Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Model Game for Testing Programs....... Position at 21

Author: blass uri

Date: 22:39:13 03/26/00

Go up one level in this thread


On March 26, 2000 at 18:03:33, Djordje Vidanovic wrote:

>Hello Chessfun,
>
>Exactly.  The first position (21.e4!!) is _very_ difficult.  We always tend to
>insist on programs being very good tacticians and the position in this case is
>based on a positional sac.  Thus the sac is tactical, but the grounds for it are
>positional.  The combination of the two is why 21.e4!! is one hell of a move to
>find.  I believe that many humans would be able to find it, but very few
>computer programs.  You are about right, I believe, that at least ten minutes of
>pondering are needed.  Would be interesting to test a program on a quad box -
>Deep Junior or Crafty, perhaps?  Or Ernst's Dark Thought?
>
>***  Djordje

I think that Deep Junior is not the right program for this position because it
probably needs a long time even only to find Nxe4 with a negative score(I have
not Deep Junior but only Junior5.9 but I assume that these programs are
similiar).

Junior has problems to find sacrifices of a piece for pawns because of the fact
that the value of the pawn of Junior is reduced(it cannot find Junior's famous
and good sacrifice against nimzo for the same reason).
I will not be surprised if people find that Junior5 is faster than Deep Junior
in finding Nxe4.

one of Fritz5 versions could find Nxe4 in 1 second on p200MMX but see a negative
score in the first minutes and expect fxe4 so I think it will also have problems
to find e4.

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.