Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Something wrong with my futility pruning?

Author: Severi Salminen

Date: 02:03:11 12/06/00

Go up one level in this thread


On December 05, 2000 at 17:42:47, Jon Dart wrote:

>On December 05, 2000 at 16:46:17, Severi Salminen wrote:
>
>
>>BTW. Does my node counts look reasonable? I'm using MVV/LVA ordering in qsearch
>>+ nullmoves (R=2).
>
>No, they don't reasonable. I'm getting less than 1/8 of your node count
>at ply 8.

Are you using hash tables, killers or history heuristics? I'm not. I'm not
sorting my root moves also anyhow. Maybe this is one of the reasons.

>I highly recommend instrumenting your program so that it prints out
>a search tree showing node alpha/beta windows, scores, and cutoffs. A
>close examination of this can show you what's going wrong (of course,
>looking at 4 million nodes isn't feasible, but if you have a basic
>problem then it will be apparent at low search depths). You can compare
>your output to another program like Crafty : Crafty's trace command
>will produce this information during a search (you will need to
>compile it with -DTRACE to enable this feature).

Have to try, thanks.

Severi



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.