Komodo64 3 see gauntlet 60/90' +30'

[Hugo]

Hello everybody.

The tournament is close to the end. Komodo has got a nice peak at the moment. As expected, Komodo made a good performance in the possitional games.
We are now at startpos number 25.


Regards, Clemens Keck

[Tomcass]

Excellent work, Clemens!.

Thank you very much for running this complete and clarifying test.

Tom.

[lkaufman]

Hello everybody.

The tournament is close to the end. Komodo has got a nice peak at the moment. As expected, Komodo made a good performance in the possitional games.
We are now at startpos number 25.


Regards, Clemens Keck

I see we finished at 2977, 28 elo above the blitz rating, and we won all the matches except against Houdini. This appears to confirm my belief that Komodo gains on other engines with more time. If so this means that our MP version should be relatively stronger against other engines than our SP version, assuming we get the same speedup as the others.
Thanks for running the test, it is quite helpful to know that we "scale" well. Now we don't need to waste testing time confirming this ourselves.

[Hugo]

Hello Larry

it was an exiting test. Komodo did verry well. The ratings are not 100% exact because the used ratings are bayesElo ratings, and shredder gui produces an EloStat rating. I estimate the tournament performance a little lower that 2977. Lets say 2970, which is great!!
Since along time now, I suspect testsets as an influence of the rating in my (also others?) list. I always used the same set for the "pure" list. I think I will change the set for some tests in future.

best regards, Clemens