Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Rebel vs Ferret on ICS

Author: Robert Hyatt

Date: 18:38:35 12/11/97

Go up one level in this thread


On December 11, 1997 at 19:55:29, Bruce Moreland wrote:

>
>On December 11, 1997 at 19:04:44, Robert Hyatt wrote:
>
>>I can remember about 2 years or 2 1/2 years ago that when I logged on to
>>ICC the thing was buzzing...  Crafty had been playing a GM, and in a
>>blitz
>>game announced a mate in 10... and it was big news.  Now everyone
>>ho-hums
>>when it announces mate in 30-40-50...   Bruce's is now the new
>>high-water
>>mark, although it probably won't stand for a long time.  But when you
>>think
>>about it, mate in N where N is > 10 is pretty amazing... :)
>
>I don't think it's a big deal since it's just precomputed data.  Sure,
>they are powerful, and they are good to have, and they take some time to
>build, but mate in 50 out of a database is no big deal, the program just
>looks up a number and prints it on the screen.  Might as well get
>excited about a !! move in someone's opening book.
>
>Mates found in a search are a little more impressive in my opinion.
>
>bruce


I agree to a point...  but I've been impressed with mates in way over 50
from Crafty, when none of the databases I use are anywhere within 10
moves
of that...  because it is able to search very deeply thru a lot of
captures
and enter a won ending...

IE in a GM game I saw a mate in 45 due to a hit on KBNK, while at the
initial position, there were still 12 pieces on the board.  So it
managed
to find a forced sequence of moves that removed 8 pieces...  that is
still amazing.

I agree that just hitting at the root is not so neat...



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.