Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: More on Rigged Opening Books

Author: Dann Corbit

Date: 14:14:23 05/30/01

Go up one level in this thread


On May 30, 2001 at 16:04:42, Alan Grotier wrote:

>
>Once again I agree with you.
>Of course a chess program must have an opening book repertoire to be able
>to compete.But it seems to me with my limited knowledge that too much emphasis
>is put on developing an opening book to give the program it's strength.

When it is the best possible source of strength, the programmers would be idiots
not to use it.

>From a consumers stand point I would prefer to have a chess engine that can
>play good chess with-out an opening book because it understands or at least
>attempts to understand the position / situation at any given moment.

Delete your opening book then.
;-)

>This does not mean precluding learning features or other enhancements!
>Just no book included.

Easily accomplished.

>Then one must be able to add any opening book of choice including the program's
>author book which would presumably be the strongest.
>
>I own F5.32 and it does not make opening stupid moves with book-off.
>With book-off as black it plays: e4 e5, d4 d5, c5 e5 and b3 e5.
>     book-off as white it plays: d4 or e4.

I notice that chess engines have preferences.  When I used to play crafty
without an opening book, it almost always played the sicilian dragon.

But I guarantee you that without an opening book, chess engines will play really
stupid moves from time to time.  The faster the time control, the worse it will
play.

An experiment:
1.  Remove your opening book.
2.  Set the board to this position:
[D]r1bqk1nr/pppp1ppp/2n5/b7/2BpP3/2P2N2/P4PPP/RNBQ1RK1 b kq -
3.  Have it analyze away.

If your engine says: dxc3 then it's wrong.  Nge7 is much better.

A computer with a good opening book won't play that.  A computer without an
opening book is very likely to play dxc3.  Let it buzz away until it finds the
right move.  Consider how it would turn out in a blitz game.

On many occasions, Vincent has shown severe weaknesses even in 16 ply or more
computer analysis for opening positions.



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.