Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Question on EGTB generation theory

Author: GuyHaworth

Date: 17:39:47 07/03/02

Go up one level in this thread



Let us suppose you are trying to pick up the (depth of) wins for White.

a)  you are looking at a btm loss for Black:  backing this up to a wtm position,
you must be looking at a win for White.  You can give it a 'provisional' depth
of d+1 - but be prepared to reduce this depth if one of its successors turns out
to be a quicker win ... later in the algorithm.

This is possible with DTM(ate) EGTs but not DTC.

b)  you are looking at a wtm win for White.  Backing this up to a btm position
is fine - but you now _must_ check all successors of that btm position to make
sure it is a forced loss for Black.

Ken Thompson had a 'counting down the number of successors' idea - but found it
was not efficient.

g



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.