Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: If programs are only IM strength, why bother to play Kramnik??

Author: Kurt Utzinger

Date: 15:57:46 05/12/01

Go up one level in this thread


On May 12, 2001 at 18:34:29, Jonas Cohonas wrote:

>On May 12, 2001 at 18:26:31, Kurt Utzinger wrote:
>
>>On May 12, 2001 at 18:08:38, Jonas Cohonas wrote:
>>
>>>On May 12, 2001 at 16:28:51, Kurt Utzinger wrote:
>>>
>>>>On May 12, 2001 at 16:23:21, Jonas Cohonas wrote:
>>>>
>>>>>On May 12, 2001 at 16:11:09, Kurt Utzinger wrote:
>>>>>
>>>>>>In my opinion no program can - on tournament level 40/120 - get more than 2300
>>>>>>ELO against solid human players. The shorter the time control the higher the
>>>>>>perfomance of the computer programs. And by the way, the SSDF list is at least
>>>>>>about 200-300 ELO too high.
>>>>>>Kurt
>>>>>
>>>>>Then how did DJ6 score 2703 against Human Gm's???
>>>>>
>>>>>Regards
>>>>>Jonas
>>>>
>>>>Study the games, and you will not much see of what I understand about soldid
>>>>playing. As already mentioned earlier I still get about 30 % against the
>>>>monsters without using any anti computer strategy. And this should simply not be
>>>>possible for a player with about 2000 ELO?
>>>>
>>>>Regards
>>>>Kurt
>>>
>>>What hardware do you run the progs on when you get a 30% score? is that a 30% of
>>>ALL the games you have played against progs? if so could you please post ALL the
>>>games so i can see the 30% score with my own eyes? To say that the games DJ6
>>>played against GM's where not what you concider "solid" chess is to say that the
>>>GM's played weak against DJ6 if they had played "solid" they would have
>>>capitalized on DJ6 "unsolid" play, to play for a draw against any of the top 10
>>>prog's, atleast 1 year ago, is not so hard if that is your main goal from the
>>>word go, but to win is whole other kedle of fish!!! and since GM's play to win
>>>and not to DRAW! they wil get stomped once in a while.
>>>
>>>Regards
>>>Jonas
>>
>>I think you have mentioned something very important: the GM's almost always play
>>to win and unfortunately even in situations where positions are equal. This does
>>not function against computers. And furthermore we should not forget one thing,
>>namely the psychological factor: if GM's are playing vs machines in public, the
>>are under a severe pressure to demonstrate their strenght. Please look at the
>>games Rebel Century3 vs John van der Wiel: It is hardly understandable how many
>>times van der Wiel made tactical blunders. Would the GM have made the same
>>mistakes against a human opponent? I am not at all sure. My opinion is the
>>following: If John van der Wiel played against computers at home without being
>>"disturbed" by any kind of public pressure, he would at 40/120 win against Rebel
>>Century3 without much difficulty.
>>Regards
>>Kurt
>
>Is that a fact? do you know that he would win or are you doing a wee bit of
>speculation here? you say that it  does not work to play for a win in equal
>positions against a program, is that because of their "unsolid" play? and when a
>2700+ human GM tries to win and fails is that because the program is unsolid??
>If anything todays topten programs run on the best hardware are very SOLID!!!
>
>Regard
>Jonas

Of course it's not a fact but merely a speculation based on some experience.
Some years ago I played with board computers at our club tournament and noticed
that in most cases the "strong" players showed a comparatively bad chess (quite
below their standard) whereas the weaker players had often not much difficulty
in drawing or even winning. Now it's late, I need some sleep (1.00).
Regards
Kurt




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.