Author: Uri Blass
Date: 04:05:41 04/17/02
Go up one level in this thread
On April 17, 2002 at 06:39:34, Miguel A. Ballicora wrote: >On April 17, 2002 at 04:12:12, Steve Maughan wrote: > >>Martin, >> >>>this is the reason i asked - if a program believes this position to be +1.4 or >>>+2.2, then it will go for this position instead of another promising >>>continuation. when you analyse games and the computer tells you you should have >>>played this and won, it's funny and not funny at the same time :-) >>>talk about comps being GMs... >> >>I would say that this is quite an extreme position and a classically tough one >>for computers to deal with. Yes it will come up - especially in anti-computer >>style chess but the code and heuristics to handle this position are not trivial. >> In general I would say that Shredder is excellent at the endgame and this is >>just one hole in the evaluation. >> >>Regards, > >Do the programs avoid Ng4 in this position throwing away black's advantage? >(I just modified it) >My program cannot avoid it. >Miguel > >[D]8/1k6/5np1/p1p4p/1pP1p3/1P2PpPB/P1K2P2/8 b - - My program has no problem here because the only knowledge that it has about pawn structure is that 2 pawns on the same file is bad. It has no knowledge about pawn endgames. A lot of other programs also avoid it(for example Junior7 or shredder5.32) They do not know like my program that Ng4 is a draw but they know that the score for black after other moves is better(shredder5.32 says g5 with a score of +2.80 for black after few seconds) Uri
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.