Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Further testing of Genius 6.5: +0 =4 -46 against Ruffian

Author: Ulrich Tuerke

Date: 06:16:57 10/25/02

Go up one level in this thread


On October 25, 2002 at 08:34:26, Uri Blass wrote:

>On October 25, 2002 at 07:53:23, Ulrich Tuerke wrote:
>
>>On October 25, 2002 at 04:55:37, Uri Blass wrote:
>>
>>>On October 25, 2002 at 04:43:52, Aloisio Ponti Lopes wrote:
>>>
>>>>On October 25, 2002 at 00:52:08, Dann Corbit wrote:
>>>>
>>>>>Try at something other than breakneck blitz speed and see what happens.
>>>>>
>>>>>I expect Genius to do much better at longer time control.
>>>>
>>>>My list is a bullet / blitz list. I'm not searching for standard time controls.
>>>>I agree with you that Genius might score better with longer time controls. The
>>>>same happens to older-than-ninety-years-old-players if you give them the hole
>>>>day to think about 1 move.
>>>>
>>>>Genius is _too_weak at bullet and blitz comparing with recent engines. Only 4
>>>>draws out of 50 games against Ruffian!
>>>
>>>I suspect that something is wrong with the games.
>>>Genius is known to be strong at fast time control.
>>>
>>>I know that Genius3  beated kasaprov at 25 minutes per game 1.5-.5
>>>
>>>only 4 draws against ruffian out of 50 games is results of engines that are
>>>200-300 elo weaker than Genius.
>>>It cannot be the result of genius in fair conditions.
>>>
>>>I suspect that the interface of Genoius6.5 is buggy and does not give genius
>>>fair conditions(I do not have this interface)
>>
>>I think it's possible that Ruffian outsearches Genius completely on modern
>>hardware.
>>However, in Blitz games on 80486 architecture, Genius would do much better than
>>this.
>>
>>Uli
>
>48-2 seem too much for me.
>
>ruffian even could not get more than 37.5-2.5 against my private movei at 7
>minutes per game and you can be sure that movei was outsearched completely by
>ruffian.
>
>Movei even did not use book in these games and to be sure of different games it
>simply always changed the first move.
>
>Ruffian could not keep movei winless and lost 2 games.
>
>That movei is not only outsearched by ruffian but has no knowledge about king
>safety or passed pawns or weak pawns.
>
>Genius is at similiar level to Crafty or Comet(certainly not more than 100 elo
>weaker than them.
>
>I expect it not to lose 48-2 against ruffian unless something strange
>happens(for example no learning and ruffian always repeat the same game)

Uri, I think that the result of a match of many games between 2 computer
opponents will often exaggerate the real difference in playing strength. This
could have happended here, too.

Nevertheless, after thinking twice, I agree that this result looks a bit
peculiar.

Uli

>
>Uri



This page took 0.01 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.