Author: Dieter Buerssner
Date: 13:46:14 05/04/04
Go up one level in this thread
On May 04, 2004 at 10:44:43, Tord Romstad wrote:
>When there is little thinking time left, my engine annoyingly often fails to win
>KQKR endgames.
Does it really happen often? I cannot remember to have seen this endgame in
enginematches. Of course, I will have forgotten, but it may indicate, that it
does not happen often (or that typically both opponents have TBs, and resigned
long before).
>Are there any simple but effective eval tricks which can help to
>play such endgames well without searching very deeply?
Ernst A. Heinz suggests (w = Q/winning side, l = losing side)
score = 400 + 1/8 * dist(K_l,R) - 1/8 * edge_dist(K_l)
- 1/8 * corner_dist(K_l) - 1/8 * dist(K_l, K_w)
in his paper about knowledgable encoding and querying of endgame databases. In
that framework, already W/D/L info is available, but I would guess for this
specific endgame, it won't help much. You know, that you have won, but you have
to find moves that make progress, anyway. Perhaps the above factors can be tuned
a bit. I didn't try it myself, yet.
Regards,
Dieter
This page took 0.01 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.