Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Fantastic attack by Junior

Author: Dann Corbit

Date: 16:51:46 07/06/04

Go up one level in this thread


On July 06, 2004 at 19:26:15, Rolf Tueschen wrote:

>On July 06, 2004 at 19:17:22, Dan Honeycutt wrote:
>
>>On July 06, 2004 at 18:51:03, Rolf Tueschen wrote:
>>
>>>On July 06, 2004 at 18:46:10, James T. Walker wrote:
>>>
>>>>On July 06, 2004 at 18:43:21, Rolf Tueschen wrote:
>>>>
>>>>>On July 06, 2004 at 18:36:43, James T. Walker wrote:
>>>>>
>>>>>>On July 06, 2004 at 18:15:53, Rolf Tueschen wrote:
>>>>>>
>>>>>>>On July 06, 2004 at 18:06:06, Amir Ban wrote:
>>>>>>>
>>>>>>>>On July 06, 2004 at 17:41:05, Dieter Buerssner wrote:
>>>>>>>>
>>>>>>>>>Amir, congratulations for the nice game!
>>>>>>>>>
>>>>>>>>>I have not seen any info about the hardware beeing used by Junior at the current
>>>>>>>>>WCCC. Please tell it to the observers!
>>>>>>>>>
>>>>>>>>>With best regards,
>>>>>>>>>Dieter
>>>>>>>>
>>>>>>>>It's an HP Proliant server. This is all HP is allowing me to say :(
>>>>>>>>
>>>>>>>>Amir
>>>>>>>
>>>>>>>
>>>>>>>And this is allowed in accordance with the ICGA rules?? - Mhhhh. Strange.
>>>>>>
>>>>>>Why not?  Since there is no limit on hardware someone could show up with the
>>>>>>worlds fastest computer and it makes no difference.  I would be interested in
>>>>>>the NPS Junior is hitting.  While analyzing the game Diep-Junior I saw Junior 8
>>>>>>hit over 1.6M nps on my XP2400.  Bob says he is hitting 8M minimum!
>>>>>>Jim
>>>>>
>>>>>
>>>>>So it would be possible that I am connected with the strongest computer in
>>>>>Japan, the second strongest was just bought by the university of
>>>>>BadenWürtemberg, and my fictive Rorybest is completely ok within the rules???
>>>>>Fine! Thanks.
>>>>
>>>>If you can write a chess program to take advantage of it's speed I would surely
>>>>love to see it play chess.
>>>
>>>Ok, thanks, although I don't know exactly what this means or what it takes, at
>>>least I can understand that there is no automatical stronger output. Anyway it'a
>>>also a gambler motif I can see here. It's the typical trick we can expect, when
>>>Amir begins to play out his different personality joker in the last rounds -
>>>then you can no longer "prepare" or hope that your counter-tuning has success.
>>>Ok, all from a lay's heart. :)
>>
>>
>>So?  More power to Amir.
>>Dan H.
>
>I didn't expect that I had to explain what fair play means in sports. It's that
>at least you know in advance what opponents you have. Or if one is a spider or a
>flying eagle. I mean, I must know how I must tune my own baby. If I have a
>chance to equalize the hardware or if I'm so far backwards that it could be
>better to remain in my observer's hut.
>
>Or the other way round: what can I win if I win on the base of my unknown
>hardware advantages? Or the claim of a possible advantage...? Even in boxing
>super heavy weight, where you can have a weight you prefer, everybody can see
>your body as soon as you appear in the ring.

This is not a uniform platform event.  The title of this contest is computer
chess world champion, and Deep Blue could enter, despite a titanic hardware
advantage.  If you want to find what program works best on a certain piece of
harware, then SSDF is a much better choice for you anyway.  With any contest of
under 30 matches against each opponent, the uncertainty of who really is
stronger is simply enormous.  We proclaim a champion here.  It is like a soccer
match or a basketball game.  Chances are above average that the strongest
combination of hardware and software wins.  That is all.  If all the contestants
were of equal strength, then it would be a pure coin toss who wins.  If most of
the top contestants are of similar strength, then there is a huge uncertainty as
to which program really is stronger.  If one program is several hundred Elo
better, we would see it quickly.

>No, I don't buy that this here is fair. Also, that HP forbids to say must be a
>joke, because a company wants PR.

Suppose it takes 3rd place?  Will they want the exact hardware disclosed?  It is
obvious, like the Fritz entries called "Quest" or something like that.  If it
loses, then "Quest" lost.  If it wins, then "Fritz" wins.



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.