Sochi RUS-tch, 04/20/2005
Round 2 [Svidler]
1.d4 P.Svidler 1...Nf6
2.c4 g6 3.Nc3
d5  4.Bg5 The weapon of choice of an increasing number of Grunfeld-slayers these days. To me this is the best advertisement of the rude good health of my beloved (well, only anyway) opening. I used to play solely 4...Bg7 here, but as I have written in these pages before, such youthful indiscretions are mostly in the past. 4...Ne4 5.Bh4 Nxc3 6.bxc3 dxc4 This is not very ambitious in itself - more of a recognition of the fact that after | ||
| [6...Bg7 7.e3 c5 8.cxd5 cxd4 9.cxd4 Qxd5 10.Nf3 Nc6 11.Be2 , which I have tried with reasonable results before, White will probably enjoy himself rather more than Black.] | ||
7.e3  Be6 I already played this line against Dreev in Poikovsky earlier this year. There after 8.Rb1 b6 9.Nf3 I tried 9...c6, but despite getting a decent game in the end, I did not like the character of that position too much. 8.Nf3 Bg7 9.Rb1 b6 10.Nd2  11.Be2 A seemingly innocous line which nonetheless has been scoring very well for White recently. 11...c5!? Not exactly a novelty. During my preparation I was initially looking for a clear way to equalize - this was, after all, a team event, and I was brought up with the understanding that a draw on a Black board is good for the team - but, luckily for me, I could not find an entirely satisfactory solution. Less than an hour remained when I was hit by an inspiration particle - which was fortunate, because with more time I would probably refute and never play the whole thing. | ||
|
[11...Qd7 12. | ||
| [11...Qd7 P.Svidler] | ||
|
The text is an attempt to save on the unnecessary Qd7 - but doesn't it lose material? 12.Bf3  Taking up the gauntlet. | ||
|
[In an Internet game between Nogueiras and Sasikiran this year Jesus played 12. | ||
|
12...cxd4 '!?' P.Svidler. This is the
bright spark that has occured to me right before the game - Black cannot save the exchange, so why not
go the whole hog and offer the rook? Then I was mainly concerned about the position arising after 13.Bxa8
dxc3, followed by 14...Qd3. A brief check indicated that Black dioes not seem to lose on the spot, which
was good enough for me. And off I went. | ||
|
[The stem game Steinhoff-Henao, Sao Paulo 1992 went 12...Bd5 13.Bxe7 Bxf3 14.Qxf3 Qxe7 15.Qxa8 cxd4 16.cxd4 c3 17.Nb3 c2 18.Rc1 . Black managed to survive in the end, but it is decidedly dodgy.] | ||
| 13.cxd4 | ||
|
[A sample line that illustrates Black's chances after 13.Bxa8 dxc3 could be 14.Nf3 Qd3 15.Rc1 Na6 16.Bb7 | ||
| (16.Bc6 Nb4 17.Bxe7 Nxc6 18.Bxf8 Kxf8) | ||
| 16...Nc5 17.Bxe7 Nxb7 18.Bxf8 Bxf8!? | ||
|
(18...Kxf8 19.Nd4
Nc5 ' ' P.Svidler. is also possible) | ||
| 19.Ne5 | ||
(19.Qxd3 cxd3
20.Rxc3 Bb4
21.Kd2 Nc5 ) | ||
| (19.Nd4 Bb4!) | ||
19...Qe4 20.Nf3
Qd3 ] | ||
|
[13.exd4 Bd5 is
not a ve ry good choice - now 14.Bxe7 runs into 14...Qxe7 check, and therefore Black is fine. | ||
| (13...Bd5 P.Svidler 14.Bxe7?? Qxe7+)] | ||
|
13...Nd7 I also looked at this position
- of course, Black did not play 12...cxd4 only to return to 13... Bd5 if White fails to take on a8 -
but here my 'analysis' was confined to a solitary line in which White was so cooperative that I won't
even reproduce it here. A good dose of unfounded optimism has rarely done anyone any harm. | ||
| [13...Bd5 P.Svidler 14.Bxe7] | ||
| 14.Bxa8 | ||
|
[14.d5 Bf5
15.e4 also looks ive at first, but in
fact the bishop can by freed by g6-g5 at any point, e.g. 15...Rc8!? | ||
(15...c3 16.Nc4
Rc8 17.Ne3
g5 18.Bxg5
Bg6 19.d6
Bf6 ) | ||
| 16.Qc2 | ||
| (16.exf5 c3) | ||
| 16...c3 17.Nb3 g5 18.Bxg5 Bg6 19.Be2 | ||
| (19.d6 Nf6) | ||
19...Nf6 20.Bd3
Nxd5 21.exd5
Qxd5 ] | ||
14...Qxa8  15. 15...c3 I felt it was very important to drive the knight to f3, where it will be very cut off from the Queenside action. The pawn on c3 will have to fend for itself for a while, though. 16.Nf3 16.Nb3 loses to 16...Qe4, of course. | ||
[16.Nb3? P.Svidler 16...Qe4 ] | ||
| 16...Rc8 | ||
|
[16...Bxa2 17.Rc1 costs
too much time: 17...Rc8 18.Qa4 Be6 19.Bxe7 '+/=' P.Svidler. , and the pawn trade should favor White.] | ||
|
17.Qa4 Nf6 This
was a pity - White's bishop on h4 is his worst piece. But I could not see any choice - if White is able
to play Rfc1 and Qa3 unmolested he will have the upper hand. There was also a small matter of protecting
the e7 pawn - the pawn in itself is not worth too much, but allowing the bishop back into the fray would
be a mistake. 18.Rfc1 | ||
|
[I expected 18.Bxf6 exf6 19.Rfc1 Bd5 20.Ne1 , but looking at it now I can see why it did not quite appeal to Alexander - after the possible 20...f5 '|^' P.Svidler. Black's play is much easier - Qa3 will always be met by Bf8, and there is no other comfortable way to attack the passer. | ||
|
(20...f5 P.Svidler 21.Qa3?!
Bf8)] | ||
|
[I was planning to meet 18.Qa3 with Nd5 , and now the e7 pawn is indirectly protected: | ||
(18...Ne4!? P.Svidler 19.Bxe7
Bd5 ) | ||
| 19.Bxe7 Nxe7 20.Qxe7 Bf8 21.Qh4 c2 22.Rbc1 Ba3 However, 18...c2 or even 18...Ne4!? are also possible.] | ||
|
18...Nd5 Very ambitious - if White's
pawn center becomes mobile Black will be in grave danger. But I was fascinated with the idea of using
this knight as an equine shield. P.Svidler: '/\Bf5' | ||
| [18...Bd5 19.Ne1 , planning f3 and Bf2, seemed too slow.] | ||
[18...Bf5?! P.Svidler 19.Bxf6!
exf6 20.Rb3
c2 21.Rb2 ] | ||
|
19.Qc2 Played quickly and confidently.
Of course, allowing 19...Bf5 with total control over the light squares was not really an option. | ||
| [19.e4 Nf4 20.Qc2 Bc4 is also inadvisable.] | ||
19...a5 20.a3  Qa6! Black calmly starts the avalanche, for a while ignoring the centre completely. | ||
|
[I briefly toyed with 20...Bf5 21.e4 Nb4 22.axb4 Bxe4 which is less stupid than it looks, but the text was much more attractive. Here is another nice sample line: 23.Qa4 Bxb1 24.Rxb1 c2 25.Rc1 axb4 26.Qxb4 Bh6 27.Bg5 Qxf3! 28.Bxh6 | ||
| (28.gxf3 Bxg5) | ||
| 28...Qd1+ 29.Qe1 Qxe1+ 30.Rxe1 b5 31.Kf1 b4 32.Bc1 Rc4 | ||
(32...e6 33.Ke2
b3 34.Kd2
Rc4 35.Re3
Rxd4+ 36.Rd3 ) | ||
| 33.Bb2 e6 34.Ke2 '+/-' P.Svidler. , and White is on top. Of course, I hadn't seen 27...Qxf3 during the game - but my feeling was that the compensation for the piece might not be enough .] | ||
|
21.Bg3 Covering f4 - White is now ready
for e3-e4. 21...Bh6 There is a certain harmony in Black's setup - all his pieces are ready to support the pawn march. 22.Re1 b5 23.e4 Now the knight will have to go, allowing d5 & Nd4. Or will it?  23...b4! Not necessarily. Black managed to connect his passers, and material becomes less important. 24.exd5 loses to Bf5, breaking the levee on c2. It took me a while to work out what my threat here was, since it is a very unappetizing move - but if Black gets to play 24...bxa3 he will not need a beautiful pawn chain anymore. The knight will arrive on b4, effectively sealing the game. P.Svidler: '/\ 24...bxa3, 25...Nb4-+' 24.Ne5 '!' P.Svidler . A good move. White needs to start some counterplay, and this move prepares the capture on d5. | ||
|
[24.exd5 P.Svidler 24...Bf5] | ||
| 24...Bd2 | ||
| [24...bxa3 25.exd5 Bf5 26.Qb3 Bxb1 27.d6 e6 28.Nxf7 '+-' P.Svidler. is an illustration of the dangers of complacency.] | ||
|
25.Rf1 '[]' P.Svidler. The only move
again. I spent more than 20 minutes here, at first struggling to believe there was no forced win for
Black. But as soon as I convinced myself that this is indeed the case, the correct solution appeared.
If the pawns cannot be promoted just yet I should bring reinforcements. | ||
|
[25.Red1 was bad: 25...Qe2 (threatening
26...Ne3) 26.exd5 Bf5 27.Qb3 Bxb1 28.d6 Qxd1+ 29.Qxd1 c2 ] | ||
|
[25.exd5 also loses: 25...Bf5 26.Qb3 Bxb1 27.d6 e6 28.d7 Bxe1 29.dxc8=Q+ Qxc8 30.Qxb1 c2 31.Qc1 Bd2 '-+' P.Svidler.] | ||
|
25...Qb5! The best move of the game
- and I would be even prouder of it if I saw the idea behind it. P.Svidler: '26...Ne3 27.fxe3 b3' | ||
|
[Lines like 25...Qe2 26.exd5 Bf5 27.Qb3 | ||
|
(27.Qd1 Qxd1
28.Rbxd1 c2
29.Nc6 cxd1=Q
30.Rxd1 ' ' P.Svidler. is also far from clear, but White can aspire to more) | ||
| 27...c2 28.d6 cxb1=Q 29.Rxb1 Bxb1 30.Qxf7+ Kh8 31.h3! Be4 32.Kh2! '+/-!' P.Svidler. with a crushing attack are much easier to find with Fritz of course, but I was loath to part with the f7 pawn - I could tell that it might be hazardous.] | ||
| [25...Ne3 26.fxe3 b3 27.Rxb3 Bxb3 28.Qxb3 Bxe3+ 29.Bf2 '+-' P.Svidler. fails as well] | ||
| [25...bxa3 26.exd5 Bf5 27.Qb3 c2 28.Rbc1 is also unconvincing] | ||
|
Tthe text creates a very strong threat of 26...Ne3 27.fxe3 b3. 26.a4 After this move White is doomed. Black no longer needs to keep the knight on d5 at all costs - his pawns are strong enough to win him the game on their own. But why not | ||
|
[26.exd5 ? When I played 25...Qb5 I planned
to respond by the simple 26...Bxd5. After that b4-b3 can no longer be stopped, and with his light pieces
still cut off from the queenside, it will be very difficult for White to fight the onward march. Even
the computer, given enough time, begins feeling the same way. But Black has a much better solution: 26...Bf5!  27.Qb3 Bxb1 28.d6 , and now the wonderful 28...Ba2! 29.d7 Rd8 30.Qxa2 b3 '-+' P.Svi dler. , and the pawns decide. Depressingly, even 26...Qxd5 is very strong, leaving my own idea a distant third. At first I was very disappointed when I found all that after getting back to my room - but since 25...Qb5 is not too obvious, and it does win as I thought it would, maybe I should not kick myself too much for playing it - even if I played it for the wrong reasons.] | ||
|
26...Qb7 '-+' P.Svidler. Now all Black
needs is a bit of patience. 27.Nf3 Or | ||
| [27.Nd3 Nf6! | ||
| (27...Ne3 28.fxe3 Qxe4 29.Nxb4 is premature) | ||
| 28.Nc5 | ||
| (28.d5 Nxd5! 29.exd5 Bf5 '-+!' P.Svidler. is a familiar theme) | ||
| 28...Rxc5 29.dxc5 Nxe4 , and Black won't be kept waiting for too long.] | ||
| 27...Bh6 28.h4 Nb6 | ||
|
[28...Nf6 29.d5
Nxd5 looked tempting, but then I realized
that after 30.Nd4 '-/+' P.Svidler. I will have to part with one of the bishops, and quickly shelved the idea.] | ||
|
29.d5 Bd7 White
has finally pushed the pawn to d5 - but it is too little much too late. 30.Ra1  Nxa4 | ||
| [30...Bxa4 31.Rxa4 Nxa4 32.Qxa4 c2 '-+' P.Svidler. was cleaner, but I missed White's 33rd.] | ||
|
31.Nd4 Qb6
32.Nb3 Bb5 ...with
33...Nb2 and 34...Bd3 to follow, or so I thought. P.Svidler: '/\ 33...Nb2' 33.Rfb1 Now the game will be prolonged a bit. 33...Nc5 | ||
| [33...Nb2? P.Svidler 34.Rxb2] | ||
34.Nxc5 Rxc5
35.Be5 Rc8  36.Rd1? '(+)?' P.Svidler. The final mistake in serious time-trouble. After | ||
| [36.h5 Bg7 37.Bxg7 Kxg7 38.hxg6 hxg6 it would take a while for the pawns to roll.] | ||
|
36...a4 Now it's all over. 37.d6 b3 38.d7 Bxd7 39.Qd3 Bb5 40.Qh3 Qe6 We passed the time control, and Alexander resigned. 0-1 | ||
