Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Smirin match conclusion?!

Author: Mark Young

Date: 04:16:29 04/30/02

Go up one level in this thread


On April 30, 2002 at 06:54:25, Uri Blass wrote:

>On April 30, 2002 at 06:44:25, Mark Young wrote:
>
>>On April 30, 2002 at 06:22:31, Otello Gnaramori wrote:
>>
>>>On April 30, 2002 at 04:44:41, Mark Young wrote:
>>>
>>>>On April 29, 2002 at 12:04:09, Chris Carson wrote:
>>>>
>>>>>On April 29, 2002 at 07:14:39, Jouni Uski wrote:
>>>>>
>>>>>>I think we can make two conclusion, even if one game is played later:
>>>>>>
>>>>>>1) Motivated super GM can still beat all PC programs
>>>>>>2) There isn't any program, which is specially good against humans
>>>>>>
>>>>>>Point 1) is actually good - there is still room for improvement, even if it's
>>>>>>unnecessary for 99,9999% of chess players.
>>>>>>Point 2) Fritz is as good as any program to play Kramnik. And with 8 CPU system
>>>>>>it can at least make interesting games.
>>>>>>
>>>>>>Jouni
>>>>>
>>>>>Number 1 is not true:
>>>>>Junior and Hiarcs have not lost a game to GM Smirin.  Fritz and Rebel did not
>>>>>get to play GM Smirin.  I guess guess you you could say that the motivated 2700+
>>>>>Super GM can win one game against Shredder and Tiger.
>>>>>
>>>>>Point 2 is not true:
>>>>>Rebel, Junior and Hiarcs all have euqal or positve records against 2700+ GM and
>>>>>Fritz did not play against the 2700 GM
>>>>
>>>>
>>>>I don't understand his match conclusions?
>>>>This was not 1 computer program playing, it is better to understand the match as
>>>>it was, 5 matches played vs the same human Grandmaster.
>>>>
>>>>1. The Hiarcs 8 match was evenly played, and +1 vs Gulko.
>>>>
>>>>2. The Junior 7 match was 1 draw, 1 null game were the computer's position was
>>>>won, even score vs GM Gulko
>>>>
>>>>3. The Shredder match was 1 loss, 1 draw, and an even score vs GM Gulko.
>>>>
>>>>4. Chess Tiger match was a joke. It did not play...
>>>>
>>>>5. The Gambit Tiger(anti-human) match. 1 draw.
>>>>
>>>>6. The Gambit Tiger(default) match. 1 loss.
>>>>
>>>>Conclusion: GM Smirin won match 3. GM Smirin should of lost match 2. GM Smirin
>>>>drew match 1. GM Smirin drew match 5 in a one game match, and won match 6 in a 1
>>>>game match.
>>>>
>>>>Conclusion #2 PC programs have once again shown they can play at a High GM
>>>>level.
>>>>
>>>>Conclusion #3. PC programs need to lose the human element. SEE match 2 & 5
>>>
>>>Mark,
>>>I think that the final right result if Junior will win tomorrow, should be
>>>4.5-4.5 ,in consideration of the fact that DJ won de facto against Smirin in the
>>>nulled game.
>>>
>>>Otello
>>
>>
>>I can't really argue, since the position was won for DJ, and GM Smirin did
>>resign! See Conclusion #3
>>
>>It must be noted that DJ, and Hiarcs 8 have played two very strong GM's. And
>>have not be beat yet.
>>
>>Gambit Tiger was not allowed to play a fair match by human mistakes.
>
>I do not see what was unfair in the game that GambitTiger lost.

Nothing is unfair in the GambitTiger(default) loss, unless you look at it as a 2
game match, because it was not.

>
>People said that ChessTiger14 could reproduce the same moves so even
>if you say that ChessTiger14 should be used then it seems that the
>choice of the program changed nothing.

Not true. I know it game one (anti-human) it would not have played the same as
CT.

CT did not play so I don't know what it would have played in all moves in both
games.

When CT plays we can eval CT....this was Gambit Tiger and I know human error
changed game 1 for sure, Game 2 stands as a loss in a 1 game match.

>>
>>Shredder was the only program that was beaten fairly, but still is only -1 and
>>played with a GM level result.
>
>Shredder used Junior's book in the games so you can claim that
>it was unfair if you look for excuses for the computers.

Using Junior's book is not unfair in my mind, changing the programs eval
function is unfair when it is known that anti-human mode weakens the program
greatly as was seen in game 1. GT was lucky to draw that game.

>
>This was the choice of the programmer to use Junior's book so
>I think that nothing was unfair

argue with yourself on this, I never said it was unfair.

>
>Uri



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.