Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Goliath Light and small selectivity test (240 games)

Author: Robert Hyatt

Date: 08:45:05 10/02/00

Go up one level in this thread


On October 02, 2000 at 03:09:36, Jouni Uski wrote:

>I try to test if selectivity=1 is better than selectivity=2 (default) as
>claimed. I played 120 60/5 games with both settings in my AMD450Mhz against
>Fritz, Junior, SOS and Crafty. Result:
>
>selectivity = 2     58/120
>selectivity = 1     57/120
>
>Conclusion: selectivity = 1 is NOT better than default. So I stop testing.
>Besides You cannot use special engine settings automatically in Fritz interface,
>so it's too complicate to use them...
>
>Jouni

Assuming that this is the null-move R value, this is pretty well expected
to be the result.  R=2 is a bit better.  Not overwhelmingly, but a bit...

New crafty version will have the ability to set this as well, although you have
to set the upper and lower bound since it uses a dynamic R value.




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.