Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: To build a book or not?

Author: Russell Reagan

Date: 19:56:37 07/17/02

Go up one level in this thread


On July 17, 2002 at 20:02:33, Uri Blass wrote:

>I think that in order to teach a computer to understand something you need first
>to define it.
>
>You do not give a definition why it is good to castle but you expect programs
>to know the definition that you seem not to know.
>
>I think that you expect too much.
>Programmers have problems to explain to computers without bugs things that they
>know to define and you expect them to explain to computers things that they even
>do not know.
>
>Uri

Let me make sure I understand you. You are saying that to use the common method
of piece tables and other penalties and bonuses it doesn't require the
programmer to have to understand a great deal about chess. To use the other
approach, you would have to have extensive chess knowledge yourself, right? I
think that is fair, and it certainly would be more difficult to use the harder
method, but I think that we should strive for that and not just settle for
something because it is the easiest.

Russell



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.