Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Super Kibitzer

Author: Uri Blass

Date: 02:45:56 11/30/00

Go up one level in this thread


On November 29, 2000 at 22:10:14, Joshua Lee wrote:

> Imagine Having a CB program that chooses the best score out of several engines
>Fritz Gandalf Nimzo Shredder Chess Tiger Rebel and any other program that is
>good at something that other programs aren't or a program that is unique in a
>special way like in it's programing or evaluation. I think if Fritz managed 2457
>on a 500Mhz Pentium
>on this "monster" running at 750Mhz *4 (or what cpu speeds compaq offers)
>This Super Kibitzer would be stronger than Deep Junior at Dortmund for several
>reasons
>1 it would have the speed that Deep Junior had and then some
>2 Have stronger Endgame play
>3 More precise Tactics Fritz Junior Nimzo and Little Goliath working together i
>wouldn't be suprised to here mate anounced by move 15 in 20 moves( ok that's a
>bit crazy)
>4 Positional Play ...Gandalf is great but i noticed something when i was looking
>at the Test your positional play positions and  the test at Tony Hedlund's
>webpage. Other program's scored higher but i bet the way to explain this is that
>even though one program may have more chess knowledge not every program uses the
>same piece values ect. So some luck comes into play, I wonder if anyone uses
>default piece values 9 for queen, 5 for rook, 3 for Knight's and Bishops, 1 for
>pawns.

I do not trust positional test suites because I guess that there are a lot of
mistakes in them.

Uri



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.