Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: what is the value of theory books for computer programs?

Author: Robert Hyatt

Date: 13:35:42 08/21/98

Go up one level in this thread


On August 21, 1998 at 07:45:14, blass uri wrote:

>
>On August 21, 1998 at 06:55:45, Amir Ban wrote:
>
>>On August 20, 1998 at 23:10:34, blass uri wrote:
>>
>>>I want to know if the book of Junior that I have(with the same probabilities) is
>>>identical to the book that was used in paris.
>>>
>>
>>Yes, but I interfered with the choices (in a legal way).
>>
>>
>>>If the answer is positive then Junior was lucky not to do a draw against a weak
>>>opponent like it did in the last round against chesstiger11.2.
>>>
>>
>>"Lucky" is an exaggeration. This is a rare occurrence.
>>
>>
>>>I think it is better to use books with probabilities of only 0 or 1 in events
>>>like paris.
>>>
>>>did Junior decide to play 1...e5 against fritz5 in the last round because of the
>>>opponent(In the other games it played 1...c5) or maybe it did not like to play
>>>1...c5 because of the only game it lost against virtualchess.
>>>
>>
>>I wanted to play the Spanish against Fritz. I wanted to play it against Shredder
>>too. Both did not cooperate. You are right that after Virtual (and Eugen) I
>>didn't want to see the Sicilian again as black. In the blitz, I let the regular
>>book play. The only game I lost in the blitz, against Nimzo, was you guessed it
>>a Sicilian as black.
>>
>>I don't believe much in books, and I think their importance is limited. I know
>>that this is a minority opinion.
>
>Books are not important if a program has good understanding of the early
>stage of the opening.
>
>It is interesting to see the results of a version without a book to see the
>value of books.
>Is there a version without a book that play in ICC or another tournament in
>order to see the difference in the elo between a version with a book and without
>a book?
>
>Uri
>>
>>Amir


This was done on ICC by someone using crafty...  he was interested in seeing how
much each of the "optional" pieces affected Crafty, so he played without the
book at all, with no "start book" (just a big wide normal book that let it play
lots of oddball openings) and without tablebases.  With no book, it does well
for a few games, but gets into trouble with the lack of variability I mentioned
before.  The wide book is quite good for blitz, because the best thing you can
do is put a human into a line he's never seen before, when he is playing 5 3
or some such fast time control.  :)



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.