Author: Christophe Theron
Date: 12:53:58 05/09/98
Go up one level in this thread
On May 09, 1998 at 15:23:14, Bruce Moreland wrote:
>
>On May 09, 1998 at 11:06:59, Amir Ban wrote:
>
>>All true, but more easily believable if the opponent is CG 3.0 than
>>probably any other program in existence.
>
>Yes, I wanted to work a comment like this in, too, but didn't.
>
>People say a lot of nasty things about Genius, but it is monster-strong,
>especially at fast time controls.
>
>I've noticed it's the program most likely to have a high ICC rating when
>run on dumpy hardware.
>
>bruce
Genius comes from programs designed to play very well on the hardware
available in the 80s. At that time, Richard's creature was unbeatable
(well, almost).
It seems to be designed especially to catch quickly any 8-12 plies
combination.
I would like to know how its selective search is done. But I don't care
too much anymore, because I noticed that on fast computers this
algorithm is not that dangerous. Always good, but not as efficient,
compared to today's well know search techniques.
But still dangerous in blitz games, even on, say, a P200. I know that
very well. I have played hundreds of Tiger/Genius blitz games by hand!
And lost more than 50% of them...
Christophe
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.