Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Comments on delta pruning

Author: Jon Dart

Date: 06:30:37 10/05/99

Go up one level in this thread


On October 05, 1999 at 05:14:23, Bas Hamstra wrote:


>(Material - Margin) is a cheap estimation of the current eval. If positional
>bonusses get frequently more than a pawn (or even more) "Margin" should be set
>to near the maximum the positional part of the score can be, or else this
>estimation can become very inaccurate and introduces much errors.

Yes. I have a compile-time switch that calculates what the pruning code
is assuming the score will be, then goes ahead and computes it. If the
score is bigger than the estimate, it prints a warning.

>
>If positional bonusses get frequently much more than a pawn, maybe it is more
>accurate to use this:
>
>	Delta = Alpha - Eval;
>
>Because with one move the Eval won't change *that* much. Reaction?

Depends a lot on what your eval code looks like. For example, I disable
qsearch pruning in the endgame, because one small move can create a big
change in the score (example: the king moves out of the "square of the
pawn" and can no longer catch a passer). If you are clever, you may be
able to quickly check for such cases without doing a full eval.

--Jon



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.