Page 6 of 7

Re: Komodo 10.2 released.

Posted: Sat Nov 19, 2016 7:04 pm
by mclane
I don't see that Komodo underperforms against stockfish,
I have an old Komodo and it beats all stockfish versions.
Only Houdini is ahead of Komodo. So I maybe have to update my Komodo version now.

Re: Komodo 10.2 released.

Posted: Sat Nov 19, 2016 7:42 pm
by Andre
So you obviously have an old version of Stockfish

Re: Komodo 10.2 released.

Posted: Sat Nov 19, 2016 10:40 pm
by mclane
Is 8 old ?

Re: Komodo 10.2 released.

Posted: Sun Nov 20, 2016 5:43 am
by Mike S.
mclane wrote:I have an old Komodo and it beats all stockfish versions.
Results from CEGT 40/20 1CPU:

http://www.husvankempen.de/nunn/40_40%2 ... ion/2.html

Among those, against

13 Komodo 10.1 x64 1CPU - 3264 100 + 31 = 66 - 3 64.0 % 3363
14 Komodo 10.2 x64 1CPU - 3259 100 + 38 = 58 - 4 67.0 % 3382

Brutal results.

The smallest gaps I found between SF8 and K10.x are at CCRL. - I did not yet find a list where H5 is ahead of SF8.

Re: Komodo 10.2 released.

Posted: Sun Nov 20, 2016 6:01 am
by Graham Banks
Mike S. wrote:
mclane wrote: I did not yet find a list where H5 is ahead of SF8.
Early days, but look here:

http://www.computerchess.org.uk/ccrl/40 ... ons_only=1

Houdini could well go top of the 40/4 4CPU list as well.

Re: Komodo 10.2 released.

Posted: Sun Nov 20, 2016 6:38 am
by lkaufman
CCRL 40/40 list now shows Komodo 10.1 tied for lead on 4 cpu with SF8, and Komodo 10.2 only 3 elo behind SF8 on one cpu. But we are not competitive in direct matches. I think this shows beyond much doubt that Komodo underperforms playing directly against SF compared to how they each do against a range of opponents. I expect Komodo 10.3 will top the 40/40 lists when it comes out, since we're virtually tied now.

Re: Komodo 10.2 released.

Posted: Sun Nov 20, 2016 6:41 am
by JJJ
lkaufman wrote:CCRL 40/40 list now shows Komodo 10.1 tied for lead on 4 cpu with SF8, and Komodo 10.2 only 3 elo behind SF8 on one cpu. But we are not competitive in direct matches. I think this shows beyond much doubt that Komodo underperforms playing directly against SF compared to how they each do against a range of opponents. I expect Komodo 10.3 will top the 40/40 lists when it comes out, since we're virtually tied now.
Komodo is not tied on every list and is behind in http://fastgm.de/ for exemple.

Also Komodo seems to loose against Houdini 5, would you says it's underperform against it ?

I ask with no offense here, just want to know you opinion about it. I truly respect your work and your engine and I m happy you re still on top with your very small team :)

Re: Komodo 10.2 released.

Posted: Sun Nov 20, 2016 8:13 am
by mclane
That's why measuring elo via match between a few top programs is not helping.
You always need a big variety of opponents to measure elo.
The smaller the group of opponents, the less real your results are.

Second match with many games against one opponent gives learning effects that influence result.

Re: Komodo 10.2 released.

Posted: Sun Nov 20, 2016 8:14 am
by mclane
I guess Houdini is the better killer.

Re: Komodo 10.2 released.

Posted: Sun Nov 20, 2016 1:34 pm
by lkaufman
JJJ wrote:
lkaufman wrote:CCRL 40/40 list now shows Komodo 10.1 tied for lead on 4 cpu with SF8, and Komodo 10.2 only 3 elo behind SF8 on one cpu. But we are not competitive in direct matches. I think this shows beyond much doubt that Komodo underperforms playing directly against SF compared to how they each do against a range of opponents. I expect Komodo 10.3 will top the 40/40 lists when it comes out, since we're virtually tied now.
Komodo is not tied on every list and is behind in http://fastgm.de/ for exemple.

Also Komodo seems to loose against Houdini 5, would you says it's underperform against it ?

I ask with no offense here, just want to know you opinion about it. I truly respect your work and your engine and I m happy you re still on top with your very small team :)
Both Stockfish 8 and Houdini 5 are stronger than Komodo 10.2 at blitz or near-blitz speeds. But at 40/40' it remains to be seen how Houdini 5 will fare; my guess is that it will not do better than Stockfish 8 did. Until we have significant data at that level it's too early to say much about Houdini 5.