1.d4
Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.f3 0-0 6.Be3 Nc6 7.Qd2 a6 8.Nge2 Rb8 9.h4
h5 10.0-0-0 b5 11.Bh6 e5 12.Nd5!?
Previously, I had studied only 12.Bxg7 Kxg7 13.dxe5
and 12.Bxg7 Kxg7 13.d5.
While OTB I was very skeptical that such an unusual
move could cause Black problems, my current assessment is that 12.Nd5 is
a valid theoretical attempt for an advantage, as I do not see a clear path
for Black to equality.
Gallagher's analysis of 12 Nd5 (from "The Samish
King's Indian") is worth reprinting in full: "12.Nd5!? bxc4 (after a move
like 12...Re8, White can simply play 13.Nxf6+ Bxf6 14.d5 as the a5-square
is not available to the knight; as Black is desperately in need of an improvement
on what follows, perhaps he could investigate 12...Bxh6 13.Qxh6 bxc4) 13.Nxf6+
Bxf6 14.d5! (14.Bxf8 exd4! gives Black a very dangerous attack) 14...Nb4
(14...Nd4 15.Nxd4 exd4 16.Bxc4 [+-]) 15.Nc3 Re8 16.Bxc4! a5 17.g4! This
move is possible due to the precarious position of the black knight. After
17...hxg4 18.fxg4 Bxg4 19.Rdg1 Bh5 20.a3 the knight has no retreat. Nevertheless,
Black should certainly try this, as he might get some swindling chances..."
Interestingly enough, Watson's "6...Nc6 Sämish
Variation" does not mention 12.Nd5. Apparently, 9...h5 has only become
popular since Watson's book was published (in 1982), as he has the following
to say about it: "This move, weakening the king's wing, has a poor reputation
..."
12...bxc4?
The contortions that Black has to go through after
12...bxc4 13.Nxf6+ Bxf6 14.d5 Nb4, and the ease with which White enforces
g4, make it hardly surprising that White stands better in that line. What
about 14...Nd4 15.Nxd4 exd4 16.Bxc4? Is that really considerably better
for White, as Gallagher claims? I believe so: White will have a much easier
time getting to Black's king than Black will to White's, i.e. 16...Rfe8
17.Rdg1 c5 (17...d3 18.b3 only costs Black his d-pawn) 18.g4 and White's
attack is in full bloom.
In fact, I have noticed that the relative ineffectiveness
of Black's queenside attack in the 9.h4 Main-Line Sämish lines, even
in positions in which it appears to be formidable, is a frequently recurring
theme (being highlighted, for one thing, in this game). Black needs to
be constantly on the watch for positions in which his attack is ineffective.
What about attempting to improve on the 14...Nd4
line given above with the swischenzug 15...c3? After 16.Qxc3, 15...c3 would
certainly be justified, though I do not believe that the improvement would
be significant enough to swing the assessment of the position after 12...bxc4
in Black's favor.
Unfortunately, 16.bxc3, the more critical test
of 15...c3, forcing Black into a sacrificial frenzy, while extremely omplicated,
appears to favor white even more than the position after 15...exd4. Hence,
rather than being a minor improvement, 15...c3 is actually a further error.
(The analysis of this, the game continuation, can be found below.)
Since after 12.Nd5 White threatens 13.Nxf6 and
14.d5, when, as Gallagher mentions, Black doesn't have 14...Na5; and since
after 14...exd4 15.Bxg7 Kxg7 16.Nxd4 Black has ceded the center without
compensating return; Black should probably play 12...Bxh6 13.Qxh6 bxc4,
as Gallagher suggests.
13.Nxf6+ Bxf6 14.d5!
14.Bxf8 exd4 15.Bh6 (After any other move Black
will play 15...Kxf8 and will have incredible compensation for only an exchange)
15...Nb4 16.Nxd4 (16.Kb1 c3 17.bxc3 {else 17...c2} 17...dxc3 18.Nxc3 Nd5+)
16...c5 is in fact quite brutal, as Gallagher mentions; while after 14.dxe5
Bxe5 (Normally, ...dxe5 is the preferable recapture, keeping control over
f4 and d4; but in this case Black's pressure along the a1-h8 diagonal is
tough to answer) 15.Nc3 Qf6 (Stopping Bxf8 long enough) 16.Bxc4 Be6 Black
threatens 17...Rxb2 and again stands better.
14...Nd4
Not worse than 14...Nb4, since both 14...Nd4 and
14...Nb4 result in positions favorable for White to an approximately equal
degree.
15.Nxd4 c3?
The complications resulting from 15...c3 16.bxc3
do appear to favor White more than the position after 15...exd4 16.Bxc4.
16.bxc3!
As mentioned above, also playable was 16.Qxc3
exd4 17.Qd2 (Of course not 17.Qc2 d3 18.Bxd3 Rxb2) 17...Re8, when a position
very similar to the position reached after 15...exd4 16.Bxc4 which Gallagher
assesses as [+-] would be reached, and which is likewise good for White,
as Black's attack is again a good step behind. For example, 18.Bc4 (18.Bd3,
attempting to refuse to justify 15...c3, protecting e4 and thereby facilitating
g2-g4, runs into 18...c5 and 19....c4) 18.c5 19.Rdg1 Rb4 20.b3 Qb6 21.g4
hxg4 22.h5 Reb8 23.Qd3 and White's attack is again stronger.
After 16.bxc3, however, the sacrifices to which
Black committed with his 15...c3 appear to be unsound.
16...exd4 17.cxd4 c5
After three obvious half-moves, the position critical
for the evaluation of White's 16.bxc3 (and hence Black's 15...c3) has arisen.
Black certainly appears to have a ton of play;
and, OTB, in the absence of the sort of exhaustive analysis that I have
since done and that alone is adequate grounds for a conclusive assessment
of such a position, I was very pleased to have reached this position. Nevertheless,
it is now quite clear that White has considerable defensive resources at
his disposal, and I believe that they ensure him of the advantage.
|
White could have demonstrated this advantage with
18.Bxf8!. After 18...Kxf8 (18...Qxf8 would be strictly worse, as Black
needs to activate the queen via either b6 or a5) 19.dxc5 (The greedy approach
not only works here but is in fact necessary. The supposedly standard antidote
to gambits - returning some material for fast development - would give
Wite no better than equality: 19.Bc4 {aiming for 19...Bd7 20.Qc2 Qa5 21.Rd3!,
with an excellent game for white} 19...Bxd4! 20.Qc2 Rb2 21.Qxb2 Bxb2+ 22.Kxb2
when Black's better bishop and strong c-pawn, along with White's awkward
rook cooperation, should be enough to offset White's slight material advantage)
19...dxc5 (19...Rb2 20.Qxb2 Bxb2+ 21.Kxb2 dxc5 22.Bc4 Qa5 23.a3 should,
on the other hand, favor White, as his material advantage is greater than
in the 19.Bc4 Bxd4 line given above, and as his central pawn mass will
be relevant) 20.Bc4 Bd7 (20...Rb2 21.Qxb2 Bxb2+ 22.Kxb2 would transpose
to the line favorable for white given above) 21.Qd3, when 21...Rb2 22.Rd2,
21...Qb6 22.Rd2, and 21...Bb2+ 22.Kd2 are all inadequate for Black.
As mentioned previously, that this position favors
white is another somewhat surprising example of the frequent impotence
of black attacks in this line. For some reason, white is often quite capable
of withstanding an attack which is restricted to the b-file. In these lines,
Black is never able to do considerable damage along the b-file. Instead,
White is able to quickly and effectively replace his missing pawn cover
with an equally-effective piece cover, and he is able to force further
trades in a position which is already somewhat simplified.
18.dxc6?
After this move, however, White is lost. Black's
sacrifices may have been unsound, but even a single inaccuracy was enough
to swing the balance.
18...Qb6 19.Qc2
Best. 19.Kc2 Qb2+ 20.Kd3 Qxd4+ and 21...Rb2 [-+]
and 19.Bd3 Qxd4 20.Bb1 (The only move that doesn't lose immediately to
20... Qa1+) 20...Qc5+ 21.Qc2 Qa3+ 22.Kd2 Rb2 [-+] would both lose immediately.
19...Bxd4!
It is hardly surprising that White, having allowed
both of Black's bishops to participate in the attack, cannot subsequently
defend the position.
20.Bc4
Also best, as the alternatives were even worse:
A) 20.Bxf8 Be3+ 21.Rd2
Qd4 22.Kd1 (22.Qd3 Qa1+ 23.Kc2 Rb2+ 24.Kc3 Rxd2+) 22...Rb1+ 23.Ke2 Bxe2
24.Qxe2 Rb2 [-+].
B) 20.c7 Rb7 21B.c4
Qxb7 [-+].
C) 20.Rd3 Be6 21.Bxf8
(Else 21...Rfc8) 21...Kxf8 22.Be2 (22.Rxd4 Qxd4 and White can't stop ...Qa1+
and ...Rb2 without giving up the KR) 22...Be3+ 23.Rxe3 (23.Kd1 Qb1+ 24.Qxb1
Rxb1+ 25.Kc2 Rxh1) 23...Qxe3+ 24.Kd1 Qd4+ and ...Qa1+, ...Qxh1.
D) 20.Rxd4 Qxd4 21.Bxf8
Kxf8 and again White can't stop ...Qa1+ and ...Rb2 without giving up the
KR.
20...Be6 21.Bxe6
Necessary, since after 21.Bb3 Rfc8 White has no
defense against ...Rxc7.
21...fxe6 22.Rxd4
Again, the best of White's three attempts to stop
22...Rfc8, inferior being:
A) 22.Bxf8 Be3+ 23.Rd2
Kxf8 24.Kd1 Bxd2 25.Qxd2 Rb1+ 26.Ke2 Rb2 [-+].
B) 22.c7 Rfc8 23.Rxd4
(23.cxb8 Qf2 [-+]) 23...Rxc7 24.Rxc4 Rbc8 and the resulting Q vs R+B endgame
is similar to that reached in the game and almost certainly winning for
Black. Queens are notoriously strong in endgames.
22...Qxd4 23.Rd1
Preferable to 23.Bxf8 Qa1+ 24.Kd2 Qxh1 25.Bxd6
Qxg2+ 26.Kc1 Kf1+ 27.Kd2 Qf2+ 28.Kd3 (28.Kc1 Qe1+ [-+]) 28...Qxf3 [-+].
23...Qa1+ 24.Kd2 Qh8 25.Bxf8
Rb2 26.c7
26.Bxd6 Qd4+ was no better.
26...Qd4+ 27.Kc1 Rxc2+ 28.Kxc2
Qc4+ 29.Kb2 Qxc7
Now, an endgame which is the correct consequence
of 18.dxc6 has arisen. In order to hold this position, White would have
to construct a fortress; and he would have to be able to restructure this
fortress in response to any of the various pawn breaks which Black has
at his disposal. I do not believe that such fortress gymnastics are possible;
and I do not believe that White could have improved on his subsequent play
- at several junctures, he could have posed different problems for Black,
but never insoluble problems. Hence, the rest of the game will be given
without comment.
This endgame is not, however, trivial, as evinced
by the fact that this game ended up going into the last half-hour of play.
30.Bxd6 Qb6+ 31.Kc2 Qc6+ 32.Kb2
Qb5+ 33.Kc3 Qa5+ 34.Kb2 Qb6+ 35.Kc2 Qf2+ 36.Rd2 Qh4+ 37.Be5 Qe7 38.Bc3
Qa3 39.Kd3 Kf7 40.Kc4 g5 41.Rb2 Kg6 42.Rd2 h4 43.Rb2 Kh5 44.Rd2 h3 45.gxh3
Kh4 46.Rg2 Qc1 47.Rd2 Kxh3 48.Rd3 Kg2 49.e5 Qa3 50.Kd4 Qa4+ 51.Ke3 Qf4+
52.Ke2 Qc4 53.Ke3 a5 54.Bd4 Qd5 55.Rd2+ Kg3 56.Rd3 Qxa2 57.Bc3 Qf2+ 58.Ke4
Qe2+ 59.Re3 Qc4+ 60.Bd4 Qd5+ 61.Kd3 a4 62.Re4 Kxf3 63.Re3+ Kf4 64.Re1 g4
65.Rf1+ Kg3 66.Rg1+ Kh2 67.Ra1 g3 68.Re1 g2 69.Ra1 a3 70.Re1 a2 71.Ra1
g1=Q 72.Rxg1 Qxd4+ 0-1. |