Computer Chess Club Archives


Search

Terms

Messages

Subject: Christophe Theron respond to this post of Hyatt! I want your response.

Author: Charles Unruh

Date: 17:41:08 11/01/99

Go up one level in this thread


Inquiring minds want to know!

>On November 01, 1999 at 10:56:41, Bas Hamstra wrote:
>
>>Hi, Vincent,
>>
>>Why you always say piece square programs are worth nothing? Not that I use them,
>>but I am still considering it.
>>
>>Because you can put much more knowledge in your program than in a ordinary
>>endnode processor. You can build enormous expertsystem-like evals and still have
>>a reasonable speed.
>>
>>
>
>
>They really don't work very well.  Because what you evaluate at the root is
>_nowhere_ near what you reach at the tips.  I, too, believe that piece/square
>tables offer some good things...  ie what better way to determine that a piece
>is centralized, or on the edge of the board.  But that ends up being < 1% of
>what you want to know about a position.
>
>Berliner described 'discontinuities' in the evaluation function some 20 years
>ago.  This is a perfect description of what piece/square tables do today.  Every
>new move has an eval that is discontinuous from the previous move.  Which is not
>the way to win chess games over the long haul.  Frans showed that they can be
>effective.  But over time, others have shown that (at least against humans) this
>is not the way to go.  Computer vs Computer?  Who knows or cares...
>
>
>
>>Succes op het NK!
>>
>>Regards,
>>Bas.
>>
>>
>>
>>On November 01, 1999 at 09:03:20, Vincent Diepeveen wrote:
>>
>>>On November 01, 1999 at 05:07:36, blass uri wrote:
>>>
>>>>On November 01, 1999 at 03:34:38, Jeroen Noomen wrote:
>>>>
>>>>>On November 01, 1999 at 03:00:04, Harald Faber wrote:
>>>>>
>>>>>
>>>>>>Isn't it amazing, the great Tiger far behind? :-)
>>>>>
>>>>>Chess Tiger is very unlucky in this tournament. It had great positions
>>>>>against Nimzo, Diep and Insomniac, but failed to win them.
>>>>>
>>>>>Best regards, Jeroen Noomen
>>>>
>>>>I did not see a clear advantage for tiger in the games against diep and
>>>>Insomniac
>>>
>>>
>>>DIEP was desparate against Tigre short after book. Then lucky Tigre
>>>allowed to exchange pieces. Then score got zero and the endgame tigre nearly
>>>played f6?? at a point but lucky for tigre failed low in time and played
>>>a much better move.
>>>
>>>I'm very impressed by the good preparing of Jeroen. I'm not impressed by
>>>the Tigre engine.
>>>
>>>Piece Square tables in the way tigre uses it don't work anymore.
>>>
>>>It was having the fritz5.16 phenomena bigtime. Before exchanging queen
>>>very happy. Then after it was exchanged suddenly seeing it was a more or
>>>less drawn position.
>>>
>>>>
>>>>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.