SF 160916 VS Komodo 10.1 long time control

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

User avatar
Laskos
Posts: 10948
Joined: Wed Jul 26, 2006 10:21 pm
Full name: Kai Laskos

Re: Games 83 and 84 , end of tournament

Post by Laskos »

mjlef wrote:You can naturally test however you want, but in science, once you decide on the rules of the tests, you should stick with it until the end. Stopping ealry skews results.
Yes, he should have tested to the end of 100 games. But I have some reserves about this test, draw rate seems too low (openings are not particularly sharp), and +20 -2 is a bit strange. But it is probably an unwarranted skepticism.
KaLiKoBa
Posts: 135
Joined: Tue Mar 15, 2016 9:35 am

Re: Games 83 and 84 , end of tournament

Post by KaLiKoBa »

+20-2 is a clear win for stockfish.
A 84 games test is of course shorter than a 100 games test, but no skepticism should be raised because 16 games have not been played....
Come on ! Don't debate about what has not been played, but about what has been played !
Don't forget I'm using Stockfish BMI version from http://chess.ultimaiq.net/stockfish.html which is 8-9 % faster than abrok.eu site, which means my results will always show a better performance for stockfish against komodo, compared to testers who run tests with stockfish from abrok.eu site !! This has to be taken into account when you judge the 20-2 result . If you ask I can run a quick match between SF from ultimaiq and stockfish from abrok, you will realize the difference in strength...
User avatar
Laskos
Posts: 10948
Joined: Wed Jul 26, 2006 10:21 pm
Full name: Kai Laskos

Re: Games 83 and 84 , end of tournament

Post by Laskos »

KaLiKoBa wrote:+20-2 is a clear win for stockfish.
A 84 games test is of course shorter than a 100 games test, but no skepticism should be raised because 16 games have not been played....
Come on ! Don't debate about what has not been played, but about what has been played !
Don't forget I'm using Stockfish BMI version from http://chess.ultimaiq.net/stockfish.html which is 8-9 % faster than abrok.eu site, which means my results will always show a better performance for stockfish against komodo, compared to testers who run tests with stockfish from abrok.eu site !! This has to be taken into account when you judge the 20-2 result . If you ask I can run a quick match between SF from ultimaiq and stockfish from abrok, you will realize the difference in strength...
Maybe my skepticism stems form my different expectations :). With Noomen Short Lines I expected a higher draw rate in these conditions, and 20-2 is a bit more than I expected. Anyway, I do not contest the results.
KaLiKoBa
Posts: 135
Joined: Tue Mar 15, 2016 9:35 am

Re: Games 83 and 84 , end of tournament

Post by KaLiKoBa »

how can I post the .pgn file for the 84 games ?
Guenther
Posts: 4718
Joined: Wed Oct 01, 2008 6:33 am
Location: Regensburg, Germany
Full name: Guenther Simon

Re: Games 83 and 84 , end of tournament

Post by Guenther »

KaLiKoBa wrote:how can I post the .pgn file for the 84 games ?
Pgn files are usually not posted here, but uploaded to a site/webspace and you just post the link to it then.
Doing otherwise would of course clutter the forum and be against the forum rules.
(Imagine CCRL and CEGT and all others would really post all pgn files in ascii here...)

Edit:
Most people who don't have own webspace use e.g. Googledrive, Mediafire, Zippyshare etc. for interim uploads...
KaLiKoBa
Posts: 135
Joined: Tue Mar 15, 2016 9:35 am

Re: Games 83 and 84 , end of tournament

Post by KaLiKoBa »

Hehe understood, I will of course run other tournaments between the top engines and I hope the results will show an almost equal strength between these !
KaLiKoBa
Posts: 135
Joined: Tue Mar 15, 2016 9:35 am

Re: Games 83 and 84 , end of tournament

Post by KaLiKoBa »

thank you for the tips, I'll try these sites then !
APassionForCriminalJustic
Posts: 417
Joined: Sat May 24, 2014 9:16 am

Re: Games 83 and 84 , end of tournament

Post by APassionForCriminalJustic »

KaLiKoBa wrote:+20-2 is a clear win for stockfish.
A 84 games test is of course shorter than a 100 games test, but no skepticism should be raised because 16 games have not been played....
Come on ! Don't debate about what has not been played, but about what has been played !
Don't forget I'm using Stockfish BMI version from http://chess.ultimaiq.net/stockfish.html which is 8-9 % faster than abrok.eu site, which means my results will always show a better performance for stockfish against komodo, compared to testers who run tests with stockfish from abrok.eu site !! This has to be taken into account when you judge the 20-2 result . If you ask I can run a quick match between SF from ultimaiq and stockfish from abrok, you will realize the difference in strength...
20-2 is insane. In fact this type of a smashing is probably a bad sign for Houdini when it plays in the superfinal. If my speculation is correct and Houdini development still is not at Komodo's level then it will be absolutely pounded. You do not need to run Stockfish Abrok versus the Ultimaiq compiles. I have access to another tournament (private) where Komodo is being crushed quite severely. You should run this same tournament but with asmFish. The asmFish compile is 16 percent faster than the Ultimaiq compiles.

Stockfish is just too strong now. It looks like all of those patches over the last few months have effectively taken their toll; also I would speculate that Lazy SMP is better than what it was when it was first coded into Stockfish during last season's TCEC.
mwyoung
Posts: 2727
Joined: Wed May 12, 2010 10:00 pm

Re: Games 83 and 84 , end of tournament

Post by mwyoung »

Laskos wrote:
KaLiKoBa wrote:+20-2 is a clear win for stockfish.
A 84 games test is of course shorter than a 100 games test, but no skepticism should be raised because 16 games have not been played....
Come on ! Don't debate about what has not been played, but about what has been played !
Don't forget I'm using Stockfish BMI version from http://chess.ultimaiq.net/stockfish.html which is 8-9 % faster than abrok.eu site, which means my results will always show a better performance for stockfish against komodo, compared to testers who run tests with stockfish from abrok.eu site !! This has to be taken into account when you judge the 20-2 result . If you ask I can run a quick match between SF from ultimaiq and stockfish from abrok, you will realize the difference in strength...
Maybe my skepticism stems form my different expectations :). With Noomen Short Lines I expected a higher draw rate in these conditions, and 20-2 is a bit more than I expected. Anyway, I do not contest the results.
I am not skeptical, my test are also showing very similar results at fast and long time controls. This could change with a newer update for Komodo, but as of today Stockfish wins against Komodo and by a good margin.
"The worst thing that can happen to a forum is a running wild attacking moderator(HGM) who is not corrected by the community." - Ed Schröder
But my words like silent raindrops fell. And echoed in the wells of silence.
KaLiKoBa
Posts: 135
Joined: Tue Mar 15, 2016 9:35 am

Re: Games 83 and 84 , end of tournament

Post by KaLiKoBa »

glad to read that similar results are happening under other testing conditions, I feel more comfortable :)