standard position of the Kengis well-placed on b6, and Black often
Variation arises after the further wishes to advance his b-pawn; a sub
moves 6 . . . c6 7 0-0 i.g7 (D) : sequent . . . tt:lb6-d7 leaves Black a
tempo down on the standard lines.
White's bishop will generally drop
back from c4 anyway, to stop silly
w
threats like . . . tt:\xc3 (and . . . i.e6xc4)
which often win a pawn in ftve-min
ute chess, and even in postal games
(hard to believe, but there are several
examples).
Another point worth considera
tion is precisely how Black should
arrange the possible queenside pawn
advance. The move . . a5 is often
.
Black will begin the charge with the l:tf2 appears awkward for Black
b-pawn ( . . . b5) . This is often pre where is his counterplay?) 16 ... 'ii'xg2
ferred when the . . . a5 advance has lit 17 0-0-0 exd4 1 8 i.d2 'ii'd 5 19 'it'b3
tle sting, and White's a4 thrust does l:td8 gave Black good compensation
not seriously damage B lack's queen for the exchange in Godena-Vagan
side - Black would prefer to meet ian, Reggio Emilia 1 994/5 . White
this with . . . a6, rather than . . . b4. In may also play other 7th moves, e.g. 7
some cases B lack is prepared to re �b3 ..tg7 8 0-0 0-0 9 liJf3 ! ? ..tg4
capture with the c-pawn on b5 ; this is (9 . . . c6 would most likely transpose
generally after a considerable sim to a normal line, e.g. 10 l:te1 lL!d7 1 1
plification of the position, so the po liJbd2) 1 0 liJbd2 a5 ( 1 0. . ...txd4?? is
tential outside passed pawn is more no good: 1 1 h3 ! +- - this indicates
relevant than central control. one drawback of omitting ... c6) 1 1 a3
If this discussion gives you a c6 1 2 h3 ..txf3 1 3 ltJxf3 e6 14 ..tg5
headache, then studying the games lL!e7 1 5 'it'd2 ;!; Psakhis-Vaganian,
of Kengis will provide the details Rostov-on-Don 1 993 .
(and quite a few more insights ! ) in a In practice, a subsequent . . . c6 will
more palatable form. often transpose to variations follow
ing 6 . . . c6 in any case, so why not
Game 20 play the move immediately, and
Short - Vaganian keep the queen's bishop flexible?
Riga Tal mem 1 995 7 0-0
7 lL!c3 is likely to transpose to
1 e4 liJf6 2 e5 liJd5 3 d4 d6 4 liJf3 lines considered under 6 'it'f3 ..te6 7
dxe5 5 ltJxe5 g6 ..tc4 ..tg7 8 lL!c3 c6.
6 �c4 c6 7 liJd2 is an important move-or
6 �e6 has been regarded as
.•• der subtlety, intending to have liJdf3
rather dubious owing to the reply 7 ready as a reply to . . . ltJd7 . See the
lL!c3. Then 7 � g7 8 lL!xe4 �xe5 9
.•• next game for details.
dxe5 lL!c6 certainly is bad in view of 7 .•. ..tg7
1 0 b3 ! (see TCA), but 7 c6 is play
.•• We should note that the last ever
able: 8 lL!e4 lL!c7 9 �xe6 (9 �b3 USSR Championship witnessed a
was the old move, reckoned to give couple of experiments with 7 ltJd7 •••
hxg5 23 lLld4 it.g6 24 'ti'g4 lLlb6 25 23 lLlxf5+ 'C!Vxf5 24 it.c2 'ii'g5 25 'C!Vd3
it.e2 c5 26 ltJf3 .l:txd 1 27 .:.xd 1 .l:txd 1 + 'if g6 26 li'd2 'ii' h 6 intending . . . .:.g8
28 it.xd l it.h6 29 'ir'h3 it.g7 30 'li'g4 and ... rJi;h8) 23 l:te2 'ii'd7 24 .l:tae 1 e6
'li'd8 3 1 ltJxg5 it.f5 32 'ii' h5 it.g6 3 3 25 lDe4 it.xe4 26 l:.xe4 l:.g8 27 ltJf3? !
'ii'g4 it.f5 3 4 'ifh5 1h- 1h. (27 it.d 1 'fkf7 2 8 'ii'e 2 i s correct -
b) 8 lLlf3 (in view of the discus B agirov) 27 . . . 'it?h8 ! + Ulybin-B agi
sion above, this seems logical) was rov, Moscow 1 992.
subsequently preferred . The game 8 it.b3 (D)
I.Zaitsev-A.G.Panchenko, USSR Ch
(Moscow) 199 1 then saw 8 lD7b6 9 ••.
29 'ii' b5 .:.c8 30 'ii'x b7 .:.c7 3 1 'ii' b5 when 9 'C!Ve2 gives rise to a position
.:.c 1 and Black managed to hold this that has tended to be reached via 8
ending. B agirov later provided a re 'ii'e2 0-0 9 it.b3. Black may then try:
finement: 8 ... it.g7 (I have already a) 9 it.e6 may be reasonable, al
..•
Game 2 1
Adams - Agdestein
Oslo (4) 1994
Now a postal game (believe it or
1 e4 ltJf6 2 e5 lZJd5 3 d4 d6 4 lZJf3 not), Timmermans-Etmans, corr.
dxe5 5 lZJxe5 g6 6 i.c4 c6 7 0-0 i.g7 1 985 continued 10 c3 lZJxe5 1 1 lZJxe5
8 lZJd2 lZJd7 i.e6 1 2 lZJd3?? lZJxc3 1 3 bxc3 i.xc4
Agdestein chooses to avoid the 14 l:te1 i.xd3 15 'ii'xd3 e5 -+.
line 8 . . . 0-0 9 lZJdf3 lZJd7 10 lZJd3, White has two serious possibili
which, though not too fearsome for ties :
Black, Adams had played success b 1 ) 10 l:tel lZJxe5 1 1 lZJxe5 trans
fully before. Actually, Black may not poses to the game.
be able to avoid this possibility if b2) 10 lZJd3:
The Main Line: 6 .tc4 1 1 7
b2 1 ) 10 b5 1 1 .tb3 a5 1 2 a4 b4
••• b23) 1 0 a5 1 1 a4 ( 1 1 a3 is maybe
••.
has its own drawbacks, and 12 i.. fl gested by Tisdall, but then 20 i. f3 is
has already been tried without suc Jess clear than Black might like. In
cess by Conquest - see TCA . stead 20 c4? exd4 ! is very good for
Note that 12 c3 allows 12 . . . i.xe5 ! Black.
1 3 dxe5 ( 1 3 Itxe5 tbxc3 is the stand a2) 18 . e5 ! is an important re
. .
1 6 a4, which gave White a clear edge tbb6 and Black bags the material.
in Kuporosov-A .G.Panchenko, Bu b) 17 i.. c4 is suspect in view of
dapest 1 99 1 . 17 . . .tbb6.
1 4 tbd3 i.. f5 c) 17 i.c2 looks best, when 1 7 ... a3
15 i.. g5 Itfe8 1 8 b3 .l:tad8 gives Black very active
16 'ii'd2 a4 play.
The Main Line: 6 i.c4 1 19
break without leaving his position 1:txe6 fxe6 24 'ii'xe6+ <iii>h 8 25 lDe5
too weak, then he should normally and B lack must find some 'only'
do so. moves to avoid instant disaster:
16 dxc5 'ii'xc5 a) 25 i.. xe5? 26 'it'xeS+ <ili>g8 27
..•
"i¥f3 after which B lack quickly got b6? ! (22 . . .l:.fd8 23 .l:.xa5 lLlxd4 keeps
himself into a terrible mess: 11 .'ti'c7
.. Black well in the game) 23 i.e3 'iie2
1 2 'ii'g3 lD7f6 1 3 c4 lbh5 14 'ilf3 lLlb6 24 .l:.d7 .l:.ac8? ! (24 . . . .l:.ab8 is more
1 5 'ilc3 l:.ad8 1 6 lLldf3 lLlf6 1 7 h3 resilient since 25 .l:.xe7 lbxd4 limps
'ilc8 1 8 'i/ia5 'ila8. ll lbxe5 12 dxe5
... on) 25 'iix b6 +- Milos-Llanos, San
a5 springs to mind as more 'natural' . Luis 1 995, 21 1i'c6 22 cxd4 .l:.fd8
...
b3) 1 1 lLlef3 is the latest attempt: 23 l:txa5 (23 l:.xd8+ l:txd8 24 i.xa5
b3 1 ) l l lbc7!? was Kengis's
... is the same) 23 .. .lha5 24 i.xa5 l:.xd4
novelty: looks good for Black, for example 25
b3 1 1 ) In Christiansen-Kengis, 'fibS lbc5 26 'iixc6 (26 i.c3 .l:.xa4)
Manila OL 1 992, play was similar to 26 . . . bxc6 27 b4 lbb3 -+.
the main game: 12 c3 c5 13 i.xe6 b32) 11 ... i.g4 12 h3 i.xf3 . 13
(Wang Zili-Deng Kongliang, Bei iLlxf3 (D).
jing Z 1 993 featured the insipid nov
elty 1 3 i.c2, and following 1 3 . . . cxd4
1 4 lbxd4 i.d5 1 5 lLlfl e5 Black was
very much OK) 1 3 ... lLlxe6 14 d5 lbc7
1 5 lbe4 lLlf6 1 6 c4 lbxe4 17 l:txe4 e6
1 8 i.g5 f6 1 9 d6 lbe8 20 l:txe6 fxg5
2 1 'ild5 'it>h8 22 l:tae1 lLlf6 23 'ikxb7
'ilb6 24 l:te7 'iVxb7 25 l:txb7 l:tfe8 26
l:txe8+ .l:txe8 27 h3 lbe4 28 d7 l:td8
29 l:txa7 i.xb2 30 a4 lLlf6 3 1 lLlxg5
l:txd7 32 l:txd7 lbxd7 reached an
ending in which Black's extra piece
triumphed in the end, although this Now B agirov suggests 13 a5 14 ...
was mainly due to White trying too i.g5 l:te8 1 5 c4 lLl5b6 as unclear, be
hard to win. cause in Ulybin-B agirov, USSR Ch
The Main Line: 6 .tc4 125
19 'ii'h4 llfe8 ! leaves White with 29 i.e1 .l:.e8 30 ltld2 ltld6! 31 �e2
problems over the defence of a2. 3 1 ltlb3 fails to 3 1 . . .ltdxe3 while
19 ... ltlcxd5 3 1 b3 ltlc3 32 l:tbc l hardly solves
20 i.g5 White's problems either.
Since this allows Black to play a 31 .l::t dxe3+! 32 fxe3 ltlf5 33 i.f2
••
1 e4 lL'lf6 c
2 e5 lL'ld5
A: White avoids 2 e5 lL'ld5 3 d4 3 d4 d6
B : 2 e5 lL'ld5 3 d4 d6 without 4 lL'lf3 4 lL'lf3 �g4
C: 2 e5 lL'ld5 3 d4 d6 4 lL'lf3 without 4 . . . c6 55; 4 . . . lL'lc6 56; 4 . . . lL'lb6 56
4 . . . dxe5 4 ... g6 5 �c4 lL'lb6 (5 ... c6 62) 6 �b3
D: 2 e5 lL'ld5 3 d4 d6 4 lL'lf3 dxe5 �g7 : 7 exd6 60; 7 0-0 63; 7 lL'lg5 63; 7
'ilfe2 66; 7 a4 70
A 5 �e2 c6
2 e5 5 . . . e6 6 0-0 �e7 79; 5 . . . lL'lc6 78
2 lL'lc3 d5 7 6 c4
2 ••• lL'ld5 6 lL'lg5 88; 6 0-0 99
3 c4 6 ••• lL'lb6
3 lL'lc3 1 5 ; 3 b3 1 5 ; 3 lL'lf3 d6 4 �c4 6 . . . lL'lc7 9 1
1 5 ; 3 �c4 1 5 ; 3 g3 1 6 7 lL'lb d2 9 6
3 ••• lL'lb6 . 7 exd6 exd6 92
4 c5
4 d4 d6 - see 3 d4 d6 4 c4 lL'lb6 D
4 lL'ld5 2 1 2 e5 lL'ld5
3 d4 d6
B 4 lL'lf3 dxe5
2 e5 lL'ld5 5 lL'lxe5 g6
3 d4 d6 5 . . . c6 57; 5 . . . lL'ld7 57
4 c4 6 �c4
4 �e2 27; 4 f4 27 ; 4 �c4 27 6 g3 1 05 ; 6 ..i.d3 1 05 ; 6 h4 1 05 ; 6 lL'ld2
4 ••• lL'lb6 1 05 ; 6 'ilff3 1 05 ; 6 c4 lL'lb6 1 07
5 f4 6 ••• c6
5 exd6: 6.. .�e6 1 1 3
5 . . . exd6 6 lL'lc3 �e7 34 7 0-0
5 . . . cxd6 6 lL'lc3 g6 36 7 lL'lc3 1 1 3 ; 7 lL'ld2 1 1 3
5 ••• dxe5 7 ..• �g7
5 . . .�f5 42 7 . . . lL'ld7 1 1 3
6 fxe5 lL'lc6 8 l:el
6 . . . c5 7 d5 e6 8 lL'lc3 exd5 9 cxd5 c4 43 8 lL'ld2 1 1 6; 8 ..i.b3 1 1 4; 8 'ilfe2 0-0 9
7 �e3 �f5 �b3 1 1 4
8 lL'lc3 e6 8 ••• 0-0
9 lL'lf3 8 . . . lL'ld7 1 20
9 ..i.e2 46 9 �b3
9
••• ..i.g4 5 1 9 ..i.fl 1 20; 9 c3 1 20
Or: 9 . . ...i.b4 46; 9. . .'i1Vd7 46; 9 ••• ..i.e6 1 23
9 . . . �e7 47 Or: 9 . . . a5 1 2 1 ; 9 . . . lL'ld7 1 2 1