Documente Academic
Documente Profesional
Documente Cultură
the
dark knight
system
a repertoire with 1 Nc6
...
EVERYMAN CHESS
Gloucester Publishers plc www.everymanchess.com
First published in 2013 by Gloucester Publishers Limited, North burgh House,
10 Northburgh Street, London EC1V OAT
The right of James Schuyler to be identified as the author of this work has been
asserted in accordance with the Copyrights, Designs and Patents Act 1988.
Distributed in North America by The Globe Pequot Press, P.O Box 480,
246 Goose Lane, Guilford, CT 0643 7-0480.
All other sales enquiries should be directed to Everym an Chess, Northburgh House,
10 Northburgh Street, London EC1V OAT
tel : 020 7 2 5 3 7887 fax: 020 7490 3 708
email : info@everymanchess.com; website: www.everymanchess.com
Everyman is the registered trade mark of Random House Inc. and is used in this
work under licence from Random House Inc.
James Schuyler is a F I D E Master. H e was Nevada State Champion in 2007 and won
the Virginia State Champion ship in both 2011 and 2012. He has been teaching
chess for over 2 5 years.
Conte nts
Introduction 7
Weak Colour Complex 13
1 ...
ctJc6 and the Kevitz System it is a black knight venturing out from a
Why another repertoire book on 1 ...l2Jc6 dark square to initiate a strategy of
- ? Didn't one come out just a few years dark-square control... need I go on ?
ago? As it turn s out, the subject matter
here is completely different, as are the Does the Dark
types of positions reached. 1...lZ:'lc6 isn't Knight System work?
really an opening yet, but the starting lt works wonderfully, and in two ways.
point for many openings. This book Firstly, it can throw White on his own
generally concerns itself with Black's devices as early as move one ! When
plans for using 1...lZ:'lc6 to force ... e7-e5, forced to improvise, even titled players
as played by Tony Miles, for instance. can play shockingly weak moves or ex
This idea is properly known as the pend their time and energy in the
Kevitz System. (Wisnewski's repertoire opening. Secondly, the opening is fully
book on ... l2Jc6 was all about the sound and playable against all calibre
Nimzowitsch and Chigorin Defences, in of opposition . Black is fundamentally
which Black plays l...ds or 2 .. ds.)
. okay so there will be no need to aban
don the repertoire just because your
The Dark Knight System? opponents are no longer surprised.
For reasons I will lay out, I will b e rec Furthermore, this book is intended to
ommending a fianchetto of the king's leave you a step or two ahead of even
bishop if White resists the ...e7-e5 ad very well-prepared opponents.
vance. These are in fact the most com I would like to take the opportunity
mon positions, and they have a much to say that, in general, the quality of
different feel from a typical Kevitz Sys play in Dark Knight variations has been
tem. Furthermore, I believe that the fi low for both colours. In many common
anchetto is a substantial improvement positions, as early as moves six, five,
over the commonly played moves, and four, and three (!) the unquestionably
therefore a new name is in order. Since best moves have been rarely or never
7
Th e Dark Kn ight System
8
In troduction
9
Th e Dark Kn ight System
10
In troduction
Houdini is not the final arbiter of I h ave not found it necessary to in
anything - especially since it is people clude in the repertoire positions worse
who must play the positions - but he is than a quarter pawn disadvantage, but
a far stronger player than I am, and he is they are not uncommon in the notes. I
nothing if not objective, so when look have not tried to be as precise in my
ing for what passes for the truth, his descriptions of theoretically unimpor
assessments carry considerable weight. tant positions, but the unadorned
When his opinions have not made sense words "advantage" and "better" mean
to me, I have looked deeper. Typically, I approximately 0.26 to 0.39 pawn s,
have become convinced, but sometimes while 0.4 to 0.6 is a "comfortable ad
I am able to convince him - rarely do we vantage" and more would be "clearly
continue to disagree. better" or some such, while more than
Assessments in this book are in one pawn would be "nearly winning".
tended to apply to n arrow ranges.
"Equal" corresponds to an advantage Personal history (with 1 ctJc6)
...
for one player of n o more than 0.09 My love affair with 1 ... tLlc6 goes back to
pawns. "Comfortably equal" is the the late '8os, and my trusty old Batsford
more pleasant half of that range. "Tiny Chess Openings. I had owned it for
advantage", "tiny edge", or "slightly some time before I came across a single
better" is an advantage of 0.10 to 0.17 line by Bogoljubow concerning the
pawns, while "nearly equal" would be a amazing 1 d4 tt:Jc6 !?. Should White
similar disadvantage. In most chess "take the bait" and try to play a kind of
works, such positions are simply la mirrored Alekhine's, a wonderfully in
belled as equal, but I believe that there teresting position may be reached: 2 d5
is far too big a difference between tt:Je5 3 f4 tt:Jg6 4 e4 e5 5 f5 (??) 'i1Yh4+ 6
+0.15 and -0.15 (two to three inaccura �d2 'ifxe4(?) 7 fxg6 'ti'xd5+ 8 �e1
cies or even two to three tempi in many 'fkxd1+ 9 'it>xd1 hxg6.
positions) to let it go without mention.
An "edge" or "small advantage" is be
tween 0.18 and 0.25 pawns. In other
works, such positions are often called
"approximately equal" or ±1=. I under
stand that the style of assessment I am
using implies a degree of precision that
is difficult to attain, but I would rather
strive for precision and risk falling
short than strive for vagueness in the
hopes of evading criticism.
11
Th e Dark Kn ight System
12
Weak Colour Complex
There is no way to play chess well while t The player's king is in the vicinity
adhering to a single idea, or even two of the weak squares.
or three ideas - the game is far too t The player's opponent still h as
complicated - but I have noticed that, the bishop that can infiltrate on
in the Dark Knight System, one concept the weak colour.
assumes far greater than n ormal im
portance, and that is the n otion of the A player will usually experience
weak colour complex. In many of the problems if three of the conditions are
high-level games that Black wins, it is met. Furthermore, if you notice two
by taking advantage of White's weak conditions in your opponent's position,
nesses on the dark squares. I would it is worth seeing if you can aggravate
assume that most readers are familiar his situation.
with the idea of a weak colour complex, A few more observations:
but since it is especially important in
the DKS, I will prattle on about it any t Obviously, if a player's opponent
way. has no access to the "weak"
In some positions, a player is more squares, there is no weak colour
likely to have problems because of complex, regardless of what
weaknesses on a single colour. The con other conditions are present.
ditions: t A knight is a handy piece for the
invader to h ave, because it al
t Most of the player's pawns are lows him to extend the attack to
on a single colour (at least in a the other colour. (A knight sit
certain area of the board). N atu ting on a weak dark square at
rally, the weak colour complex tacks light squares).
will occur on the opposite colour. t A space advantage is no protec
t The player is missing the bishop tion against a weak colour com
that he would need to guard the plex - it can even be a vulner
weak colour. ability.
13
t Weak colour complexes are com has extra protection against be
mon in fianchetto openings - for ing traded off for a knight.
the opponent of the player that
fianchettos, that is. This is be The weak colour complex will not
cause of how each player is likely come up often in the theoretical sec
to set up their pawns, and also tion because it does not usually appear
because the fianchettoed bishop in a full-blown form until the middle
(the one that is likely to be infil game, but the idea permeates the
trating the enemy weaknesses) g ames section.
14
Section On e
1 d 4 ti:Jc6
Others:
a) 2 e3 e5 (or 2 ... d6 and 3 ... g6, to keep
things interesting - White's pawn on
e3 makes a poor impression in this
King's Indian type of position) 3 t2Jf3
sees White try to play a French with an
The Dark Knight first caught my at extra move, but this move order gives
tention as a defence to 1 d4. Although Black a few good options. 3 ... e4 4 t2Jfd2
playing it against 1 e4 requires very f5 5 c4 t2Jf6 6 t2Jc3 �e7 ! transposes to
little additional knowledge, Black may line B2 in Chapter Seven. Also possible
find it useful that his 1 d4 opponents is 3 ... exd4 4 exd4 d5 (an Exchange
are unlikely to know the ins and outs of French) to dry up the game. Instead, 3
the Scotch and the Pirc. Besides, since it c4 transposes to A2; whereas 3 d5 is
is not so easy (for most of us) to meet 1 schizophrenic nonsen se: 3 ...t2Jce7 4 c4
d4, it is especially nice to find an effec d6 5 t2Jc3 f5 6 �d3 l2Jf6 7 �c2 g6 gave
tive defence. lt even neutralizes the Black an extremely comfortable version
London, Trompowsky, and other white of a King's Indian in J.Paasikangas
"easy" systems. Tella-T. Lindqvist, Finnish Team Cham
Apart from 2 e4 which is covered via pion ship 1996 (see Game 1).
1 e4 in Section Two, White has three b) 2 t2Jc3 e5 will soon transpose, af
main moves: ter 3 d5 t2Jce7 4 e4 or 3 dxe5 l2Jxe5 4 e4,
15
Th e Dark Kn ight System
16
1 d4 lLlc6
17
Chapte r On e
1 d4 lt:Jc6 2 lt:Jf3
This is the most common reply, and a find a convenient opportunity to relo
very logical one for a 1 d4 player. White cate for some time, leaving the c-pawn
stops Black's planned 2 ... e5, while try out of play.
ing to maintain a familiar position White h as:
(unlike 2 d5).
2 d6
••• A: 3 C4 1 9
B: 3 ds 22
(: 3 .if4 26
D: 3 g3 30
E: 3 .tgs 33
Instead:
a) 3 e4 is covered via 1 e4 - see
Chapter Five.
b) 3 tt::l c 3 tt::lf6 4 e4 also reaches
Chapter Five.
c) 3 e3 can be met by 3 ... g6.
There may be nothing wrong with d) 3 h 3 ! ? may be a trick to induce
2 .. d5, but with 2 ... d6 Black continues to
. 3 ... e5 4 e4, which is now some sort of
fight for the e5-square. Also, after 2 ... d6 Philidor. 4 ... exd4 5 tt::l xd4 g 6 ! ? is not
Black's tt::l c 6 will usually be presented really bad, but 3 ... tt::lf6 4 tt::l c 3 g 6 5 e4
with an excuse to move soon, freeing will transpose into lines we are more
the c-pawn to join in the battle for the familiar with - see line Cl in Chapter
centre. While the knight is n ot exactly Five again.
glued to the board in Chigorin-type e) 3 c3 was used to good effect in
positions (i.e. after 2 ... d5), it may not B.Kurajica-Z.Mestrovic, Bosnian Team
18
1 d4 ltJc6 2 lDf3
Until I started researching this book, This is now an obscure King's Indian
always played 3 ... e 5 ? ! here, but if variation that can arise via 1 d4 l2Jf6 2
White follows up correctly, 4 d5! will c4 g 6 3 l2Jc3 i.g7 4 e4 d6 5 l2Jf3 o-o 6
lead to unpleasant positions for Black i.e2 l2Jc6 ! ? (6 ... e5 is "normal") 7 d5
(though this is not widely known). The l2Jb8. lt is also .. .
text move is in keeping with the idea of
provoking d4-d5 while leaving the a1- Position One
h8 diagonal open, a common theme in
the Dark Knight.
4d5!
White should play this n ow, while
Black is mid-fianchetto, or it will be l ess
effective.
4 l2Jc3 i.g7 5 e4 i.g4 6 i.e3 e5 7 d5
..'2id4 8 i.e2 i.xf3 9 i.xf3 c5 is a com
fortable variation of the Modern De
fence in which Black has outscored
White, though the game should be
equal after 10 dxc6 bxc6 11 o-o l2Je7 12 White's position certainly is large
c5! 0-0. Another option is 9 ... l2Je7 ! ?, and and, indeed, he has the advantage, but
19
The Dark Kn ight System
Black has counterplay on the dark We will need to look carefully at:
squares. One important thing to realize
is that both ... e7-e5 and ... c7-c5 are poor Al: 8 h3 20
at this stage in the game (though they A2: 8 0-0 21
are frequently played). Both m oves take
squares away from Black that he will Al: 8 h3l2Jbd7
enjoy using, while neither m ove puts
any pressure on White's centre, or
anywhere else for that matter. If that's
not clear enough, notice that ... e7-e5 or
... c7-c5 will place Black two tempi down
in a main line King's Indian (Petrosian
System) or Benoni. Meanwhile, ... c7-c6
or ... e7-e6 actually does pressure
White's centre while reserving the cs
and eS-squares for Black's pieces. Even
so, there's no rush, since White has no
convenient pawn break. Because Black With h 2-h 3 included, it becomes
has good control over when and how possible to start con sidering ... e7-e5,
much the board opens up, it makes though the immediate 8 ... es 9 i.g s has
sense for him to wait for a particularly scored 100% for White.
good opportunity. 9..ie3
Since Black's play is on the dark Logically, White takes measures
squares, White has a space advantage, against 9 ...l2Jcs.
and the game is not open, the trade 9...l2Jcsl
... �g4xf3 suggests itself - but in spite
of the favourable factors, it is still no
bargain to part with the bishop pair. In
any case, Black is not the one in control
of the trade, since White could have
played h 2-h 3 at any point going back to
move five. Indeed, GM Neverov and IM
Bonin did choose 5 h 3 when confronted
with this situation . Personally, that
would please me, as it costs White a
tempo and saves me from a difficult
decision. If you prefer ... �g4, play it as But we play it anyway! White's dark
soon as White plays ..te2. bishop is way too valuable to trade.
20
1 d4 ltJc6 2 tt:'lf3
and clamps down on the dark squares 9 ... e 5 ? ! is more common, but this
�fore engaging in pawn play. White move of Stefano Rosselli del Turco's is
should also be concerned about the the most accurate. Piece play before
!Xlsitional damage he could suffer after pawn play in this variation !
for instance) 12 .l:r.ad1 �xc3 ! ? 13 bxc3, 10 �e3 tt:'lcs 11 f3 es 1 12 tt:'lb3 tt:'lfd7
:hough this is obviously a double- The two-time Italian Champion's
edged sword. 12 ...b6 13 'ii'd 2 tt:'le8? ! 14 l:!.ae1 f5 1 5
Now if 12 tt:'ld4 e5 13 tt:'ldb 5 f5 with exf5 tt:'lxb3 16 axb3 gxfs 17 f4 was not
good play, or 12 �d4 �xd4 13 tt:'lxd4 e5 very successful in S.Flohr-S.Rosselli del
14 dxe6 fxe6 1 5l:tad1 e 5 16 tt:'lf3 b6 17 Turco, Zurich 1934.
:fe1 �b7 18 �f1 �g 7 19 .l:r.e3 tt:'le6 13 'i'd2 b6
with equality. Black may consider White may be slightly better be-
21
Th e Dark Kn ight System
cause of his extra space, but he has no early. I f Black must play for a win,
pawn play. 3 ... lt:'lb8 is better, probably transposing
to line A above.
4 lt:'lxes
If White delays this capture, pre
sumably with 4 e4, Black plays 4 ...ltJf6 5
lt:'lc3 ltJxf3 6 'ii'xf3 g 6 and the fianchet
toed bishop will enjoy its open diago
nal, while White's queen will soon need
to move again (but to no particular
effect); as for example in R.Fischer
J.Schuyler, Richmond 2008 (see Game
6}.
Meanwhile, we can advance .. .f7-f5 Sometimes in blitz White tries to
at any time; we are also threatening avoid trading my "problem" knight by
14 ... lt:'lxb3 1 5 axb3 lt:'lc5, which will force playing 4 ltJd4, but White's knight is no
White to play an awkward m ove to de better off after 4... c5. If 5 dxc6 ltJxc6 6
fend the b-pawn (or else give up his e4, we are in a Sicilian with Black hav
good bishop). ing slipped in an extra move while
White's attention was diverted. Focus
B: 3 dS on the board, White!
4... dxes 5 e4
This is nearly automatic, but not
obligatory - in L.Altounian-J .Schuyler,
Las Vegas 2008, White played 5 c4 e6 6
lt:'lc3 lt:'lf6 7 g 3 ! ? (see Game 7).
s ...lt:'lf6!
22
1 d4 liJc6 2 liJf3
The move Black wants and needs to White is doing well with 7 f4! .
play is ... e7-e6, but after s . . .e 6 ? ? Black is
already lost! Then 6 i..b S+! i.. d 7 7 dxe6!
l.xbs 8 'ifh s ! was brutal in J .Bonin
J.Schuyler, New York 1988 (see Game
8). I shouldn't feel too badly, I suppose
- Mestrovic has made this blunder, and
Miles made it twice! ! (Since Miles was
�iles, he actually lost neither game.)
White now has:
81: 6li'Jc3 23
82: 6 i.. b S + 24 7 i.. b S+ i.. d 7 8 dxe6 i..x bs!
Previously, both Jonathan Speelman
6 i..d 3 is less common and less logi and Emmanuel Bricard have tried
cal. H.Keskar-J.Schuyler, Norfolk 2008, 8 .. .fxe6, each securing a draw - though
continued 6 ... e6 7 c4 i.. c s 8 o-o o-o 9 if you see the gam es, you may not be so
::X3 exds 10 cxds, when I set about eager to repeat their methods.
blockading the d-pawn with 10 ...li'Je8 Instead, after 8 ... i.. xbs, White h as
11 i..e 3 i..x e3 12 fxe3 i.. d 7 13 Yi'f3 cs 14 several paths to the endgame, but al
ifg 3 f6 15 .Uad1li'Jd6 and Black is a bit most no chance of extracting anything
better already. The game concluded 16 from the position.
b3 bS 17 'ifh4 �as 18 �cl b4 19 li'Ja4
=ac8 20 l::tf3 c4 21 .iifl cxb3 22 .:!xc8
.:.Xc8 23 axb3 i.. xa4 24 bxa4 b3 25 'ii'g 4
b2 26 Yi'e6+ 'it>h8 2 7 llxf6 'i¥d8 0-1.
Three years later, H.Keskar-
J .Schuyler, Hampton 2011, went 7 dxe6
l.xe6 with comfortable equality, albeit
with a long struggle for a win (see
Game 9).
23
Th e Dark Kn ig h t System
played four times, but it is slightly 18 ...l:.d6 19 ci;e4 .l:!.xf7 20 f3 ci;d7 - al
weaker than the text. Then 9 .. .fxe6 ! 11 though White can't win, he can try to
tbxc7+ ci;f7 12 ttJbs .l:.c8 ! (better than lose if he likes with 2 1 �xes l:.e7+ 22
12 ... ttJxe4?!, as in I. Kreitner-H.Stenzel, ci;f4 .:!.e2.
Long Island 1997) 13 tbc3 i.b4 14 i.d2
.i.xc3 15 i.xc3 ttJxe4 16 .i.xe s :xc2 is 82: 6 i.bS+ Ji.d7
equal .
b) 9 exf7+ costs a move (as opposed
to waiting for .. .f7xe6) and improves
Black's king position, so it is n ot a good
idea. 9 ... ci;xf7 10 ttJxbs (or 10 'ii'x d8+
.l::tx d8 11 ttJxbs ttJxe4 12 f3 a6 13 fxes
axbs and Black has slightly the better
of the probable draw) 10 .. .'ilt'xd1+ 1 1
ci;xd1 ttJxe4 12 ci;e2 c 6 13 tb c 3 ttJxc3 1 4
bxc3 is roughly equal again.
9 1\Vxd1+ 10 ci;xd1 0-0-0+ 11 �e2 a6
•••
24
1 d4 tLlc6 2 tLlf3
One important point of Black's certainly had the right idea and was a
queenside expan sion is that it secures bit better following 9 ... e6 10 i.. xf6 gxf6
the c5-square for his bishop, ensuring 11 dxe6 fxe6 (11 .. .'iWxe6 is also good} 12
that it won't get shut out of play (as it o-o?! (12 'irh 5+ 'i¥f7) 12 ... 0-o-o 13 'ii'c4
would if White were allowed to seize l:tg 8 14 ti:Jc3, even if he went on to lose
space on the queenside with a2-a3 and after the passive 14 ...l:te8? ! (here
b2-b4}. White's knight h as also been 14.. .'�i'b8 15 .l:r.fd1 i.d6 16 b4 f5 was bet
taken out of contact with the impor ter).
tant d5-square. Notice that if White For White, the most accurate con
ever plays d5xe6, and for some reason tinuation may be 10 ti:Jc3 i.b4 11 0-0
Black doesn't feel like recapturing with i.xc3 12 bxc3 exd5 13 i.. xf6 gxf6 14
a piece, .. .f7xe6 is position ally sound l:tfd1 'i�Vh5 15 'ii'f3 o-o-o, though he has
because the f-file is valuable and the nothing to show for it - instead he
e6-pawn controls critical squares. should have been accurate on move
Hold on ! Couldn't 12 ...b4 have been three!
prevented? Indeed, 12 a3 is possible for Alternatively, the untried 9 ... h6!? is
White and not a bad move, but 1 1 ...b5 playable, although in this case 10 i.xf6
was not played with only 12 ...b4 in exf6 11 o-o f5 12 exf5 'ii'xf5 13 ti:Jc3
mind. After 12 a3 o-o 13 .l:!.d1 exd5 14 i.d6 14 tt:Je4 o-o 15 c4 is a tiny edge for
cxd5 l:.fd8, Black's idea is to pl ay 1 5 ... c5 White.
and, if White doesn't capture, 16 ...tt:Je8 9 e6 10 c4 i.e7!
...
and 17 ... ti:Jd6. White's edge is tiny. Allowing Black to castle short. In
7 a6
... stead 10 ... i.c5 ? ! 11 i.g 5 spells trouble.
I'm not crazy about 7 ... g 6 ! ? with 11 ti:Jc3 o-o 12 1:td1 exds 13 cxds
that silly pawn sitting on e5, but it does
avoid spending a tempo on 7 ... a6, and
Mestrovic is 2-0 with it, which suggests
it is worth a try in a must-win game.
Check out D. Rasic-Z.Mestrovic, Croatian
Team Championship 2001, and J.Barle
Z.Mestrovic, Slovenian Championship
1997, in the games section (Games 11
and 12).
8 i.xd7 'i¥xd7 9 o-o
Somehow White h as done well with
9 i.g 5, though it is not a move that White's d5-pawn is currently an as
should cause problems. In P.Staniszew set, and Black h as two possible ways to
ski-H.Kaulfuss, Darmstadt 1996, Black neutralize it. First, he can attack it di-
25
Th e Dark Kn ight System
redly with ... c7-c6. This is definitely Therefore, 13 . . ..l:!.ad8 1 4 i.. g 5 lt:le8 15
worth considering in some similar po i..x e7 'ifxe7 16 .l:f.ac1 lt:ld6 17 b3 fs and
sitions, but it doesn't work so well here; with all Black has accomplished, per
e.g. 13 ... c6?! 14 i.. g s ! .:fd8 15 i..xf6 haps White should resign? Unfortu
i..xf6 16 dxc6 'ii'x c6 and the eternal nately, chess is not quite that simple,
knight sets in with 17 lt:lds. but Black can now start fighting on
If Black tries to prepare this with equal terms.
... h7-h 6 (in fact not a bad move at all) it
is unlikely he will be fully ready for C: 3 i..f4
14 ... c6 anyway; e.g. 13 ...h 6 14 a3 (14
i.. e 3 lt:lg4!) 14 ... c6 1 5 i..e 3 cxds 16
lt:lxds lt:lxds 17 l:!.xds ii'e6 18 :ad1 and
while Black's position is playable, it is
very dull, and White has a small but
clear advantage.
The second plan is very appropriate
here, which is a timely ... lt:le8 and
... lt:ld6. This idea is useful in many
variations of the Dark Knight, but espe
cially the ones starting with 3 d5. The
point of the knight transfer is fourfold: White plays the London System,
the knight is safe, as White's own d which stops the ... e7-e5 break for now,
pawn shields it from attack; the knight but the bishop bites a granite pawn on
is active - centralized and controlling d6.
the important e4-, fs-, c4-, and b7- 3 lt:lf6!
...
26
1 d4 ltJc6 2 lLlf3
27
Th e Dark Kn ig h t System
28
1 d4 lbc6 2 lbf3
exd4 9 exd4 lt:le4! in K.Gunasekaran i..xf4 i.. xf4 1 5 'ifh3 d5 16 i.. d 3 .l:txd1 17
K.Akshayraj, Dhaka 2005, and went on .:.xd1 lt:le7 and although White is n ot
to win after 10 lt:lc3 ? (but if 10 i..f4! worse, he is once again without his fa
.:e8 11 o-o g 5 ! 12 i.. e 3 g4! 13 hxg4 vourite bishop. I've played the London
8g 3 ! 14 .l.:.e1 lt:lxe2 15 .l:!.xe2 i.. x g4 with System quite a lot, so I know exactly
a big edge to Black) 10 ... lt:lg 5 ! 11 o-o how annoying these ideas can be .
.:::.xf 3 + 12 i..xf3 lt:lxd4 with a free pawn. 8 0-0 es 9 i.. h 2 'ii'e 7
7 lt:ld7
...
29
Th e Dark Kn ight System
1997 (see Games 13 and 14); whereas i n g to m y database, 4 d5 has never ac
V.Golod-E.Sutovsky, Netanya (rapid) tually been played.
2009 (Game 15) shows Black, a strong
GM, fail utterly to contain th e London
bishop.
0: 3 g3
4...ttJb 8
I have selected this retreat over
other options in many variations, and
the more I think about it, the more I
like it. Like General MacArthur, the
A subtle (read: boring) move. knight shall return, likely settling on
Mestrovic has responded 3 ... e 5 ! ? here the newly soft c5-square. Incidentally,
five times, drawing all corners from if ... ttJb4 is played, it is with a similar
expert to GM. This is a good bet for idea: ... a7-a5, ... ttJa6 and ... ttJc5.
Black theoretically, since the endgame 5 i.g2 ..t g 7 6 0-0 ttJf6
after 4 dxe5 ttJxe5 5 ttJxe5 dxe5 6 'ii'x d8 I also like 6 ... e 5 ! ? 7 dxe6 (7 e4 ttJd7 8
'it>xd8 gives only a tiny edge for White. c4 tbe7 is likely to transpose to D2; e.g.
Perhaps your opponent will turn away 9 tbc3 0-0) 7 .. .fxe6 8 e4 ttJh 6 ! 9 c4 tDf7.
from this Mutually Assured Dullness, 7 c4 o-o 8 ttJc3
but I would as soon not give him the This line was topical in the 1920s (!)
opportunity. Thus, the usual solution : with Frederick Yates seen frequently
3 g6!?
..• behind the black pieces again st the
When we must look at: best players of his day, while Alekhine
and Grunfeld championed White. Rich
01: 4 d5 30 ard Reti played both colours. Yates was
02: 4 i.g 2 31 able to defeat Reti, Kmoch, and
Alekhine, the last of these games win
01: 4 d5 ning a brilliancy prize at Carlsbad 1923.
If White is going to kick the knight, 8 ttJbd7
...
he should do it now - though, accord- The most common move, 8 ... e5?!,
30
1 d4 tt:ic6 2 tt:lf3
31
Th e Dark Kn ight System
9 d5
These other moves are almost as Things are still complicated, but
common : Black is slightly better due t o White's
a) 9 �g 5 �f6 ! ? (9 .. .f6) 10 �xf6 tbxf6 looser position .
32
1 d4 t'Llc6 2 t'Llf3
33
Chapte r Two
1 d4 t2Jc6 2 c4
34
1 d4 lLlc6 2 c4
35
Th e Dark Kn ight System
A1: 6 d6!?
5 'ii'xd2
Alternatively, 5 t2Jxd2 t2Jce7 6 d6 ! ? (or
6 e4 d6 7 ..id3 t2Jf6 ! - in this particular
position it is too costly to play 7 .. .f5?!,
activating both White's ..id3 and his 6 cxd6
••.
36
1 d4 lbc6 2 c4
o-o-o i.. d 7 ! 11 lbf3 :res 12 cxd5? :es 13 plain to me why people didn't play
"ii'd4 lbexd5 14 'it>b1 lbxc3+ 1 5 bxc3 3 ... i..b4+. lt seemed obvious at the time
-'Lle4, winning) 10 ...lbexd5 11 lbxd5 that Black's position was bad - after all,
-'Llxd5 12 a3 (this sad move is necessary the d-pawn is hopelessly backward. How
to prevent 12 .. .'i!Va5+; castling just loses can Black possibly evict White's queen
after 12 0-0-0?? i.. e 6, with 13 ...l:tc8+ 14 and achieve the ... d7-d5 advance?
ittb 1 lbc3+ coming) 12 ...l:te8 13 'fi'd4 Often we can't, but there are other
(after thirteen moves, White's kingside options and, depending on how White
is hilarious - the most plausible expla continues, it is actually possible to
nation is that he rolled very bad dice in "play around" White's queen and leave
his Chaturanga game) 13 ... b 5 ! 14 lbf3 the d7-pawn alone for the foreseeable
..i..b 7 15 e3 "fic7 16 i.. e 2 l:tad8 gives future. After all, it is securely guarded
Black plenty for the pawn. and Black's pieces do have other ways
8 0-0
... to develop. This idea is demon strated
Which brings us to: by the variation 9 e4 lbc6 10 lbf3 'ii'a 5 !
11 o-o-o (11 i.. d 3 :es 12 i.. c 2 b6 13 a3
Position Two 'ii'c 5 or 12 o-o J::r. e 6 13 'fi'a3 'ti'xa3 14
bxa3 and White can stop bragging
about his superior structure) 1 1 ... a6 12
'it>b1 .:te8 13 a3 b5 14 cxb5 axb5 15
i..xb5 i.. a 6.
37
Th e Dark Kn ight System
weakens White's control over d5, and once again a target, and again leaves
opens lines against White's king. lt is the dark squares weak. On the other
powerful enough that it can sometimes hand, if Black doesn't find a plan, he
be used even when White has not cas may find himself statically worse with
tled queenside (and ... b7-b5 can occa his backward d-pawn, despite White's
sionally be played without ... a7-a6). "bad" bishop. lt is Black's idle l2Je7 that
Was White's play too co-operative? will save the day by repositioning to
Presumably 9 e4 is the culprit, provid target those soft dark squares: 15 ... t2Jg 6
ing a target for Black's ... t2Jf6 and 16 g 3 t2Jf8 ! 17 l:tad1 l2Je6 ! 18 iVxd6 l2Jd4
.. ."i¥a5. lt also leaves a hole on the d4- (or 18 .. ."ifxd6 19 l:txd6 l2Jc5) 19 'ii'x d8
square which could turn into a long t2Jxe2+ 20 t2Jxe2 .l:.axd8, which is at
term problem, although it did stop least equal for Black.
... l2Jf5, a useful move for Black. There Can White save a tempo by keeping
fore 9 e3 I:te8 10 t2Jf3 l2Jf5 11 11t'd2 d6 ! his e-pawn flexible? Not unless he
(11 ...b 5 ! ? is a great try, but 12 t2Jxb5 wants an e-pawn shoved up his king's
i.b7 13 i.e2 l2Je4 14 'i!i'c2 'ii'a 5+ 1 5 t2Jc3 file: 9 t2Jf3 .:r.e8 (or the wild 9 ... e4 10
t2Jxc3 16 bxc3 t2Jd6 17 J:txd7 .l::t ab8 is a l2Jd4 e3 11 fxe3 b 5 ! 12 t2Jdxb5 a6 13
little better for White) 12 i.e2 (12 l:.d1 tbc7 .l:.a7 14 l2J7d5 t2Jexd5 15 cxd5 llb7
'ilt'b6 13 .ie2 i.e6 14 o-o h6 or with excellent play) 10 l:.d1 l2Jf5 11 'i¥a3
14 ... .l:tad8 ! ?; or 12 e4 l2Je7 13 0-0-0 'iVb6 (11 �d2 e4 12 'iic 2 d 5 ! with advantage)
14 'ii'x d6 'ii'xf2 15 t2Jxe5 and 15 ... i.f5, 11 ... e4 12 l2Jd4 t2Jxd4 13 .:txd4 e3 and
1 5 ... i.g4 ! ?, 15 ... i.e6 ! ?, or 1 5 ... t2Jc6 ! ? although Black's d-pawn remain s,
with equal chances i n all cases) 12 ...b6 White now has a matching one on the
13 e4 l2Je7 14 o-o .ib7 1 5 'ii'd 3. lovely e2-square, and h as at least as
much to worry about.
A2: 6 t2Jc3 d6
38
1 d4 lL'Ic6 2 c4
39
Th e Dark Kn ight System
show V. Rao-J.Schuyler, New York 1986. I hxg5 25 'ii'xg 5+ 'it>h8 2 6 li:Jf5 i.xf5 2 7
had not yet started to use 1 ... li:Jc6 regu .l:txf5 .l:iaf8 28 h 6 'ii'e 8
larly, but it seemed like the perfect
choice again st the straight-laced,
booked-up senior master. We pick up
after 1 d4 li:Jc6 2 c4 e5 3 d5 i..b 4+ 4 i..d 2
i..x d2+ 5 �xd2 li:Jce7 6 li:Jc3 d6 7 li:Jf3 f5.
29 'ii'g 6??
White, short on time and frustrated
about being unable to break through,
commits a h orrible blunder, allowing
his queen and rook to be forked. In
8 g3!? li:Jf6 9 e4 h6 stead, after 29 l:txf6 ! ? li:Jxf6 30 l:i.xf6
Generally a useful move, preventing "fle7 3 1 ki.xf7 'ii'xg 5 3 2 l:lxf8+ 'it>h 7 a
li:Jg 5 and making ... g 7-g 5 possible. Here crazy endgam e arises - presumably
Black prepares to play 10 ... fxe4. White has some advantage.
10 exf5 i..xf5 11 i.. g 2 o-o 12 o-o 29 li:Jh7???
.•.
40
1 d4 lbc6 2 c4
41
Th e Dark Kn ight System
c5 18 .i.c3 lt:'Jxc3 19 bxc3 lt:'Jg6 20 .i.e2 o r 1 4.. JWd7 gives Bl ack slightly the bet
l:tb8 21 o-o lt:'Jf4 or 21 ... .i.e4 ! ?, or 15 ter chances) 9 ... lt:'Jf6 10 lt:'Jg 5 "ille 7 11 g4
.i.d4 lt:'Je4 16 g4! ? .i.g6 17 h4 h 6 with lt:'Jxg4 12 lt:'Je4 .i.f5 13 lt:'Jg 3 .id7 14 .ih 3
just a big mess. 0-0-0 15 .i.xg4 and now in D.Justo
For the cowardly it is n ot strictly P.German, Buenos Aires 1995, Bl ack
necessary to sacrifice a pawn : 8 ... e4 spoiled some fine work with
(instead of 8 ... lt:'Je4!) 9 lt:'Jd4 0-0 is play 1 5 ... .i.xg4?! 16 lt:'Je4 .i.f5 17 lt:'Jf6 which
able, though White has a small advan was approximately equal, whereas
tage after 10 'i¥c2 (or 10 e3 d6 11 .i.e2 15 ... lldg 8 ! 16 .i.xd7+ �xd7 17 lt:'Jf5
.i.g4 12 f3, or 10 g3 c5 11 lt:'Jb5 lt:'Jf5 12 Uxg 7 18 lt:'Jxe7 lt:'Jxe7 would have left
e3 b6 intending 13 ....i.a6, 14 ... .i.xb5 him with chances to convert his advan
and 1 5 ... lt:'Jd6) 10 ... d6 1 1 g 3 c6 12 dxc6 tage.
ltJxc6 13 lt:'Jxc6 bxc6 14 .i.g2 d5 15 cxd5 6 ... d6 7 e4 fs 8 exfs
cxd5 16 .i.c3. Since the main line offers White
c) 5 'ii'c 2 ! ? f5 6 lt:'Jf3 .i.xd2 7 lt:'Jxd2 d6 nothing, it makes sense that in
8 e4 lt:'Jf6 9 .i.d3 was tried in S.Ariste Cas M.Gurevich-D.Zoler, Antwerp 1998,
tano-J.Salgado Gonzalez, Saragossa White looked for and found something
1998, at which point Black panicked and else - 8 'ti'h 5 + ! ? - and was immediately
played 9 .. .f4?, a decision which led to rewarded by 8 .. 5.fi>f8 ? ! , after which
difficulties since he lacked counterplay Black's static king provided a nice tar
against White's big queenside (10 c5! is get for Gurevich's king side storm (see
best). Instead, 9 ... 0-0! 10 exf5 'i!Ve8 is ab Game 21). Instead, 8 ... g6 9 'ii'h 4 fxe4 10
solutely fine for Black; e.g. 11 f3 'iWh 5 12 f3 exf3 11 lt:'Jxf3 lt:'Jf5 12 �xd8 �xd8 13
g4?! 'ir'h4+ 13 'ito>d1 h5 14 g5 'ifxg 5 15 .i.d3 lt:'Jf6 leaves White with enough for
lt:'Je4 'ifxf5 16 lt:'Jg 5 e4!, or 11 h3 'ifh 5 12 the pawn, but no more than that; e.g.
g4?! ltJxg4 13 .i.e2 .i.xf5 14 �d1 'iWh4 15 14 o-o .l:!.f8 15 .U.ae1 'ito>e7 (15 ... 'iit e 8 ! ?,
.i.xg4 .i.xg4 16 'ifxg4 'ifxf2+ 17 'ito>d1 lt:'Jf5 1 5 ... a5 ! ?) and now Mr. H sees nothing
18 lt:'Je4 'i¥d4+ and the suffering of better than 16 .ia5 'it>d7 17 .i.d2 'ito>e7,
White's king is far greater than our repeating position.
small material investment. s ... .i.xfs 9 lt:'Je2 lt:'Jf6 10 lt:'Jg3
s ... .i.xd2+ 6 .i.xd2 Either 10 ... .ig6, as in E.Arlandi
6 'i¥xd2 d6 7 ii'g 5 is pointless be M.Lanzani, San Marino 1998 (see Game
cause of 7 ... h 6 ! 8 'i!Vxg7 ? ! lt:'Jg6 9 lt:'Jf3 22), or 10 ... 0-o, as in G.Grigore
(White should probably prefer 9 h4! P.Brochet, Creon 1999 (see Game 23), is
lt:'Jf6 10 h5 �h7 11 �xh 7 lt:'Jxh 7 12 hxg6 good enough for equality, though in
fxg6 13 l:txh6 lt:'Jf8 14 lt:'Jf3, when 14 ... c6 the latter g am e Black contrived to lose.
42
C h a pter Th ree
1 d4 'Dc6 2 dS
White takes the bull by the horns - a velopment, space, and open lines in
surprisingly uncommon reaction. lt is exchange for his anti-positional cap
worth noting that this is how Miles ture. Simply 3 ... d6 transposes to 1 d4
played against his own specialty when tt:Jc6 2 tLlf3 d6 3 d5 tLle5 in Chapter One.
he faced Zvonimir Mestrovic. b) 3 i.f4 tt:Jg6 4 i.g 3 - does the
2 ttJes
•.• bishop dominate the knight, or does
the knight dominate the bishop? In the
blitz g ame E.Bacrot-C.Bauer, Ajaccio
2007, it looked like the latter after
4... tt:Jf6 5 tt:Jc3 e5 6 e4?! i.b4 7 i.d3 d6 8
f3 tLlh 5 9 .if2 c6! 10 dxc6 bxc6 11 g 3
i.a5 12 f4?! tLlf6 13 f5 tLle7 1 4 tt:Jge2
tLlg4, when Black h as a huge advantage
(though he went on to lose). Bauer's
method of meeting 3 i.f4 should be
copied exactly. 6 dxe6 is an improve
ment for White, but Black has nothing
White normally chooses between : to fear after 6 .. .fxe6 and 7 ...i.b4.
c) 3 tt:Jc3 e6 will soon transpose to
A: 3 e4 43 other lines in this chapter; e.g. 4 dxe6
8: 3 f4 46 fxe6 5 e4 (see line A) or 4 f4 tt:Jg6 5 dxe6
fxe6 6 e4 (see 6 tLlc3!? in line B).
Others:
a) After 3 lLlf3 Black may wish to try A: 3 e4 e6
3 ....�:Jxf3 4 exf3 e5 (or 4... e6 or 4 ... g6), Mestrovic played 3 ... d6!? 4 f4 lLld7 5
but I don't care for it - White has de- tt:Jc3 c6 and went on to draw with GM
43
Th e Dark Kn ight System
Drasko (Game 24). Black has also scored sponse is 5 lZ:lc3! lZ:lg 6 6 'i!Vxd5 lZ:lf6 7
well with 3 ... lZ:lg6?!, which can transpose i¥d3 SLc5 8 SLe3 'Wie7 9 o-o-o o-o 10 lZ:lf3
into lines considered below, but his po with a small advantage for White.
sition is highly suspect after 4 h4!. Diebl s's 4 ... exd5 was tested again in
V.Erdos-R.Rapport, Hungarian Team
Champion ship 2012, to produce an
other entertaining draw (see Game 2 5).
If it is Black who is keen to avoid the
draw, he needs to forego 4 ... exd5 and
pl ay 4 ... lZ:lg 6, transposing to position s
considered in line B below.
4 dxe6!
4 f4? ! is seen here most often and
seemed fine the first 5 5 times it was
used, but on the 56th, the German
master Lutz Diebl played 4 ... exd5 ! and
drew with GM Gyimesi in the
Bundesliga. Bravo! If White takes the
knight, Black draws by perpetual : 5 4 fxe6
...
fxe5 'ili'h4+ 6 'it>e2 'Ml 5 + 7 'it>d2 'ii'h 6+ 8 This recapture is certainly dynamic,
'iii> c 3 'ifc6+ 9 'it>d3 'ifa6+ etc. although the somewhat exposed posi
If 6 g 3 ?, as Gyimesi actually played, tion of Black's king requires careful
then 6 ... 'ifxe4+ 7 'ife2 'ifxh 1 8 lZ:lf3 b6 9 treatment.
lZ:lc3 and now, rather than Diebl's pre Many players prefer 4 ... dxe6 5
mature 9 ... SLa6?, the preliminary 9 ... c6! 'ifxd8+ 'it>xd8, but Black's results have
maintains Black's nearly winning ad been poor - pretty much draws and
vantage (10 ... SLa6 is still coming to ex losses. Nonetheless, Short m ade it look
tricate the queen). easy to defend in S.Gordon-N.Short,
Meanwhile, White doesn't need to British Championship 2011, so his
go in for Gyimesi's contortion s (Gyim m ethod could certainly be tried (see
n astics?) to avoid a forced draw; he can Game 26); and M.Gurevich-M.Rohde,
play 5 exd5 or 5 'ifxd5, although he Philadelphia {blitz) 1989, shows that it
cannot h ope for an opening advantage is possible for Black to win if White
after such concessions. The best re- overextends (see Game 27).
44
1 d4 t'Llc6 2 d5
5 t'Llc3
This flexible move makes it h ard for
Black to activate his f8-bishop, since
s .. ..ic5?? loses to 6 'ifh 5 + and 5 ... i..b 4?
.
45
Th e Dark Kn ight System
..ib4 10 l:te1 lt:\g4! 1 1 h3 lt:\ge 5 12 lt:\xe5 forward 9 ... ..ib4 can also lead to com
lt:\xe5 13 'iih 5+ lt:\f7 14 ..if4 i.. x c3 1 5 plication s - e.g. 10 o-o lt:\f6 11 l:i.d1 fiie 7
bxc3 g 6 16 'ii'g 4 'ir'e7 White h as a small 12 e5 ..ixc3 13 bxc3 lt:\e4 14 .l:!.d4 lt:\xc3
advantage, though he has long-term 15 'ifd3 lt:\d5 - but here Bl ack is fine.
concerns about his pawn structure.
Black still has a useful choice as to 8: 3 f4 lLlg6
where he should castle. 17 ... 'it'cs is
usually a good move.
4 e4
Other moves:
7 ... ..ib7 8 iie2 a6 a) 4 h4 is an interesting attempt to
This useful little move prevents lt:\b5 take advantage of Black's inflexible
and ..ia6, while preparing ... b6-b5-b4. knight. In the expert section of the
9 ..ib3 i.. b4 1985 New York Open, I fell for White's
trap and played 4... e5? 5 h S ! lt:\xf4 ! ? 6
e3, though I went on to win an ugly
miniature with 6 ... 'it'g 5 ! 7 iif3 lt:\xds 8
'ii'x d5 iig 3+ 9 'iii>d 1 d6 10 ..ibS+ 'it>d8 11
'ii'xf7 lt:\f6 12 lt:\f3 ?? \i'xg 2 13 :f1 ..ig4
14 ..ie2 'ii'xf1+! 15 i.. xf1 ..ixf3+ 16 ..ie2
..tds (trapping White's queen) o-1. Hi
larious!
Instead, 4 ... e6! 5 h 5 lt:\6e7 6 c4! ? (6
dxe6 fxe6 7 e4 d5 transposes to 6 h4 d5
7 h5 lt:\6e7 in the notes to Position
Here 9 ... lt:\f6 ? ! was played in C.Crouch Three below) 6 ... lt:\f6 7 lt:\c3 lt:\f5 8 'ii'd 3
A.Karpatchev, Cappelle la Grande, (if 8 dxe6 fxe6 9 g4, then 9 ... lt:\xg4! ? 10
1993, starting complications which e4 lt:\fe3 or 10 ... lt:\fh 6 is possible, but
objectively favour White. The straight- relatively simplest is 9 ...lt:\g 3 10 l:th 3
46
1 d4 tbc6 2 d5
tbxf1 11 g 5 tbxh 5 12 .l:txh 5 tbg 3 13 .:i.h 3 c) 4 lDf3 is legal and was in fact the
tbf5 14 e4 tbd6 1 5 tbf3 tbxc4 16 tbh4 move order for Onischuk-Shkuro men
l:i.g 8 17 'ifh 5+ g 6 18 'i!Vxh 7 .U.g 7 19 'ifh6 tioned below (see note 'e' to Position
l:i.g 8 with a draw) 8 ... i.c5 ! 9 e4 tbg4 is Three). 4 ... e6 5 dxe6 fxe6 6 e4 trans
fine for Black, according to Mr. H, poses to the main line, while 5 c4 i.c5
though there are some crazy variations will not bring White any happiness.
to con sider: 4 es
...
47
The Dark Kn ight System
point out to me last year. 'i¥xf4 "ii'f6 8 "ir'xf6 tt:Jxf6 leaves Black
with a superior pawn structure and
White with an acute shortage of dark
squared bishops) 6 ... i.cs 7 i.xf4 tt:Jxf4 8
'ifxf4 �f6 9 'iifg 3 d6 10 i.bs+ �f8 1 1
lt:Jf3 'ili'g6 and Black can look forward to
a long and pleasant endgame.
d) s lt:Je2 (or s lt:Jh 3) s ... exf4 6 tt:Jxf4
�d6! already puts the enemy kingside
under pressure: 7 tt:Jxg6?! hxg6 8 'iff3
�4+ 9 �d1 lt:Jf6 10 i.d3 i.es is obvi
ously not satisfactory for White, but 7
a1) 7 fxg6 ? tt:Jxe4+ 8 �e2 'i¥f2+ 9 'iif3 tt:Jf6 8 lt:Jc3 0-0 9 i.d3 leaves Black
�d3 tt:Jcs+ 10 �c3 lt:Ja4+ 11 �b3 'iib 6+ with several good ideas, the simplest
12 �c4 'ii'a 6+! 13 �b3 tt:Jcs+ 14 'iiic 3 being 9 ..�es (9 ... c6! ? 9 ...tt:Jes ! ?) 10 o-o
.
lt:Je4+ 1 S �b3 'i¥b6+ 16 �c4 'i¥h4+ 17 d6 11 h3 cs, when Black's activity and
�d3 lt:Jf2+ wins White's queen. strong es-point give him the advantage.
a2) 7 lt:Jc3 �b4 8 fxg 6 lt:Jxe4+ 9 �e2 s fxe6
...
'i¥f2+ 10 �d3 fs (10 ... �xc3 11 'iitx e4! The endgame after s ... dxe6 6 'i¥xd8+
�as 12 c3 hxg6 13 �d3 d6 14 �c4 i.fs 'iitx d8 is playable in theory, but with
1 S 'i¥e2 'ii'h 6 may be better, but it's far Black's slightly misplaced g6-knight, it
more complicated, as Black will still be is less appealing than the similar end
down material for some time) 11 tt:Jxe4 ing in line A (without 3 f4 tt:Jg 6). In
�d4+ 12 �e2 'ili'xe4+ 13 i.e3 f4 14 'i¥d3 practice, Black's results are quite poor.
'ili'xe3+ also leaves Black much better. Instead, S .. .fxe6 brings us to:
a3) 7 'i¥f3 lt:Jxe4+ 8 �e2 lt:Jf4+ 9
i.xf4 'i¥xf4 10 Vxf4 exf4 11 �f3 tt:Jf6 12 Position Three
lt:Jc3 �b4 13 tt:Jge2 (13 .l:te1+ 'iiif8 14
i.c4 b S ! 1 S i.xbs �b7) 13 ... d6 14 �xf4
h S ! is relatively best, but still very good
for Black.
b) S lt:Jf3 ? is apparently tempting (it
h as been played several times), but af
ter S ... exf4 6 lt:Jc3 �CS 7 i.d3 'ife7 ! 8
'i¥e2 d6 9 lt:Ja4 (9 �d2? ! a6 ! is even
worse) 9 ... �b6 10 �d2 tt:Jf6 White has
far too little for the pawn.
c) S 'ii'f3 exf4 6 lt:Jc3 (6 i.xf4 tt:Jxf4 7
48
1 d4 li:Jc6 2 ds
Not to put too fine a point on it, but �xe4, even if this wins a pawn .
the players who have reached this posi Let's get to the analysis.
tion with the black pieces need collec 6 li:Jf3
tively to have their heads examined. Other moves:
White's main asset is his powerful a) 6 �d3 is less popular and less
pawn duo on e4 and f4. Black's main successful. Then Black has done fine
asset is his central pawn majority. One with 6 ...�c5, but 6 ... d5! is more accu
need notice only one of these two rate, as in B.Wood-J. Penrose, Southend
things in order to come up with the 1957(!),
correct plan (or at least the correct
sixth move) for Black. In fact, I'm not
even going to insult my readers by say
ing it out loud, so if you still don't
know, see Wood-Penrose below and
then read Pawn Power in Chess by Hans
Kmoch.
Instead of taking the opportunity to
strike in the centre, Black h as generally
been seduced by 6 ... �c5, occupying the
a7-g 1 diagonal, presumably to stop
White from castling. N aturally this is which continued 7 e5 li:Jh 6 8 .i.e3 ? ! ,
less important than the central battle and n o w not the game's 8 . . .li:Jf5 ? ! 9
and, what's worse, it doesn't even �xf5 exf5 10 li:Jf3 .i.e7? ! (1o ... c 5 ! ) 11 c4
work. If White finds the bishop trouble which is good for White, but instead
some, he can trade it off with a timely 8 ... li:Jh4! 9 'ii'e 2 c 5 ! and Black is better.
tt:Ja4 or 'ii'e 2 and �e3 . To add insult to No better is 7 li:Jc3 �b4 8 �d2 li:Jh6
injury, White is usually better off cas 9 li:Jf3 o-o 10 g3, when Black gently
tled long anyway. Sometimes 7 ... �c5 is plays 10 ... e5 ! ! and lets White try to
a good idea (or 7 ...�b4+ 8 c3 �c5), but work out the details with his king in
if there is a knight on c3, it is much bet the centre. The correct solution is 11
ter to put pressure on White's centre li:Jxd5 �xd2+ 12 'ii'x d2 c6 13 li:Je3 exf4
with 7 ... �b4, which prepares ...li:Jf6-e4. 14 �c4+ 'it>h8 1 5 'ii'x d8 :xd8 16 gxf4
Black's e4-knight can be a very an li:Jxf4 17 :g1 b5 with equality. Accord
noying piece. As we see in the following ing to Houdini, 11 f5 dxe4 12 li:Jxe4
analysis, White's light squares are usu .i.xd2 13 'ii'x d2 �xf5 14 o-o-o 'ilie7 is
ally too weak (because he h as h ad to also equal, but to me it looks a lot like
play g 2-g3) to allow him to eliminate White is down a pawn for nothing. For
the knight comfortably with .i.d3 and tunately, it's not really our problem.
49
Th e Dark Kn ight System
50
1 d4 tLlc6 2 dS
51
The Dark Kn ight System
h appy about. There is also 8 h 4 ! ? ..tcs 9 the knight to h 6 avoids a traffic jam on
h S (or 9 ..td3 l2Jg4 when, according to the e7-square, which may be needed
Houdini, White h as nothing better than for Black's other knight, not to mention
10 l2Jd4 l2Jh 6 11 l2Jf3, with a repetition) the bishop and queen . (You may now
9 ... lbe7 10 lLlg s lLlhfs 11 'ii'd 3 h 6 12 g4 resume your normal programming.)
l2Jd4 13 c3 l2Jdc6 14 l2Jf3 as ! with equal c) The immediate 7 h4! ? is also pos
chances. Black intends ...b7-b6, ... .i.a6 sible, but Black is already fine after
( ... ..tb7), ... �d7 (or ... dS-d4, .. .'�ds), and 7 ... .i.b4+! 8 c3 (not 8 .i.d2 ? dxe4! )
will have the choice of which side to 8 . . ..i. c s 9 h s l2J6e7; e . g . 10 b 4 .i.b6 11
castle. White may enjoy his space, but c4! ? dxc4 12 'it'xd8+ 'it>xd8 13 .i.b2 (not
may also find himself overextended. 13 ttJg s ? ..ltd4!) 13 ... exf3 14 ..txg 7 lLlfs
Hold on : what's so great about 15 ..txh8 f2+ 16 Wd2 l2Jg 3 17 cs l2Jxh 1
7 ... l2Jh 6 - ? 18 cxb6 axb6 19 lLlc3 lbe7, which is cer
tainly no worse for Black.
d) 7 .i.d3 allows us to surprise White
with 7 ... dxe4! 8 ..txe4 'i!t'xd1+ 9 'it>xd1
l2Jf6 10 ..txg6 hxg6.
52
1 d4 tL\c6 2 dS
11 tLic3 i.. d 7 12 tLie5 (12 �e2 i..d 6! 13 bishop. Unsurprisingly Black has many
tLie5 i.. xe5 14 fxe5 tLig4 nets a pawn) ways to play: 8 ... ..ig4 9 o-o i.c5+ 10 �h l
12 ... 0-0-0! ! 13 tLif7 i..c 6+ 14 ttJxd8 �xd8 tiJ8e7 11 h 3 i.xf3 12 'i!Vxf3 o-o seems
15 h4 i..x g 2 16 l:i.h 2 i..f3+ 17 �el i.. c 5 simplest, or Black can enter an equal
and how is White going to untangle endgame with 8 .....ib4+ and 9 .. .'iVe7+.
himself without shedding any mate 7 ..ib4
...
rial ?
e) 7 c4! ? does force 7 ... dxe4, but
weaken s White's position as well : 8
"O'xd8 �xd8 9 ttJg 5 'ilo>e8 10 ttJxe4 tLif6
11 i.. d 3 b6 12 ttJbc3 ..ib7 13 o-o l:td8
and Black is comfortably equal.
f) 7 g 3
8 g3
White usually finds it necessary to
play this sooner or later. Otherwise:
a) 8 i.d3 i.. x c3 9 bxc3 dxe4 10 i.. xe4
li'xdl+ 11 �xdl tiJf6 12 i.. d 3 o-o 13
.l:tel tL'lg4 14 i.. x g6 hxg6 15 h 3 tiJf6 16
tLie5 tLih 5 17 ttJxg6 l:!.f6 18 tLie7+ �f7 19
7 . . ...ic5 ( 7. . ...ib4+ 8 c3 ..i c 5 9 'ii'e 2 ttJxc8 ttJxf4! with equal chances.
·iie7 10 tiJbd2 tLih6 1 1 tLig 5 o-o 12 h4 b) 8 e5 tiJ8e7 (8 ... c5 isn't bad either)
.:e8 is okay, too) 8 li'e2 (8 tiJc3 tiJf6 ! 9 9 g3 0-0 10 ..td3 c5 and Black's good
e5 tLig4 10 tLid4 0-0 11 'ii'x g4 ..ixd4 12 centre and rapid deployment ensure
�d2 c5! 13 o-o-o li'b6 with equal that he will not be rolled up on the
chances) 8 ...tiJf6 9 e5 tLie4 10 ..ie3 ..ixe3 kingside and that his chances are not
11 'ifxe3 tLie7 ! 12 i..d 3 i.. d 7 13 ttJbd2 worse.
�xd2 14 ttJxd2 (14 'ii'x d2 c5) 14 ... 0-0 1 5 c) 8 'ii'd 3 ! ? i.. xc3+! 9 'ii'x c3 tiJf6 10 e5
o-o tLif5 and with 16 . . .'ii'e 7 (or 16 . . .b6) tLie4 1 1 �a3 'ii'e 7 12 �xe7 ttJxe7 13
and 17 ... c5 coming, Black h as sufficient i..d 3 b6 14 i..x e4 dxe4 1 5 tLig 5 h6 16
counterplay. ttJxe4 i.b7 17 tL'lc3 tLif5 18 o-o o-o-o and
g) 7 exd5?! exd5 8 i..d 3 makes no Black h as the d-file, while White's cl
sense - it surrenders the centre and ac bishop is a huge problem, a situation
tivates Black's problem piece, the c8- which fully compensates for the pawn.
53
The Dark Kn ight System
54
Section Two
1 e4 tt:Jc6
Other moves:
a) 2 ..ic4 commits the bishop way
too early: 2 ... tLlf6 3 lLlc3 e6! 4 d3 ..ib4 5
.id2 d5 6 exd5 exd5 7 ..ib5 o-o 8 tLlge2
One might argue that it m akes tLle 5 ! 9 a3 ..id6 10 ..ig 5 c6! was P.Cruz
more sense to play the Dark Knight A.Kogan, Lisbon 2000. Black has a big
ag ain st 1 e4 than 1 d4, because the advantage, largely due to h aving shut
dangerous m ove 2 dS is n o l on g er White's light-squared bishop out of
available. On the other h an d, White, as play.
a 1 e4 pl ayer, is more likely to be fa b) 2 f4 is not as bad as it looks.
miliar with the possible tran spositions 2 ... d5 ! 3 exd5 'ii'x d5 4 tLlc3 leads to a
to the Scotch and the Pirc. Personally, I bizarre kind of Scandin avian. Fortu
find the second argument to be m ore nately, 4 .. .'ii' e 6+! is more awkward for
compelling th an the first (for a decade White than it is for Black; e.g. 5 ..ie2
I played 1 d4 Nc6 with no plan for de tLld4! or 5 'ife2 tLlb4! or 5 tLlge2 tLld4! 6
fendin g the Scotch), but th e reader d3 tLlf6 7 l2Je4 tLlf5 8 c3 g 6 (8 .. tLlxe4 9
.
can m ake up his own mind, or better 'ii'a4+) with some advantage to Black in
yet just pl ay 1...l2Jc6 ag ain st every all cases, according to Houdini.
thing ! 3 e5 is a good Nimzowitsch Defence
55
Th e Dark Kn ig h t System
for Black since White's pawn does not with a small advantage according to
belong on f4 {at least until .. .f7-f6 is New In Chess. Actually, I don't believe
played). In R.Barkman-S.Lejlic, Karls Black is better, but the position is in
krona 1997, Black used this to good teresting and Black has the makings of
effect: 3 ... tt:lh6 4 d4 .ltg4 5 .lte2 .ltxe2 6 a kingside attack. Wisnewski's recom
tt:lxe2 tt:lf5 7 c3 e6 8 'ii'd 3 h 5 {8 .. .'ifh4+!) mendation is 3 ... d4! ? - one idea is to
9 tt:ld2? ! �4+ with advantage due to clear the d5-square for Bla�k's queen,
White's light-square problem s - White but I am n ot keen on relinquishing con
decided to pitch a pawn with 10 tt:lg 3 trol over e4.
'ii'xf4 11 tt:lxf5 'ii'xf5 12 'ii'xf5 exf5, but For those who are interested, 2 ... e5
he had no compensation . 3 tt:lf3 f5 ! ? is the Adelaide Counter
Alternatively, Black can throw a clog Gambit, a wild and fascinating defence
in White's machinery with 3 ... g 5 ! ? 4 d4 to the King's Gambit, but truly outside
gxf4 5 .ltxf4 .ltf5 6 c3 'i!Vd7 7 tt:lf3 h5 8 the scope of this book. Besides, there is
ie2 0-0-0 9 tt:lh4 ig4 10 tt:ld2 ih6 as also 3 tt:lc3, with wildness that White is
in H.Gohlke-S.Wiezer, Gorlitz 1985, presumably prepared for.
56
C h a pte r Fo u r
1 e 4 'Dc6 2 d4 es
A: 3 d s ll:Jce7
57
Th e Dark Kn ig h t System
58
1 e4 tbc6 2 d4 e5
59
Th e Dark Kn ight System
d5 is great for Black) 13 ... i.b7 14 tt:'lc3 9 ... i.xc3 10 bxc3 cxd5 11 i.xh 5 (11
i.e7, followed by ... tt:'lf8-e6 - Black's exd5 ! ?) 11...tt:'Jf4 12 i.f3 tt:'le6 13 i.xf6
extra pawn may not be much of an as 'ifxf6 14 exd5 tt:'lc5 15 g 3 (15 h 5 ! ?)
set, but he is not suffering for h aving it. 1 5 ... d6 with only a tiny disadvantage,
However, the correct respon se to but White missed some good chances
7 ... ..tc5, namely 8 tt:'ld2 !, puts Black un along the way.
der some pressure. After the text, we're in uncharted
8 tt:'ld2 ! territory again, which is just how we
Instead: like it. Right? ! Right.
a) 8 a3 i.xc3+ 9 bxc3 c6 10 c4 d6 11
tt:'ld2 'i¥a5 was L.Christiansen-J . Ben
jamin, U S Championship 2000 (see
Game 3 3). Black h ad nothing to com
plain about and went on to win .
b ) 8 i.e2 allows u s t o preserve the
bishop with 8 ... d6 9 o-o i.d7 10 a3 i.c5
11 'ii'd 3 a6 12 g3 i.h 3 13 .:fd1 and
break the annoying pin by 13 ...'ii'c 8 (or
13 ...'ii'b 8! ?), when we're just about
equal. (This queen manoeuvre should
be kept in mind throughout this sec a) 9 i.b5+ is certainly n ot a big deal
tion .) 14 ... tt:'lg4 is likely, targetin g the f2- after 9 ... i.d7 10 i.xd7+ 'i¥xd7. Sure, we
square. liked that bishop, but White has solved
The text move is more challenging. our space problem and helped us break
the annoying pin on our f6-knight.
Then 11 i.xf6 gxf6 12 'iif3 'i¥g4
(12 ... i.xc3 13 'iVxc3 f5 14 g3 o-o-o is
okay, too) 13 'ifxg4 (or 13 'ifxf6 i.xc3 14
bxc3 tt:'lf4! 15 tt:'lf3 'i¥xg 2 16 tt:'lg 5 .l:.f8 17
o-o-o 'ili'xf2 18 tt:'lh7 tt:'le2+ 19 'ifi>d2 'ii'xf6
20 tt:'lxf6+ 'ifi>e7 2 1 tt:'lxh 5 �h 8 22 'iti>xe2
.l:i.xh 5 23 l'!dg 1 with a draw, presuma
bly, though only Black can pretend to
play for a win) 13 ... hxg4 14 g3 i.xc3 1 5
bxc3 tt:'le7 16 c 4 f5 is level.
8 d6!
... b) 9 i.e2 tt:'lf4! (as in many similar
In E.Morten sen-C. Hoi, Ostrava 1992, positions, White does not want to take
Black played 8 ... c6 9 i.e2 (9 g 3 ! ?) this knight, especially since his bishop
60
1 e4 lbc6 2 d4 es
�8 (this time 10 ...1i.d7 doesn't work Not 6 lbbd2 c6! 7 c4? ! (this is n o
well - White will gain the very useful good, but the alternative is t o give up
g2-g3 with tempo) 11 0-0 ii.g4 12 f3 (or the centre) 7 ... lbg4! 8 Ji.g 5 ? �6 9
12 �e1 Ji.d7 ! - the situation has �e2 ? ! ii.c5 and White needs C PR.
ch anged again already; the bishop ex 6 .lbg41 7 Ji.d2
..
change now brings equality: 13 ii.xd7 Or 7 ii.g 5 1i.e7 8 1i.d2 1i.c5, tran spos
·o-xd7 and Black will follow with ing.
14 .. .'�g4 if possible, otherwise 14 ... c6, 1 ...ii.cs s o-o as 9 h 3 lDf6
or similarly 13 Ji.d3 ! ? c6) 1 2 ... c6! 13
..i.a4 ii.xc3 14 bxc3 ..td7 15 dxc6 ii.xc6
16 Ji.xc6 �6+ 17 'it?h 1 �xc6 18 Ji.xf4
exf4 19 lbb3 d5 20 e5 lbd7 2 1 l:le1 .:e8
22 "ikd4 .:i.h 6 is roughly equal.
There are many other ways the
game could go, and White does h ave
chances for an advantage, but the pre
ceding lines give a good indication of
Black's resources.
61
The Dark Kn ight System
62
1 e4 ltJc6 2 d4 es
63
The Dark Kn ig h t System
i.f4 �ae8 (or 16 ... g s ! ? 17 i.xes dxes 18 b) 6 f3 i.b4+ (6 ... b6 is still interest
exfs .l:!.a6 with more than enough for ing, but after 7 tt::l c 3 i.cs 8 i.xcs bxcs 9
the pawn) 17 b4 (not 17 i.xe s ? .l:!.xes 18 d6 ! White has balanced the chances) 7
i.f3 f4 19 i.g 2 'ii'h s 20 b4 i.h 3 2 1 bxcs tt::l d 2 'ii'e 7 (7 ... b 6 ! ?). Miles scored 4-0
i.xg 2, winning) 17 ... axb4 18 axb4 from this position, beating GM Kaida
i.xb4 with clearly the better game for nov and IMs Shirazi and Langeweg; e.g.
Black. 8 g 3 o-o 9 ..th 3 c6 (or 9 ... as or 9 ... i.cs)
and Black is better, going on to win in
A22: 5 ..te3 tt::lf6 G.Kaidanov-A.Miles, Palma de Mallorca
1989 (see Game 37).
c) 6 i.g S ? ! is ridiculous here and in
all positions where Black can play ...h7-
h6. The bishop is way too valuable to
trade, and even when it's possible to
retreat, ... h 7-h 6 is always a useful move
for Black. So 6 ...h 6 ! 7 i.xf6 'fllxf6 and
White h as even more dark-square
problems than usual. Black's queen is
also very happy on the f6-square.
6 .tb4 7 f3
...
g 3 tt::l g 6 9 Wg 2 i.cs is similar, but White Black's results are excellent (+21 -12
is certainly not suffering from his in =8) with this move, so there is little
ability to castle) 7 tt::l c 3 i.. c s (in V.Vilkov need to worry about alternatives.
A.Provotorov, Kalug a 1996, the only Sometimes Black plays 7 ...'ile7 with the
time this position has been reached, idea of trading the dark-squared bish
Black played the highly inconsistent ops, but this is time-consuming com
7 ... i.b4?! and went on to lose) 8 tt::lf3 pared to the text, and hasn't worked
0-0 9 0-0 and there are many roads for well in real life.
both players, but they all lead to equal Quoting Joel Benjamin: "This is im
positions - the imminent and posi portant: do not hang your bishop with
tionally favourable bishop trade bal 7 ... d6 8 'ili'a4+." (Actually, according to
ances White's space advantage; e.g. my database, 7 ... d6? ? h as been played
9 ... as 10 'ii'd 2 'flle 7 1 1 .l:!.fe1 d6 12 .txcs three times and White h as yet to play 8
bxcs 13 g3 a4 and the game is still bal 'ili'a4+, even in the g am e where White
anced. was a GM. I won't name the GM. Can
64
1 e4 tt:lc6 2 d4 es
this possibly be right? ! Can you explain on the queen side, a strategy which is
yourself, Pablo?) far less effective, as we will see.
8 bxc3 d6 9 'if'd2!? Can things really be that simple for
In fact, just about anything is better Black? Not quite, but if White wants to
than the lemon 9 c S ? ! that White get any play, he pretty much needs to
squeezes out most of the time, which go crazy like a Grandmaster (in I.Jelen
leads to: E.Dizdarevic, Ljubljana 1992 - see
Game 40 - White went crazy like an I M,
Position Four which did not work), as in E. Rozentalis
M.Lazic, Genoa 2004, which went 9 cs
o-o 10 cxd6 cxd6 11 i.d3
65
Th e Dark Kn ight System
break with .. .f7-fS, White's strategy dangerous for Black, who must b e ex
would h ave failed. N otice, too, that tremely resourceful in a dizzying array
Black was never losing - in fact, he was of variations in order to avoid disaster.
never significantly worse - until (See for yourself - I won't bore you with
28 ... 1:!.bc7??. Instead 28 . .J::!.f7 ! (preparing the details. Frankly, I wouldn't even
the escape 29 .. .'�e7 if necessary) would know where to start.) After the text
have pre-empted White's sacrificial move, Black can bail out to the queen
attack and held the balance. White's side if things get too hairy too quickly -
ideal position wasn't so dangerous af a surprising but valuable option !
ter all ! 10 h4
Let's go back to move 11, set up the Not forced, obviously, but if this is
proper defence without fear, and try to what White is about, he is likely to get
play a little more quickly and actively right to it.
on the queen side: 11 ... 'iia s ! (or If 10 ..id3 (or 10 lt:\e2, for that mat
11 ... lt:ld7 first) 12 lt:\e2 lt:ld7 13 h4 lt:le7 ter), Black could castle, h aving lost
14 hs h6 1s g4 f6 16 'i!Vd2 lt:lcs 17 ..ic2 nothing - but I would prefer to stay
b6 18 lt:lg 3 (18 c4! 'ii'a6 is equal) flexible with 10 ... lt:ld7; for example, 10
18 ... ..ia6 19 lt:lfs lt:lxfs 20 gxfs .l:.f7 2 1 lt:\e2 lLld7 11 h4 h6 12 h S lt:lgf8 ! (in
llg 1 'ilo>f8, followed by 22 . . ..:!. c 8 with Z. Koczka-Zsu.Simon, Hungarian Team
pressure on the c3-pawn. Champion ship 2003, 12 ... lt:le7? was 1-o
We now return to our regular pro in 30 moves) and 13 ...lt:lh 7 will be like
gramming. the main variation. If 10 ..id3 lt:ld7 11
lt:\e2, Black can still delay ... o-o in favour
of 1 1 ... lt:lcs ! . Can White keep waiting
before committing to either castling or
h 2-h4 - ?
10... h6
This stops White's pawn from going
to h6, which would leave Black with a
big cramp and weaknesses on the dark
squares.
11 g3
Someone will try 11 hS lt:lf8 ! (I just
9 ... b6 love this move - the knight heads for
Palliser concludes that there is no the open cs-square, whereas a knight
point delaying castling, but I am find on e7 would be going nowhere for a
ing that 9 ... 0-0 10 h4! (intending long time) 12 g4 lt:\6h 7 ! (preventing 13
10 ... lt:lh s 11 ..if2 or 11 lt:\e2) is very g S and preparing an eventual ...lt:lg s !,
66
1 e4 lDc6 2 d4 e5
67
Th e Dark Kn ight System
since we will need to know the posi This is not good, but i t is always
tions after 6 ... i.cs anyway because of played, if it hasn't been played already.
the different move orders White can Even Houdini likes it! Black's attacking
use (see line A2 1 for this). Although ideas should be familiar by now:
Black's bishop is very strong on the a7- 1o tt:Jhs
...
68
1 e4 tt:lc6 2 d4 e5
8 a6
...
69
Th e Dark Kn ight System
sessed by De Firmian as better for White ity), and ...c7-c6 may be on the cards.
- a case of annotating by result. The One interesting idea is ...ltJf4. One
game continued preposterously with thing we absolutely do not need to
9 ...'ii'f6?! 10 .l:th2 d6? ! (10 .. .'i!Vd8! is equal) worry about is �xf4 exf4 - White's
11 ltJa4 �b4+?? 12 ltJec3 ! and Black's good bishop is too valuable to squan
h S-knight is toast. der in such a fashion, and our pawn on
Going back to move nine, it is true f4 is strong, not weak. We will very
that White has more space, but his po much enjoy the use of the es-square
sition is very loose. Simply 9 ... a6 ! fol for our knight, bishop, rook, or queen.
lowed by ... d7-d6 and ... c7-c6, and what Even if ... g7-g 5 is necessary to defend
moves does White imagine that he is f4, Black's kingside will remain safe.
going to be playing ? For example, 9 ... a6 Another idea is ... �d4 and ... c7-c5,
10 o-o d6 11 Wh 2 liJf6 12 I;lb1 c6 13 b4 anchoring the bishop. In this closed
'ifd7 14 ltJg 1 �d4 15 ltJce2 il.. a 7 and position, we will not mind if our bishop
White's position is slowly degrading. gets traded off, especially since we will
9 0-0 d6 10 ltJg31 wind up with a protected passed pawn
on the d4-square (if the bishop is taken,
either recapture can be considered,
though ... esxd4 is the typical answer).
Naturally, White will not be eager to
play dsxc6 (en passant) because of the
loss of space and centre. 10 ...�d4 is not
the most accurate, but 10 ... h6 11 ltJa4
can definitely be met by 11 ... �d4 and
12 ... cs.
A25 : 5 g3
otherwise, the nasty 10...ltJh S ! is
coming; e.g. 10 .l:!.b1 liJh S ! 11 b4 �a7 12
Wh 1 ltJhf4 and .. .'ii'h 4 or ...ltJh4 with
threats again st g2, h3, and f2.
(9 ... ltJh S ? ! was no good because of 10
ltJa4! �a7 11 d6 ! with a small edge and
a large disruption of Black's plans.)
The position after 10 ltJg3 has never
been reached, but we have m any rea
sonable and sensible moves such as
10 ...�d7, 10 ....l:!.e8, or 10 ...h6 (with equal-
70
1 e4 tLlc6 2 d4 es
is slow, and the white bishop is not go The popular 7 tLle2 allows Black to
ing to be active on g2. take the initiative immediately with
s .tbf6
.. 7 ... h S !, when 8 ... h4 cannot be stopped
because 8 h4? tLlg4 9 o-o? (or 9 l:i.f1
'ii'f6 ! ) 9 ... tLlxh4! is awful for White, as is
8 il.g s ? il.xf2+.
7 0-0
...
6 il.g2
If 6 tLlc3, we have that choice again
- to take the strong diagonal or to
double White's c-pawns. I prefer to
have a queenside target, particularly 7 ... d6, 7 ... as and 7 ... hs are also in
since White has taken measures teresting, but castling is flexible and
against our kingside play. Also, White's leads into the amusing miniature
fianchetto leaves the c4-pawn without H.Titz-C.Rossi, Austrian Team Champi
protection. So 6 ... il.b4 7 il.g 2 il.xc3 8 onship 2001: 8 tLlf3 d6 9 0-0 a6 10 'ii'd 3
bxc3 d6 9 tLle2 (or 9 h4 h 6 with a com bS 11 b3 il.d7 12 a4 b4 13 lL:ld1 'i!Vc8 14
fortably equal position very similar to il.e3 tLlxe4 15 tLlxes tLlxes 16 'ifxe4 il.fs
A.Karpov-D.Chevallier in A22 - if 10 17 'i!Vh4 tLlg6 0-1. White resigned be
h s ? ! then 10 ... tLlf8 ! with advantage) cause 18 'ii'h s .tg4 19 'i!Vg s h6 traps the
9 ... 0-0 and Black has won every g ame. queen .
White's plan to play f2-f4 is far too As seen in this g ame, ... a7-a6 is of
weakening - after ... esxf4 g 3xf4, White ten the best way to preserve the bishop
has problems on c4, e4, f4, g4, and h4. in this variation, because with White's
While Black is waiting for f2-f4, he can bishop fianchettoed, the advance ... b7-
play ... b7-b6, ... tLld7, .. .tbcs, ...f7-f5 and bS is easy to achieve and likely to be
perhaps ... a7-a5 and ... il.a6. E.Schien effective.
dorfer-D.Recuero Guerra, Herceg Novi
2006, is a typical disaster for White (see A26: 5 a3 tLlf6 6 tLlc3 il.cs
71
Th e Dark Kn ight System
72
1 e4 ltJc6 2 d4 es
White has:
The usual move is 5 ..."fif6, but then
81: 4 lL'lf3 73 White can either allow or avoid the
8 2 : 4 f4 76 queen trade as he pleases, with a
pleasant position in either case. Al
Instead, 4 lL'lc3 .tc5 5 lL'lf3 ? (5 f4 l2Jc6 though the text move has had no seri
transposes to line B22 below) 5 ... l2Jg4! 6 ous trials, analysis demonstrates its
l2Jd4 is like a Two Knights Defence (1 e4 viability and none of the ideas previ
e5 2 lL'lf3 l2Jc6 3 �c4 lL'lf6 4 lL'lg 5 d5 5 ously tried h ave proven themselves
exd5 l2Jxd5) with reversed colours, ex reliable. In most cases, we intend to
cept that Black's king's knight is at play like a Philidor, counting on the
home instead of his queen's knight. This knight trade to ease our space disad
makes it harder to pressure the l2Jd4, so vantage, while trying to show that
the standard idea of 6 ...l2Jxf2 is less ef White's queen is awkward on f3; and
fective; e.g. 7 'iitxf2 'ii'f6+ 8 'iiie 3 lL'le7 9 there are other possibilities depending
lL'lb5 'iie 5 10 c3 f5 11 'iiid 2 fxe4 12 'iitc 2 on how White reacts. lt is not possible
o-o 13 g3 d5 14 �f4 does not give Black to analyse exhaustively, but the follow-
73
Th e Dark Kn ight System
811: 6 lLlc3 74
812: 6 ..tc4 74
8 1 3 : 6 e s 75
814: 6 ..id 3 75
8 1 5 : 6 ..te2 75
816: 6 ..tgs 76
74
1 e4 lZJc6 2 d4 es
75
Th e Dark Kn ig h t System
82: 4 f4
This is far less popular than 4 4Jf3
among the strongest players, but gives
excellent results. Fortunately, it's easy
to see where Black has been going
wrong.
4... tt:Jc6
White has:
821: 5 .tc4 77
822: 5 4Jf3 78
Alternatively:
a) 5 tt:Jc3 .tcs 6 4Jf3 transposes to
line B22.
b) 5 .ie3 prevents s ... .tcs, but Black
In principle, we can already notice is okay after s ... .tb4+ 6 c3 .tas
that, compared to a regular Philidor, (R.Hilbner-V.Hort, German League
White's .tg s and 'ti'f3 are awkwardly 1984, see Game 47). Even better is the
placed, so White cannot expect much unplayed novelty s ... d S ! which brings
from the opening; e.g. 8 h 3 0-0 9 0-o-o equality; e.g. 6 exds tt:Jb4 7 .tbs+ .id7 8
c6 10 'ili'g 3 (10 es tt:Je8 doesn't do any .ixd7+ 'i!i'xd7 9 4Jf3 tt:Jxds 10 'i!i'e2 o-o-o
thing, while after 10 .te2 �as 11 .td2 - come to think of it, White is much
'ili'c7 12 �g 3 bS 13 .th 6 tt:Je8 Black is worse here, so 8 1We2 tt:Jxc2+ 9 'ii'x c2
certainly no worse) 10 ... �as 11 f4 .ie6 .ixbs 10 4Jc3 .ta6 11 4Jf3 .1i.d6 12 o-o-o
leads to a wild race with approximately tt:Jf6 13 llhel o-o, when Black's bishop
even chances. pair balances White's space advantage.
76
1 e 4 lb c 6 2 d4 e s
77
Th e Dark Kn ig h t System
6 ... d6
In stead:
a) 6 ... a6? ! 7 'iNe2 ! d6 8 .i.e3 .i.xe3 9
'iHxe3 ll:\f6 10 o-o-o is pleasant for
White.
b) 6 ... ll:\f6 ! ? leads to wild complica
tions and is fully playable if you enjoy
such positions. Here are some sample
variations: 7 e s ! ll:\g4 8 .i.c4 (or 8 ll:\e4!
.i.b6 9 i.c4 ds 10 'i!Vxds 'ii'e 7
{10 ...'ii'x dS ! ?} 1 1 h 3 i.e6 12 �5 a6 13
14...d4?! 'iWa4 .i.xc4 14 'ifxc4 ll:\e3 15 i.xe3 i.xe3
14 ... .l::[e 8! allows less counterplay: 15 16 g3 0-0-0, when Black has compensa
.i.e3 d4! 16 ll:\xd4 .l:txe3 17 'ji'xe3 ll:\xd4 tion for most of a pawn) 8 ... d6 9 ll:\g s
18 'ili'd3 lLlf3+ 19 .l:.xf3 .i.xf3 . (or 9 lLle4 i.e3 10 exd6 0-0 {10 ... i.xc 1 ! ?}
15 fs ! dxc3 16 ll:\f4! 'jjff7 17 ll:\e6+ 'iitg8 11 i.xe3 ll:\xe3 12 'ife2 ll:\xc4 13 'i!Vxc4
18 bxc3 .i.d6 19 'ji'e4 ll:\es 20 'i!i'xb7 lieS cxd6 14 o-o-o i.e6 15 'iih s dS 16 fS a6
21 ll:\d4 cs 22 'iVa6 cxd4 23 'ii'x d6 ll:\f3+ 17 'iic s i.xfs 18 l:.xds 'iie 7 19 ll:\d6 i.e6
24 'iitg 2 'ii'b 7 20 �d2 i.xa2 ! 21 b3 'ii'f6 22 ll:\d4 as)
White has had enough. Did I men 9 ...0-0 10 h3 ll:\h 6 11 ll:\a4 i.b6 12 ll:\xb6
tion that Godena is my new h ero? axb6 13 o-o dxes 14 'i!Vxd8 l:txd8 1 5
fxes ll:\xes 16 i.b3 cs 17 .i.f4 c 4 with
B22: s lLlf3 equality.
Rare at a high level, this lazy-looking
move is White's best, apparently doing
nothing to contest the a7-g 1 diagonal.
s ... .i.cs 6 lLlc3
78
1 e4 'Llc6 2 d4 es
merit the exclamation mark bestowed 16 .ll c 3 'ilfc5+ 17 �h 1 'ii'c4 with equal
by Kalinin - the simple text is best. ity.
10 o-o o-o 11 .:!.e1 White has also tried 11 b3 ! ? .ll d7 12
.llb 2 'Llb4! 13 �d2 'Llxd3 14 cxd3 c5 15
f5 Si.c6 16 �g 5 h 6 17 Si.xf6 'ii'xf6 18
�xf6 gxf6, when Black had equalized in
J.R.Capablanca(!)-M.H.McGuire, New
Orlean s (simul) 1911. Black went on to
win, outmanoeuvring Capablanca with
his better minor and queenside pawn
majority. Outrageous !
C: 3 'Llf3 exd4
79
Th e Dark Kn ight System
really be okay, but has had serious ... �b8, ....l::!. c 8, ... c6-cs is time consum
problems in practice. One good idea for ing).
Black is ... .tas-b6 to pressure White's
d4-pawn and shore up the queenside.
80
1 e4 tt:lc6 2 d4 es
(111: 7 tt:ld2 82
c112: 1 tt:Jbs 84
Others:
a) The unlikely-looking 7 'i!Ve2 has
also been popular. White hopes for
7 ...ilixe4?? 8 tt:Jxc6 i..x e3 9 tt:ld4, win
6 C3 ning (this has yet to work, but hope
White can try 6 tt:lbs here, but after springs eternal). Instead, 7 ... tt:lxd4! 8
6 ... i.. x e3 7 fxe3 'jjVh 4+ (this intermezzo cxd4 (or 8 i..xd4 i.. xd4 9 cxd4 lbe7 10
forces a concession) 8 g3 'ii'd 8 9 'i¥g4 tt:lc3 o-o 11 o-o-o c6 and with 12 ... ds
g S ! 10 tt:l1c3 tt:Jes 11 'i¥e2 d6 12 h 3 c6 coming, Black is slightly better)
13 lbd4 tt:lf6 14 o-o-o 'iie 7 15 'ii'f2 i.. e 6 8 ... i..b4+ 9 i.. d 2 i..x d2+ 10 lbxd2 tt:le7
16 i.. e 2 o-o-o Black was obviously fine 11 g 3 ! (to discourage ll...dS) 11...0-0 12
and went on to win a marathon g ame i.. g 2 �6 and White will have compen-
81
Th e Dark Kn ight System
sation for the pawn he is losing, but no deed dangerous to try to win a pawn,
more than that. but in E.Stavropoulou-M. Ikonomo
b) 7 f3 a6 ! ? (this is a novelty - it's poulou, Athen s 2003, Black found an
time to put a stop to lZ:lbs once and for excellent alternative: 7 ...lZ:lf6 ! 8 lLld2 (8
all) 8 lZ:ld2 d6 9 'i¥c2 lLlge7 10 0-0-0 il.. a7 0-0 d6 9 'it>h 1 ii.d7 10 lLld2 o-o is equal)
11 <it>b1 0-0 is equal; or 8 'ii'd 2 lZ:lge7, 8 ... ds ! 9 exds lZ:lxds 10 il..f3 lZ:lxe3 11
tran sposing to S.Vajda-S.Skembris, fxe3 lLleS ! with a clear advantage.
Naujac 1999, which continued 9 lZ:lc2
il..x e3 10 lZ:lxe3 d6 11 il.. e 2 and Black (111: 7 lLld2
found plenty of activity with 11 .. .fs, This is pathetic, but it's played most
going on to win (see Game 52). of the time, so I guess that makes it the
c) Dembo and Palliser (D&P) like 7 main line! I can't imagine why White
'fif3 pretty well, would play the Scotch if this is his plan
for dealing with 6 ... 'i¥g6. The following
draws significantly on Dembo and Pal
liser.
7 lZ:lf6
...
82
1 e4 tt:\c6 2 d4 es
lbxe4 14 fxe4 lbes and White needs ing, Leeuwarden 2001, which was
both to survive and to extricate the equal after 14 .. J!kxc6 15 o-o (D&P), but
knight, which is more hard than easy) Black can keep some pressure with
9 .. .'�xg2 10 l:!.g 1 �xh 2 11 J::t xg 7. Up un 14 ...'i¥a6 !, stubbornly denying White's
til now we've been following E. Berg castling rights. After 15 �e2 �xc6 16
I .Morovic Fernandez, European Cham 0-0 tt:\g4 17 tt:\d4 'iUd6 White will soon
pionship, Saint Vincent 2000, which have an isolated e-pawn and a some
continued 11 ... 1i.d7 12 'ifl>3 �4+ 13 what worse position.
�d1 and D& P says favours Black, but f) 8 f4 has worked very well for
Houdini calls even, and Black did go on White in practice, so be careful ! The
to lose. Instead, I am recommending careful respon se is 8 ... lbxd4 9 cxd4 ..ltb4
11 ...�f8 ! ? 12 .l:.g 1 lbxd4 13 cxd4 1i.e7 14 10 fs 'ii'g 4 11 'ii'x g4 lbxg4 12 li.f4 ds 13
'ifh3 c6 15 e s ! (after 15 1i.d3 dxe4 1 5 h3 lbf6 14 es lbe4 15 g4 g6 16 fxg 6
fxe4 'i!Vh4+ 16 1i.f2 'ii'f4 17 l:i.fl 1i.e6, fxg6, which is slightly better for Black
White's compensation is insufficient) because of his imminent f-file control,
1S ... cxbs 16 exf6 'i!Vh4+ 17 �e2 'ii'xf6 18 though it was not enough to win in
'it'xds 'li'fs 19 'ii'e 4 l!Vxe4 20 fxe4 hs 2 1 A.Motylev-S.Gligoric, Yugoslav Team
lbf3 li.g4 with approximate equality i n Champion ship 2000 (see Game 5 3).
an unbalanced endgame. g) 8 h4 hs! is not helpful to White in
b) 8 1i.e2 dS ! transposes to Stavro any variation .
poulou-lkonomopoulou a few para s lbg4! g ttJfs
...
graphs above. I must say it's interesting The unplayed 9 liJbS ! ? is a better try,
that Black so rarely captures on e4 or though after 9 ...lbxe3 10 fxe3 �d8, the
g2. However, 6 .. .'ikg 6 has other points position is balanced. And imbalanced.
to it, and White can't leave those Go figure.
pawns hanging forever. g lbxe3 10 lbxe3 o-o 11 lbds lbes 12
...
83
Th e Dark Kn ig h t System
c112: 1 tt:Jbs!?
84
1 e4 t'Dc6 2 d4 es
an endgame, two bishops are normally shows that in the endgames that are
equal to a rook and two pawns. normally reached, White sometimes
Black h ad a chance to show the wins and just about never loses. Yuck!
power of points #2 and #4 in Don't worry, it's all taken care of.
P.Hromada-L.Ostrowski, Moravian
Team Champion ship 2002 (see Game
54).
In D.Campora-V.Tkachiev, Biel 199S,
White tried 11 iifs 'ii'xfs 12 exfs b6 13
tLla3 i..b 7. As I just mentioned, a queen
trade generally favours White, but the
price was too high: a tempo, a crippled
pawn majority, and a weakened centre
position. Black had no problems after
pocketing the knight, and had winning
chances, although the game ended in a s 'iff6 6 11Vd2
...
85
Th e Dark Kn ight System
86
1 e4 lL\c6 2 d4 es
10...ttJes
The knight is strong, but if White
tries to dislodge it with 11 f4, then
11...t2Jg4 12 'ii'e 1 (12 l2Jd1? ? t2Jxh 2)
12 ... l2Jxh 2 (in B.Sultimov-N .Pokazanjev,
Russia 2007, Black tried for and got
more with the risky 12 ... i.d7 ! ? - see 14 c3 i.f6 15 'it'xd6 cxd6 is about
Game 58) 13 cJo>xh 2 'iih 6+ 14 cJo>g 3 'i¥g6+ equal, while the common 14 'i¥g 5 'it'g6
is a draw by perpetual. is a little better for Black.
87
Th e Dark Kn ig h t System
88
1 e4 lbc6 2 d4 e5
C22: s es lbg4!?
(221: 6 'ike2 89
(22 2 : 6 0-0 90
C221: 6 'ii'e 2
Most often played, but White strug
gles to equalize in the main lines.
89
Th e Dark Kn ig h t System
There is still chess to be played, but then 9 b3 ! o-o 10 .i.. a 3 'ii'd 8 11 i.. x e7
the position is equal after (among oth- ttJxe7 12 'ifxd4 'ifxd4 13 ttJxd4 gives
ers) 10 ttJbd2 d6 11 ttJb3 d3 12 i..xd3 White a good endgame.
ttJb4 13 o-o-o ttJxd3 14 .l:!.xd3 h 6 and
1 S ... �f7, though Black has man aged to
stir things up a bit. Worse for White is
10 i.. x c7? ! d6 ! 11 .i..b s i..d 7 1 2 i..x c6
i.. x c6 13 ttJxd4 i.. x g 2 14 .l:.g 1, as in
R.Stranz-K.Neumeier, Austrian Team
Championship 2004, when Black
should have continued 14 ... i..h 3 with
an edge.
90
1 e4 Ci:,c6 2 d4 e5
you won't be seeing the position for couple draws, but 15 .l:!.e8 ? ! is not one
the first time, unlike your poor oppo of them : 1S ...'il¥xe8 16 Ci:,xe8 .i.e6 ! 17
nent. As it turns out, White has several � 5 .i.g4 18 t£:,f6+ .txf6 19 'iVds .l:tad8
acceptable routes to a draw, but no 20 Wxf2 .l::!. x ds 21 tt:Jxds .i.d4+ is better
good way to play for a win : for Black, who is more active, and
White's king is still a problem. Instead,
15 'ifxf7+ l:!.xf7 16 t£:,xf7 Ci:,d3+ 17 'iii>h 1
Cf:,f2+, and now i t i s Black who is best
advised to repeat moves; or 1 5 t£:,xf7
'Lld3+ 16 .ie3 'ifxds 17 .txds Ci:,xe1 18
.l:txe1 .i.fs 19 .ixd4 Ci:,xd4 20 .l:.e7 .i.xc2
21 l:txb7 Cf:,fs and the usual repetition
follows.
b3) 14 .txf7+ �h 8 15 �xf2 .id4+ 16
�g 2 Vi'xd6 17 .if4 Vies 18 Ci:,e4 'iffs and
the complications are not over, but
b 1 } 1 4 t£:,xf7 'iVxd1 1 5 Ci:,xd1 Ci:,xd1 White's king position is far too fragile
and White can repeat moves now with for him to have serious thoughts about
16 Ci:,d6+ 'it>h 8 17 Cf:,f7+, or play 17 l:f.xd1 winning - losing is quite attainable
.id4+ 18 .l:!.xd4 (18 �h 1 .ig4! is trouble though !
for White, as is 18 �g 2 :f2+ 19 �h 1 b4} 14 'it>xf2 ? ! doesn't make sense .
.ig4) 18 ... Ci:,xd4 and repeat moves here 14 ... .i.d4+ and 1 S ... 'it'xd6 is not terrible
with 19 Cf:,f7+ 'it>g 8 20 Ci:,g s+ etc. for White, but he h as nothing to com
b2} 14 'i!Vds .i.d4!, when White h as a pensate for his loose king.
91
C h a pter Five
1 e4 t2Jc6 2 tiJf3
92
1 e4 ti:Jc6 2 tl:Jf3
and 61). Mestrovic is also 2-0 with 4 ... g 6 J.Schuyler, Washington 2012: 8 ... d5! 9
here. T o continue: 5 .i.d3 .il.g 7 6 o-o o-o .i.g 5 ? ! (9 exd5 ti:Jxd5 10 dxe5 ti:Jxe5 11
7 h 3 e5 8 .l:le1. ti:Jxe5 .i.xe5 12 .l:le1 .i.g7 is equal)
9 ... dxe4 10 .i.xe4 exd4 11 .i.xc6? ! (11
cxd4 is still fairly level) 11 ... dxc3! with
great complications favouring Black
(see Game 62).
b) 4 ti:Jbd2 is similarly met by 4 ... g6.
93
The Dark Kn ight System
A: 5 .ib5 94
B: 5 d 5 9 6
C: Others (without d4-d 5) 102
A: 5 .ib5 a6
94
1 e4 l2Jc6 2 lZ'lf3
u ..tgs?!
The strong White and Black players
to hold this position were focused on
the ... e7-e5 break when they should be
preparing for ... c6-c5.
8 h3 11 ..te3 is a bit better, though it still
White usually plays this move doesn 't stop 11 ... c 5 ! ; e.g. 12 dxc5 .l:txb 2 !
sooner or later in the Dark Knight Pirc, 13 .l:!.xb2 ..txc3 1 4 .Ub3 ..txe1 1 5 'i!i'xe1
as he gets tired of worrying about l2Jxc5 16 ..txc5 dxc5 equal, or 12 Vi'd2
... ..tg4. Others: cxd4 13 ..txd4 l2Jf6 14 e5 lZ'lh 5 15 exd6
a) 8 e 5 ! ? lZ'ld5 9 lZ'le4 o-o 10 l:te1 was cxd6 16 ..txg7 lZ'lxg 7 with equality.
95
Th e Dark Kn ight System
after all, when Black plays ... c7-c6 9 i..h 6 i.. g 4 10 i..xg7 �xg7 11 i.e2
96
1 e4 l'Llc6 2 l'Llf3
l'Llbd7 - White's "attack" is going no position to B2. Of course White could
where; e.g. 12 h3 i..xf3 13 i..xf3 'irb6 14 have played 6 h3, insisting on B1.
h4? ! tt:Jes, threatening ...l'Llc4) 8 ...i..xh6 ! 9 7 0-0 0-0
�xh6 'i!Vb6 10 o-o-o i.. g 4 is an interest
ing position, not at all unfavourable for
Black; e.g. 11 l:.d2 i..xf3 12 gxf3 l'Llbd7 13
f4 cxds 14 exds .l:.c8 (14...0-o-o! ? and
15 ...<;i;>b8) with great interest in ....l:txc3;
or 11 i..e 2 'ii'xf2 12 .l:.hf1 'ii'c s and while
White is certainly well developed, he has
nothing concrete for the pawn. In
N.Sulava-M.Muse, Croatian Team
Championship 2002, White changed the
course of the game with 11 e s ? ! dxes 12
d6, but had Black spotted 12 ...l'Llbd7 13 White has:
dxe7 'ii'h4, followed by 14...'ili'xe7 and
15 ...0-0-0, White would have found him 81: 8 h 3 9 7
self without sufficient compensation. 82: Others 99
c) 6 i.. e 3 is very similar to 6 i.. g s,
into which it often transposes (i.e. after Or 8 a4 as ! - Black cannot allow
i..e 3-h6). One time that didn't happen himself to become further cramped.
was in K.Nemcova-F.Olafsson, Marian The insertion of the two a-pawn moves
ske Lazne 2008, which went 6 ... i.. g 7 7 is helpful to Black though, since it helps
'iWd2 c6 (by transposition) 8 h 3 b S ? ! him to establish knight outposts on the
( 8. . .0-o 9 i.. e 2 b s ! , a s i n J. Hjartarson cs- and b4-squares.
F.Olafsson, Reykjavik 1995, was a better
move order - see Game 68) 9 a3 ? ! (9 81: 8 h 3
dxc6 ! b4 10 i..b s ! is unpleasant for
Black) 9 ... a6 10 dxc6 l'Llxc6 1 1 i..d 3 o-o
12 o-o i..b 7, resultin g in an equal
Dragodorf type of position which Black
went on to win (see Game 67).
d) 6 h 3 ! is the most accurate move,
reaching line B1 below after 6 ... i.. g 7 7
i.. e 2 o-o 8 o-o.
6 i.. g 7
...
97
The Dark Kn ig h t System
This is played most often, spending a sition where his normal play (with c2-
tempo to stop 8 ... .ig4. As we will see, c4-c5) is blocked by the c3-knight. This
8 ... .ig4 was indeed Black's intention, problem is serious for White, more so
but a tempo is a tempo. There is some than Black's funny knight on b8 (which
disagreement about the merits of 8 h 3 is n ormally on the e7-square). Black will
- Alburt and Chemin adorn i t with an play carefully for .. .f7-fS. Part of being
{!), while Nunn says it is "not really careful is considering ...h 7-h 6 to pre
necessary". I think it is the best move in vent White's tt:lg s-e6.
the position . g .ixe6
...
s es!
...
811: 10 i.gs 98
812: 10 tt:ld4 99
811: 10 .igs !
98
1 e4 tLlc6 2 lDf3
99
Th e Dark Kn ight System
15 ...tt:\a5 ? ! 16 lLlf3, which was good for If White is going to refrain from 8
White in K.Hulak-S.Marangunic, Yugo h3, it makes very little difference which
slavia 1977, Black should play 15 ... tt:\d4! move he chooses, but we'll take the
16 tt:\f3 tt:\xf3+ 17 'ii'xf3 i.c6 with an following as the main line:
equal position. 8 i.e3
12 lLlf3 :es 13 i.c4 .i.e6 14 tt:\ds 'it'd7 Alternatively:
15 c3 i.xds 16 exds .l:txe1+ 17 'ili'xe1 a) 8 i.g 5 should be treated simi
tt:\es ! 18 tt:\xes dxes larly: 8 ... i.g4! 9 h 3 i.xf3 10 i.xf3 c6 11
'ii'd 2 lLlbd7 and Black follows with
...'ii'b 6 , ... a7-a5, ... tt:\c5 and ... tt:\fd7, with
typical dark-square play, even if White
is still slightly better.
White can al so try to do without h2-
h 3 . For in stance, 9 'ir'd2 c6 10 l:lfe1
tt:\bd7 transposes to L.Vajda-M.Marin,
Rumanian Champion ship, Bucharest
1998, and M.Kolosowski-Dan .Fraczek,
Legnica 2011, which both continued 11
dxc6 ! ? bxc6 12 tt:\d4 :c8? ! 13 f3 i.e6 14
The position is equal, White's bish tt:\xe6 fxe6 with advantage to White.
ops being offset by his poor develop According to Houdini, Black can equal
ment and the inconvenience of guard ize in this line with 12 ...'ii'b 6 ! 13 tt:\a4
ing the d5-pawn. If 19 'it'xe5, then after 'iWc7 14 i.xg4 tt:\xg4 15 i.xe7 d5 16 f4
19 ... l:te8 and 20 ...l:te1+, White will never (16 lLlf3 ? ! tt:\de 5 ! 17 i.xf8 lLlxf3+ 18
complete his development. gxf3 'ili'xh 2+ 19 �f1 .l:.xf8 20 fxg4 'ili'h 1+
2 1 �e2 'ili'xe4+ 22 �1 'ir'xa4 is good for
82: Others (besides8 h3) Black) 16 ... l:tfe8 17 exd5 cxd5 18 h 3
lLl g e 5 19 dxe5 l:!.xe7.
b) 8 l:!.e1 i.g4 (of course) 9 h3 i.xf3
10 i.xf3 tt:\fd7! 11 i.e3 c6 was
D. Rogozenco-V.Nevednichy, Rumanian
Team Championship 2005, where
White went wrong immediately with
12 i.d4?!, initiating an exchange bene
ficial to Black; e.g. 12 ...i.xd4 13 'ii'xd4
'ii'b 6 is equal. Without this mistake
White is a little better, but Black's mi
nor pieces are all good, and he h as the
1 00
1 e4 t'Dc6 2 tbf3
simple plan of ... a7-a5, ...t'Da6, ... t'Dacs, great success in the 1990s (three
... as-a4-a3 (if possible), and ...'i!Vb6 or draws, three wins, all against GMs). The
.. .'�C7. normal m ove order to reach this posi
8 ..tg4
... tion is 1 e4 d6 2 d4 ttJf6 3 t'Dc3 g6 4 l'Df3
Black has done very well with 8 ... c6, ..tg7 5 ..te2 0-0 6 0-0 ..tg4 7 ..te3 t'Dc6 8
but Joel Benjamin (who played 1...l'Dc6 ds t'Db8. (8 ... ..txf3 is the traditional
frequently in the 1990s) warned m e main line, but it is not stronger.)
that Black's position is difficult after 9
a4! as 10 h 3 (otherwise 10 ... ..tg4)
10 ... l'Da6 11 ..txa6 ! (an idea I have not
been able to find in any published
games - the point is to stop Black's ac
tive 1 1...l'Db4) 11...l:.xa6, when it is hard
to find a con structive plan for Black,
whereas White can still build; e.g. 12
'i!Vd2 .l:.a8 13 l':!.ad1 l:.e8 14 ..td4 'ilc7 15
.:tfe1. Black lacks space, development
and pawn play, while his only "asset",
the light-squared bishop, is more of a 9 h3
problem than anything else. Indeed, White must play this sooner or later
this type of position acts m ore closed unless he is intending to allow the ex
than open, in part because nobody change of light-squared bishops. White
wants to relieve the tension between should at least retain the bishop pair if
the ds- and c6-pawns - for White to he is h oping to keep an advantage:
trade would assist Black greatly in the a) 9 t'Dd2 ..txe2 10 'ilxe2, and now
central battle, while if Black trades, he Chemin's recommendation 10 .. J:te8 11
has accessible weaknesses on the bs f4 e6 has been tested only once, in
and e7-squares. A.Czebe-N .Resika, Budapest 2000 -
A possible antidote is 8 ... as ! ? 9 a4 Black, an FM, held the draw against the
tt:Ja6, when White shoul d be less eager GM.
to snap off the knight. N otice that since b) 9 l'Dd4 ..txe2 10 'i/xe2 cs 11 l2Jf3
Black h as not yet played ... c7-c6, he can 'iib 6 12 l1ab1 'ila6 13 'i/d2 l'Dg4 14 ..tg s
later try ... e7-e6 or ... e7-e5 instead. l:te8 with equality in B.Ch atalbashev
However, there are other moves to M.Popchev, Cacak 1991 (see Game 70).
worry about besides 9 a4. c) 9 t'Dg s ..txe2 10 'ilxe2 c6 11 .l:.ad1
8 ... ..tg4 is simplest, transposing to a 'i/as 12 f4 'i/a6 13 'ilf3 t'Dbd7 with
favourite line of the great Pirc expert equality. Black, with more experience
Alexander Chemin, who used it with in this type of position, went on to win
1 01
Th e Dark Kn ight System
1 02
1 e4 ltJc6 2 lDj3
s .i.g7 6 .i.e3
... Then 6 ... 0-o 7 0-0 (7 .i.e3 tran sposes to
By far the most common, but also 7 .i.e2 in the notes to the main line)
seen are: 7 ... es (7 ... a6 ! ? is more combative, and
a) 6 .i.g s o-o 7 'iii'd 2, when both was tried successfully in N.Ryba
7 ... a6 and the surprising 7 ... ds have J.Schuyler, Washington 2012 - see
worked extremely well for Black. Game 74) and now:
b1) 8 dxes
1 03
Tn e D a rk Kn ig h t Sys tem
normal choice. The reader would be Black can decline the gift: 10 ....l:tf8 11 c3
well advised to take special n ote of any (11 .if4 'ii'e 6+ 12 .ie5 �g8 13 c3 lt::l x e5
exceptions. 14 lt::l dxe 5 'ir'd6 is equal) 11 .. .<�g8 12 o-o
Here 8 ....�)xe5 9 lt::lx e5 dxe5 10 .ig 5 h6 13 l:te1 g 5, though White is slightly
c6 is equal, when 11 .ic4 b5 allows better here.
Black free expansion on the queen side. d) 6 Ji.b5 looks silly since Black can
b2) 8 .ie3 is a bad version of 7 .ie2 castle out of the pin, but after 6 ... 0-0 7
e5 in the main line - if White is castled o-o Black has nothing better than 7 ... a6
kingside he has no attack to compen 8 .ixc6 bxc6, transposing to line A
sate for his troubles; i.e. 8 ... exd4 9 lt::lxd4 above .
.l:te8 and White already lacks a comfort 6 0-0
...
1 04
1 e4 lL\c6 2 lZJf3
7-.eS 8 dS
Others:
a) 8 0-0-0 exd4 9 lZJxd4 l!e8 10 f3 -
3lack wins nearly every game from this
!Xlsition. Indeed, White's pawns on h3
and f3 make a ludicrous impression (he
's essentially down a full tempo in a
Philidor Defence Larsen Variation),
though White should not actually be
worse. 10 ...lZJxd4 11 .ixd4 i.e6 12 g4 (12
l.f2 a6 13 'iii>b 1 bS 14 h4 cs! was about
equal in G.Bastrikov-E.Geller, Tashkent 8 ... l2Jb8 is also quite reasonable. The
19 S8 - see Game 76) 12 ... cs ! 13 .ie3 e7-knight supports the ... c7-c6 break,
IUrXd6i'
waS .li'. (1 3 ... d S I. equa1·lZeS ) 14 'B
·;M; . .l (14 but with White likely to castle long and
l.h6! i.xh6 1S 'ii'xh6 i.xa2 16 l:f.xd6 lle6 form a battery on the d-file, the useful
is slightly better for White) 14...l2Jxe4! 1S ness of this option is reduced.
fxe4 .ixc3 16 'ifxcs ? i.xb2+ 17 'it>xb2 9 0-0-0
·«xa2+ 18 'iii>c 1 l:lac8 and White soon Instead:
called it quits in D.Bescos Cortes-S.Garza a) 9 .ih 6 is statistically the most
Marco, San Jose 1998. 16 bxc3? 'ii'a 3+ 17 dangerous, scoring 6/7 for White, but
'it>d2 l:!.ad8 wouldn't have worked either; Black was heavily outrated and his play
instead after 16 i.d3, closing the d-file, quite uninspired. Artashes Minasian
the game still continues, though Black is shows us the way in D.Saduakassova
clearly better. Art.Minasian, Dubai 2011: 9 ... .id7 ! 10
b) 8 dxes seems like it is headed for .id3 (10 o-o-o? ! b S ! ) 10 ... c6 (or 1o ... bs
dullness, but things could get interest 11 a3 aS 12 0-0 b4 intending ... C7-C6
ing if nobody trades those queens; e.g. with counterplay) 11 dxc6 .ixc6 (taking
8 ... l2Jxes 9 lZJxes dxes 10 o-o-o .ie6 11 firm control over the dS-square; back
g4 c6 12 g S (12 'ifxd8 .Ufxd8 13 �xd8 ward pawn ? what backward pawn ?) 12
l:i.xd8 14 Si.xa7 .ih6+ 1S .ie3 .ixe3+ 16 .ixg 7 'it>xg 7 13 o-o "fic7 (13 ...bs ! ?) with
fxe3 hs 17 g s lLih7 18 h4 f6 19 gxf6 easy equality (see Game 77).
lL\xf6 is equal despite the slight pawn b) 9 Si.e2 .id7 (9 ... lZJxe4 ! ? 10 l2Jxe4 fs
deficit: Black's activity and future 11 lLic3 f4 12 o-o-o is slightly better for
passed h-pawn are sufficient) 12 ...tZ:'ld7 White) 10 g4 (not 10 .ih 6 ? ! c6 ! 11 dxc6
13 h4 'it'as and the race is on. Black has .ixc6 with advantage) 10 ...b s ! 11 g S
no reason to be pessimistic about his (11 Si.d3 ? ! b4 12 tZ:'le2 c 6 13 dxc6 .ixc6
prospects. 14 tZ:'lg 3 lZJxe4! is good for Black; while
8 tZ:'le7
•.• 11 a3 l:tb8 is equal, since 12 .ixa7? runs
1 05
Th e Dark Kn ight System
into 12 ... .l::t a 8 ! 13 j_e3 b4) 1 1 ... b4 12 White has a hard time evicting the h s
gxf6 bxc3 13 �xc3 j_xf6 with a level knight, and will also fi n d i t hard to
position. profit even if the h-file opens. One con
c) 9 .U.d1 tt:'lh s 10 g4 lL'lf4 (this ag tinuation of many: 12 ..ti>b1 ..td7 13
gression is called for because White lL'lh 2 ltJc8 ! ? 14 lL'lg4 lL'lb6 15 b3 b4 16
can no longer tuck his king away on the lbe2 as, still with approximately even
queenside) 11 j_xf4 exf4 12 �xf4 f5 13 chances.
exfs gxfs 14 g s j_xc3+ 15 bxc3 .l:te8 16
j_e2 lbg6 with compensation . C2: 5 j_e2 i.g7 6 o-o o-o
d ) 9 g4 a 6 should transpose t o the
main line once White castles long.
9 a6 10 g4 bS 11 gS
...
1 06
1 e4 lbc6 2 lbf3
lbe1 fs most closely resembles a King's to take! ) 8 ... exd4 9 lbxd4 .l::!. e 8 10 f3 (10
Indian, where White's attack will be lbxc6 bxc6 11 f3 dS tran sposes;
greatly delayed by his pawn stuck on 11 ...'i*'e7 is also fine) 10 ... d s ! 11 lbxc6
the C2-square. bxc6 12 llad1 'i¥e7 13 il.. d4 dxe4 14
c) 8 il..f4 lbg4! 9 h3 (if White does fxe4 lbxe4 15 lbxe4 'ii'x e4 16 .ixg 7
not play h 2-h 3, the game will transpose c.t>xg 7 17 .if3 i¥e3+ 18 'it'xe3 .l:!.xe3 19
to note 'b' just above; e.g. 9 dS es 10 .ixc6 l:f.b8 20 �d8 �b6 with a level
.i.d2 lbe7, or 9 .id2 es, or even 9 .ic1 endgame.
es 10 ds etc) 9 ... es 10 dxes ttJgxes 11 s lbxes
...
lbxe s ? ! (11 i¥d2 is better, retaining a 8 ... dxes is also playable. Z.Bratanov
tiny edge after 1 1 ...c.t>h 7; Black can con- B.Chatalbashev, Bulgarian Champion
sider ... lbxf3+ and ...lbd4, or ... .i.e6, or ship 2004, continued 9 'ii'x d8 �xd8 10
.. .f7-f5, or ... l:te8) 11 ... dxes was level in il..c4 h6 11 h3 b6 12 lbds lbas ! 13
A.Sakharov-A.Adorjan, Sochi 1976, lbxf6+ �xf6 14 .ie2 .ib7 1 5 b4 lbc6 16
though Black went on to win a wild c3 as 17 a3 lbe7 18 lbd2 .ig S ! and
game (see Game 78). Black went on to win (see Game 79).
7 es!
••• g lbxes dxes 10 'i!Vxd8 J:!.xd8
As usual, 7 ... a6 can be tried - the
main line is a bit drawish - but then 8
dS lbb8 9 a4 is an excellent answer.
Alternatively, 7 ... il..g 4 8 dS lbb8 trans
poses to line B2 (8 ... .ixf3 9 .ixf3 lbes
10 �e2 c6 is a main line Classical Pirc
which will not be covered).
8 dxes
Or:
a) 8 dS lbe7 9 1i'd2 lbg4 10 �g s h 6
11 il..h 4 g s 12 .ig 3 f s 13 h 3 lbf6 1 4 exfs
lbxfs 15 .ih 2 'ii'e 8 with a tiny edge as In practice this is equal, but White
Black considers a kingside attack based can try to extract a little something:
on his space advantage there, or the a) By far the most common is 11
... eS-e4-e3 lunge, or ...'ii'f7 and ... lbe7, �fd1 (or 11 .laad1 - it doesn't much
building pressure on White's d-pawn. matter) 11 ... �e6 12 lbbs, when the
White is not well situated to use his new move 12 .. J:tdc8 ! keeps things level;
asset - the e4-square. e.g. 1 3 lbxa7 .l:.xa7 14 �xa7 b6 1 5 .l:.d3
b) 8 i¥d2 (as usual, it is a bad idea .l:Ia8 16 .l:.a3 .if8 17 :a6 �c8 18 :a4
for White to try to maintain the tension .id7 etc.
- this only works if Black is not h appy b) 11 lbbs .id7 12 f3 looks scary,
107
� ,.., e :J a rtc Knig h t System
and both 12 ...b6 and 12 ... ..1i.xb5 give 7 h 3 is C 1 again, and 7 d5 ti:Jb8 8 h 3
White something to work with. The c6 is i n the notes t o line B. Others:
novelty 12 ... a6 13 ti:Jxc7 llac8 14 i.b6 a) 7 o-o-o h as not scored well, but
i.c6 1 5 li:Jxa6 l:!.d2 16 .itd3 bxa6 17 .ite3 White's set-up is challenging - he h as
llxd3 18 cxd3 is a tiny edge for White, man aged to omit h 2-h 3 and stop ... e7-
but the imbalances should provide e5, while 7 ... li:Jg4? ! 8 ..li.f4 e5 9 dxe5
Black some winning chances as well. ti:Jgxe5 10 ti:Jxe5 lt:\xe5 11 h4 is still not
c) 11 ..tc4 c6 12 ltad1 l:f.e8 1 3 a4 i..f8 convincing.
14 f3 Wg 7 1 5 .l::t d 2 i..b 4 16 J::f.fd1 :e7 17
�f2 l:td7 and White is running out of
things to play for.
1 08
1 e4 ltJc6 2 lLlf3
not the time for the exchange varia A.Safranska, Grenoble 2003, when
tion, and 9 o-o-o 'Wie7 (or 9 .. .tL'lg4) 10 12 ... c5 13 c3 'i!Va5 ! 14 lt:Ja3 (14 cxd4?
�g 5 �e6 11 lt:Jd5 ? ! i.xd5 12 exd5 .t'!.fd8 cxd4 15 'ii'a 3 .l:.ac8+ 16 Wb1 'ifxa3 17
just makes matters worse for White. tt:Jxa3 lt:Jxe4 18 .l:te1 lt:Jxf2 19 l:f.g 1 �xf3
c) 7 �c4 lt:Jg4 8 i.g 5 h6 9 i.h4 g5 10 20 gxf3 d3 21 :1xe5 d2 22 �e2 .l:tfe8 2 3
�g 3 e5 11 d5 lt:Jd4 is about equal, but lt:Jc4 b5 is nearly winning) 1 4. . .�xf3 1 5
not a very rational position - the tac gxf3 lt:Je6 would h ave given Black a
tics would take pages. Instead, large positional advantage.
7 ... i.g4!? keeps things under control;
e.g. 8 d5 i.xf3 9 gxf3 lt:Je5 10 i.e2 c5 11 C4: 5 �C4 � g 7
o-o-o 'i!Va5 12 ..t?b1 l:tab8, or 8 o-o-o
tt:Jxe4 9 tt:Jxe4 d5 10 �d3 dxe4 11 �xe4
li'd7 12 d5 �xf3 13 �xf3 tt:Je5 14 �e2
lt:Jg4 15 �d4 e5 16 dxe6 'i!Vxe6 17 �xg 7
'it>xg 7 with near equality.
7 esl
...
Instead:
a) 6 i.g 5 ? ! tt:Jxe4! 7 tt:Jxe4 (or 7
�xf7+ '>t>xf7 8 t2Jxe4 d5 9 lt:Jc5 lif8)
7 ... d5 8 c3 ! ? dxc4 9 d5 f5 ! 10 dxc6
'ili'xdl+ 11 ltxd1 fxe4 12 lt:Jd2 bxc6 13
tt:Jxe4 l:.b8 is good for Black.
b) 6 �f4 - As a novice, I remember
This position has led to three draws, thinking that this was some kind of
but Black h as a slight advantage based ideal position for White. N owadays, it
on White's weak bishop. For instance, looks like White is begging for trouble,
11 'ili'e3 lt:Jd4 12 lt:Jb5 ? ! was F.Saez- as the bishops are vulnerable and do
1 09
Th e Dark Kn ight System
not defend his centre. Indeed, my re only a little worse for White) 9 ... dxe4 10
search uncovered a large collection of i.. x e4 l2Jxd4 is clearly unsatisfactory for
games by some of the lowest-rated White.
players I have ever seen in any data 8 ...l:txf7 9 tt::lxe4 ds!
base. - 6 ... tt::l xe4 (6 ... i.. g 4 is m ore com
bative) 7 tt::lxe4 (7 i.. xf7+ <tlxf7 8 tt::lx e4
:fs 9 dS 'it>g 8 ! 10 'iVd2 'ii'd7 ! ! 11 dxc6? !
�g4! o r 1 0 i.. g s tt::l e s with a tiny edge)
7 ... ds 8 i.. d 3 dxe4 9 i..xe4 l2Jxd4 10
tt::l xd4 'i!i'xd4 11 'it'xd4 i..xd4 12 i..x c7
i.. xb2 13 l:tbl fS 14 i..f3 i.. d4 with a
level endgame.
(41: 6 0-0
equality.
7 h3?!
This is no good, but it is the most
common, and other moves are either
met by ... i.. g 4 or else are covered else
where: 7 dS tt::lb 8 is line B again, while 7
i.. e 3 tt::l g 4 is C43 below.
7 ..tt::lxe4 8 i..xf 7+
.
110
1 e4 ltJc6 2 tLlf3
es and won (against much weaker op 8 d S ? ! tLld4! 9 .i.xd4 (or 9 'i¥d1 .ltxf3
ponents). If Black takes the bull by the 10 gxf3 o-o) 9 .. exd4 10 tLlbs (10 e s ?
.
horns with 7 ... .i.xf3 8 gxf3 liJxd4 9 exf6 0-0! 1 1 exf6 dxc3 wins) 1 0. . .0-0 11 0-0
ttJxe2 10 fxg 7 l:tg8 11 �xe2 ! ? (or simi l::te 8 was much better for Black in
larly 11 .i.xe2 - White is 2-0 here as M.Strubreiter-K.Rogetzer, Austrian
well) 11 ...l:i.xg 7 we reach this mess: Team Champion ship 2004, and after 12
liJbxd4? ! ltJxe4 13 �d3 ttJcs 14 �d2
Black should have cashed out: 14 ....txf3
15 ./tJxf3 .i.xb2, with a squeaky-clean
extra pawn to go with his positional
advantages.
8 ./tJxes 9 .itb3 o-o 10 o-o-o
...
111
The Dark Knight System
C43: 6 i..e 3 o-o tural advantage with ... .ixc3) 11...i.. xh6
12 tZ'lf3 c6 and White is only slightly
worse.
9 ... gs
This potentially king-weakening
move is acceptable because White has
also castled short.
10 i.. g 3 tZ'lf6
7 0-0
7 'ii'd 2 transposes to 7 .ic4 in line
C3 (see note 'c' to White's 7th m ove).
1 .. .tbg4 s .igs
In stead:
a) 8 .if4?! tbxd4! (fork tricks every
where !) 9 tbxd4 es 10 .ie3 tbxe3 11
fxe3 exd4 12 exd4 .ie6 was H .Hughes Chances are equal; e.g. 11 h 3 tbxe4,
K.Richardson, British League 2004. or 11 .ib3 .ig4 12 dS tbas 13 h3 .ih s
Black had the edge and went on to win 14 'ii'd 3 .ig6.
(see Game 82), although 12 ... c6! was
simpler with a comfortable advantage. CS: 5 .igS .i g 7 6 11i'd 2 h61
b) I don't know who would play it,
but Houdini likes 8 .ic1, when 8 ... es
allows White a small advantage after 9
.ig s ! i..f6 10 .ixf6 tbxf6 11 ds tbb8.
Instead, 8 ...tbf6 returns to 6 ...0-o in line
C41 and offers (or bluffs) a repetition.
s ... h6 9 .ih4
On 9 .if4?! tbxd4! is best, as in the
previous note, even though White has
the extra possibility 10 tbxd4 es 11
.ixh 6 ! (after 1 1 tbe6 ! ? fxe6 12 'ii'x g4
exf4 13 i..x e6+ �h 7 14 .ixc8 'ir'xc8 This bishop hunt does work.
Black's active bishop allows him a small 7 .if4
plus - which he m ay convert to a struc- Others:
112
1 e4 lL\c6 2 ltJf3
because White could not control the 10 �xc7 is too greedy; e.g. 10 ...�g4
dark squares (see Game 83). 11 c3 (11 i.e2 .l:.c8 12 i.g 3 �xf3 13
b) 7 �e3 l2Jg4 (8 �f4 t2Jxd4! 9 t2Jxd4 i.xf3 'i!Ve6+ 14 ii'e3 t2Jxd4 1 5 �xe6
e5 equalizes; e.g. 10 h3 exd4 11 l2Jb5 t2Jxf3+ 16 gxf3 fxe6 is equal) 11...�xf3
l2Je5 12 t2Jxd4 'Wie7 13 �b5+ �d7 14 12 gxf3 :l.c8 13 �f4 'i!Vxf3 14 .l:tg 1 e5 1 5
�xd7+ t2Jxd7 1 5 0-0-0 0-0-0) 8 0-0-0 dxe5 �xe5 16 �xe5 t2Jxe5 17 i.b5+
t2Jxe3 9 'i!Vxe3 0-0 10 h3 a6 11 g4 b5 was 'ifi>e7 and Black h as slightly fewer king
equal in D.Janowski-F.Yates, Marienbad problem s than White.
1925 (!), a wild game that ended in a 10 �g4 11 �e2 o-o-o
...
113
C h a pte r S ix
1 e4 l2Jc6 2 l2Jc3
114
1 e4 lt:Jc6 2 lt:Jc3
115
Th e Dark Kn ig h t System
a) 5 lt:Jf3 exd4 6 lt:Jxd4 is some sort met by ...h7-h6 so that the enemy pawn
of Philidor where White's 4 f4 is prema does not reach the h6-square. The posi
ture. tion is equal.
c) 5 fxe5 dxe 5 6 d5 lt:Jd4! (6 ...lt:Je7
transposes to 6 fxe5 in n ote 'b') 7 lt:Jf3
.i.c5 8 lt:Ja4 ! ? (or 8 .i.g 5 o-o 9 'ii'd 3 1Wd6
and since 10 o-o-o? ! lt:Jg4! is strong,
White is clearly worse) 8 ...lt:Jxe4 9 lt:Jxc5
lt:Jxc5 10 lt:Jxd4 'ii'h 4+ 11 g3 1Wxd4 12
1Wxd4 exd4 and although White's bish
ops are nice, it is not easy for him to
recover the pawn with a good position
- Black is a little better.
d) 5 .i.b5 ? ! exd4 6 'ii'xd4 .i.e7 7 lt:Jf3
Black must be alert in order to prove 0-0 8 .i.xc6 bxc6 9 o-o l:te8 10 .:td1 .i.b7
this - and it is worth noting that, until with two bishops and a big bull's-eye
n ow, he has not been up to the task: on the e4-pawn - more than enough to
6 ... �e7 7 .i.e2 o-o 8 .i.e3 lt:Jxd4 9 .i.xd4 make up for our space disadvantage.
and here the new move 9 ... d5 ! 10 .i.xf6 s ... dxes
.i.xf6 11 'ii'x d5 .i.xc3+ 12 bxC3 1i'f6 13 Black often throws in 5 ... .i.g4 6 lt:Jf3,
'ii'd 2 �d8 14 .i.d3 (14 'ii'e 3 is met by but this does not help the situation.
14 ... 'ifh4+ 15 'ii'f2 1i'f6 16 e5 \i'c6, or 1 5 6 1Wxd8 �xd8 7 lt:Jf3 �b4 8 .i.d3 .:te8
g 3 'ifh 3 and White h a s some trouble
getting his king safe and activating his
h 1-rook; e.g. 16 .i.f3 b6 17 e 5 11b8 18
'ili'e2 .i.b7 19 .i.xb7 .l::t xb7 20 :d1 l:tbb8
with compensation) 14...'ii'b 6 15 1Wf2
'ii'b 2 16 0-0 1Wxc3 17 e5 1i'd4 18 1Wxd4
�xd4 19 �ad1 �f8 20 c3 �d7 2 1 .i.e4
�b8 and Black unravels safely.
b) 5 d5 lt:Je7 6 lt:Jf3 (6 fxe5 dxe5 7 .i.e3
lt:Jg6 8 a3 .i.d6 is fine for Black; or 6 f5
c6 ! 7 dxc6 bxc6 8 lt:Jf3 1i'c7 with usually
... d6-d5 coming soon, and sometimes This position has been reached only
...�b8 and/or ... g7-g6) 6 ...exf4 7 .i.xf4 once: in Ma.Tseitlin-C.Barlocco, World
lt:Jg6 8 .i.g 3 .i.e7 9 'ii'd 2 0-0 10 0-0-0 lt:Jd7 Seniors Championship, Port Erin 2004,
intending ... .i.f6, ... l:te8, ... a7-a6, ... b7-b5, which continued 9 o-o exf4 10 lt:Jd5
...lt:Jc5, ....i.b7. White's h 2-h4 must be lt:Jxd5 11 exd5 lt:Je7, and White went on
116
1 e4 ltJc6 2 ltJc3
to win . However, Black can equalize only one obscure trial) s ... c6 6 "i¥d2
with 9 ... h6! 10 fxes ltJg4 11 h3 (or 11 ltJbd7 7 o-o-o cxds 8 exd5 a6 should
l:td1 ..tcs+ 12 �f1 ..td7) 1 1 ... ltJgxes 12 make for a lively game, with ...b7-b5
..tf4 (or 12 ..tbs ..td7 13 ttJds ltJxf3+ 14 and ... ..tb7 coming, and probably ... g7-
gxf3 ..td6) 12 ... ltJxd3 13 cxd3 �e6. g6 and ...�g7, with maybe ....:c8, ... "iic 7
and/or ...ltJb6. White will be unable to
B: 4 d S liJbB organize the rapid pressure on the e
file that would make this uncomfort
able for Black.
s c6
...
117
Th e Dark Knig h t System
Black did go on to lose. He can improve 'ii'x c3 14 bxc3 tZ:lf6 with slightly the
slightly by 7 ... i.g 7 and 8 ... 0-0, rather more comfortable g ame for Black.
than committing the queen so early. 8 ... i.g7 9 o-o o-o 10 'ith1 i. g4 11 i.e3
6 ... tZ:lxc6 7 tLlf3 g 6 .l:tc8 12 i.b3
118
Section Th ree
Others
Miscellaneous Topics - Chapter Ten like 1 ... g6. I choose to believe that this is
not too much of a departure, since the
This section deals with almost eve fianchetto is so common in the Dark
rything except the opening moves 1 e4 Knight System.
and 1 d4, including some quite un I will warn the reader that the cov
usual openings. Except for 1 c4, which erage in section 3 is less detailed than
needs to be taken seriously, the chal in the other sections. This is because
lenge for Black is to give himself these openings are far less common,
chances to wrest the initiative early, and also because you may already have
and/or make sure that White does not your own systems of defence, which
reach the type of position with which there is n o need to abandon.
119
C h a pter Seve n
1 c4 ltJc6
120
1 c4 lbc6
121
Th e Dark Kn ight System
122
1 c4 lbc6
But also:
a) 5 ltJes - only larry Christiansen
(and Houdini) seems interested in play
ing this move, though he has an im
pressive 3-0 with it. After s ... l2Jf6 6 ..if4,
as in l.Christiansen-S.Conquest, Oviedo
(rapid) 1992, the new move 6 ... !Dh s ! ? 7
The only time this position has been ..id2 l2Jf6 is fine for Black if he is con
reached, Black started losing the thread tent with a draw, because 8 e3 d6 9
with 11 ... ..ie6? ! 12 lbd4 l2Jxd4?! 13 l2Jxc6 bxc6 10 ..ie2 ..ie7 1 1 o-o o-o is
'ir'xd4 l2Jxc3 ? ! 14 l:txc3 cs 1S "ii'd 2, when nothing for White. Otherwise Black can
White had a pleasant long-term advan try 6 ... d6 7 l2Jxc6 bxc6 8 e3 ..ie7 9 ..ie2
tage and went on to win in the game 0-0 10 0-0 ..ie6, with a slight edge for
Har.Becker-Joa. Franz, G erman league White.
1997. b) s ..ig s ..ie7 6 ..ixe7 l2Jgxe7 7 !Dd2
Instead, after 11 ...1fe7 12 l2Jd4 ..id7 l2Jxd4 8 l2Jdxe4 !De6 9 l2Jd2 b6! 10 e3
13 "ifh3 l2Jcs 14 'ii'c 2 aS, White's edge is ..ib7 1 1 liJf3 0-0 with easy equality and
tiny. chances for more; e.g. 12 ..ie2 f4! is
123
good for Black, is 12 g 3 ? ! f4! 13 exf4
as again resembles a reversed Grand Prix
�4 14 gxf4 :!.xf4 1 5 i.g 2 {or 1 5 i.e2) Attack. White's king's knight will soon
15 ... 'iff8 ! . reach the excellent f4-square, but the
four moves it takes to get there is a
B1: s lLlgs i.b4 high price, even in a blocked position.
6 ... lLlf6 7 e3 i.xc3+ 8 bxc3 d6 g l2Jf4 o-o
6 lLlh 3
This retreat is necessary. White's c4-pawn is weak and his
a) G.McKenna-J.Schuyler, Richmond bishops are not yet working, but he
2008, instead continued 6 f3 ? ! h 6 ! 7 should be able to force through the c4-
lLlh 3 exf3, threatening 8 ...'ifh4+ and c5 break. If Black is not careful, this will
9 ...'iixd4. My opponent found 8 e3! bring White a significant advantage.
which keeps White in the game, However, White needs to be careful as
though 8 ...'ifh4+ 9 g3 i.xc3+ 10 bxc3 well, because if the position opens up
'iih 5 11 i.g 2 'ii'f7 12 i.xf3 'ii'xc4 13 at the wrong time or in the wrong way,
.ih 5+ �d8 would have allowed Black he will be punished for his slow devel
an edge. {A. Beliavsky-V.Bagirov, Minsk opment.
1983, also saw White play 6 f3 ?!, but 10 h4 b6 11 i.a 3 lif7!
Black did not punish him and went on This is a new move, and it makes us
to lose.) more comfortable by stepping out of
b) 6 g 3 ? ! , as in J .Timman-I.Sokolov, the influence of White's a3-bishop and
Dortmund 1999, is n ot a good idea ei avoiding lLle6 forks. lt is also useful to
ther - the bishop should not be caught guard the seventh rank and shield our
dead on the g2-square. After 6 ... lLlf6 7 king . The similar 11 ... 1ie8 was tried
d5 lLle5 8 "ifb3 'ii'e 7 9 i.g 2 ? ! h 6 10 lLlh 3 successfully in R.Koch-M.Wiedenkeller,
'iic 5 11 i.f4? ! lLlxc4 White had little for Reggio Emilia 1982/83, but it accom
the pawn. plishes a bit less and leaves Black vul
After the text move, the position nerable to the il.b5 pin . lt is worth not-
124
1 c4 lbc6
ing that Black's results have been excel tives, but White's "bad" bishop is very
lent in this variation even without the strong . Black will not be better unless
text move, though a few high-level he can arrange to exchange the other
players have successfully advocated pair of rooks.
White.
12 C5 B2: s lDd2
If White does not play this now,
12 ... i.. a 6 is coming, and the c4-c5 break
will be forever ineffective.
12 ... bxcs
12 ... d5 ! ?.
13 i..c4
Naturally, if 13 dxc5 then 13 ... d5! 14
c4? d4 and White's structure is awful.
13 ... ds 14 i.. b s lbbs 15 i..x cs c6 16 i.. e 2
i.a6
1 25
Th e Dark Kn ig h t System
i.d2 filc7
6 ... i.e7l
6 ... g6 is just about always played,
but the fianchetto is slow and some
what accommodating, releasing the
b4-square. Sokolov and Sigurjonsson
both got squashed by White's huge
queenside after l:tbl and b2-b4. The
text move keeps Black focused on
where the play will actually take place.
7 i.e2 o-o 8 o-o d S I
Previously 8 . . .'it'e8 and 8 ... d 6 have Intending 13 ... i.e6, 14 ...l:tfd8, and
been tried. The text is an unplayed Black is fine.
126
C h a pte r E ight
1 t2Jf3 t2Jc6
4 if.. g 2 g61
An excellent recipe. White's extra
tempo amounts to little.
5 o-o i.g7 6 e4
White challenges the centre with 6
c4 less often. In Bu Xiangzhi
V.Ivanchuk, Khanty-Man siysk 2011,
Black played 6 ... dxc4 7 dxc4 'ii'x d1 8
.!:Ixd1 e4 9 tbfd2 fS, when White needed
10 f3 to equalize. Instead, after 10 tbc3
if.. e 6 11 tbdS ? ! o-o-o lvanchuk went on
If you play 1 ... tbc6 against 1 e4 and to win (see Game 9 5).
1 d4, you may as well play it against 1 6...tbge7 7 tbbd2 o-o 8 c3 a s 9 a4 h6
tbf3, since White will usually respond
with 2 d4, 2 e4, or 2 C4, transposing to
Chapters One, Five and Seven respec
tively. The only continuation with in
dependent significance is:
2 g3 e5 3 d 3 d S I
A reversed Pirc is quite safe for
Black, as long as he is careful not to
choose a sharp variation. I 've played
3 .. .fs here in the past because it is ex
tremely effective against the King's
Indian Attack, but its value is question Black h as outscored White here at
able against other set-ups. all levels of play. K.Arakhamia Grant-
127
Th e Dark Kn ight System
A. Raetsky, Bern 1995, showed Black A.Baburin, San Francisco 1997, saw
expanding and using his space advan Black's central control turning into a
tage after 10 .l:.e1 ..te6 11 exds ..txdS ! powerful tactical strike against White's
(see Game 96); while V. Frias Pablaza- king after 10 exds ll'lxds (see Game 97).
128
C h a pte r N i n e
Oth e rs
I used to "pre-move" 1 ... lt:Jc6 in online dini and I have cooked up something
blitz games - after all, it's playable new and fun against Larsen's Attack: 1
against everything, right? Eventually I b3 g6 2 ii.b2 lt:Jf6 3 e4 .i.g 7 ! ? (3 ... d6 is
was embarrassed (several times!) by a much more common) 4 es lt:Jds 5 c4
player who sometimes opened 1 b4. lt:Jf4!
After 2 bS, I did not feel like I had found
the refutation to the Orang-utan !
2 ... lt:Jes 3 ii.b2 ? ! lt:Jc4 ! ? is not bad, but 3
e4 or 3 d4 leaves Black with no excuse
for his knight placement.
In all seriousness, 1 .. lt:Jc6 is just not
.
129
tion with .ifs ) 9 l2Jcd7 10 f4 o-o 11
... ... In the games I can find, Black keeps
lDf3 c 6 ! 12 'it'b3 cxds 13 cxds dxes 14 allowing the enemy pawn to a6, which
fxes aS 1S bxas t'Llcs, White's position accelerates White's queenside initia
is approaching the later stages of a tive. The reader is better off studying
long and painful decline. lt is easy to the classic King's Indian Attacks by
criticize White's play, but he was al Fischer, Petrosian, Bronstein, Larsen,
ready worse very early; for instance, 8 etc. They go (in reverse - I have trans
t'Llf3 o-o 9 t'Llc3 dxes 10 dxes t'Lld7 11 lated here) ... t'Llf8, ...h 7-h S, ... .ifs, ... a7-
t'Lle4 b6 12 �d3 .i.b7 13 'i!Ve3 fS ! 14 exf6 a6 (when White's pawn reaches as),
exf6 1 S 0-0-0 still leaves White strug ... t'Ll8h7, ... h S-h4 (and ... h4-h 3 if possi
gling after 1S ... .:e8 or 1 S .. .fs. ble), ... t'Llg s with scary invasions coming
on the light squares or (if White has
Sokolsky's Opening (aka played h 2-h 3) scary sacrifices on the
The Orang-utan, a ka The Polish) h 3 -pawn.
1 b4 g6 2 .i.b2 t'iJf6 3 C4 .ig 7 4 t'iJf3 0-0
S e3 d6 6 d4 t'Llbd7 7 .ie2 es 8 o-o e4 9 Bird's Opening
t'Llfd2 .l:1e8 Against 1 f4, I would warn against play
ing ... d7-dS, which gives White what he
wants: the eS-square. In stead, after
1 ... g6 2 t'Llf3 .ig7 3 e3 d6 4 d4 t'Lld7,
130
Oth ers
ii.d3 e5 6 c3 'flie7 was no fun at all for animal is bad", but here the Hippo
White in A.Capaliku-J.Gombac, Nova potamus is well-suited to combat both
Gorica 2010 (see Game 98); while 5 White's kingside expansion and his
3Lc4 e6 has scored 100% for Black, most fianchettoed bishop. Black certainly
notably in A.Spichkin-D. Reinderman, made it look easy in P.Auchenberg
Rijeka 2010 (see Game 99). To.Christensen, Helsingor 1997 (see
Of course, White does not have to Game 100). An important idea is to
play a Stonewall - just as popular is a play .. .f7-f5 just when White is ready
reversed Leningrad Dutch with 3 g3, for his f4-f5 advance (though Christen
when one system has performed ex sen did not find this necessary).
tremely well for Black: 3 ... b6 ! ? 4 .ltg 2
�b7 5 0-0 e6 6 d3 ti:Je7 7 e4 d6 Other others
As for 1 ti:Jc3 and 1 g3, 1 ... ti:Jc6 should
transpose to our repertoire at some
point.
As I m entioned already, 1 ... g6 is a
good an swer to 1 e3, and also to 1 a3,
when White is challenged to find any
use for his opening move. If 1 h3, 1 h4,
1 ti:Jh 3, 1 lLla3, 1 a4, or 1 f3, you're on
your own . 1 f3 is particularly troubling
because it is not clear whether White is
intending 2 ti:Jh 3 and 3 ti:Jf2, or 2 g4!, or
with 8 ... l2Jd7, 9 ... c5, 10 ...'i!Vc7, and 2 �f2 ! and 3 �g 3 ! ! . May you be con-
11 ... 0-0 to follow. Dzindzi pithily states fronted with these problem s fre
that "any opening named after an quently.
131
C h a pter Te n
132
M iscella n eous Topics
The Light Knight Attack!? perience Black usually plays 2 ... e6,
If the Dark Knight System is so great, 2 ... c6, or 2 ... lL'lf6 in stead, transposing to
why not play it with an extra tempo? his favourite defence (French, Caro
Well, the good news is that nobody can Kann, and Alekhine's, respectively), or
stop you; e.g. 1 lL'lc3 ! ? e5 2 d3 d5 3 g 3 just 2 ... dxe4 3 t2Jxe4 and 3 ... �f5,
d4 4 lL'lb1 lL'lc6 5 i.g 2 with a reversed 3 ... lL'lf6, or 3 ... lL'ld7, with a good version
Dark Knight Pirc. Furthermore, there of Black's favourite Caro-Kann varia
are many other possible move orders tion. While these positions are all play
White can use. The bad news is that it able for White, there is little chance of
will often be Black fighting for the ad getting Black out of his comfort zone.
vantage, not White. How can a system Of course, real chess is not the same
be good for Black and not for White? as theory, and for the right player
The answer is that Black and White and/or the right opponent, the "Light
have different opening objectives. In Knight Attack" could be the perfect
the D KS Black often loses some time weapon. The grandmaster Jorg Hickl
with his knight, but this is mitigated by has played 1 g 3 e5 2 �g 2 d5 3 d3 lL'lf6 4
the fact that he has reached a position lL'lf3 i.d6 5 lL'lc3 c6 6 e4 (or 6 o-o o-o 7
where tempi are relatively unimpor e4) m any times with good results.
tant. Therefore, playing such a position
a tempo up represents only a small
gain, sometimes not enough even to
equalize.
As for 1 lL'lc3 d5 2 e4,
133
I l l u st rative Ga mes
134
Illustrative Games
An a nti-Stonewall -
rare light-squared play
Gam e 2
J.Vialatte-F .Giroux
Pa ris 2006
1 d4 lt:lc6 2 f4 d5
17 lt:ld2 lt:lc5 18 ..ic2 l:!.a6 19 .i.b2 .i.g7 Black has not signed a contract to
20 e4 f4?! 21 'it>h1 'i*'d7 22 f3?! play on the dark squares, and White
White weaken s his dark squares has already given himself a big hole on
unnecessarily. Black should have to e4. Furthermore, with Black not having
work for this. played ... lt:lf6, White is not guaranteed
22 ....l:!.fa8 2 3 lt:lab1 lt:lc8 24 lt:lc3 lt:lb6 25 an outpost on es.
..ib3 "iVf7 26 ..ic2 i.f8 27 i.a3 lt:lcxa4 28 3 lt:lf3 i.g4 4 e3 f6!?
lt:lxa4 lt:lxa4 29 i.xd6 .i.xd6 30 .l:txa4 Black may or may not get ... e7-e5 in,
lba4 31 i.xa4 "fic7 but he has certainly stopped lt:les !
5 ..ib5 'ii'd 6 6 o-o a 6 7 ..ixc6+?!
This is not going to help the situa
tion on the light squares.
7 .. JWxc6 8 c3?! lt:lh6 9 lt:lbd2
135
Th e Dark Kn ight System
d6-square as soon as possible - 9 ...lZ'lfS ! 29 ... bxcs 3 0 �xcs 'i'a s 31 'i'c6 l::!. d 6 3 2
10 .l:!.e1 lZ'ld6 ! with a pleasant advan 'ifb7+ 'it>d7 3 3 .l:!.ec1 .l:!.a8 3 4 'fibS+ �xbs
tage. Black eventually gets the position 3 S .l:.xbs .l:.xa4 36 l:tb8 lt:'lxe3 3 7 l:te1
he should have, but only with some co l:te6 38 .l:.g8 l:te7 o-1
operation from his opponent.
10 �e1 .ifs 11 'ir'e2 e6 12 .l:i.e1 i.e4 13 Can White be made to
C4 i.b4!? pay for delaying d4-d 5 - ?
Black's plan, which he will soon exe
cute, is to give up both bishops and Game 3
play "good knight versus bad bishop". M. Tratar-M.Srebrnic
This fails to take into account White's S l ove n i a n C h a m pion s h i p,
development and queenside counter Lj u blja n a 2010
play.
14 cxds exds 1S a3 ..itxf3 16 'ilfxf3 i.xd2 1 d4 d6 2 lZ'lf3 g6 3 c4 .ig7 4 lZ'lc3 lt:'lc6
17 ..itxd2 fs 18 .l:.ac1 'ife6 19 'ife2 l:!.d7 We would reach this by 1 d4 lZ'lc6 2
20 b4 cJi>b8 21 bs axbs 22 'i'xbs lZ'lf7 23 lZ'lf3 d6 3 c4 g6 4 lZ'lc3 .ig7 - in this par
i.b4 lZ'ld6 ticular game White did not even have
the option of playing 4 ds.
s ds ttJes 6 lt:'lxes .ixes 7 e4 lZ'lf6
lt seems odd to cut off the retreat
for the dark-squared bishop, but the
piece is very active where it is, and the
danger is far less than it appears.
White's lame attempt to trap the
bishop on move nine goes nowhere.
Such lovely dark square control !
8 .id3 o-o 9 lZ'le2
In addition to the threat of f2-f4,
24 'ii'a 4? White makes sure Black cannot dam
An incomprehensible positional age his structure with ... .ixc3 - how
mistake. The knight that lands on c4 is ever, this retreat is still not the most
worth far more than the "tall pawn" on accurate.
b4. However, in this roundabout way 9 ... lt:'ld7 10 h4 hs 11 lZ'lf4 lt:'lcs 12 .ic2
we see the proper fruition of Black's e6!
early advantage. White has forgotten about his de
24...lZ'lc4 2 S .ics b6 26 'ii'a 6 'ii'c 6 27 l:tb1 velopment and now stands worse.
'ifa8 28 'fibs? 'it>c8! 29 a4 White is a GM, by the way. The rest of
29 i.b4 c6 traps the queen. the game is kind of brutal .
136
Illustrative Games
Be careful on the dark squares, �f4? ..txb2 16 ..th6 1i'h4 17 �xf8 tZ'lxf8
White! 18 lZ'lf4 i.xa1 19 'il'xa1 �d7
14 exds 15 cxd s c6 16 dxc6 bxc6 17
••. White has sacrificed a pawn to
tZ'le2 �a6 18 �gs 'il'c8?1 [18 ... 'ilfb6] 19 trade off Black's powerful bishop, only
�f2 .l:tb8 20 b3 .l:!.e8 21 :e1 tZ'le6 22 i.. e 3 to find that the exchange has not
dSI 2 3 exd s cxd s 24 Itc1 �d8 2 5 ..td3 helped him. This is a typical result.
'ii'x h4+ 26 �f1 �b7 2 7 'iid 2 d4 28 �g1 20 'iic 3 :e8 21 cs �bs 22 g3 "ike7 2 3
lZ'lf4 29 tZ'lxf4 �xf4 30 'it'xa s �xf3 31 lib1 � a 6 24 �a4 'it'e4 2 5 i.c2 'i!Ve3+ 26
.!:l.c2 �g3 0-1 'ifxe3 l:txe3 27 cxd6 cxd6 28 �a4 l:ta3
29 .i.b3 lZ'ld7 30 l:i.e1 �f8
Making White pay
for delaying d4-d 5
Game 4
A.lpatov-R.Antoniewski
G e rm a n League 2011
137
Th e Dark Kn ight System
l:tc8+ <J;; e 7 35 l:tc7+ 'it>f6 36 l:txh 7 .ic4 3 7 tt::lxe4 19 'ii'e 3 .ixd4 20 �xd4 tt::l c s 21
e7 .i b s 38 h 4 d s 39 h s tt::lf 3+ 4 0 <J;;g 2 :e1 'iif8 Yz-Yz
tt::l g s 41 :h6+ rj;;x e7 42 .l:tg6 tt::lf7 43 h6
.l:ta6 0-1 White avoids tt:Jxes
12 .l:.ad1
On 12 'ii'f3, I intend 12 ... tt::l e 8 fol
lowed by 13 ... c6.
12 exds 13 exd s?!
•••
12 tt::lf6
••• Instead 13 i.. xf6 ! .ixf6 13 tt::l x ds is
A valid choice, though Black could equal .
just as comfortably have all owed the 13 .. Jlae8 14 'iif3 tt::l g4 15 h3 tt::l e s 16
exchange with 12 ... as or 12 ....U.e8. li'g3?!
13 0-0 a s 14 kte1 .:res 15 .if1 .id7 16 16 �e2 is the lesser evil .
.l:.xe8+ 'ii'x e8 17 'iid 2 tt::lfe4 18 tt::lxe4 16 .'i!Vfs!
••
138
Illustrative Games
Black is better here, but there are 11 �b3 ..tb6 12 h 3 ..tf5 13 �h2 h6 14
many mistakes in the coming compli �c471
cations - White's next move really stirs This only helps me find the right
things up. plan. After this game, the knight tran s
11 lt:Jb517 h6 18 ..te3 'Yi'xc2 19 ..td471 [19 fer became the standard way for me to
..txa7 ! ?] 19 ... a6? [19 .. .'i¥c4! 20 tt:Jxa7 combat this whole variation with 3 ds.
"i!Vxds] 20 .l:.c1? 'Yi'e4 21 .l:.fe1 'ii'xd 5 22 The alternative try 14 a4 would at least
Ci:Jxc7 'ii'x d4 2 3 tt:Jxe8 .l:txe8 24 �h1 l::td 8 have given me a little problem to deal
25 :c1 'iib 6 26 �e77 �f8 0-1 with.
Came l
L.Aitounian-J.Schuyler
La s Vega s 2008
139
Th e Dark Kn ig h t System
Gam e 8
J.Bonin-J.Schuyler
N ew Yo rk 1988
Ow!
1 d4 l2Jc6 2 l2Jf3 d6 3 d5 l2Je5 4 l2Jxe5 12 ....txb2+ 13 'itt b 1 l2Je7 14 .ie3 .ic3 15
dxe5 5 e4 e6?? [s ... l2Jf6 !] 6 .ib5+ .id7 7 'i!i'xg7 .:gs 16 'i!i'xe5 1-0
dxe6!! The carnage is unspeakable.
140
Illustrative Games
141
Th e Dark Kn ight System
�e3 �d6 49 �xe4 .l:tf2 so �e3 l:.f1 ii.e4 'iii>xg3 82 �es �f2 83 ii.d s �e3 84
If 50 .. J!xg 2 51 �f3 .J::! g 1 52 �g4 and c4 :as ss �fs g3 86 �g4 �f2 87 �4
5 3 'i.t>xg 5 draws. .l::t c s 88 �es �e3 89 �5 .l:.c8 go �es
51 �e2 i:tf8 52 �e3 �es 53 ii.c4 l:td8
142
Illustrative Games
s lLlxes dxes 6 i.bS+ i.d7 7 �d3 a6 8 22 �c2 hxg4 2 3 hxg4 :h2+ 24 l:ld2 i.cs
i..x d7+ 'ii'xd7 9 lLlc3 e6 10 ..tgs i.b4 11 2S lLlc3 i.d4 26 1:lfd1 i.e3
o-o-o o-o-o 12 f3 1i'e7 13 �c4 h6 14
i.xf6 gxf6 15 g4
143
Th e Dark Kn ight System
144
Illustrative Games
145
Th e Dark Kn ight System
Game 14
R.Valenti-V.Tkachiev
Corsica (ra p id) 1997
Game 1 5
V.Golod-E.Sutovsky
N ata nya (ra pid) 2009
146
Illustrative Games
147
Th e Dark Kn ight System
14 tt::ld b5
White should have preferred 14 The losing mistake, though White
dxe6, though Black is fine after was about waist-deep in it anyway.
14... tt:Jxe6, or even 14 ... .1txe6 1 5 tt:Jxe6 37 ... tt:Jg4 38 l::tf3 �xh3+ 39 'iiig 1 .ltg7 40
�xe6 16 .ltxcs dxcs 17 .ltxb7 .l:.ad8 18 l:td3 l:i.e3 41 lbe3 Uxe3 42 .ltc2 .l:.xg3!
�f4 'ifb6 19 .ltf3 'ii'xb2, when Black's 43 t'Dxg3 .i.d4+ 44 <;t>f1 tt::l e 3+ 0-1
strong bishop and activity m ake up for
the crippled queen side. s tt:J as!? - a strong alternative
•••
148
Illustrative Games
gam e Black has other ideas. king with 24 ...h 6 or 24 ... h 5 . Against the
6 o-o c6 7 e4 lbf6 8 'ii'e 2 cxd 5 9 exd5 text White i s right back in the game
.i.d7 10 lbfd2 b5 11 b4 lbc4 12 lbxc4 after 25 .i.d4.
bxc4 13 �xc4 .l:!.c8 25 lDb3? 'i!Ve6 26 �xe6 .Uxe6 27 .i.d4 .l:!.a3
14 'it'e2?! 28 lba 5?
lt was already dangerous to win the 2 8 lbd2 was a better try, intending
pawn, and now White chooses the 28 ... .i.xc3? 29 lbb1.
wrong retreat: 14 "ii'h 3 lbxd5 ? ! 1 5 .i.xd5 28....i.xc3 29 lbc4 l:ta1! 30 .i.xc3 .l:txf1+
.i.xa1 16 .i.xf7+ �f8 17 c3 is good for 31 �xf1 l:i.c6 32 .i.d4 .:.xc4 3 3 .i.xa7
White. l:!.xb4 34 h4 �g7 3 5 �g2 �6 36 i.e3
14...lbxd5 15 .i.xd5 .i.xa1 16 c3 o-o 17 �e5 37 �3 h5 38 �g2 �e4 39 i.c5 .l:tb5
.i.g5 .i.c6?! [17 ... .i.f5 !] 18 .i.xc6 l:txc6 19 40 i.e3 �d3 41 .i.f4 �e2 42 .i.e3 l:!.b3 43
.i.xe7 l:te8 20 'iff3 �d7 21 .i.f6 .l:.a6 22 i.c5 l:!.f3 44 .i.d4 f5 45 �g1 .l:.d3 o-1
lbd2 .Uxa2 23 �f4 �f5 24 'ifxd6
Black, given plenty of rope,
tries to hang himself but fails
Game 18
R.Aghasaryan-A.Chibukhchian
Kaj a ra n 2 0 1 1
149
Th e Dark Kn ight System
4... ttJf6 5 tiJc3 .tb4 6 .td2 o-o 7 ..te2 c6 Crushed on the d-file -
8 tiJf3 d6 9 o-o .l:!.e8 10 �b3 ..tas 11 don't let this happen to you!
.:tad1 ..tc7 12 ttJd4 'iie 7 13 h3 ttJg6 14
.l::i.fe1 ttJh4 15 g3 ttJg6 16 ..tf1 hs 11 f4?! Game 1 9
A.Hoffman-A.Ferna ndez
Ma r d e l Plata 1996
1 50
Illustrative Gam es
Gam e 2 0
H. Titz-C.Ba rlocco
Dresd e n 2 004
151
Th e Dark Kn ight System
Gam e 2 1
M.Gu revich-D.Zoler
Antwe rp 1998
8 ...'it>f8?1
8 ... g 6 9 'ii'h 4 fxe4 i s not the most Game 22
fun ever, but it does leave Black with a E.Arlandi-M.Lanzani
pawn for his trouble. The text m ove is Sa n Ma ri n o 1998
less accurate and demands m ore preci
sion from Black in the coming moves in 1 d4 l2Jc6 2 C4 e5 3 d S .i.b4+ 4 l2Jd2
order to avoid a disaster like the one in lbce7 5 a 3 .i.xd2+ 6 .i.xd2 d6 7 e4 f5 8
the game. exfs ..txf5 9 lbe2 lbf6 10 lbg3 .i.g6 11
9 f3 [9 exfS ! ] g...lbf6 10 'ii'h 4 lbg6 11 ..te2 o-o 12 o-o
1 52
Illustrative Games
Came 23
G.Grigore-P.Brochet
C reon 1999
1 53
Th e Dark Kn ig h t System
28 ... exf4 29 'ii'xf6+ �xf6 30 �xh 6 g4 ..txg4 44 'ifxg4+ �b8 45 l:tg7 �f8 46
llxh6 3 1 l:txh 6 �g8 holds the balance. 'Wd7 'ii'x d7 47 l:txd7 �f4 48 .l:td8+ �a7
29 fxgs 'it'xgs 30 .l:.hs! 'it'xg4 31 �e3 49 a4 e4 so .:tg8 .l:.h4 51 ..tfs .:tf4 52
�7 3 2 llsh4 'ii'g 7 3 3 .l:!.xh6 :xh6 34 ..td7 e3 5 3 l:.e8 1-0
�xh6 �h8 35 �g6 'ir'f8 36 'ii'g s �e8 37
�f6 'ii'e 7 3 ... d6!? an interesting
-
a lternative to 3 . . e6.
Came 24
M.Drasko-Z.Mestrovic
Bos n i a n Tea m
C h a m pio n s h i p 2003
1 d4 tt::l c6 2 d S tt::l e s 3 e4 d6
Previously Mestrovic h ad success
fully played 3 ... e6, but m aybe he no
longer believed in it? The text m ove
38 'ii'g 6+?! winds up leading to a position consid
38 .tfs maintains more pressure. ered in Chapter Six (see 6 tt::lf3 in line B).
Presumably the mistakes here are due 4 f4 tt::ld 7 5 tt::l c 3 c6 6 tt::lf3 cxds 7 exds
to a tim e scramble. tt::l gf6 8 ..te3 g6 9 ..te2 ..tg7 10 o-o o-o
38 ...�d8 11 'i!i'd2 a6 12 a4 'ikc7 13 ..td4
1 54
Illustrative Games
13 ...4Jb6?!
The position was equal, until this
time-waster hands White an advan
tage. 13 ... 4Jcs i s better.
14 as 4Jbd7 15 4Ja4 4Je4 16 'iie 3 �xd4
17 4Jxd4 4Jef6 18 4Jc3 'ii'c s 19 l:tfd1 bS
20 axb6?! [20 b4] 20 ... 4Jxb6 21 b4 Yz-Yz
Perhaps nobody was in the m ood
for a fight - the position is equal again
anyway.
GM uses Diebl's novelty 4 exd s ! •.• 1S ... .tb4+ [1S ... 'ii'x e4 ! ?] 16 �d3 dxe4+
17 'iixe4 'ii'xe4+ 18 'it>xe4 4Jf6+ 19 �e3
Game 25 .tcs+ 20 �f3 bs 2 1 .td3?!
V.Erdos-R.Ra pport According to Houdini, White is bet
H u n ga ri a n Tea m ter after 21 �e3 ! .tb7+ 22 �e2 �xe3
C h a m pions h i p 2012 2 3 �xe3, though it hardly looks like a
safe advantage. Then again, the text
1 d 4 4Jc6 2 d s tZJes 3 e 4 e 6 4 f4 exd s 5 move doesn't look safe either, and
fxes 'ifh4+ 6 'it>e2 'ii'h s+ 7 �d2 'ii'h 6+ 8 White clearly has no advantage there.
�c3 'ii'c 6+ 9 �d2 'iVh6+ 10 �d3 'ii'a 6+ 23 ... .tb7+ 22 �g3 .td6+ 23 �f2 4Jg4+
11 �d2 'i'h6+ 12 �e1 'i'h4+ 13 �d2 24 �1 4Jxh2+ 25 �f2 4Jg4+ 26 �1
4Jh2+ 2 7 �f2 4Jg4+ Yz-Yz
An alternative to 4 fxe6!? -
...
Gam e 2 6
S.Gordon-N.Short
B riti s h C h a m pio n s h i p,
S h effie l d 2011
1 d4 4Jc6
I can't help but get the feeling these I must say, this warms my heart.
guys are m essing around. Black's next 2 dS tZJes 3 e4 e6 4 dxe6 dxe6 5 'i'xd8+
constitutes a risky attempt to play for 'it>xd8 6 f4 4Jc6 7 c3 .tcs 8 4Jf3 as 9 a4
the win . 4Jh6
13 . .Ji'f4+ 1 4 � c 3 'ii'xes+ 15 'iVd4 In this position, 9 ... 4Jf6 is also good.
1 55
Th e Dark Kn ight System
However, it is nice for this knight to 42 l:.td3 l:!.ab8 43 lhd2 l:ta8 44 'iii>d 1 lt:lg7
have a stable home, controlling impor 45 ri;c2 lt:le8 46 'iio> b2 lt:lg7 4 7 lt:lc4 ..te8 48
tant squares. Short's 10 .. .f6 en sures 'it>a3 l:.dc8 49 lt:lbs i.c6 so l!td4 lt:le8 51
th at he will not be squeezed on the lt:lbd6 �d8 S2 lt:lxe8 .l:.xd4 53 �xd4 Yz-Yz
kingside, as does 14 ...h s. Wow, that was boring ! Are you sold
on 4 .. .fxe6 yet?
A better advertisement
for 4 .. dxe6
.
Game 2 7
M.Gu revich-M.Rohde
Ph i l a d e l p h i a ( b l itz) 1989
156
Illustrative Games
�e2 tt::lf6 12 0-0-0 rJ;;e 7 13 f4?! dxe6 fxe6 6 e4 ii.c5?! [6 ... ds ! ] 7 ii.d3?!
Again 13 f3. lbh6 8 'i!Ve2 o-o 9 g3 a6?! [9 . dS ! ] 10
. .
13 ii.c6
..• tt::lc 3 b5 11 e5 ii.b7 12 tt::le4 ii.b6 13 ii.d2
tt::lfs 14 0-0-0 h6?! 15 .l::t hf1 c5?!
17 c4?
Game 2 8 White should play 17 ii.xfs !, main
A.Onischuk-I.Shkuro taining the wonderful d6-knight.
U kra i n i a n Tea m 17 bxc4 18 tt::lxc4 tt::lfe7?1 [18 ... l:tb8] 19
..•
157
Th e Dark Kn ight System
hxg7 .l:.f7 29 ltJd6 i.xd6 30 exd6 i.e4 31 The great Beliavsky has not suc
.l::tf4 i.xd3+ 3 2 'ili'xd3 'iff6? 33 l:Ih1 l:txg7 ceeded in casting any doubt on Black's
34 .l::txf5 �e6 35 �h3 1-0 idea. Now Miles should just get on with
In spite of the inaccuracies, there is his comfortable King's Indian : 7 ... g6.
feeling of inevitability about the result The c-pawn can wait.
of this gam e due to White's uncon 7 ... c6 8 c4 c5 9 g3 g6 10 ltJc3 i.g7 11 g4
tested superiority in the centre and on f4 12 i.d2 g5 13 b4 b6 14 bxc5 bxc5 15
the kingside. i.d3?!
Doubtful . 1 5 h4 looks more to the
4 i.e3 fs!? - take one point.
15 ... h5! 16 h3 �f7 17 �e2 ltJg6 18 'i!Va4
Came 29 i.f8 19 liab1 i.e7 20 i.c2
A.Beliavsky-A.Miles
E u ro pea n C h a m pio n s h i p
Sa i nt Vi n ce nt 2000
20 ... hxg4?!
I suppose it's a draw either way, but
there is no disadvantage to maintain
ing the favourable tension . After this,
I 'm not sure who's playing for the win,
5 f3 or why.
White resigns himself to allowing 21 hxg4 .:.xh1 22 l::tx h1 i.d7 23 "ii'a 3
Black a good King's Indian with an l:i.b8 2 4 tt:lb1 'ili'b6 25 i.c3 l:l h 8 26 .l:g1
early and easy .. .f7-fS. In stead, if 5 ltJc3 i.c8 27 ltJd2 'ii'a 6 28 'ii'x a6 i.xa6 29
tt:lf6 6 tt:lf3 d6 7 exfs c6 (7 ... a6 ! ?), l:tb1 i.d8 30 i.a4 i.c7 31 i.b5 i.c8 3 2
White's centre disintegrates. (In fact, ltJf1 <;t;e7 3 3 i.c6 ltJh4 34 ltJd2 ltJg2 3 5
White can still play for an advantage, i.b7 i.d7 36 i.a6 i.b6 37 ltJf1 ltJh4 3 8
but it is easy to see why thi s was not i.e1 ltJg2 39 i.d2 ltJh4 4 0 i.b5 i.c8 41
appealing.) i.e1 ltJg2 42 i.d2 ltJh4 43 l:f.b3 'it>d8 44
5 ...tt:lf6 6 liJh 3 d6 7 ltJf2 i.c6 Yz-Yz
1 58
Illustrative Gam es
1 59
Th e Dark Kn ight System
1 60
Illustrative Games
11 o-o bs 18 a3 t2Jc8 19 .l:[fd1 t2Jh6 20 Yikes! Black sure has a lot of centre
tbf1 rri;;e 7 21 t2Je3 pawn s !
21 a4. 30 l:ta1 rri;;d 6 31 l:!.e2 d4 3 2 'ri;;f2 cs 3 3 c 3
21 tbb6 22 l:tac1 t2Jg4 23 i.xg4 i.xg4
••• li.d7 34 l:!. d 2 'lii' e 6 3 5 rri;;e 2 l:.ab8 36 'lii'e 1
24 .l':!.d3 li.a4
ganize a pawn break, but he is showing 40 .l:txb2 :xb2 41 a4 .l:i.g2 42 .l:.a3 rri;;d 6
no inclination to do so. 43 a s rri;;c 7 44 a6 '>t>b8 45 a7+ rri;;a 8 46
24 t2Jc4 25 t2Jxc4?!
••• .l:ta6 .l:i.xg3 47 !:txf6 rri;;x a7 0-1
Now Black h as b2 as a target. The
immediate 2 5 :Lb1 was better.
25 bxc4 26 l:!.e3 .l:i.hb8 27 .l:i.b1 l:.b7 28
••• Came 33
f3 i.e6 29 tbd1 d S L.Christiansen-J.Benjamin
U S C h a m pion s h i p,
Seattle 2000
1 61
Th e Dark Kn ight System
1 62
Illustrative Games
1 63
Th e Dark Kn ig h t System
'it>d6 47 <;itd3 wcs 48 'it>d2 i.c6 49 'iii>d 3 of worms he opened up on the a-file,
i.bS+ 50 'it>d2 i.f1 51 g3 fxg3 52 hxg3 but his remaining pieces are not very
hs 53 tbd1 gs 54 tbe3 i.bs ss tbfs i.d7 well co-ordinated. Suddenly Black
56 lbe3 bs 57 Wd3 h4 58 gxh4 gxh4 59 starts attacking. All the attacking
a3 h3 60 tbf1 i.fS+ 61 'it>c3 b4+ 62 axb4+ m oves are typical of the variation, and
axb4+ 63 Wd2 Wd4 64 tbg3 i.d3 65 tbh1 the rooks are not needed.
h2 66 tbg3 i.f1 67 tbfS+ 'it>cs 0-1 tbg4 21 i.e1 tbf4 22 i.c2?
The only satisfactory defence is 2 2
i.f1, though Black will still have some
Gam e 35 initiative.
B.Perrusset-I.Moul l ier 22 .. .'�f6!
Pa ris 2005
1 64
Illustrative Games
20 I;iae1?
White decides that with three f
pawns, he won't miss one. However,
even tripled pawns control squares,
and the f4-pawn is particularly impor
tant since it covers weak dark squares.
Far better to give up the fs-pawn which
is one m ore obstruction to White's sad
bishop (though in fact White should
not be eager to give up any of his
13 lLle3?! pawns). After this his gam e slowly de
Here 13 g3 i s best, which certainly teriorates.
makes it clear why White needs to omit 20...'ii'xf4 21 'il¥e4 'il¥g5+ 22 'il¥g4 'il¥f6 23
h2-h 3 . l:[e4 l:f.ae8 24 .l:i.fe1 l:.xe4 25 .l:i.xe4
13 ...lt::lgf4 14 tt:Jf5?! Vi'f6?! Since the f4-pawn's disappearance
At some cost, White has arranged White h as had to be careful not to al
not to be checkmated on the kingside. low ... i.d4 and ... i.es, but he can't de
Black should slow down and play fend everything.
14 ... i.xfs 15 exfs e4 or 1 5 . J:tfe8 with
. 25 .. .'ii' b 2 26 i.e2?! Vi'xa3 27 f6?
positional advantages based on space, Tim e trouble?
activity, and White's weak fs-pawn. 27 ...'ii'a 1+ 28 �g2 Vi'xf6 29 f3 'i!Vd8 30
15 g3! i.xfS 16 exf5 e4 17 i.xf4 lt:Jxf4 'i!Vf4 g6 31 i.d3 �g7 32 h4 h 5 3 3 'ili'g5
18 gxf4 exf3 19 i.xf3 .U.e8 34 'ii'x d8 .l:r.xd8 3 5 .l:!.e7 i.b8 36 b5
You don't see this every day. Black i s axb5 3 7 cxb5 �f6 38 l:te4 i.a7 39 �f4+
still slightly better, and his position is �g7 40 .Ue4 i.b6 41 �g3 .l:ra8 42 �4
simpler to play, as we will see. l:.a3 43 i.e2 i.f2 44 .l:r.e7 lla4+ 45 .l:!.e4
1 65
Th e Dark Kn ig h t System
Gam e 3 7
G.Kaidanov-A.Miles
Pa l m a d e M a l l o rca 1989
Game 38
The strongest players are much S.Brudno-J.Benjamin
more likely to do thi s than look to Boston 2001
blockade the queen side.
10 a3 ..tcs 11 lt'lf1?! bs! 1 d4 lt'lf6 2 c4 lt'lc6 3 lt'lc3 es 4 dS lt'le7 5
White's slow m anoeuvring does not e4 lt'lg6 6 .i.e3 .i.b4 7 f3 .i.xc3+ 8 bxc3
take into account Black's option to d6 9 cs o-o 10 .i.d3 lt'ld7! 11 cxd6 cxd6
open the position up. Black was better 12 lt'le2 'if as 13 o-o lt'lcs 14 ..tc4 .id7 15
already, but now White is in serious .i.b3 l:[ac8 16 g3 fs
trouble. After con structing an ideal position
12 b4 .i.d4! on the queen side, Benjamin opens up
That square again ! When will White the second front. The m ore I look at the
learn not to play c2-c4 ? - Dark Knight System, the more I'm
13 .i.xd4 exd4 14 'ifxd4 bxc4 15 d6 'ii'e s struck by how often it is Black who h as
Understandably, White tries to close the convenient pawn breaks, and how
things again. Miles has correctly seen useful that is.
1 66
Illustrative Games
23 ... bxcs?
This g ame was a long time ago, but
I remember that I never even con sid
ered the right move, which is 2 3 ... as !,
even if Black i s actually not much bet-
167
Th e Dark Kn ig h t System
ter after 24 lLlg 3 ! lLJh4 2 5 lLlf1 h6 26 53 Wc2 lLJxf7 54 'ifilxb3 e4 5 5 Wc3 Wf2
lLle3. Still, White must find these s6 lLle7 gs 57 lLlfs e3 0-1
moves. After the text, Black gets all
tangled up and is lucky to extricate
himself. The rest of the gam e is full of Game 40
inaccuracies, most of which I will not I.Jelen-E.Dizdarevic
bother to mention. Lj u b lj a n a 1992
24 .:!.a6 .l:td8 2 5 .l:tca2 [25 b4! c4 26 l:l.c6]
2S ....l:Ib7 26 .U2a3 .l::.d d7 27 lLJc3 [27 1 d4 lLJf6 2 c4 lLJc6 3 lLJc3 es 4 dS lLJe7 5
.l:.c6 !] 2 7 ...'it>f8 28 lLla4 lLJf4 29 lLlb2 We7 e4 lLJg6 6 i.. e 3 i.. b4 7 f3 i..x c3+ 8 bxc3
30 �f2 fs 31 g3 lLJhs 32 lLJc4 fxe4 3 3 d6 9 cs 0-0 10 h4?!
fxe4 lLlf6 34 lLJa s lib6 3 5 lLJc6+ 'iii>f7 36 White must attack or Black will
We3 lLJg4+ 3 7 �d3 lLJxh2 38 .:.xa7 .l:!.xa7 tighten the screws one by one. How
39 l:!.xa7+ Wf6 40 'ifilc4 lLlf1 41 l:td7 ever, thi s is too soon and too crude.
lLJxg3 42 ltxd6+ �gs 43 l:.d7 g6 44 1o ... lLJhs 11 ..tf2 c6!
.l:.xh7 lLJxe4 45 l:.e7 �f4 46 b4 cxb4 47
lLJxb4 lLlf2 48 l':!.f7+ �g3 49 WcS l:.b8 so
lLJc6??
1 68
Illustrative Games
18 .. ."i!Vxe4+, winning the rook. move, but 13 .. .fs can be played, and so I
11 gs 18 'ii b 3 i.d7 19 li:Je2
••• believe it should be played.
14 .l:.b1 as 1s f4?!
In his annotations, Palliser gives thi s
an ' ! ' and indicates that i t i s the begin
ning of the end for Chevallier.
1S exf4 16 gxf4
•••
1 9 lt:Jeg6?
..•
lt:Je2 'ifa s+ 30 lt:Jc3 lt:Je2 3 1 .l:i.fc1 lt:Jxc1 After this Black i s a little worse.
32 J::tx c1 Wg8 33 f4? Meanwhile, should Black find 16 ...h s,
White isn't having any fun at all, he is a little better! In one stroke, Black
but thi s only m akes things worse. stabilizes his knight, clamps down on
33 exf4 34 es? fxes 3S d6+ Wf8 36
.•• White's weak g4-square, and immobi
"Yi'b1 'it>g8 3 7 1lt'b3+ 'ifi>f8 38 �b1 'ilkcs o-1 lizes White's i solated h-pawn. Of
course, with a small edge (or even a
Black loses because small disadvantage) Karpov will proba
he's playing Karpov bly win - he does outrate Chevallier by
430 points.
Gam e 4 1 17 .i.xcs ! bxcs 18 hs lt:Jf8 19 .l::!.g 1 f6 20
A.Karpov-D.Chevallier .i.d3 li:Jd7 21 'ii'e 2 i.a6?
Fra nce 1993 Black's situation will not improve
with his king in the centre. For better or
1 d4 li:Jf6 2 c4 lt:Jc6 3 lt:Jc3 es 4 ds lt:Je7 s for worse, he must castle and try for
e4 lt:Jg6 6 i.e3 i.b4 7 f3 i.xc3+ 8 bxc3 .. .f6-fs.
d6 9 �d2 lt:Jd7 10 h4 h6 11 g3 b6 12 22 tt:lg4 lt:Jb6 2 3 lt:Je3 i.c8 24 lli'g2 l:t.g8
tt:lh3 tt:Jcs 13 lt:Jf2 'i!Vd7 2S 'i!Vg6+ 'iii>d 8?! [2S ... Wf8] 26 'it>d2 .i.d7
There i s nothing wrong with this 27 �h7 'i!Vf8 28 .l:f.g2 Wc8 29 l:i.bg1 'it>b7
1 69
Th e Dark Kn ight System
Gam e 4 2
W.M.Buehi-J.Benjam i n This i s why I like doubling White's
Re no 1999 pawn s so much - it provides a ready
m ade target on the other front if the
1 d4 lt::lf6 2 c4 lbc6 3 lt::l c 3 es 4 d S lt::le 7 S king side play is running out of steam .
e4 lt::lg 6 6 lt::lf3 .i.b4 7 i.. d 3 i..xc3+ 8 bxc3 17 fxes 'it'xc4 18 .l:!.d1 tt:Jxes 19 l:d4 Vies
d6 9 o-o o-o 10 h3? tt:Jhs 11 'it>h2!? 20 lbe2 .l:!.ae8 21 lbf4 lt:Jg6 22 tt:Jxg6
i..xg6 23 .i.gs .l:i.e2 24 a4?! .:f.ff2 2S l:.g1
.i.e4 0-1
Game 43
A.Hahn-J.Bonin
New Yo rk (ra pid) 2003
1 70
Illustrative Games
Cam e 44
E.Schiendorfer
D.Recuero Guerra
E u ro pea n J u n ior C h ' s h i ps,
H e rceg N ovi 2006
1 71
Th e Dark Kn ight System
17 �bs?!
..•
1 72
Illustrative Games
position, trading off White's l ame light tion where the opponent has put most
squared bishop and allowing White to of his pawns on one colour (especially
castle without dropping the h 3 -pawn . in the vicinity of his king).
After the text m ove, White has equal 24 'itc2 .l:i.c8
ized - though not for long. Houdini points out 24 ... g 5 ! 25 tt:Jg 2
(the knight's idiot square; but if 2 5 tt:Je2
h 5 !) 25 .. Jlc8 and 26 ...tt:Jxe4. In the
game White's knight becomes a real
pain in the keister.
25 'itb2 tt:Jxe4 26 tZ:Ih5 tt:Jc3 27 l:tbc1 g6
28 "ir'd2
H ad White played 27 :fell he could
have gone for 2 8 .l:!.xc3 �xc3 29 'itxc3
gxh 5 30 gxh 5 with no fear of 30 ....txh 3.
28...tt:Je2+ 29 Wg2 .l:!.xc1 30 'i¥xh6 tZ:If4+
31 'ii'xf4 .l:!.xf1 3 2 'itxf1 gxh 5 3 3 'ti'h6 f5
14 .tg5?! 34 'ti'g6+ 'ith8?!
14 'itf1! 'ii'c 8 1 5 Wg 2 was correct. Christiansen, who has been defend
14 ...iVc8 15 g4 c6 16 tt:Je2 h6 17 .td2 ing perfectly until now, slips. lt is not
cxd5 18 cxd 5 'ilr'd8 19 tt:Jg3 .tb6 20 o-o good to allow 'i!Vxh 5 with check, and by
tZ:If4! 21 .txf4 exf4 22 tZ:Ih5 tZ:If6 23 tZ:Ixf4 winning the pawn White defends his
.l:!.e8 own peon on h 3 .
35 gxf5 .l:.f8 36 'itxh 5+ 'itg8 3 7 1i'g6+
'ith8 38 'ili'h6+ 'itg8 39 f6
1 73
Th e Dark Kn ight System
and lets Hi.ibner off the hook 19 ... tt::lfs ! is better, forcing 20 tt::lf1
(or 20 tt::l e4, which is n ot very good ei
Gam e 4 7 ther), because 20 .l:.g 1? loses to tt::l x g 3 !
R.H i.ibner-V.Hort 2 1 l:txg 3 l:txd2 ! 22 <iii>x d2 i.xf4+ and
Germa n League 1984 2 3 ... i.xg 3 .
20 tt::ld c4 tt::lxc4 21 tt::lxc4 i.d3
1 e4 tt::lc 6 2 d4 e5 3 dxe5 tt::lxe5 4 f4 tt::lc6 There is still some play, but Black
5 i.e3 i.b4+ [s ... ds!] 6 c3 i.a5 1 tt::lf3 has no real advantage.
tt::lf6 s e5 [8 'i¥a4 ! ?] s tt::l g4 9 i.c5? d6!
••• 22 b3 .l:.ab8 2 3 Wf2 b5 24 tt::l e 3 c4 25
10 exd6 o-o tt::l d 5 i.b6+ 26 'it>g2 wfs 2 7 1Ihe1 .l:te8 28
b4 i.f5 29 .l:lxe8+ l1xe8 30 a4 bxa4 31
.l:.xa4 h6 3 2 gxh6 gxh6 3 3 tt::lx b6 axb6
34 'it>f2 lidS
By now it is White who is trying for
a win, though it is unlikely.
35 .l:la8 .:txa8 36 i.xa8 We7 37 'it>e3 'it>d6
38 'it>d4 b5 39 i.d 5 f6 40 i.t1 i.d7 41
'it>e4 cJ;;e 7 42 i.g6 i.c6+ 43 'it>d4 'iii>d 6 44
i.h5 f5 45 i.g6 i.d7 46 i.f7 i.cs 47
i.eS i.a6 48 'it>e3 'it>e7 49 i.g6 i.c8 50
'it>d4 'iii>d 6 51 i.e8 i.a6 Yz-Yz
1 74
Illustrative Games
.l:!.he1?
Game 4 8 25 ..ixc 3 ! i.xc3 26 lld1+ and Black
M.Orso-G.Bordas still has to show good technique.
B u d a pest 2000 25 .. Jla1+ 26 Wc2 .l:l1a2+ 27 rJitd3 l:!.d2+
28 'iii>e 4 f5+ 29 'iii>xf5 .l::!.x d4 30 c5 .i.xc5
1 e4 lt:lc6 2 d4 e5 3 dxe5 lt:lxe5 4 f4 lt:lc6 31 l::tx c3 l:tf8+ 32 rJitg5 i.e7+ o-1
5 i.c4 l'Llf6 6 l'Llc3 i.b4 1 e5 d5 8 exf6
dxc4 9 �e2+ 'iii>f8!? [9 ... i.e6 ! ] 10 .ie3
Had White been tempted by 10 Gam e 49
fxg7+? <it>xg7, Black's l ast move would S.Fedorchuk-A.Miles
have been amply rewarded. The rook's E u ro pea n C h a m pio n s h i p,
rapid arrival on the e-file is a serious Oh rid 2001
problem for White.
10 .. .'i¥xf6 11 o-o-o i.e6 1 e4 l'Llc6 2 d4 e5 3 dxe5 lt:lxe5 4 l'Llc3
The inconvenience suffered by i.c5 5 f4 lt:lc6 6 lLlf3 d6 1 lt:la4 i.b6 8
Black's king is not enough to make up .id3 d5?!
for White's missing pawn and shortage
of light-squared bishops.
12 lt:le4 'iWe7?! [12 .. .'i¥fs] 13 l'Llf3 c3 14
b3?! [14 bxc3] 14 ... a5 15 'it>b1 a4 16
'iWb5??
White is now lost. 16 .i.f2 is essen
tial, though Black has a clear edge.
16 ...axb3 17 axb3
1 75
Th e Dark Kn ight System
Gam e 50
Goh Wei Ming-F.Bellini
T u r i n O lym p i a d 2006
1 76
Illustrative Games
26 ...ttJcd7?!
Black is reluctant to put any pawns
on light squares for long-term posi
tional reasons, but White won't reach
the long term after 26 ... ds ! 27 .:te1
tbce4+ 2 8 i.xe4 tLlxe4+ 29 lLlxe4 dxe4,
This Basmaniac-esque move i s a bit which exposes the ludicrous placement
funny-looking at first, but if we think of White's king ; e.g. 30 �c1 �as 3 1
about it, it is the dark squares that �b1 :i.g 2 3 2 'iWf1 'i¥d2 3 3 l:tc1 �xe3 34
need coverage since we have a light �xg 2 'ii'xf4 with a winning position.
squared bishop and we are about to Instead, soon comes an avalanche of
have a knight posted on the lovely es blunders that is presumably time
outpost. To be a little more concrete, it induced.
is beneficial to take the f4-square from You know what? You don't want to
White's queen. see the rest of this. lt's ridiculous and
10 ttJ1c3 lLle5 11 'ii'e 2 d6 12 h3 c6 13 embarrassing.
lLld4 tLlf6 14 o-o-o 'ii'e 7 15 i¥f2 ..te6 16 Black eventually won the game on
..te2 o-o-o 17 tLlf5 ..txf5 18 exf5 move 81.
1 77
Th e Dark Kn ight System
12 exfs .i.xfs
12 ... ltJxfS is equally good.
13 0-0 0-0
13 ... 0-0-0 is at least as good too. 22 ... tLlf2?!
14 tLla3 .l:tae8 15 .l:tae1 'it>h8 Black keeps an edge with 22 ... ltJxd2
it's time to take away White's op 23 .l:.xd2 .l:t.af8 or 23 �xd2 .i.xg4.
tion to snap off the bishop; i.e. 23 .l:ldf1 .Uaf8 24 .i.gs ltJxg4 25 l:.xf8
1S ... .i.e6. .l::!.xf8 26 .i.xd s .i.xds 27 .l::!.xg4 .l:t.f1+ 28
16 ltJac2 'if'h6 17 .i.d3 .i.xd3 18 'if'xd3 'it>d2 l:tf2+ 29 'iii> c 3 a s 30 .i.h4 l:tf3+ 31
ltJg6 19 ttJd s?! ltJces 20 �d4 c6 21 �d2 .i.f7 3 2 d s .i.xds 3 3 l:txg6 l:.b3
ltJde3?! [21 f4!] 21 ...ltJf4 22 �h1 'ii'g 6?! Yz-Yz
1 78
Illustrative Games
1 79
Th e Dark Kn ight System
17 lba3
If 17 l:txds?! lbf6 18 .l:.d1 .l:f.e8+ 19 Game 5 6
�e2 �a6, White regrets his pawn grab. I.Grynfeld-A.Bisguier
11 ... lbf6 18 lDbS+ �d7 19 �e2 lieS 20 H e l s i n ki O lym p i a d 1 9 5 2
�f1 l:r.es 21 h4 lbe4 22 l1h3 lbd2+ 23
�e1 �a6 24 a4 lbc4 2 5 b3 lbd6 26 1 e4 e5
.l::r. h d3 d4! 2 7 cxd4?! Once, when GM Bisguier was ana
White's rooks will not enjoy lan lysing one of my games, he said he'd
guishing behind the i solated d-pawn . known all about ...lbc6 and ... e7-e5
However, the position i s n ot rich since long before I'd been born. Thi s
enough to offer Black good winning m akes sense because he was coached
chances (in GM play). by Alexander Kevitz, by whose nam e
27 .. J!e4 28 g4 �f4 29 �f1 lbxbs 30 this system i s known .
axbs �xbs 31 �3d2! 2 lbf3 tbc6 3 d4 exd4 4 lbxd4 �cs 5
lbxc6 'iVf6 6 lli'f3 'iVxf3 7 gxf3 bxc6 8
i.e3 �xe3 9 fxe3
1 80
Illustrative Gam es
1 81
Th e Dark Kn ight System
20 ...l:thg8?
Black is nearly wmnmg with
20 ... tt:Jxd3 2 1 cxd3 'il¥f2 22 lt:Jf4 .l:!.hg8,
because White's queen, knight, and
king's rook are stuck in a very un
healthy eo-dependent relation ship.
21 �f1 l:!.d6 22 'ir'e7 'il¥a 5?! 23 llxh 5 l::[d 7
24 'ii'f6 .l:td6 2 5 'fie7 .l:i.d7 26 'ii'f6 J::r.d 6 27
'fie7 .l:!.d7 Yz-Yz 12 ... 1i.d7?! 13 lt:Ja4?!
Instead of repeating moves, Black Out of book, White m akes a poor
was much better after 27 ... tt:Jxd3 28 choice. As we will see, Black does not
cxd3 (or 2 8 'ii'xf7 lt:Jf2+!) 28 .. .'fVxa2 29 even have to allow his bishop to be
lt:Jf4 �xd3 ! (and not 30 tt:Jxd3 ?? 'ii'xg 2 forced off the strong diagonal.
m ate).
Cam e 58
B.Sultimov-N.Pokazanjev
Russia 2007
1 82
Illustrative Games
Game 59
D.Von Wa ntoch
Rekowski-J.Peric
Yugoslav J u n ior C h ' s h i ps,
Tivat 2001
10 i.. h 2?!
10 hxg4 hxg 3 11 fxg 3 d6 is relatively
best, though still with advantage to
Black.
1o tt:Jxh2 11 'it>xh2 d6 12 i.. b 5 dxe5 13
.••
1 83
Th e Dark Knight System
1 84
Illustrative Games
story of the old lady who swallowed a s ... a6 i s al so fine, or even s ...Si.g 7 6
fly, and a spider to catch the fly, and a dS a6 7 Si.a4 bS 8 dxc6 bxa4 9 'ii'x a4 o-o
bird to catch the spider, etc. White pre with compensation .
vents ... .i.h6 for now, but the new prob 6 1i'e2 i.. g 7 7 0-0 0-0 8 l:td1 'i!ke8 9 C4 e5
lems are bigger than the old ones. 10 dxe5 tt::lx e5 11 ..txd7?! tt::l xf3+ 12
23 ... exf4 24 'iixf4 b5 2 5 tt::l e 3 Si.e5 'ifxf3 tt::l x d7
Game 6 1
R.Zelcic-A.Miles
P u l a 1994
1 85
Th e Dark Kn ight System
1 86
Illustrative Games
Game 63
A.Zapata-A.Miles
Mata nzas 199 5
187
Th e Dark Kn ig h t System
1 88
Illustrative Games
46 ... exf3 47 .l::r.x d6+ �e7 48 'ii'a 7+ �f6 i.c4 li:Jc6 6 d S li:Jb8 7 h3 o-o 8 o-o c6
49 l:!.xe6+ l:.xe6 so 'ifxcs �g6 51 'ifd4 [8 ... e s ! ] 9 a4! as 10 l:!.e1 tt::lfd71
'ii'fs 52 g4 hxg4 5 3 hxg4 'iff7? 54 .l:Ul?l
And h ere 54 b4! .
5 4...l::.f6 s s 'it>f2?1 'it'b3 56 'ii'd 3+ Wg7 57
.:d1 �xb2+ 58 lld2 'fibS 59 'i!id7+ 'iii> h 6
60 �f1 'ii'b 1+ 61 l:!.d1 'ii'e 4 62 'ii'd 2??
After thi s move White i s losing. 62
:d2 held the bal ance.
62 ...f2 63 �d s 'ii'xg4 64 'ii'h 1+ 'it>g6
[64 ... �g 7 ! ] 6S l:d4
Gam e 6 6 Game 6 7
N.Praznik-A.Beliavsky K.Nemcova-F .Oiafsson
B l ed 1999 Mari a n s ke La zne 2008
1 89
Th e Dark Kn ight System
190
Illustrative Gam es
17 'i!Vc1
Very soon Black has all the play he
needs. White can make him work
harder with 17 a4.
17 .. -'i!Vcs 18 a3 bs 19 tZ:ld4 tZ:lxd4 20
.i.xd4 't!Vb7 21 �f4 tbd7 22 .i.xg7 'it>xg7
White has nothing better than this 23 �d2 'i!Yb6 24 .l:.e3 �adS 25 g3 tZ:lf6 26
l am e try. Black h as equalized already. .i.g2 a s 27 b3 .l::. b 8 28 a4 bxa4?1 [28 ...b4]
10 a6 11 o-o .i.b7 12 I:l.fe1 tbbd7 13
••• 29 tZ:lxa4 'ii'b4 30 tl:\c3 tl:\d7 31 Wh2 tZ:lcs
dxc6 .i.xc6 14 .i.d3 .i.b7 15 .i.h6 .:!.c8 16 32 f4 f6?1 33 'i!Ve2?l [33 fs !] 33 ....i.f7 34
�ad1 1i'b6 17 ..txg7 Wxg7 18 :e3 tbes tZ:lds .i.xds 35 exd s .l:be3 36 'ili'xe3 l:tf8
19 tbxes dxes 20 .l::i.e 2 .:!.fd8 37 fs gs 38 �e2 tbd7 39 l:ta1 tbes 40
By now Black is a little better, but l:!.a4 �c3 41 l:!.e4 .:!.b8 42 �h s .l::!. b4 43
the players soon decided to call it a day. .:!.e2 a4 44 'iVeS axb3 Yz-Yz
21 'i!Ve3 'ifd4 22 .:!.de1 tbhs Yz-Yz
Game 70
Game 69 B.Chatalbashev-M.Popchev
Y.Gruenfeld-I.Smirin Caca k 1991
I s ra e l i Tea m
C h a m pion s h i p 1997 1 d 4 d6 2 e 4 tbf6 3 tZ:lc3 g6 4 tZ:lf3 i..g 7 5
.i.e2 o-o 6 o-o i.. g4 7 i..e 3 tZ:lc6 8 d S
1 e 4 d6 2 d 4 tbf6 3 tbc3 g 6 4 tbf3 .i.g7 5 tbb8 9 tZ:ld4 .i.xe2 1 0 �xe2 cs 11 tbf3
.i.e2 o-o 6 o-o tbc6 7 d s tbbs 8 h3 es 9 'i'ib6 12 I:tab1 1lia6
dxe6 .i.xe6 10 ..tgs h6 11 .i.e3 tbc6 12 Chemin points out that this queen
191
The Dark Knight System
m anoeuvre is both typical and strong. tt'lc3 o-o 6 o-o ..ig4 7 ..te3 tt'lc6 8 d 5 tt'lb8
13 'iVd2 tt'lg4 14 ..tg5 .:tes 15 'ii'f4 tt'le5 9 tt'lg5 ..txe2 10 'ii'xe2 c6 11 �ad1 'ii'a 5
16 tt'lxe5 ..txe5 17 "YWh4 "iWc4 12 f4 'ii'a 6 13 'ii'f3 tt'lbd7 14 li'h3 h61 15
tt'lf3 h51 16 ..id4?1 c5 17 ..ixf6 tt'lxf6 18
'ii' h 4?1 b5
18 .l::!. b e1
White gains nothing by grabbing
the pawn : 18 ..txe7 tt'ld7 19 f4 ..id4+ 20 White hasn't gotten anything done
Wh 1 ..txc3 21 bxc3 "YWxe4 is equal . on the kingside, but he has serious
18 ..txc3 19 bxc3 f6 20 ..ih6 tt'ld7 21
•.. problem s on the queen's wing.
'fi'g4 e6 22 f4 .Uad8 2 3 dxe6 �xe6 24 19 tt'le2?1 'ii'a 4 20 tt'lg3 'ii'x c2 21 e5 dxe5
'i!Vg3 'ii'x a2 2 5 f5 'i!Vf7 26 fxg6 hxg6 27 22 fxe5 tt'lg4 2 3 .l:l.d2 'ili'c4 24 "iix e7 I;Iae8
'i!Vxd6 �h7 28 'iff4?1 25 d6?1
lt i s tim e to extricate the bishop. A desperate try for complication s
28 l:.e5 29 ..tg5 .:tfs 30 "YWh4+ 'ifilgs 31
••• which fails after Chemin's accurate
..if4 �e6 32 llf3 g5 33 'ii'g4 lUeS 34 treatment .
..tg3 tt'le5 35 ..txe5 .l:.xe5 36 h4?1 'ii'e 6 2 5 ... ..th61 26 b3 'ifb4 2 7 e6 l:Ixe7 28
37 'ii'h 5 �g7 38 .l::tef1?? dxe7 ..txd2 29 exf7+ �xf7 30 exf8'ii'+
Oops! Black had an advantage, but 'iitxf8 31 tt'lxd2+ 'lt>g7 3 2 tt'lge4 'i!Vd4+ 3 3
thi s throws the game. 'it> h 1 c 4 34 bxc4 bxc4 3 5 h3 tt'le3 36 .l:r.f4
38 .l:th8 39 'ii'x h8+ 'ifilxh8 40 .Uxf6 'ir'c4
••. c3 3 7 tt'lf3 'ii'd 1+ 38 'ifilh2 tt'lf1+ 39 'iii> g 1
0-1 tt'ld2+ 40 'it>f2 tt'lxe4+ 41 .l:f.xe4 c2 0-1
Gam e 71 Game 72
R.Ziatdinov-A.Chernin So.Polgar-J.Fries Nielsen
N ew York Open 1998 R i m a vs ka Sobota 1991
192
Illustrative Games
19 �d1?!
Asking for trouble. 19 .l:!.b2 was cor
rect.
19 ... a4 20 'ii'h 6 f6 21 i.c4+ e6 22 dS
tbe7?
Since White has counterplay in ei
ther case, it does not help to give up
the pawn. Hence 2 2 ... tt:Jd8 ! .
2 3 dxe6 i.c6 2 4 tt:Jd4 axb3 2S tt:Jxc6
bxc6 26 i.xb3 .l:tab8 27 :e1?! The knight can stay here as long as
The way to extricate the king i s 27 it wants, but it does not have impor
We2 ! l:txb3 2 8 axb3 'itxb3 29 l:thd1. tant targets in Black's position.
27 ....l:i.xb3 28 axb3 'itxb3+ 29 'ite2 'itc4+ 16 ... c6 17 i.c4 'ii' b 6 18 b3 i.fs 19 exd6
30 l:td3? exd6 20 g4 i.xg4 21 .l:.dg1?! i.fs 2 2
Walking into a dangerous pin . 30 tb h 3 'itb4 2 3 'itxb4 .l:lxb4 24 .l:tg3 as 2 S
193
Th e Dark Kn ig h t System
..ixd s cxd s 26 .l:.d1 :c8 21 .l:!.d2 a4 28 bled! The half-open b-file co-ordinates
�b2 axb3 29 cxb3 .l:.cc4 30 ltJf4 .l:.xd4 well with the dark-squared bishop, and
31 .l:!.xd4 .l:!.xd4 32 ltJxh5+ 'iitf8 3 3 lt:lf6 it i s nice to have the ds-square securely
.Ud1 34 h s gxhs 3 5 lt:lxh s d4 36 l:i.gs?! guarded.
.l:!.d2+ 3 7 'iii>c 1 .l:!.xf2 38 lt:lg3 ..ig6 39 .:td s
I:. g 2 40 ltJhs .l:.f2 41 ltJg 3 l:!.f3 42 lbe2 d3
43 lt:lc3 �e7 44 b4 �e6 45 l:i.d4 .U.f2 46
ltJbs .l:.c2+ 47 'it>d1 ..ih5+ 48 'iit e 1 .l:!.e2+
49 Wf1 d2 o-1
Gam e 74
N.Ryba-J.Schuyler
Wa s h i n gto n 2012
1 e4 ltJc6 2 lbf3 d6 3 d4 ltJf6 4 ltJc3 g6 5 11 ltJxc6 bxc6 12 ..if3 ltJd7 13 'il¥d2 .Ub8
i.e2 ..ig7 6 o-o o-o 7 h 3 a6!? 14 .l:tab1 �f6 15 i.e2 li'es 16 l:!.fd1 'i!Va s
Against a lower-rated opponent lt i s often deflating to computer
there is some concern that the natural check your own games. This queen
m ove, 7 ... es, will lead to a position with m anoeuvre, which I was proud of dur
too few winning chances, in which case ing the g ame, accomplishes less than
7 ... a6 is often a good alternative. nothing after 17 b4! l:.xb4?! 18 lt:lds !
8 a4 e5 .Uxa4 19 'i!Vxas .Uxas 2 0 lt:lxc7 l:!.xe4 2 1
lt is nice for m e to have the bS l:i.xd6 with an edge for White.
square covered and his b4-square 17 i.d4 ..ixd4 18 'ifxd4 .l::t b4 19 'ifd2
weak. Therefore, if 9 dxes, Black will ttJcs
play 9 ... dxes, keeping the extra pair of Thi s will not get anything done ei
knights on the board. ther. Black is not even threatening
9 ..ie3?! 20 ... ltJxa4 because of 2 1 1:ta1 and 2 2 b3.
lt i s alm ost always a bad idea for 20 b3 ltJxe4 21 ltJxe4 .l:!.exe4 22 .i.f3 J::.e s
White to try to preserve the central 2 3 ..ixc6 ..ifs 24 i.f3 'i!Vcs 2 5 c3 l:!.b8 26
ten sion in the Dark Knight Pirc. This is b4 'il¥c4 2 7 .:!.b2 l:!.be8 28 'it'd4 'it'e6 29
especially true when he has played h2- �h2 .i.e4!
h 3 . Black i s already at least equal. lt isn't much, but I will have a little
9 ... exd4 10 ltJxd4 J:le8 something to play for after stripping
One Pirc author recomm ends White's king of its best defender. Who
10 ... .i.d7 first in this type of position, knows? The e-file might even be worth
but I prefer to have my c-pawns dou- something.
1 94
Illustrative Games
Threatening 4S ....l:!.xh 3+ 46 �g l
.l:!.xg 2+ and wins ! Somehow I over
looked both of White's defences.
45 'ii'x h7
Or 45 h4+! 'it>xh4 46 'i!Vxh 7+ 'i.t>g s 47
.lhe4, when Black has to find
47 .. Jixg 2+! 48 �xg 2 l:td2+ 49 l::t e 2 lid4!
and draws (apparently).
4S ...l:!.dd2
Thi s i s the m ove that would have
won l ast turn.
30 .i.xe4 l:f.xe4 31 Wkd3 a s 3 2 'ilid s 'if'f6 46 .l:.g1??
33 1!i'xa s 1!i'xc3 34 .l:.bb1 cs?! Snatching defeat from the jaws of a
This is another move that seemed draw. My opponent saw the correct
very strong to me when I played it, but continuation but misevaluated it: 46
White can equalize starting with 3 5 h4+ �fs 47 l:t8es+ "il/xes (47 .. .fxes 48
ft 6 ! (the text m ove is not bad, either). "ilif7+ �g4 49 .l:!e4+ .l:!f4 so 1Wxg 6+ 'i.t>xh4
34 .. Jie2 i s a better try. But by now we 51 'ii'h 6+ draws) 48 I!.xes fxe s 49 'ii'f7+
were running low on tim e to reach We4 so 1!i'b7+ is no worse for White.
move 40. 46 ...Wkc6 47 h4+ 'it>fs o-1
35 �bs cxb4 36 .:txd6 l:!.8es 37 'ii' b 8+ Not the most beautiful game, but
�g7 38 .:td8 .:tfs 39 l::tg 8+ �h6 4o 'iWf8+ the opening was a success.
'lt>g s 41 �d8+ f6 42 l:te8 l:td4 43 "ille 7?
Houdini finds 43 f4+! :fxf4 44 l:teS+
with equality. White i s lost after any Gam e 75
other move. N.J h u njhnuwala-S.Giigoric
43 ...l:Ixf2 44 :e1 lld3?? Luce r n e O lym piad 1982
195
Th e Dark Kn ig h t System
14 ... cxds 15 c3 i.. b 7 16 i.. h 6 lL'le4 17 d4 exd4 6 ttJxd4 i.. g7 7 i.e3 lL'lf6 8 'ii'd 2
'iWc1 lUeS 18 i..xg 7 �xg 7 19 ltJd2 lL'lcs o-o 9 o-o-o .:tes 10 f3 ttJxd4 11 i.xd4
20 lL'lb3 lL'le6 21 .l:tfe1 as i.. e 6 12 i..f2 a6 13 �b1 bS 14 h4 cs 15
g4?
White can play the aggressive 15
h S ! ? or the defensive 15 a3 - the text
move is too slow.
15 ... b4 16 lL'le2?!
22 i.. b s .l:.e7?!
Black does not h ave a good reason
to disconnect his rooks. However,
White's next makes it easy to fix the
problem.
2 3 i..f1? [23 a4] 2 3 ... a4 24 lL'la1 l::t d 8 2 5 16 ...1Was?
lL'lc2 f6 26 lL'lb4 1Wb6 2 7 1Wc2 ttJ c s 28 a3 Black g ets a big advantage with
l:.ed7 29 .l:te3 lL'lb3 30 .l:.ee1 :Ld6 31 1We2 16 ... ttJxe4! 17 fxe4 'ii'f6, threatening
d4 3 2 'iib s 'tWxbs 3 3 i..x bs dxc3 34 m ate and White's bishop .
.l:.xd6 .:txd6 35 bxc3 ttJcs 36 .1:te2 .l:f.d1+ 17 lL'lc1 i.xg4?! 18 i..g 2?
37 �h2 e4 38 �g3 fs 39 h4 'iW6 40 Black's sacrifice is thematic, but not
i.. e s? h6? [40 ...l:.h 1!) 41 i.. b s? [41 �h 2 ! ) quite sound. 18 fxg4 lt:'Jxe4 19 'iWf4
4 1. .Jih1 42 f4 gs 43 hxgS+ hxgs 44 lL'lc3+?! 20 bxc3 bxc3 2 1 i.. c4 is defence
fxgS+?! 'ii?xg s 45 �f2 f4 46 i.. c 6? i..xc6 and counterattack.
0-1 18 ... i.. e 6 19 'ii'x d6 .:!.ac8 20 lL'lb3 'Yi'a4 21
Since after 47 lt:'Jxc6 the reply i.. h 3? i..f8?
47 ... lL'ld3 is mate. Unnecessarily removing the bishop
from the m ain diagonal and misevalu
ating the m ost direct continuation :
Gam e 76 2 1 ... i.. xh 3 2 2 l:txh 3 c4 2 3 lL'lc5 'ifh5
G.Bastrikov-E.Gel ler threatens 24 .. J:tc6.
Ta s h kent 1958 22 'ii'h 2?! i..x h3 23 'i!Vxh 3 c4 24 lL'lc1 c3
25 i.. d 4 cxb2 26 lL'lb3 i.. g 7 27 i.xb2 'tWc6
1 e 4 es 2 lL'lf3 lL'lc6 3 lL'lc3 d6 4 h 3 g6 s 28 .l:.h2 J:.ed8 29 .l:thd�? .l:!.xd2 30 .l:.xd2
196
Illustrative Games
Game 7 7
D.Saduakassova
Art.Minasian
D u ba i 2011
197
Th e Dark Kn ight System
'iii>xg4 60 'iii> b 2 :as o-1 [22 ... ltJfs !] 23 'ti'h2 'i¥xh2+ 24 <;i;>xh2
lbxf3+ 2 5 �h1 lbd2 26 lbe8?? [26 lbe6 ! ]
26 ...ltJxc4 27 ltJxg 7 Wxg7 28 l::tfe1 f s 2 9
Gam e 78 .l:.ad1 1:1f7 30 b3 ttJes 31 lld6 lbf3 3 2
A.Sakha rov-A.Adorjan J:1ed1 .l:.e7 3 3 l:!. d 7 <3;;f7 34 <3;; g 2 f4 3 5
Soch i 1976 l:tld6 lbh4+ 36 �f1 e 3 3 7 c 4 e2+ 0-1
198
Illustrative Games
best is the first step on the road to a where 7 ... e s ! works a little better than
win. usual and 7 ... a6 works a little worse
19 .i.xgs hxgs 20 l:tad1 axb4 21 axb4 than usual.
'i.t> g 7 22 f3 l:!.a2 23 tt:'lc4 Z:.xd1 24 .i.xd1 8 dS! tt:'lb8 9 a4!
.i.a6 2 5 .i.b3 .i.xc4 26 .i.xa2 .i.xa2 27
.l:!.a1 .i.c4 28 l:i.a7 bs 29 .l:!.xc7 'it>f6
Perhaps White should hold this, but
he has a long and thankless defensive
task ahead of him. Black's task is to
penetrate with his king .
3 0 .l:!.a7 'it>e6 31 .l::t a 6+ W d 7 3 2 1:1a7+ Wd6
33 .l:!.a6+ tt:'lc6 34 <;t>f2 <;t>d7 35 Wg3 tt:'ld8
36 h4 gxh4+ 37 �xh4 �e7 38 �g3 tt:'le6
39 �f2 tt:'lf4 40 g3 tt:'le2 41 lla3 �f6 42
'lt>g 2 �gs
And now White is lost. How did that Since a2-a4 is normally m et by ... a7-
happen ? as, it stands to reason that Black's a
43 'it>f2 fs 44 exfs gxfs 45 'iio> g 2 f4 46 pawn is misplaced.
gxf4+ �xf4 47 '>W2 tt:'lc1 48 .l::ta 8 tt:'ld3+ 9 c6
•••
49 <;t>g2 �e3 so <;t>g3 Wd2 51 l:!.a3 'iii>c 2 I n ow believe that 9 ... b6 is the best
52 <;t>g4 <;t>b2 53 .l:!.a7 �xc3 54 'it>fs Wxb4 move in the position, intending to fol
0-1 low with ... c7-c6, ... .i.b7, ... tt:Jbd7. At the
time I was reluctant to try thi s since
White h ad not spent a tempo on h 2-h 3
Game Bo to prevent ... .i.g4.
K. Wang-J.Schuyler
Wa s h i ngton 2012
1 99
The Dark Kn ight System
Came 8 1
R.Zelcic-M.Djurkovic
P u l a 2001
200
Illustrative Games
201
The Dark Kn ight System
Gam e 83
A.Grilc-G.Moh r
Slove n ia n Tea m
C h a m pion s h i p 2008
2 02
Illustrative Games
203
The Dark Kn ig h t System
Game 85 Game 8 6
A.Mista-M.Szelag I.Jakic-Z.Mestrovic
Kosza li n 1999 Zad a r 2001
204
Illustrative Games
Game 8 7 Game BB
R.Zelcic-Z.Mestrovic L.Gofshtein-N.Mitkov
Nova Gorica 2003 Lis bo n 1999
19 ... h6?!
After this, White has an edge. lt is
simpler and better to keep the knight
When will they ever learn ? From out with 19 ... b6.
now on White has a severe weak colour 20 tDa s i.xg 2 21 �xg2 gs?l [2 1...b6] 22
complex on the dark squares, and h3 1i'g6? 23 tbxb7 hs 24 .:tad1 g4 2 5
Mestrovic's play from thi s point on is tt'lxd6 cxd6 2 6 .l:.xd6 1-0
impeccable. Black h as no real counterplay.
15 .l:tb1 .ie6 16 i.e2 :labS 17 o-o 'iif4
18 �h1 i.xc3 19 bxc3 .l:.xb1 20 'ii'x b1 Punishing the doubled c-pawns
'ii'e 3 21 .id3 c4 22 .:te1 1i'd2 23 .tf1
'it'xc3 24 h3 .:tds 25 'it'c1 .:tbs 26 .:te3 Game 89
'ii'e s 27 l:a3 .l:i.b2 28 'it'd1? as 29 'ii'c 1 c3 J.Ramirez-J.Sch uyler
30 .l:.a4 'ii'f4 31 'ii'xf4 gxf4 3 2 i.c4 .l:txc2 La s Vega s 2007
33 .td s i.xds 34 exds .l:.b2 3 5 .l:.c4 c2
36 �h2 :xa2 0-1 1 c4 tt'lc6 2 tt'lc3 es 3 g 3 fs 4 i.g2 tt'lf6 5
2 05
Th e Dark Kn ig h t System
value of the bishop pair and h alf-open 'itd2 'i!Vxf8 17 h 3 lt:Jxf2 18 'fia4 b5 19
b-file, spending a tempo to reach a po 'ii'a 5 tt:Jxh1 20 i.xh1 'fif7 21 .:tb2 'fih5
sition Black would happily play with a 0-1
tempo less. White's play was poor, but not un
6...i.xc3+ 1 bxc3 d6 8 l:b1 o-o 9 e3? common for a club player. That was
how to win.
Game 90
O.Foisor-J.M.Degraeve
Le Tou q uet 1996
206
Illustrative Games
26 'ilt'd3??
26 'it>g 1 ! �4 27 Wfl holds on . The
text just loses.
26 ... g5 27 'ii'd 4 gxf4 0-1 18 ...lt:Jxe4?
After 18 .. .fxe4, Black keeps his
pawn, and with it a large advantage.
Came 9 1 19 lt:Jxe4 fxe4 20 lt:Jc3 i.b8 21 ctJxe4 iVc6
J.lruzubieta 22 lt:Jd2 ct:Jfs 23 1We2 i.c7 24 bs 'ii'd 7 25
Villal uenga-B.Gu l ko .txa7 b6? 26 as bxas 27 b6 ctJd4 28 �e4
Sa n Sebasti a n 1996 lt:Jc6 29 bxc7?! lt:Jxa7 30 l:;Ixas lt:Jc6 Yz-Yz
207
The Dark Kn ight System
Gam e 93
F Bruno
.
-
B . Kurajica
Luga n o 1985
208
Illustrative Games
2 09
Th e Dark Kn ight System
1 lLlf3 ds 2 d3 g 6 3 g3 i. g 7 4 i.g2 es s
Gam e 9 6 o-o lbe7 6 e4 o-o 7 tLlbd2 tLlbc6 8 c3 a s
K.Arakhamia G rant 9 a 4 h6 1 0 exd s tLlxd s 11 lbc4 ..tfs 12
A.Raetsky .l:.e1 :es 13 lLlh4 i.e6 14 i.d2 'ii'd 7 1S
Bern 1 9 9 5 �b3 tLlde7 16 i.f1 .l:.ad8 17 .l:.ad1 b6 18
i.c1 gs 19 lbg2 i.g4 20 i.e2 i.xe2 21
1 lZJf3 d s 2 g3 g6 3 i.g2 i.g7 4 o-o e s s .l:.xe2 lLlfs 22 .l:.ee1??
d3 lLJc6 6 lZJbd2 lLJge7 7 e4 0-0 8 C3 aS 9
a4 h6 10 l:le1 i.e6 11 exds i.xd s !
210
Illustrative Games
211
The Dark Kn ight System
'ii'xa1 ttJbs 3 2 i..a 4 i..e 7 33 i.. e 1 i.. d 7 34 Black needs t o start using the c-file
ttJes i..e 8 35 l:.c1 i..d 6 36 tiJf3 h6 37 as soon as possible. His next few m oves
i..c 2 tiJa3 0-1 do not work towards that, and he starts
Did White's flag fall ? Black hasn't drifting.
made any progress on the board since 14....i.a6?! 15 i.. e 3 ttJcs 16 .U.c1 'ii'd 7 17
winning the a-pawn. tiJe1 tiJc6 18 'iVd2 i..d 4?! 19 .txd4 ttJxd4
Gam e 1 00
P .Auchenberg
To.Christensen
H e l s i ngor 1997
Adieu
With this, I bid my readers adieu. I hope
this book was more fun for you to read
than it was for me to write. Har! I wish
you many successes with 1...tiJc6 and
the Dark Knight System .
212
I nd ex of Va riatio n s
Chapter One
1 d4 t2Jc6 2 t2Jf3 (others - 15) 2 ... d6
A1: 8 h 3 - 20
A2: 8 0-0 - 2 1
8 : 3 d5 t2Je5 4 t2Jxe5 dxe 5 5 e 4 t2Jf6 - 2 2
81: 6 t2Jc3 - 2 3
8 2 : 6 i.b5+ - 2 4
C: 3 i.f4 t2Jf6 4 e3 g6 5 i.e2 i.g7 - 2 6
213
Th e Dark Kn ight System
Cl: 6 0-0 - 27
(2: 6 h 3 - 2 8
D: 3 g 3 g 6
D l : 4 ds - 3o
D2: 4 .i.g 2 - 3 1
E : 3 ..tg s - 3 3
Chapter Two
1 d4 ltJc6 2 c4 es 3 ds ..tb4+
214
Index of Variations
A l : 6 d6 - 3 6
A2: 6 ti'J c 3 - 3 8
B: 4 ti'Jd2 - 40
Chapter Three
1 d4 ti'Jc6 2 ds ti'Jes
A: 3 e4 - 43
B: 3 f4 - 46
215
Th e Dark Kn ight System
Chapter Four
1 e4 l2Jc6 2 d4 (others - 5 5) 2 ... e5
A: 3 d5 l2Jce7 - 57
Al: 4 l2Jf3 l2Jg 6 - 5 8
All: 5 h 4 ! - 59
A12: 5 .te3 - 6 1
A 2 : 4 C4? ! l2Jg6 - 62
A21: 5 lLlc3 - 62
216
Index of Variations
A22: 5 �e3 - 64
A2 3 : 5 tLlf3 - 67
A24: 5 �d3 - 68
A2 5 : 5 g 3 - 70
A26: 5 a3 - 71
B: 3 dxe5 t2Jxe5 - 73
B11: 6 tL::l c 3 - 74
812: 6 �c4 - 74
B13: 6 e 5 - 75
814: 6 �d3 - 75
B15: 6 �e2 - 75
B16: 6 �9 5 - 76
B2: 4 f4 tL::l c 6 - 76
821: 5 �C4 - 77
B22: 5 tLlf3 - 78
21 7
Th e Dark Knig h t System
C: 3 tLlf3 exd4 - 79
(21: 5 0-0 - 8 8
218
Index of Variations
C22: 5 e 5 tt:\g4 - 89
C221: 6 'ii'e 2 - 89
(222: 6 0-0 - 90
Chapter Five
1 e4 tt:\c6 2 tt:\f3 d6 3 d4 tt:\f6 4 tt:\c3 g 6
A : 5 i.b5 - 94
B: 5 d5 tt:\b8 6 i.e2 i.g 7 7 o-o o-o - 96
219
The Dark Kn ig h t System
Chapter Six
1 e4 t'Llc6 2 t'Llc3 t'Llf6 3 d4 d6
A: 4 f4 - 115
8: 4 d5 - 117
Chapter Seven
1 C4 t2Jc6 2 t2Jc3 e S
A : 3 g 3 - 121
8: 3 t'Llf3 fs 4 d4 e4 - 123
81: s t'Llg s - 124
82: 5 t'Lld2 - 1 2 5
Chapter Eight
1 t'Llf3 t'Llc6 - 127
220
In dex of Varia tions
Chapter Nine
Oth ers - 1 3 1
1 b 3 - 129
1 b4 - 1 3 0
1 f4 - 1 3 0
221
I n dex of Ga mes
222
Index of Games
Golod.V-Sutovsky.E, Nata nya (ra pid) 2009 ... .. . ..... .. . . . ............... 146 ...... ..................................
Gordon.S-Short.N, British C h a m pion s h i p, Sheffield 2011 . . . . . 155..... ..... ................. ........ . ...
Grilc.A-Mohr.G, Slove n i a n Tea m C h a m pionsh i p 2008 . . . ......... 202 .................. ...... ...............
223
Th e Dark Kn ig h t System
Paasikan gas Tella.J-Lindqvist.T, F i n n ish Tea m Cham pio n s h i p 1996 . 134 ...................... ..
Pedersen.P.B-Bekker Jensen.D, Da nish Tea m C h a m pions h i p 2008 .. ... . ... 145 ......... ...... . . .
Rasic.D-Mestrovic.Z, C roatia n Tea m Cha m pions h i p 2001.. .. . . ...... . . . . . 144 ... .... . ..... ... . . ........ .
Zelcic.R-Miles.A, P u l a 1994 1 85
....................................................................................................
224