Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Computer's are Grandmasters....Until next loss

Author: Robert Hyatt

Date: 06:00:11 06/12/01

Go up one level in this thread


On June 12, 2001 at 06:47:59, odell hall wrote:

>On June 11, 2001 at 09:05:21, Chessfun wrote:
>
>>On June 10, 2001 at 02:55:54, odell hall wrote:
>>
>>>On June 10, 2001 at 02:52:50, Slater Wold wrote:
>>>
>>>>On June 10, 2001 at 02:14:50, odell hall wrote:
>>>>
>>>>>On June 09, 2001 at 23:56:07, Slater Wold wrote:
>>>>>
>>>>>>No.  I don't think so.  It was only 2 games, and not even tournament length
>>>>>>games at that.
>>>>>>
>>>>>>When I saw how easily Petr defeated Deep Shredder, I was amazed.  He was
>>>>>>kibitz'ing most of the game.  Like he didn't even have to concentrate.
>>>>>>
>>>>>>Remember, he said that Deep Shredder performance was approx. 2100-2200 FIDE in
>>>>>>the game against him?  (Which it played 100x worse against him, I agree.)  Guess
>>>>>>what Petr said about these games?
>>>>>>
>>>>>>Thoughts on ending of game 1:
>>>>>>
>>>>>>Petrovich(GM) kibitzes: it's very difficult for human to play positions with
>>>>>>compensations for a pawn
>>>>>>
>>>>>>Petrovich(GM) kibitzes: may be just white has no sufficient compensation :)
>>>>>>Project played like Machine :))
>>>>>>
>>>>>>Game 2:
>>>>>>
>>>>>>Petrovich(GM) kibitzes: 4...Bxc3?
>>>>>>
>>>>>>Petrovich(GM) kibitzes: I beat Jrlok never play Bxc3 vs human GM :)
>>>>>>
>>>>>>Petrovich(GM) kibitzes: please think about it :)
>>>>>>
>>>>>>Petrovich(GM) kibitzes: comp played opening like patzer again
>>>>>>
>>>>>>And about the series:
>>>>>>
>>>>>>Petrovich(GM) tells you: hmm, comp played like 2200 in the closed position and
>>>>>>like 2600 in open :) It's hard to evaluate his strength
>>>>>>
>>>>>>Petrovich(GM) tells you: so, 2200+2600 / 2 = 2400 :) lol
>>>>>>
>>>>>>
>>>>>>
>>>>>>Petr and I just went over quickly the games.  These are not annotated or
>>>>>>anything, just general conversation on the games.
>>>>>>
>>>>>>Also, please be aware that when I played Petr, I did _NOT_ prepare at all.  I
>>>>>>asked him for a game, and we played.  Having been singled out by JRLOK, I
>>>>>>prepared an better book.  This time when I played JRLOK, there was no in & out
>>>>>>of book problems.
>>>>>>
>>>>>>Shredder played a much better game this time around, than last.  However, I am
>>>>>>not ready to say that Shredder is GM strength.  Like Robert said, to be 2600 you
>>>>>>have to play like a 2600 every game.  Not 2100 here and 2700 here.  2600 ALL THE
>>>>>>TIME.
>>>>>>
>>>>>>I am planning more and more GM vs Shredder games.  Perhaps at the end of this
>>>>>>year, we can make a FAIR guess at Deep Shredder's ELO.
>>>>>>
>>>>>>
>>>>>>Slate
>>>>>
>>>>>
>>>>> Well if what he said is true, a 2100-22000 just beat roman , one of the
>>>>>strongest grandmasters in computer chess around, and that is impressive.
>>>>
>>>>NO NO NO!  He rated my performance against HIM as being about 2100-2200!  Just
>>>>_ONE_ game!
>>>>
>>>>There's NO doubt Deep Shredder is stronger, it just had a "bad game."
>>>>
>>>>
>>>>Slate
>>>
>>>
>>>Yep just like many humans, i was surprised though that jrlok lost, since he
>>>knows computer chess so well, i expected a win.
>>
>>
>>Just my 2 cents for what it's worth.
>>Roman may lose to your program first few times, even first ten times.
>>But once he sees what book lines you have, what positions you play moves like a3
>>or h3 in. Then he will play some more and win some. Make no changes and he will
>>pound you to pieces. This is one of the reasons I think Bob believes computers
>>are NOT GM strength.
>>
>>Sarah.
>
>
>  Hi Sarah
>
>
>  I am curios what do you believe? Since you have done alot of testing of
>programs, (and i have been to your site, very nice). Do you think they are Gm
>Strength? I realize that technically they can never be Grandmasters , since Fide
>will not give the title to a computer.


Shouldn't you be asking a human GM?  Would you ask a policeman if "XXX" is a
great brain-surgeon???  Would you ask a brain-surgeon if "YYY" is a great
programmer?  Etc...

I don't think that someone that is _not_ a GM is really qualified to answer
this question, particularly if they don't know a _great_ deal about what goes
on inside a chess program...  and even then, they will need GM input...



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.