Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: 12.Nxg5! seems OK

Author: Slater Wold

Date: 22:23:02 06/16/02

Go up one level in this thread


On June 17, 2002 at 01:16:45, Gian-Carlo Pascutto wrote:

>On June 17, 2002 at 01:11:12, Slater Wold wrote:
>
>>On June 17, 2002 at 00:51:33, Gian-Carlo Pascutto wrote:
>>
>>>On June 17, 2002 at 00:16:36, Slater Wold wrote:
>>>
>>>>I believe that this position exploits the fact that *most* programmers give a
>>>>bonus for captures.
>>>
>>>I don't know of a single program that does this.
>>>
>>>--
>>>GCP
>>
>>TSCP, and I *think* Crafty did.  (At least, ordered capturing moves higher than
>>non-capture.)
>
>That's a difference from giving them a 'bonus'...unless the scores are
>exactly equal, it won't matter. And at the root many order differently.
>
>--
>GCP

Well, not exactly.

I know that TSCP give an actual "bonus" for capturing moves.  I *think* Crafty
does/did too.  Perhaps it only orders them higher.

However, if Program A orders capturing moves higher than non-capture moves, and
it stumbles across position X where the score for a capturing and non-capturing
move are the same, it's going to play the capturing move.

Anyway, perhaps I was mistaken.  But, I could have sworn I read a thread a while
back (on move ordering) where the common voice was to give a capturing move a
"small bonus".

Maybe it was that other computer chess board I troll.  :)



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.