Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: To build a book or not?

Author: Mogens Larsen

Date: 13:23:28 07/16/02

Go up one level in this thread


On July 16, 2002 at 11:04:40, Robert Hyatt wrote:

>In general, you are going to have to do this code development _anyway_ so that
>you can combat people that play 1. h3 and 2. a3 for example.  Because if you
>play silly opening moves, white can afford to give up a couple of tempi in
>order to get you to play garbage moves.  So you _have_ to fix that.  But when
>you do, you will _still_ fall into traps that are too deep for your search to
>see, and that will be the next problem you have to face.
>
>Opening books provide two key things:
>
>1.  quick moves to save time and provide randomness so that the program
>plays a variety of different opening systems.  Without a book, it will play
>the same stuff over and over and eventually lose the same games over and
>over and over.
>
>2.  avoids known deep traps.  This is not nearly as common as the point 1
>above, but it can be critical if people are trying to "prep for you" by
>planning on playing obscure lines that you won't know, to get you into a
>position where you are lost without knowing it.

I think adding:

3 (or 1b.). Aiding (proper) development and perhaps castling. Quite a few
engines are incapable of finding solutions to either without getting into
difficulties. Repeating it over and over again doesn't make it better of course.

Regards,
Mogens



This page took 0.01 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.