Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: over zealous extensions

Author: Andrew Platt

Date: 10:16:44 09/26/04

Go up one level in this thread


On September 26, 2004 at 13:05:00, David B Weller wrote:

>check, pawn_2_7th, mate_threat are all here in abundance.
>
>I do only 1 ply total each ply, but I _was_ 'accumulating' them and then using
>them later so _every_ extension got 'done' eventually even if 1 or 2 ply
>later...
>
>so, a pawn to 7th which gave check _AND_ was in a node where there was a
>mate_threat would accumulate 3 ply of extensions to be used over the next 3
>plies
>
>I think I need to work on that ... :)

Yes. It looks like it would be a simple matter to disable that, run against some
test suite you have numbers for and compare. I bet the results are largely
better in the version where you restrict to a single ply.

It doesn't seem that there is any real difference between spreading the
extensions lower down and just increasing the depth by > 1 at the node you
generate all those extensions. You still end up extending a bunch of moves that
wouldn't qualify for any extension by themselves.

Andy.



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.