Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Looks like Crafty and Fritz tie for 4th

Author: Robert Hyatt

Date: 10:15:11 07/12/04

Go up one level in this thread


On July 12, 2004 at 10:44:11, Peter Berger wrote:

>On July 12, 2004 at 09:47:54, C McClain Morris, Jr. wrote:
>
>>Congrats to Dr Hyatt. It seems that Crafty's book was not up to par,
>
>What makes you think that?


I think the book worked pretty well.  The two openings that made me shudder,
were crafty as black in a Ruy where it played Bd6 blocking the d pawn (against
Falcon I think, could have been Be6 blocking e pawn, I don't have game handy)
and the ParSOS opening where it played the early b3.  It is not easy to cover
every possible hole in the book, and have something reasonable planned for every
possible opponent move.  It is almost impossible.

There are some things that need work.  It'll be better next year.  And next
year, regardless of whether someone else doe sthe book or if I do, there will
_still_ be an occasional bad line chosen by Crafty.  Just as surely as other bad
things happen in non-chess events. :)

For a program with a lousy search, primitive evaluation, poor parallel search
performance, crumby book, "rude quiescence search", a program that had no chance
to win, a program that would have been competing for years had the author
thought he had a chance to win, I think it did pretty well.

We'll see about better hardware for next year maybe, depending on when and
where, and if there will be any significant number of participants.




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.