Friday
Nov122010
Tal Memorial, Round 7: What Happened in Wang Hao-Gelfand?
Friday, November 12, 2010 at 1:42PM
According to the official site, Gelfand resigned. ICC, where I was watching, dutifully repeated the information, and while I wasn't listening, apparently the GM commentary agreed. That's all nice, except that the final position is drawn, not lost. It's not impossible that Gelfand panicked himself out of half a point, but before I report it I'd like to hear from someone with firsthand information whether this actually happened.
Reader Comments (5)
In video translation Gelfand did offer his hand to Wang Hao, this can only mean resigning.
You can see it for yourself here (at 21:24:35): http://video.russiachess.org/browse/index/day/19
Thanks, Andrey. I suspected I'd hear from you on this!
I downloaded the game at Chessbase and it definitely is a White win. Rybka shows White with a +5 advantage. Perhaps you have the wrong game score?
[DM: I have the right game score, and that the position is objectively drawn has been confirmed at other places. Your Rybka should not say +5!]
My mistake Dennis, I somehow got the Eljanov - Gelfand game mixed up with the Wang Hoa - Gelfand game. However, I did get the correct game score and Deep Rybka 4 (and Stockfish 1,9 ) did evaluate it at around +1.6 for White after 75..Ke6, not a clear win but Black would have work to do to try and draw it. In fact everything else seems to be a dead lose for Black.Granted chess engines are not infallible but they don't see the position as drawn. Maybe Gelfand in analysing the position saw all the other moves losing outright and Ke6 was considered to only prolong the agony? BTW my cpu is an Intel i7 that has 4 cores (8 logical) running at 3GHz and has 8 GB of memory.
[DM: Big difference between +5 and +1.6. Sure, all our engines said that, which means in a rook ending that the computer doesn't see a forced win. In this case, it's not because the win is many moves off, but because there isn't one. If you have a rook ending with lots of pawns, on both sides of the board, +1.6 is a good indication that it's an eventual win, but not here. In fact, it's closer to meaning a draw, though one has to look with one's own eyes.
As for this game, Gelfand simply missed the idea. It happens, even to super-GMs.]