Author: Uri Blass
Date: 05:02:32 11/06/01
Go up one level in this thread
On November 06, 2001 at 07:18:29, Leen Ammeraal wrote: >In the directory containing my Nalimov databases, >when I give the command > > dir kqr*.* > >I obtain the following output: > >KQRKNBB EMD 296,182 10-04-99 2:31p KQRKNBB.EMD >KQRKNBW EMD 211,121 10-04-99 2:31p KQRKNBW.EMD >KQRKQNBW EMD 21,727,039 04-11-99 11:37a KQRKQNBW.EMD >KQRKQNBB EMD 37,812,136 04-11-99 11:40a KQRKQNBB.EMD > >This suggests that I have no file for the KQRkr endgame, >which may very well be the case, since I know that >my set of 5-man files is incomplete. > >I had always thought that this should not be a serious problem, >because my program will probably deal well with KQRkr >endgames without any egtb. However, I just found that there >are a serious problems in the case of promotion moves, >as in the following trivial position: > >[D]8/8/8/8/6r1/8/3K1kp1/R7 b - - 0 3 > >My program sees that black deserves a very high score, >derived from the egtb, but fails to make the trivial >move 1. ... g1Q because it does not find an entry >for the resulting position in the egtb, and because >the computed score after this promotion move >is lower than the egtb score retrieved after the >move Rg3, so the latter move is made and the >game results in a draw instead of in a win for >black. Has anyone encountered similar problems? It is a known problem and all the commercial programs that I know had the same problem when they started to use tablebases. I was disappointed from it because I expected the commercial programs to test their programs before releasing them and not after it. The solution that was posted is also known. 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.