Author: Robert Hyatt
Date: 10:08:33 07/23/98
Go up one level in this thread
On July 23, 1998 at 12:58:20, Robert Hyatt wrote: >On July 23, 1998 at 12:52:25, Bruce Moreland wrote: > >> >>On July 23, 1998 at 12:28:13, Amir Ban wrote: >> >>>It does look bad for white, but resigning here is premature. >> >>I would have dragged it out a little more, too, but it is possible that Ed saw >>he was going to lose a pawn or two and decided to call it quits. >> >>That big trade down into an ending was interesting. Mine would have played Nxf7 >>and the rest of that as well, thinking it was doing just dandy at the start, >>then a little less dandy as the end of the exchange came closer. >> >>Two bishops will kill a rook, but there has to be some point where you add >>enough pawns to go with the rook that you'd prefer the rook. You'd think that >>three pawns would be beyond that point. >> >>I wonder what was going through Anand's mind during that game. >> >>bruce > > >The major move that Crafty didn't like was the Qh5+ move which traded the knight >on h8 for black's remaining two pawns. Crafty has specific eval code that says >if it is down a piece, even with three pawns for it, it isn't going to be happy >unless all the pieces are gone except for that one extra piece for the opponent. > >In every game I have ever played with a computer, being down a piece with a >couple of rooks bishops and queens on the board has resulted in the side that >is down a piece losing the game. I've tried to stop this. Perhaps here Qh5 >is a good move, but my eval dropped sharply after that. > >I thought that Rebel would win this after it had such a wall of pawns left on >the kingside, but it never seemed to try to get them moving, and a wall of pawns >on the 2nd/3rd ranks is not nearly so impressive as that same wall of pawns on >the 5th-6th... > >Interesting decision by Anand to start that Bd6 sequence. Two other comments I left off: 1. Obviously this was a good result by Rebel. And even though it didn't do well in the two long games, one draw and one near-win that became a loss vs Anand can't be considered a bad performance. 2. Just as obviously, computers still have some serious holes. The second game showed one, where all the programs were sure white was winning. I was too. And a GM that was logged on as a guest was also chatting with me during the game and he too thought that white would win because of the extra passed connected pawns. So the question is, did Anand (a) play Bd6 to enter a forced winning line or (b) did he play Bd6 to enter a line he was pretty sure a computer would screw up or (c) did he make a mistake in his calculations, but manage to win anyway because white blew up and made some weak moves? I still like white, but I'm not a GM. But after the trade on g6, Crafty was anxious to advance the pawns. Whether that would have worked or not, I don't know, but *obviously* they had to move. You can't afford to let black go to work with the R and 2 B's without trying to keep him off-balance and defending or stopping the pawn roller. Be a good game for analysis for quite a while, of course. And I didn't see any overwhelming reason to resign. Crafty was at -3, but that's not exactly what I call "prime resignation" territory... With those pawns, I'd want to give black plenty of room to make a one-move mistake, which might be all it took. And Anand certainly made plenty of mistakes through the match...
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.