AH_LTC Match: Houdini 3 vs Komodo 6

[Aser Huerga]

AH_LTC Match: Houdini 3 vs Komodo 6

i7-3930K CPUs 4.25 GHz
90'+30" TC
1 core for all engines
Ponder off
1024 Hash
3-4-5 EGTBs (when available) in SSDs
150 Early Starting Positions Suite, slightly tunned to avoid transpositions (checked with engine vs same-engine matches), and created as a proportional representation of the most played openings/variations on the last recent years in high quality human chess tournaments (source TWIC, only 2400+ ELO players) AH_150_Opening_Suite
All positions are played with Switched Colors for a total of 300 games per match
Games available for download, including eval/time/depth and PV for each move

Results:

Code: Select all

1 Komodo 6  +74/-58/=168 52.67% 158.0/300
2 Houdini 3 +58/-74/=168 47.33% 142.0/300

[/color]

Games: Houdini 3 - Komodo 6

[Modern Times]

Thanks for that match, that is interesting. Longer time control on a recent Intel CPU, and Komodo sneaks ahead. But, it still gets beaten by Stockfish.

[Adam Hair]

Here are the Ordo ratings computed from the three matches:

Code: Select all

   # PLAYER                   : RATING  ERROR   POINTS  PLAYED    (%)
   1 Stockfish_SZ 13110122    :   29.7   12.9    337.5     600   56.2%
   2 Komodo 6                 :   -4.7   13.0    294.0     600   49.0%
   3 Houdini 3                :  -24.9   13.0    268.5     600   44.8%

[Aser Huerga]

Thanks!
Now I will be out for a week, but there will be more matches with new engines in the future.

[lkaufman]

Thanks for that match, that is interesting. Longer time control on a recent Intel CPU, and Komodo sneaks ahead. But, it still gets beaten by Stockfish.

These results confirm my contention that both Komodo and Stockfish scale better than Houdini. When Stockfish plays Komodo, time limit doesn't seem to make much difference, but the rate of progress for both programs is such that if one of the programs tested is a month newer than the other it will win the match!

[Don]

Thanks for that match, that is interesting. Longer time control on a recent Intel CPU, and Komodo sneaks ahead. But, it still gets beaten by Stockfish.

Which version Komodo and SF are you talking about? I'll bet you are comparing the most cutting edge SF dev version against some ancient version of Komodo, right? Komodo 6 is now ancient.

So which version o these programs are you talking about? It's not fair to make comparisons like this without being specific.

Both Houdini and Komodo have advance you should realize.

Don

[beram]

Aser has used Stockfish_SZ 13110122 against Komodo 6 http://www.talkchess.com/forum/viewtopic.php?t=49944

Code: Select all

1 Stockfish_SZ 13110122 +70/-42/=188 54.67% 164.0/300 
2 Komodo 6              +42/-70/=188 45.33% 136.0/300

As far as I can see there is only 3 weeks in release date for them.
btw since when is an engine of nearly one month old, an ancient engine ?

I guess that only when you would have improved Komodo 6 by 35 ELO in that three weeks you would have been on par with it
But even then, by now SF131111 has already been improved with yet another 8 ELO..

It would be really very impressive if Komodo team could keep up that incredible pace of improvement of Stockfish worldwide framework

kind regards

[Modern Times]

I'm referring to publicly available versions of both engines. Yes of course Komodo has improved since K6. Aser is using K6, so that was the context of my observations.

[Modern Times]

I'm convinced that the latest version of Komodo will win TCEC, let's see.

[Don]

I'm referring to publicly available versions of both engines. Yes of course Komodo has improved since K6. Aser is using K6, so that was the context of my observations.

Yes that is what I thought. Komodo has made a lot of progress since Komodo 6 so the fairst compairson is whatever SF version was avilable then.