Author: Sune Fischer
Date: 16:19:06 01/10/02
Go up one level in this thread
On January 10, 2002 at 16:34:49, Uri Blass wrote: >> >>This problem is easy to fix. >> >>tell it never to resign in comp-comp games. >>It is good not only because of this reason. >> >>The other program may have a bug and miss mate in 1. >> >>Uri > >Another possible solution is to tell it never to resign based only >on mate score. >The other program may not have the right tablebases and this case is more >practical than the case of draw by the 50 move rule. > >The static score of the position on the board without tablebases should be also >a convincing score of a big advantage. > >Uri The way I see it, the computer might choose one particular endgame because it sees _that_ endgame as won, when in fact it could be drawn. If the endgame tables ignore the 50 move rule, there will be these cases (although rare I think) when the comp. is tricked. Maybe one could make a check at the root just before entering the 5 man tables, that either the position has a pawn or there will be a mate or capture within 50 moves, else it should see it as drawn no matter what the table says. It is last chance to avoid that endgame, and choose a safer one with a pawn or whatever. -S.
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.