Author: Peter Berger
Date: 08:45:13 01/02/04
Go up one level in this thread
On January 02, 2004 at 11:29:51, Dieter Buerssner wrote: >On January 02, 2004 at 09:05:05, Gareth McCaughan wrote: > >>On January 01, 2004 at 14:00:52, Dieter Buerssner wrote: >> >>> The idea of Roland: When "I" have a Q as only piece, and my K and pawns have no >>> moves, I cannot win. >> >>Could you expand on that a little? As it stands, the criterion >>you just described is met by the position 8/8/8/8/8/7k/8/6QK >>which is certainly not drawn :-). > >You are of course right. But a one ply search will be enough, to see, that this >is won, even when the above rule is used. OTOH, for the position discussed at >the start of the thread, a very long search will be needed, to see a draw (or in >other cases, to find the only drawing move). The idea cannot be used for pruning >as shown by your example. But supported by search, it should work rather >reliably (meaning no draw score in won positions and finding draw for the class >of positions it was designed for). <snip> > >I don't think, the idea has much (or any) practical relevance for games. > >Regards, >Dieter How would we know? It is a well-known idea in human games and I remember real games where it occured myself. As nearly all engines don't know about it, if it happened in SSDF games for example, it would most probably just go unnoticed. Peter
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.