I have created some new comparings, statistics and decided to share here too
Two separate Elo calculations, between both my hardwares:
Code: Select all
i7 980 3.33GHz:
---------------
# PLAYER : RATING POINTS PLAYED (%)
1 Stockfish 260214 x64 6c : 3539.5 405.5 486 83.4%
2 Houdini 4 x64 6c : 3531.8 374.5 442 84.7%
3 Komodo TCECr x64 6c : 3449.5 347.5 442 78.6%
4 Gull 3 x64 6c : 3434.2 377.5 500 75.5%
5 Gull 2.8b x64 6c : 3432.0 348.0 462 75.3%
6 Rybka 4.1 x64 6c : 3399.8 354.0 474 74.7%
7 Chiron 2 x64 6c : 3258.4 265.0 486 54.5%
8 Hannibal 1.4b x64 6c : 3253.5 262.0 486 53.9%
9 Protector 1.6.0 x64 6c : 3239.7 253.5 486 52.2%
10 Hiarcs 14 WCSC w32 6c : 3205.9 241.5 501 48.2%
11 Spike 1.4 x64 6c : 3185.3 237.5 518 45.8%
12 Senpai 1.0 x64 6c : 3177.5 243.5 545 44.7%
13 Texel 1.03 x64 6c : 3161.7 207.0 486 42.6%
14 Junior 13.3 x64 6c : 3159.9 246.5 575 42.9%
15 Shredder 12 x64 6c : 3147.0 213.0 519 41.0%
16 Zappa Mexico II x64 6c : 3136.5 205.0 517 39.7%
17 Spark 1.0 x64 6c : 3128.9 202.0 520 38.8%
18 Gaviota 1.0 x64 6c : 3074.6 156.5 487 32.1%
19 Sjeng WC2008 x64 6c : 3029.3 132.0 486 27.2%
20 Tornado 4.88 x64 6c : 3017.6 126.0 486 25.9%
21 Crafty 23.8 x64 6c : 3010.2 122.5 486 25.2%
22 Fritz 10.1 w32 6c : 3000.0 117.5 486 24.2%
i7 970 3.20GHz:
---------------
# PLAYER : RATING POINTS PLAYED (%)
1 Stockfish 260214 x64 6c : 3490.0 478.0 591 80.9%
2 Houdini 4 x64 6c : 3477.4 496.0 633 78.4%
3 Komodo TCECr x64 6c : 3467.6 490.0 632 77.5%
4 Gull 2.8b x64 6c : 3410.9 422.5 567 74.5%
5 Gull 3 x64 6c : 3406.9 371.5 500 74.3%
6 Rybka 4.1 x64 6c : 3356.7 382.5 591 64.7%
7 Chiron 2 x64 6c : 3263.5 338.0 592 57.1%
8 Hannibal 1.4b x64 6c : 3241.5 321.5 591 54.4%
9 Protector 1.6.0 x64 6c : 3223.8 308.0 591 52.1%
10 Hiarcs 14 WCSC w32 6c : 3209.6 285.0 570 50.0%
11 Senpai 1.0 x64 6c : 3197.2 244.5 505 48.4%
12 Spike 1.4 x64 6c : 3177.5 253.5 554 45.8%
13 Shredder 12 x64 6c : 3155.1 237.5 553 42.9%
14 Texel 1.03 x64 6c : 3144.1 248.5 593 41.9%
15 Junior 13.3 x64 6c : 3131.8 197.5 496 39.8%
16 Spark 1.0 x64 6c : 3125.2 217.0 553 39.2%
17 Zappa Mexico II x64 6c : 3102.7 204.0 556 36.7%
18 Gaviota 1.0 x64 6c : 3079.7 203.0 592 34.3%
19 Crafty 23.8 x64 6c : 3016.6 160.5 593 27.1%
20 Sjeng WC2008 x64 6c : 3008.9 155.5 593 26.2%
21 Tornado 4.88 x64 6c : 3000.2 150.5 592 25.4%
22 Fritz 10.1 w32 6c : 3000.0 150.0 592 25.3%*Note: The strongest top 5-6 engines perform better on my i7 980 3.33GHz machine
*********************************************************************************
Games lost on time
SCCT has only 46 games lost on time, played out of more than 11500 games
Mainly lost games on time by:
Chiron 2 = 16 games
Fritz 10.1 = 6 games
Crafty 23.8 = 5 games
Houdini 4 = 4 games
Komodo TCECr = 3 games
Sjeng WC2008 = 3 games
Gull 3 = 3 games
Hannibal 1.4b = 2 games
Hiarcs 14 WCSC = 2 games
*Note: those game forfeits are mainly by Commercial engines!
And I hope those bugs will be fixed in the next new releases!
*********************************************************************************
Record Games...
The below games are record games, impressive performance...
The fastest wins belong to Whites:
Gull 2.8b vs Tornado 4.88 = 1-0 (game finished in 18 moves)
Gaviota 1.0 vs Tornado 4.88 = 1-0 (game finished in 18 moves)
*Note: Tornado lost the above 2 unlucky games exactly with the same D45 opening
As Blacks the records belongs to Houdini:
Crafty 22.8 - Houdini 4 = 0-1 (game finished in 20 moves)
*Note: Crafty loses the above unlucky game with D46 opening
My congrats to the record winners !
*********************************************************************************
Who is stronger? Free or Commercial work for chess ??
I've calculated and compared the Elo differences between Free and Commercial engines
And here are the current new results:
Commercial vs Free + 8 Elo in favor for Commercial engines
*Note: Based on overall top 20 engines, and as we see that the commercial work is still leading over free work, but with a very small Elo difference...
But in case of comparing individually, then there is no doubt that the free Stockfish is the leader over all engines (including commercial engines too)
Note also that the comparings are based on only matches: Free vs Commercial matches
I mean no games between Free vs Free or Commercial vs Commercial
Yes...Well done to Commercial programmers!
But however:
- My BIG congratulations to free program developers too, especially to Marco Costalba!!
For more details (conditions):
http://www.sedatcanbaz.com/chess/?page_id=1224
Have fun,
Sedat