Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Diep fullwidth vs Deep Blue fullwidth

Author: blass uri

Date: 23:05:20 12/21/99

Go up one level in this thread


On December 21, 1999 at 22:50:30, Robert Hyatt wrote:











<snipped>
>Baloney.  I saw them extend beyond 30 plies with a forced variation against
>my program.  They saw something about 10 full moves before we saw it.  I saw
>this against other programs...

I was not convinced that they saw 10 fulls moves before cray blitz in this game
because a team of me+computer could not see that the variation is forced even
after some hours.

I believe that the evaluation of +2 that they saw was wrong(maybe because they
could not detect a repetition and maybe because of other reasons(some bug in the
evaluation function) and that they won because cray blitz did a mistake later in
the game.

I believe that they did a lot of extensions but
I am not sure if it was a good idea because I believe that most of the tactics
in chess are not long forced lines and there are often cases when the opponent
has 2 or 3 logical moves and there are also cases when all the moves seem almost
the same but if you search deep enough you see that you lose.

Their first Deep blue lost against kasparov in game 5 because it could not see
this quiet tactics when many moves seem to have the same value and it was too
late when they saw it.

I believe that they could see it if they did less extensions and searched
deeper(of course a better evaluation could also help in this case)

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.