Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Junior 6 vs modified settingsNimzo 8 score W26 L18

Author: Terry McCracken

Date: 20:34:46 02/27/01

Go up one level in this thread


On February 27, 2001 at 23:18:09, Terry McCracken wrote:

>On February 27, 2001 at 21:13:47, eric guttenberg wrote:
>
>>The games can be found and downloaded by going to the SSDF site, clicking on
>>"games" and then clicking on Tony Hedlund's database of test games which
>>gets you to Tony's chess site where you download NEWPGN.  That will get
>>you some 600 new test games played by SSDF, including the 42 between Deep Fritz
>>and Junior 6.
>>
>>I am not suggesting that Junior 6 is stronger than Deep Fritz but the final
>>score of that "match" is what it is. Maybe Deep Fritz was just unlucky or
>>maybe Junior 6 just happens to be a program that gives Deep
>>Fritz trouble. I agree Deep Fritz outperforms Junior 6 in their games
>>against other common opponents.
>>
>>eric
>
>I will check on these games later, I did suspect they would be under NEWPGN.
>My experience though has been contrary to this match between DF and Junior 6.
>I'm glad you're not suggesting Junior 6 is stronger than DF, but in my engine
>match games, DF won by a fairly large margin, but chance could have played a
>role.
>One thing appears clear from my own tests, that Junior 6 does'nt give DF any
>real difficulty. Maybe DF was lucky? Although, I doubt it could be that lucky,
>most of the time. I think the SSDF testing of DF vs Junior 6 was probably a bit
>lucky for Junior 6.
>You said DF was closing the gap as time progressed. I believe
>DF would have caught up and surpassed Junior 6.
>
>Regardless, they are both strong prgramms.
>
>Shredder 5 appears to be better than them all from recent tounaments, but still
>a bit of luck may exist there, as the tournaments test many programms at once.
>There's not enough games to make a final case for Shredder or Deep Fritz, Fritz
>6 etc. DF only lost by half a point, that is really meaningless.
>Shredder would have to go one on one with Fritz for at least 200 games at
>40/2hrs., before a final judgement on their repective strengths could be passed.
>I also think it would be interesting to have the top 5 programms in a
>tournament, go head to head then I think we might get a better picture, which
>programm would probably be best.
>
>Terry McCracken

Also, these 5 top programms would need to take on the top 5 humans in the world,
not only to find out which is best, but best overall!
I'd love to see something like this, but I won't hold breath!;)

Terry



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.