Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Kiriakov - Deep Shredder 1 - 0

Author: Sune Larsson

Date: 16:44:01 06/02/01

Go up one level in this thread


On June 02, 2001 at 18:49:21, Slater Wold wrote:

>I agree 100%.  While C vs C is interesting, I have came to the conclusion that C
>vs C does NOT calculate ELO.



 Right. Handled optimal they can indicate differences in playing strength
 when about competition between comps. They even say very little about
 possible performance in games vs humans... Not taken down to any extreme
 of course - I mean, Fritz for sure gives me more trouble than Ufim...;)
 For me as a player, not gamer, progs like Crafty (massive battling vs
 top humans for a long time thanks to Bob's focus) and Rebel/Gambit Tiger
 are especially interesting. Junior 7 might also have something going.
 Not much difference between Deep Fritz, Gandalf 5, Deep Shredder though.




  According to SSDF, Deep Fritz is 2600+ and
>Shredder is technically the "World Champ" and SHOULD have a higher ELO.  However
>Petr said that Shredder performed @ approx. 2100-2200 in our game.



 He's obviously a very kind and generous man...;)


>
>And I do NOT disagree.  Shredder played like a complete patzer.  I have gone
>over and over the dump file from Shredder 1,000 times.  It changed plans every
>move.  It gathered no intiative, and developed nothing.  It simple played a very
>reactive game of chess.  Reactive playing is of course NOT the way to win.



 Not to mention the waste of tempi in moves 9-14... A definite no-no.


>
>I played the game, with the SAME time controls with Deep Shredder vs Fritz 6
>from the moment I went out of book.  (Out of book for good that is.)  Is it
>surprising that Deep Shredder won?



 :)) That's my point. Comp-comp games cover up these weaknesses. But just
 one game by Petr was needed here...


>
>Petr and I will be playing again soon.  I hope to play EVER GM on ICC at LEAST 1
>30/10 game.  After the game, I told Petr, "I searched the DB, and I noticed you
>have NEVER won a game against a (C) here with a standard timer.  How did you do
>so well here."  He simply replied, "I test theory with computers.  And none of
>those games were for money."



 Right. It's spelled motivation. Something to remember when browsing
 through results.


>
>Lesson being:  Don't tempt GM's with money.  :)


 ...cause some more of your illusions might get crushed... ;)

 Good luck with the games! Keep them coming!

 Sune
>
>
>Slate



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.