UFIM 500

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.

UFIM 500

Postby Dann Corbit » 08 Jul 2004, 20:48

Geschrieben von:/Posted by: Dann Corbit at 08 July 2004 21:48:23:

Have others seen the startling improvement in Ufim recently, or are my tests an anomaly?
Ufim500 appears to be an elite engine now.



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

Re: UFIM 500

Postby Tom Likens » 08 Jul 2004, 21:28

Geschrieben von:/Posted by: Tom Likens at 08 July 2004 22:28:29:
Als Antwort auf:/In reply to: UFIM 500 geschrieben von:/posted by: Dann Corbit at 08 July 2004 21:48:23:
Have others seen the startling improvement in Ufim recently, or are my tests an anomaly?
Ufim500 appears to be an elite engine now.
Hello Dann,
I believe you're correct. Ufim 5.00 is *much* better in the tests I've run as well.
As a data point, and for what it's worth (which may not be much) Ufim kicks
Djinn's butt all over the place,- Niyas is definitely to be commended.
I commented awhile back, tongue-in-cheek, that the Russian programmers were
making a bid to take over the computer chess world (Ufim, SmarThink etc.), but
it doesn't seem as funny to me now.
regards,
--tom
Tom Likens
 

Re: UFIM 500

Postby Norm Pollock » 08 Jul 2004, 22:00

Geschrieben von:/Posted by: Norm Pollock at 08 July 2004 23:00:53:
Als Antwort auf:/In reply to: Re: UFIM 500 geschrieben von:/posted by: Tom Likens at 08 July 2004 22:28:29:
Have others seen the startling improvement in Ufim recently, or are my tests an anomaly?
Ufim500 appears to be an elite engine now.
Hello Dann,
I believe you're correct. Ufim 5.00 is *much* better in the tests I've run as well.
As a data point, and for what it's worth (which may not be much) Ufim kicks
Djinn's butt all over the place,- Niyas is definitely to be commended.
I commented awhile back, tongue-in-cheek, that the Russian programmers were
making a bid to take over the computer chess world (Ufim, SmarThink etc.), but
it doesn't seem as funny to me now.
regards,
--tom
Great improvement can also be said for new versions of Booot (Ukraine) and Wildcat (Belarus).
Norm Pollock
 

Re: UFIM 500

Postby Tom Likens » 08 Jul 2004, 22:05

Geschrieben von:/Posted by: Tom Likens at 08 July 2004 23:05:32:
Als Antwort auf:/In reply to: Re: UFIM 500 geschrieben von:/posted by: Norm Pollock at 08 July 2004 23:00:53:
Have others seen the startling improvement in Ufim recently, or are my tests an anomaly?
Ufim500 appears to be an elite engine now.
Hello Dann,
I believe you're correct. Ufim 5.00 is *much* better in the tests I've run as well.
As a data point, and for what it's worth (which may not be much) Ufim kicks
Djinn's butt all over the place,- Niyas is definitely to be commended.
I commented awhile back, tongue-in-cheek, that the Russian programmers were
making a bid to take over the computer chess world (Ufim, SmarThink etc.), but
it doesn't seem as funny to me now.
regards,
--tom
Great improvement can also be said for new versions of Booot (Ukraine) and Wildcat (Belarus).
Agreed, the improvement in Booot seems especially impressive. I suspect he may win the
6th division of Leo's tournament ;-)
--tom
Tom Likens
 

Re: UFIM 500

Postby Marcus Prewarski » 08 Jul 2004, 22:10

Geschrieben von:/Posted by: Marcus Prewarski at 08 July 2004 23:10:53:
Als Antwort auf:/In reply to: Re: UFIM 500 geschrieben von:/posted by: Tom Likens at 08 July 2004 23:05:32:

Great improvement can also be said for new versions of Booot (Ukraine) and Wildcat (Belarus).
Agreed, the improvement in Booot seems especially impressive. I suspect he may win the
6th division of Leo's tournament ;-)
--tom
Don't forget Eeyore, it is my pick to win the 5th division which I think is pretty good for a fairly new engine.
Marcus
Marcus Prewarski
 

Re: UFIM 500

Postby Igor Gorelikov » 09 Jul 2004, 10:36

Geschrieben von:/Posted by: Igor Gorelikov at 09 July 2004 11:36:49:
Als Antwort auf:/In reply to: UFIM 500 geschrieben von:/posted by: Dann Corbit at 08 July 2004 21:48:23:
Have others seen the startling improvement in Ufim recently, or are my tests an anomaly?
Ufim500 appears to be an elite engine now.
I'v got opposite results:
http://f11.parsimony.net/forum16635/messages/68238.htm
Igor Gorelikov
 

Re: UFIM 500

Postby Nicolas Normand » 09 Jul 2004, 13:27

Geschrieben von:/Posted by: Nicolas Normand at 09 July 2004 14:27:16:
Als Antwort auf:/In reply to: UFIM 500 geschrieben von:/posted by: Dann Corbit at 08 July 2004 21:48:23:
Have others seen the startling improvement in Ufim recently, or are my tests an anomaly?
Ufim500 appears to be an elite engine now.

hello,
Ufim 5.00 doesn't run a really impressive Zonal qualification tournament in the olympiads 2004. It seems to have the same level as before ....
I run a gauntlet after this tournament ( to get a greater number of games), and the version 4.04 seems at least as strong (probably it is stronger )!
Regards
Nicolas


Zonal 9 cross table
Nicolas Normand
 

Re: UFIM 500

Postby Heinz van Kempen » 09 Jul 2004, 14:08

Geschrieben von:/Posted by: Heinz van Kempen at 09 July 2004 15:08:24:
Als Antwort auf:/In reply to: Re: UFIM 500 geschrieben von:/posted by: Igor Gorelikov at 09 July 2004 11:36:49:

Hi :-),
to verify if a version is improved or not you will still need hundreds of games. After 50 or 100 games it can still look as if an older version is better.
Ufim 5.00 at this time played around 800 games in my tournaments. Rating is 80-100 points better than version 4.04 with 680 games.
This is mainly Blitz on faster hardware. If someone has the CPU ressources to run a few hundred games with more time, maybe results might differ, but I doubt this for most engines. But there are exceptions.
Of course everyone trusts most in his own results, but I have simply stopped to believe in anything with less than 400 games, regardless of the time control.
http://www.husvankempen.de/nunn/
Best Regards
Heinz
Heinz van Kempen
 

testing

Postby Igor Gorelikov » 09 Jul 2004, 14:40

Geschrieben von:/Posted by: Igor Gorelikov at 09 July 2004 15:40:32:
Als Antwort auf:/In reply to: Re: UFIM 500 geschrieben von:/posted by: Heinz van Kempen at 09 July 2004 15:08:24:

Of course everyone trusts most in his own results, but I have simply stopped to believe in anything with less than 400 games, regardless of the time control.
Best Regards
Heinz
Someones run events at short time control with a lot of games.
Someones run events at longer time control with a small amount of games.
In my opinion, both do good job.
I trust any tests if they conform to the following:
1) There is a big pool (more than 200 participants).
2) Most games are played between engines of the same level.
best regards,
Igor
Igor Gorelikov
 

Re: testing

Postby Heinz van Kempen » 09 Jul 2004, 14:47

Geschrieben von:/Posted by: Heinz van Kempen at 09 July 2004 15:47:16:
Als Antwort auf:/In reply to: testing geschrieben von:/posted by: Igor Gorelikov at 09 July 2004 15:40:32:
Someones run events at short time control with a lot of games.
Someones run events at longer time control with a small amount of games.
In my opinion, both do good job.
I trust any tests if they conform to the following:
1) There is a big pool (more than 200 participants).
2) Most games are played between engines of the same level.
best regards,
Igor
Hello Igor,
right, I would never tell that some tests are better than others. All have their stronger and weaker points. We have to see them all combined, so I just want to contribute as we have contradictory results here.
My wish would be that a lot of testers would combine their efforts and CPU power for a big amateur test on all levels over a few months. But this will be illusion, as everyone wants to do his own thing and so we get all kinds of tests, what is also good.
Best Regards
Heinz
Heinz van Kempen
 

Re: testing

Postby Igor Gorelikov » 09 Jul 2004, 15:04

Geschrieben von:/Posted by: Igor Gorelikov at 09 July 2004 16:04:17:
Als Antwort auf:/In reply to: Re: testing geschrieben von:/posted by: Heinz van Kempen at 09 July 2004 15:47:16:

My wish would be that a lot of testers would combine their efforts and CPU power for a big amateur test on all levels over a few months. But this will be illusion, as everyone wants to do his own thing and so we get all kinds of tests, what is also good.
Best Regards
Heinz
It's my wish, too.
Igor
Igor Gorelikov
 

Re: UFIM 500

Postby Tom Likens » 10 Jul 2004, 04:19

Geschrieben von:/Posted by: Tom Likens at 10 July 2004 05:19:26:
Als Antwort auf:/In reply to: Re: UFIM 500 geschrieben von:/posted by: Marcus Prewarski at 08 July 2004 23:10:53:
Great improvement can also be said for new versions of Booot (Ukraine) and Wildcat (Belarus).
Agreed, the improvement in Booot seems especially impressive. I suspect he may win the
6th division of Leo's tournament ;-)
--tom
Don't forget Eeyore, it is my pick to win the 5th division which I think is pretty good for a fairly new engine.
Marcus
Hello Marcus,
I don't know much about Eeyore, but you're right- winning *any* division
of Leo's tournament is a great accomplishment (especially, for a new engine).
Eeyore is more than likey an engine to watch. It's almost scary, how many
good engines are appearing these days!
regards,
--tom
Tom Likens
 


Return to Archive (Old Parsimony Forum)

Who is online

Users browsing this forum: No registered users and 23 guests

cron