Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: The Next #1 SSDF no doubt Fritz6a

Author: Thorsten Czub

Date: 13:21:56 03/04/00

Go up one level in this thread


On March 04, 2000 at 12:03:33, blass uri wrote:
>The question is what is the definition of being better in tactics.

easy, for me better in tactics means: it does not make mistakes that
cause losing pieces or lose whole game in a chess-game.

better in tactics does not mean for me finding key moves in test-suites
since these positions often are TOO late.
the opponent made the mistakes a few moves or 1 move before, so it is not
often the program finding the key move that was good, but the opponent doing
very weak.
the difference to a chess-game bahaviour is important. here the programs
have PB behaviour, and they plan to come into certain positions (e.g. don't
allow their pieces to be trapped).
in a test-suite positions all this has happened before, but the finder
program does not have to take care about.
you can write a test-suite program that is very good but plays weak chess
in real games.

>My definition is that a program that can see faster a big change in evaluation
>is better in tactics and it is possible to prove that program A is better than B
>in tactics by stesting them in the same positions from games and see how much
>time they need to see the change.

some extensions may help solving test-suites, and programmers have to put them
into, but do not help playing better chess or beeing better in game-tactics.



>I believe that usually good solvers are better in tactics by my definition but

i have to disagree here.

>I think that it is possible that a program is going to be better in finding key
>moves and not be better in tactics.

this is my point, yes.

>For example if you reduce the value of the pieces  you can get a better solver
>but you do not get a program that is better in tactics.

right. tuning chessmaster can make such a FINDER out of it, but the finder will
not generally play better = stronger in tactics.

>Maybe you define as tactics part of what I define as positional understanding

positional understanding is different from tactical understanding IMO.
you mix them.
i don't believe in this.
sorry. again i have to disagree.



>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.