Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Does Nimzo98 is a positional program?

Author: Robert Hyatt

Date: 18:31:02 03/11/98

Go up one level in this thread


On March 11, 1998 at 18:03:25, Torsten Schoop wrote:

>Hi all,
>
>I have tested some programs on a well known position:
>
>FEN: 3rr1k1/2p2ppp/p1nq1n2/1pbppb2/8/8/PPPPPPPP/RNBQKBNR w  - -
>
>I have made a note of the score of the programs after 1 min (on P-120):
>
>CSTal 287c = -4.32
>Zarkov 4.2b = -3.14
>Nimzo98 = -3.06
>M-Chess 7.1 = -3.03
>Shredder 2 = -2.81
>Comet A85 = -2.33
>Gromit 1.2 = -2.12
>CM5000 = -2.07
>Hiarcs 6 = -1.96
>Kallisto AEGON = -1.75
>Patzer 2.99z9 = -1.59
>Diogenes 4.11 = -1.40
>Rebel 9 = -1.36
>Genius 5 = -1.30
>Corel Chess 1.0 = -1.28
>Junior 4.6 = -1.20
>Socrates 3 = -1.13
>Fritz 5.01 = -1.06
>W-Chess 1.04 = -0.96
>TCB 0035 = -0.72
>Crafty 14.5 = -0.68
>Tjes 0.1.02 = -0.57
>
>
>Obvioulsy Nimzo98 is fast but not stupid.
>
>Crafty = -0.68 - a lot to do, Bob!
>
>Ciao
>Torsten


I get -1.5 from ply=1 thru 9...

But I really don't worry about this kind of position, because it is
impossible to get such a position against Crafty.  I have too much code
to encourage development before anything else, so it starts off in a
position that could never happen, which means it is probably going to
grossly misevaluate it.

you might try 14.12 which I tried, which is much closer to reality with
the -1.5...  but it won't ever understand bizarre positions like this.
I would suspect that Nimzo must grossly exaggerate tempi, which in this
position gives black a huge plus.  But I also suspect that will hurt it
in
other positions that are perfectly safe but somewhat undeveloped...



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.