Author: Christophe Theron
Date: 10:11:28 10/18/00
Go up one level in this thread
On October 18, 2000 at 04:06:57, Uri Blass wrote:
>kasparov played 42...g5 and some bean counters like hiarcs7.32 and Fritz6lihght
>like this move unless they have enough time.
>
>kasparov played 45...Kg6 and not 45...Re7 that was suggested by gambittiger
>because kasparov is similiar to the bean counters.
>
>kramnik also follow the line of the bean counters and did not play 47.Rc7.
>
>I do not say that gambittiger is wrong(I do not know who is right here) but only
>that the claim of trying to emulate the human brain is wrong.
>
>I believe that Christophe tries to find better evaluation function for tiger
>about king attack and not to emulate the human brain.
That's right, Uri.
I'm not trying to emulate the human brain. First I don't know how to do it, and
second I guess a computer has his own strengths and it is intersting to use
them.
My point is that every human player knows that a strong king attack is in many
cases worth much more than a pawn, even as much as a knight or a rook.
And I believe this knowledge is right both for humans and for computers. But
many chess programs do not have this knowledge (I'm not insulting anybody BTW,
my own Chess Tiger 12 was lacking this knowlege as well).
I think it is interesting to put this knowledge on top of a bean counter, it
should produce some new kind of playing style.
Christophe
>I expect programs to have better evaluation function than humans.
>It was done in backgammon.
>The problem in chess is more comlicated but I guess that it can be done also in
>chess.
>
>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.