TCEC LIVE - Division I, Season 1 - All the top engines!

[M ANSARI]

Thanks Martin for your time and effort, I am sure this effort has dramatically changed how you allocate free time in your daily life ... but just to let you know we appreciate it . Some really high quality games there. Now if you could only do that one extra step and get another super fast 6 core and we could have ponder ON games That would be the ultimate in simulating real world match competition.

[JWorcester]

Thanks Martin for your time and effort, I am sure this effort has dramatically changed how you allocate free time in your daily life ... but just to let you know we appreciate it

I'll add my thanks here too. In between watching TCEC and TATA my wife was getting quite annoyed .

[S.Taylor]

Three losses for Rybka none for Houdini and Rybka beaten by Houdini, what more did you expect?

]

Quite different if you take the second half seperately. When Rybka was beating Houdini and SF with TCEC matches, it was clearly on top in all sections.

[S.Taylor]

Houdini could easily have won 2 more games in the second half (against Ivanhoe and Stockfish), and could easily have failed to convert 2 wins in the first half of the tournament.

Robert

Thanks!
This was closer to what i wanted to know!
(If the only thing we are looking at is results, then you need many games. But if you look at other aspects, you can see alot from less games too. [this is what i have always believed, but never got much sympathy for the idea]).

[S.Taylor]

I always wish that the best engine, is clearly above the others. Not because of bias in favour of a particular engine or author (atleast irrelevant of that), but because i like to know that the top engine includes the capabilities of those below, so that i know which engine will give me the best opinion if i set up a position with it (doesn't matter if two engines give equally good continuations, as long as the top one doesn't give inferior to one which is of lower strength), and not that i would need to try all the top 20 engines.
This is why i like the top engine to atleast hardly ever lose a game vs others. Unless i know that it is because of a bad opening or a clear bug.

(obviously i know this would never be absolute, if it is not more than 200 elo above, or even much more. But i always wish it to be almost like this).

[Rubinus]

Unfortunatelly.
Ivanhoe 9.49j probabably very better than 9.47c (B47bC). Only little less, than Houdini 1.5.

[Houdini]

Unfortunatelly.
Ivanhoe 9.49j probabably very better than 9.47c (B47bC). Only little less, than Houdini 1.5.

This is based on how many games?

Robert

[Rubinus]

Unfortunatelly.
Ivanhoe 9.49j probabably very better than 9.47c (B47bC). Only little less, than Houdini 1.5.

This is based on how many games?

Robert

cca 120 games - 4cpu, 90+30 (and 18 - 2cpu)
http://www.chesspraga.cz/engine-turn/ph-2011-01/ (Houdini 1.5 and Ivanhoe 9.49j - 33.5/60)
http://www.chesspraga.cz/engine-turn/ph-2011-02/ (Houdini 1.5 - 35/60, Ivanhoe 9.49j - 31.5/60)
PH-ELO:
1. Houdini 1.5 x64_4CPU 3274 +42 -42 120 57,1% 3225 54,2%
2. Ivanhoe 9.49j x64_4CPU 3259 +38 -37 120 54,2% 3230 63,3%
And 2CPU tournament (laptop)
1.Houdini 1.5 - 13/18, 2.Stockfish 1.9.1JA - 12, 3.Rybka 4.1beta4 - 11.5, 4.-5.Deep Rybka 4, Ivanhoe 9.49j - 11, 6.Deep Junior 12 - 7.5, 7.Gull 1.1 - 7, 8.Hiarcs 13.2 - 6.5, 9.Hannibal 1.0 - 5.5, 10.Spark 1.0 - 5.
games



Ivanhoe 9.47c so far cca 21games - 4cpu and results cca any Stockfish 2.0.1 - www page ph-2011-03 probably tomorrow
2cpu - 5 games, weak 1.5/5 - games
So far (6 rounds whitout Ivanhoe-Zappa) 1.Houdini 1.5a - 5, 2.-3.Stockfish 2.0.1JA, Critter 0.90 - 4.5, 4.-5.Rybka 4.1beta4, Deep Saros 2.3gs - 3.5, 6.Naraku 1.31 - 2, 7.-8.Zappa II Mexico, Ivanhoe 9.47c - 1.5/5, 9.Smarthink 1.2 - 1.5, 10.Deep Sjeng WC2008 - 1.
Version 9.47 less probably to 50 ELO!

Best wishes Pavel Háse

[Houdini]

Pavel, thanks for the details.

[Rubinus]

Currently ph-2011-03 (44/120)

ph-elo 5.2.2011
1. Houdini 1.5 x64_4CPU 3281 +42 -42 120 57,1% 3231 54,2%
2. Ivanhoe 9.49j x64_4CPU 3266 +38 -37 120 54,2% 3237 63,3%
3. Deep Rybka 4.1 x64 Beta 4_4_CPU 3235 +49 -50 60 47,5% 3252 68,3%
- Deep Rybka 4 x64_4CPU 3222 +27 -27 210 52,4% 3205 66,7%
- Houdini 1.03a x64 4_CPU 3218 +51 -51 60 51,7% 3207 66,7%
- Deep Rybka 4.1 x64 Beta 1_4_CPU 3215 +74 -72 30 53,3% 3192 66,7%
4. Stockfish 2.0.1 JA 64bit_4_CPU 3209 +48 -49 82 46,3% 3235 58,5%
- IvanHoe 9.50b x64_4_CPU 3205 +47 -47 60 49,2% 3210 71,7%
- Stockfish 1.9.1 JA 64bit_4_CPU 3192 +31 -31 210 46,9% 3214 57,6%
- IvanHoe 9.47c x64_4_CPU 3188 +96 -95 22 52,3% 3172 59,1%
5. Fire 1.3.1 x64KLO_4_CPU 3186 +55 -55 60 50,8% 3181 61,7%
6. Deep Saros 2.3gs x64_4_CPU 3173 +77 -77 22 50,0% 3173 72,7%
- Deep Saros 2.3c3 w32_4_CPU 3163 +48 -62 30 43,3% 3210 80,0%
7. Critter 0.90 x64_4_CPU 3128 +40 -43 82 40,2% 3197 68,3%