One round down, and the Western States Open started on a positive note for me. I drew against a 2400 player named Alexander Kretchetov in only 20 moves. Before you say “grandmaster draw,” take a look at the final position.
FEN: r1bq4/ppk1bR2/2n1p2p/2p4Q/2B1N3/2PP4/PP4PP/6K1 w – - 0 20
I’ve sacrificed a piece for a pawn (soon to be two) and a dangerous attack. Kretchetov played his previous move, 19. … Kc7, and then after I had already spent about two minutes thinking about my response, he said as if it was an afterthought: “Draw?”
My immediate reaction was that my position must be pretty good if a 2400 player is offering a draw against me. But I didn’t see anything resembling a clear win. The main moves I looked at were 20. Nxc5, 20. Qxh6 and 20. Bb5. What would you do here? Agree to a draw or play on? If you’d play on, what move would you choose?
I probably thought another five minutes and then accepted. I felt really craven to do so, because I felt sure that a Tal or a Topalov would have played on here, regardless of whether they saw a win.
My decision was mostly an emotional one, not based on the position. I had played a previous game against Kretchetov (in fact, in the first round of this same tournament last year!) that was quite similar to this one, where I sacrificed two pawns and got what should have been a winning position, but I went astray and eventually lost. I wrote about it in my post, Negative two pawns are better than negative one. I didn’t want something like that to happen again. I wanted to have something to show for my efforts, so I agreed to a draw.
When we looked at the game quickly, it looked as if White is indeed winning after 20. Nxc5 Kb8 21. Nxe6 Bxe6 22. Bxe6 Qxd3 23. Rxe7. A key point that I missed at the board was that if Black takes the rook, he simply gets mated: 23. … Nxe7?? 24. Qe5+ and mate next move. Oy veh!
But when I put this position on the computer, I was surprised to see that Rybka says Black is better! The point is that Black can play a long series of checks designed to wrest control over the e5 square, beginning with 22. … Qb1+ 23. Kf2 Qxb2+. I won’t give the whole line here, but the upshot is that Black eventually gets his queen to f4 with check and then takes the rook. White still gets Qe8+, but this leads to an endgame that is slightly better for Black.
There is no other variation that leads to an advantage for White, either, so at least from Rybka’s point of view I did the right thing. Of course, in a game between humans absolutely anything could have happened in this position. Psychologically it would have been very difficult for Black. Even if I play the simple 20. Qxh6, he’s still got a lot of problems organizing his pieces, and the connected passed pawns on the kingside will be very dangerous. The computer may say Black is OK, but I think the onus is on Black to prove it. That’s how people like Tal and Topalov won (and win) their games.
Not being in their category yet, I guess I’ll settle for the half point and vow to fight a little harder next time.