|
Bucharest 2001
Mikhail Golubev
Gergely Szabo
Sicilian B87
1 e4 c5 2 Nf3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 a6 6 Bc4 e6 7 Bb3 b5 8 0-0 Be7 9 Qf3 Qb6 10 Be3 Qb7 11 Qg3
|
|
|
This is a well-known crossroads in the Fischer/Sozin Attack. Black has four popular moves here: 11...Nc6 (illogical); 11...0-0 (slightly better for White); 11...Nbd7!? (consequent, but risky) and 11...b4 12 Na4 0-0 (solid main line). I spent a lot of time on all this while playing and analysing my previous games and writing the 9 Qf3 chapter for my Sicilian Sozin book. One more option is 11...Bd7?, but here White gets a crushing attack by 12 Nf5! exf5 13 Qxg7 Rf8 14 Bg5!, as Ivanchuk played against Shakhvorostov in the USSR Junior Championship 1985 (in which I also participated). All other Black's moves are virtually untested till now.
11...h5
A surprise. At the board I could remember only that White won a game after 12 Nf5, but hardly anything else. A less ambitious move would be 12 f3, after which Black can either start to disturb the white pieces with pawns (12...h4!? and 13 Qxg7?! Rh7 14 Qg5 Rh5 is a guaranteed draw) or else develop his queenside pieces.
12 Nf5 exf5 13 Qxg7
I spent about half an hour on the last two moves. After 13 e5 dxe5!? 14 Qxg7 Rf8 White has some compensation but I didn't like 15 Rad1 Be6 at all, and 15 Bg5 (where 15...b4!? is possible) didn't attract me either. The final decision was made after I found an additional idea 16 Rad1!? and 17 Rxd6+ in the game line.
13...Rh7!
I expected this, and it is clearly the best move. After the game my database showed that Black had previously played only 13...Rf8?!. Instinct tells me that White should be much better after that; possibly 14 Bg5!? Nxe4 15 Bd5 Nxg5 16 Bxb7 Bxb7 17 Rfe1 Ne6 18 Qh7, Sofia Polgar-Enoshi, Biel 1987, is confirmation of that view.
14 Bxf7+ Kd8 15 Qg6 Nbd7
|
|
|
In this critical position Black threatens to win a bishop by 16...Ne5 or 16...Nf8, and White has to do something. 16 Nd5?! Nf8 17 Bb6+ Qxb6 18 Nxb6 Nxg6 19 Bxg6 Rg7 is clearly not good, but 16 Bd5 is probably strongest; Black then has two possibilities.
One line is 16...Nxd5 17 Nxd5 (or 17 Qxh7!?) 17...Nf8 18 Bb6+ (18 Qg3 may be of some interest) 18...Qxb6 19 Qxh7 Nxh7 20 Nxb6 Rb8, and White can play 21 Nd5 (rather than 21 Nxc8 Rxc8 22 exf5 Rxc2).
I didn't like 16...Qb8 17 Bxa8 Nf8 (17...Ne5 18 Qg5!) where Black first attacks the white queen and later takes on a8 in the best possible circumstances. Now I think that after 18 Qg5! his initiative is rather temporary. 17 Qxf5!? (instead of 17 Bxa8) deserves attention as well.
16 Rad1
After long reflection I chose this move, planning a further queen sacrifice that worked so well in the game. However, 16 Rad1 could possibly cost White a point, at least when playing against a computer.
16...Ne5
16...Nf8 17 Rxd6+ with good chances after 17...N6d7 18 Bb6+!? Qxb6 19 Rxd7+ Nxd7 20 Qxh7 or 17...Kc7 18 Qxf6!? Bxf6 19 Rxf6. If 17...Bd7, White can play as in the game.
17 Rxd6+
|
|
|
17...Bd7
If Black has any natural move here, then this is it. After 17...Bxd6, the forced line 18 Qxf6+ Qe7 19 Bb6+ Kd7 20 Qxf5+ Kc6 21 Qxh7 leaves White with four pawns for a bishop.
However, in the case of 17...Kc7! (king attacks rook!) Black can probably survive his opponent's activity with a good chance of winning: 18 Nd5+ (other moves don't work at all, as far as I can be sure about such a position) 18...Kxd6! (other moves are unconvincing: 18...Nxd5?! 19 Qxh7!? Nxe3 20 Bd5; 18...Qxd5 19 exd5!? Nxg6 20 Rc6+; 18...Kb8 19 Rxf6!? Nxg6 20 Bxg6 Bxf6 21 Bf4+) 19 Bf4 (this was my main hope; 19 Rd1? Nxg6; 19 Nxf6? Rxf7!?) 19...Rxf7! (the only move) 20 Rd1 (20 Qxf7 Nxd5! 21 Bxe5+ Kxe5, and also good is 20....Be6!?)
|
|
|
... and now, incredibly, the cold computer move 20...fxe4!! leaves White with just too few fighters on the board.
In all the lines that I was able to find, White is in big trouble. The main point is that, after 21 Nxf6+ Ke6, White's queen is attacked and he must play 22 Bxe5 where, after 22...Bxf6!? 23 Bxf6 Rxf6 24 Qe8+ Qe7 25 Qc6+ Kf7 26 Qxa8 Rd6, Black has a bishop for two pawns.
18 Qxf6!
18 Rxd7+? Qxd7 is hopeless for White; also bad is 18 Bd5?! Qxd5 19 Nxd5 Nxg6 20 Bb6+ Kc8.
18...Bxf6
18...Nxf7? 19 Rxd7+! Qxd7 20 Qg6 and White wins.
19 Bd5
|
|
|
White has only three pawns for a queen at this juncture. Also, he has no concrete threats against the adversary's king. Nevertheless, compensation is very considerable. I liked all the lines I saw for White in this position, and the most difficult part was to convince myself that I really can play without a queen. But... it's only a piece after all.
19...Qb8
19...Qc7? is impossible because of 20 Bb6!. After 19...Nc6 White plays 20 Rxf6.
20 Rxf6 Rg7
Two good illustrative lines are 20...Ng4 21 Rf8+ Be8 22 Bg5+ Re7 23 Rd1! Qxh2+ 24 Kf1 and 20...Ke7 21 Bg5! Ke8 22 exf5.
21 Rf8+ Be8 22 Rd1 Rd7?
Black should play 22...Nd7, but White is already better anyway.
23 Bg5+ Kc7 24 Bf4 Kd8
24...Kb6 loses by force after 25 Rf6+ Ka5 26 a3! b4 27 axb4+ Kxb4 28 Rd4+ Kc5 29 Be3.
25 Bg5+
The old masters would never understand such a manoeuvre, but they played without a clock...
25...Kc7 26 Bf4 Kd8 27 Rxf5
|
|
|
White is winning from now until the end. The game continued in increasingly hectic mutual time-trouble where my main aim was not to let the white pieces come under attack. So it was not easy to find such an aggressive idea as 41 Nd8+, winning a rook...
27...b4 28 Bxe5 Qa7 29 Ne2 Rc8 30 Nd4 Qb6 31 Rd3 Ke7 32 Bf6+ Qxf6 33 Rxf6 Kxf6 34 f3 Rc5 35 Kf2 Rdc7 36 Ke3 Bf7 37 Bxf7 Kxf7 38 f4 Ra5 39 Rb3 Rb7 40 Nc6 Rxa2 41 Nxb4? Ra4 42 Nc6 Rxb3+ 43 cxb3 Ra2 44 Ne5+ Ke6 45 Nd3 a5 1-0
Here I reckoned that 40 moves must have been made. My flag fell; we reconstructed the last 15 moves and Black resigned. He could play on, but White has too many pawns for the exchange.
The Sicilian Sozin
BCM Online
© Chess-Sector.odessa.ua
|
