Pseudo test

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.

Pseudo test

Postby Jan Klima » 06 Oct 2004, 18:33

Geschrieben von:/Posted by: Jan Klima at 06 October 2004 19:33:36:

Hi,
is there anyone who could do the following:)
1)match 0.6b against 0.6f(or 0.6g after released),40+ games(rather more) under some gui like arena or chessbase with third party opening book(not nunn positions),engines play the position after the last book move as both black and white
2)time control>=5+2s,=15%(!!!!!) nps somewhere and i wonder if I got something for it. It might be that there's some overtype(typist error? or how to say that in english) that spoils everything or it's just that i made it safer and safer with every new version to be sure there are no bugs during WBEC tournament. But anyway what matters here is if the new versions(0.6f+g,0.6e is buggy) get at least 55% against 0.6b.
Jan Klima
 

Re: Pseudo test

Postby Peter Eizenhammer » 06 Oct 2004, 20:48

Geschrieben von:/Posted by: Peter Eizenhammer at 06 October 2004 21:48:16:
Als Antwort auf:/In reply to: Pseudo test geschrieben von:/posted by: Jan Klima at 06 October 2004 19:33:36:

Hi Jan,
i have 0.6f (and 0.6).
If you send the 0.6b to eizenhammeratweb.de (at as usual)
a little test gets started.
Peter
Peter Eizenhammer
 

Re: Pseudo test

Postby Axel Schumacher » 07 Oct 2004, 04:32

Geschrieben von:/Posted by: Axel Schumacher at 07 October 2004 05:32:59:
Als Antwort auf:/In reply to: Pseudo test geschrieben von:/posted by: Jan Klima at 06 October 2004 19:33:36:
Hi,
is there anyone who could do the following:)
1)match 0.6b against 0.6f(or 0.6g after released),40+ games(rather more) under some gui like arena or chessbase with third party opening book(not nunn positions),engines play the position after the last book move as both black and white
2)time control>=5+2s,The goal of the test obviously is to find if 0.6b was the best of all these versions(although it had bugs that needed to be fixed). I lost >=15%(!!!!!) nps somewhere and i wonder if I got something for it. It might be that there's some overtype(typist error? or how to say that in english) that spoils everything or it's just that i made it safer and safer with every new version to be sure there are no bugs during WBEC tournament. But anyway what matters here is if the new versions(0.6f+g,0.6e is buggy) get at least 55% against 0.6b.

I don't have the time for that now, but at least I can give you my results which nearly qualify for your criteria from my tourney (5+1; Draw-book, against other engines of similar strenght, CB GUI):
Version Elo performance # games
___________________________________________
Pseudo 0.6b 2575 51
Pseudo 0.6f 2534 75

regards
Axel
Axel Schumacher
 

Re: Pseudo test

Postby Dann Corbit » 08 Oct 2004, 21:12

Geschrieben von:/Posted by: Dann Corbit at 08 October 2004 22:12:56:
Als Antwort auf:/In reply to: Pseudo test geschrieben von:/posted by: Jan Klima at 06 October 2004 19:33:36:
Hi,
is there anyone who could do the following:)
1)match 0.6b against 0.6f(or 0.6g after released),40+ games(rather more) under some gui like arena or chessbase with third party opening book(not nunn positions),engines play the position after the last book move as both black and white
2)time control>=5+2s,The goal of the test obviously is to find if 0.6b was the best of all these versions(although it had bugs that needed to be fixed). I lost >=15%(!!!!!) nps somewhere and i wonder if I got something for it. It might be that there's some overtype(typist error? or how to say that in english) that spoils everything or it's just that i made it safer and safer with every new version to be sure there are no bugs during WBEC tournament. But anyway what matters here is if the new versions(0.6f+g,0.6e is buggy) get at least 55% against 0.6b.
Here are the results of 6b and 6e against thinker:

Program Elo + - Games Score Av.Op. Draws
1 Thinker : 2685 61 84 75 72.7 % 2515 28.0 %
2 Pseudo06b : 2543 132 149 18 30.6 % 2685 38.9 %
3 Pseudo06e : 2506 103 70 57 26.3 % 2685 24.6 %




my ftp site {remove http:// unless you like error messages}
Dann Corbit
 

Re: Pseudo test {WARNING!! absurdly long post contains the P

Postby Dann Corbit » 08 Oct 2004, 21:14

Dann Corbit
 


Return to Archive (Old Parsimony Forum)

Who is online

Users browsing this forum: No registered users and 9 guests