Nunn Tournaments updated

Archive of the old Parsimony forum. Some messages couldn't be restored. Limitations: Search for authors does not work, Parsimony specific formats do not work, threaded view does not work properly. Posting is disabled.

Nunn Tournaments updated

Postby Heinz van Kempen » 18 Feb 2004, 18:39

Geschrieben von: / Posted by: Heinz van Kempen at 18 February 2004 18:39:08:

Hi all,
Nunn tournaments B and C are updated. Quark 2.35, Movei and The Baron are added to all crosstables of Tournament B to have more games for all engines here and a direct comparison. All games for Quark 2.35 and updated crosstables are downloadable with position 9. Games for Movei and The Baron were in before.
For Quark I propose just to wait until all 600 games are finished. Just to mention one unusual result: Quark 2.35 vs. Quark 2.05b 6.5:13.5. There are also test matches in progress between both Quark versions and AnMon 5.30, WildCat 3.0, Gothmog 0.4.6a and Movei 00.8.158.
Concerning The Baron I will just test one or two newer versions, including probably 1.2.0 and then finally grab the old 1.0.2 version to run a huge gauntlet with it :-).


Pyramid results:
Amy 0.84 vs. Green Light Chess 3.00  0:1
Slow Chess 2.89b vs. Delfi 4.4       0:1
Cheetah 0.71 vs. Requiem 0.53        1:0 (congrats to Ralf for promoting to the top 100!) 


http://www.husvankempen.de/nunn/
Best Regards
Heinz
Heinz van Kempen
 

Re: Nunn Tournaments updated

Postby Robert Allgeuer » 18 Feb 2004, 18:57

Geschrieben von: / Posted by: Robert Allgeuer at 18 February 2004 18:57:26:
Als Antwort auf: / In reply to: Nunn Tournaments updated geschrieben von: / posted by: Heinz van Kempen at 18 February 2004 18:39:08:
Hi all,
Nunn tournaments B and C are updated. Quark 2.35, Movei and The Baron are added to all crosstables of Tournament B to have more games for all engines here and a direct comparison. All games for Quark 2.35 and updated crosstables are downloadable with position 9. Games for Movei and The Baron were in before.
For Quark I propose just to wait until all 600 games are finished. Just to mention one unusual result: Quark 2.35 vs. Quark 2.05b 6.5:13.5. There are also test matches in progress between both Quark versions and AnMon 5.30, WildCat 3.0, Gothmog 0.4.6a and Movei 00.8.158.
Concerning The Baron I will just test one or two newer versions, including probably 1.2.0 and then finally grab the old 1.0.2 version to run a huge gauntlet with it :-).



http://www.husvankempen.de/nunn/
Best Regards
Heinz
>Pyramid results:
>Amy 0.84 vs. Green Light Chess 3.00  0:1
>Slow Chess 2.89b vs. Delfi 4.4       0:1
>Cheetah 0.71 vs. Requiem 0.53        1:0 (congrats to Ralf for promoting to the top 100!) 
>
Hi,
is Quark 2.35 the newest version of Quark and publically available, if yes, where can it be obtained.
From your experience would you generally see Quark 2.05b being stronger than 2.35, or do you think that this was a statistical anomaly?
Thanks
Robert
Robert Allgeuer
 

Re: Nunn Tournaments updated

Postby Thomas Mayer » 18 Feb 2004, 19:02

Geschrieben von: / Posted by: Thomas Mayer at 18 February 2004 19:02:09:
Als Antwort auf: / In reply to: Re: Nunn Tournaments updated geschrieben von: / posted by: Robert Allgeuer at 18 February 2004 18:57:26:

Hi Roger,
is Quark 2.35 the newest version of Quark and publically available, if yes,
where can it be obtained.
From your experience would you generally see Quark 2.05b being stronger than
2.35, or do you think that this was a statistical anomaly?
The v2.35 is not yet available, but I plan to release it... but when I look at the results from Heinz I may think about it... :) According to my own observations, the v2.05 scores in blitz a bit better then the v2.35 -> but with longer terms the thing changes completely... I think starting with time controls like 40 in 10 minutes v2.35 usually achieves results the v2.05 never can... but of course I might be wrong... Anyway, usually almost all results I have from the newer Quarks since say v2.27 shows a clearly improvement over the old v2.05... Those from Heinz seems to be different.
Greets, Thomas
Thomas Mayer
 

Re: Nunn Tournaments updated

Postby Uri Blass » 18 Feb 2004, 19:04

Geschrieben von: / Posted by: Uri Blass at 18 February 2004 19:04:22:
Als Antwort auf: / In reply to: Re: Nunn Tournaments updated geschrieben von: / posted by: Robert Allgeuer at 18 February 2004 18:57:26:
Hi all,
Nunn tournaments B and C are updated. Quark 2.35, Movei and The Baron are added to all crosstables of Tournament B to have more games for all engines here and a direct comparison. All games for Quark 2.35 and updated crosstables are downloadable with position 9. Games for Movei and The Baron were in before.
For Quark I propose just to wait until all 600 games are finished. Just to mention one unusual result: Quark 2.35 vs. Quark 2.05b 6.5:13.5. There are also test matches in progress between both Quark versions and AnMon 5.30, WildCat 3.0, Gothmog 0.4.6a and Movei 00.8.158.
Concerning The Baron I will just test one or two newer versions, including probably 1.2.0 and then finally grab the old 1.0.2 version to run a huge gauntlet with it :-).



http://www.husvankempen.de/nunn/
Best Regards
Heinz
Hi,
is Quark 2.35 the newest version of Quark and publically available, if yes, where can it be obtained.
From your experience would you generally see Quark 2.05b being stronger than 2.35, or do you think that this was a statistical anomaly?
Thanks
Robert
>>Pyramid results:
>>Amy 0.84 vs. Green Light Chess 3.00  0:1
>>Slow Chess 2.89b vs. Delfi 4.4       0:1
>>Cheetah 0.71 vs. Requiem 0.53        1:0 (congrats to Ralf for promoting to the top 100!) 
>>
From the table
Quark2.35b has 0.5 point more than Quark2.05 after 270 games
It is not clear which version is better at 4+2
Uri
Uri Blass
 

Re: Nunn Tournaments updated

Postby Thomas Mayer » 18 Feb 2004, 19:07

Geschrieben von: / Posted by: Thomas Mayer at 18 February 2004 19:07:51:
Als Antwort auf: / In reply to: Re: Nunn Tournaments updated geschrieben von: / posted by: Robert Allgeuer at 18 February 2004 18:57:26:

Just to add:
the v2.35 did score in most positions better then the old v2.05... only in position 7 and 9 it scores worser (but then WAY worser... maybe I will take a look if this was always already caused by the first few moves...) -> so when it does better in 7 of 9 I am still satisfied... it's not always the numbers that are so interesting... and playing versions of the same program against each other is IMO not such a good idea at all...
Greets, Thomas
Thomas Mayer
 

Re: Nunn Tournaments updated

Postby Heinz van Kempen » 18 Feb 2004, 19:22

Geschrieben von: / Posted by: Heinz van Kempen at 18 February 2004 19:22:22:
Als Antwort auf: / In reply to: Re: Nunn Tournaments updated geschrieben von: / posted by: Thomas Mayer at 18 February 2004 19:02:09:
Hi Roger,
is Quark 2.35 the newest version of Quark and publically available, if yes,
where can it be obtained.
From your experience would you generally see Quark 2.05b being stronger than
2.35, or do you think that this was a statistical anomaly?
The v2.35 is not yet available, but I plan to release it... but when I look at the results from Heinz I may think about it... :) According to my own observations, the v2.05 scores in blitz a bit better then the v2.35 -> but with longer terms the thing changes completely... I think starting with time controls like 40 in 10 minutes v2.35 usually achieves results the v2.05 never can... but of course I might be wrong... Anyway, usually almost all results I have from the newer Quarks since say v2.27 shows a clearly improvement over the old v2.05... Those from Heinz seems to be different.
Greets, Thomas
Hello Thomas,
I plan not to count those games between different versions of one engine for rating calculation, but usually it is the other way round, that the new version scores extrmely good against the old one. In my opinion we have a statistical anomaly here and that is so far confirmed by the test games with Nunn positions I am currently running. Here Quark 2.35 scores better than the old version. Nunn test games and gauntlets also count for rating calculation and please again wait until more games are played. I am running these tests for years now and 270 games are by far not enough, believe me.
Anyway one other thing, Thomas, you could release version 2.27 as well perhaps. I know from one of your testers that this had results around +40 ELO better than 2.05b. And anyway I still believe in the 2.35 version, too.
Best Regards
Heinz
Heinz van Kempen
 

Re: Nunn Tournaments updated

Postby Peter Fendrich » 18 Feb 2004, 22:30

Geschrieben von: / Posted by: Peter Fendrich at 18 February 2004 22:30:17:
Als Antwort auf: / In reply to: Nunn Tournaments updated geschrieben von: / posted by: Heinz van Kempen at 18 February 2004 18:39:08:
Hi all,
Nunn tournaments B and C are updated. Quark 2.35, Movei and The Baron are added to all crosstables of Tournament B to have more games for all engines here and a direct comparison. All games for Quark 2.35 and updated crosstables are downloadable with position 9. Games for Movei and The Baron were in before.
For Quark I propose just to wait until all 600 games are finished. Just to mention one unusual result: Quark 2.35 vs. Quark 2.05b 6.5:13.5. There are also test matches in progress between both Quark versions and AnMon 5.30, WildCat 3.0, Gothmog 0.4.6a and Movei 00.8.158.
Concerning The Baron I will just test one or two newer versions, including probably 1.2.0 and then finally grab the old 1.0.2 version to run a huge gauntlet with it :-).



http://www.husvankempen.de/nunn/
Best Regards
Heinz
>Pyramid results:
>Amy 0.84 vs. Green Light Chess 3.00  0:1
>Slow Chess 2.89b vs. Delfi 4.4       0:1
>Cheetah 0.71 vs. Requiem 0.53        1:0 (congrats to Ralf for promoting to the top 100!) 
>
Hello Heinz,
May I suggest that you put the short description you gave (a few themes and variants) for each Nunn position, at you home page?
I can collect the 20 texts for you and send it to you. Just tell me!
/Peter
Peter Fendrich
 

Re: Nunn Tournaments updated

Postby Heinz van Kempen » 18 Feb 2004, 23:15

Geschrieben von: / Posted by: Heinz van Kempen at 18 February 2004 23:15:59:
Als Antwort auf: / In reply to: Re: Nunn Tournaments updated geschrieben von: / posted by: Peter Fendrich at 18 February 2004 22:30:17:
Hello Heinz,
May I suggest that you put the short description you gave (a few themes and variants) for each Nunn position, at you home page?
I can collect the 20 texts for you and send it to you. Just tell me!
/Peter
Hello Peter,
yes, is only a question of having enough time. As you know I am still very busy with correspondence chess.
I will put this tests I sent to all participants from Nunn A on my webpage this weekend and also a collection of around 2000 games from IM and GM for all 20 positions, if there is interst in.
For all other things like graphics etc. you will probably have to wait until summer.
Best Regards
Heinz
Heinz van Kempen
 


Return to Archive (Old Parsimony Forum)

Who is online

Users browsing this forum: No registered users and 18 guests