Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Finding mate in the quiesce search

Author: Alessandro Scotti

Date: 08:43:16 08/24/04

Go up one level in this thread


On August 24, 2004 at 11:32:28, Tord Romstad wrote:

>On August 24, 2004 at 10:56:13, Alessandro Scotti wrote:
>
>>On August 24, 2004 at 10:50:07, Álvaro Begué wrote:
>>
>>>>I tracked the cause to the quiescence search returning mate, where it doesn't
>>>>actually explore all the possible variations.
>>>Your diagnostic is incorrect. There is no problem with returning mate scores in
>>>the quiescence search. You must be doing something else wrong.
>>
>>If I evaluate a position using the quiescence search and that returns "mate in
>>5", how can I be sure that it's really a mate in 5 if the quiescence search
>>hasn't explored all the possible moves?
>
>It sounds like you have forgotten to allow the return of "stand pat" scores
>in the qsearch, or something similar. If you are at a qsearch node where the
>side to move is not in check, and all the moves you search are refuted by a
>mate, you shouldn't return the mate score, but the stand pat score.  The only
>way the qsearch can possibly return a mate score is if the side to move at
>the first ply of the qsearch is in check, and all legal check evasions are
>refuted by a mate.  And in this case, of course, you *can* be sure that it's
>really a mate in x moves.

Tord, I think this explanation and the magic words "stand pat" have rung a bell
in my head... :-)
Can't check the code right now but I think you put me on the right track...
thanks!



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.