Testing Stockfish 11-03-13. 480 Games.

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

Tomcass
Posts: 786
Joined: Sun Apr 16, 2006 9:09 pm

Re: Testing Stockfish 11-03-13. 480 Games.

Post by Tomcass »

TESTING STOCKFISH DEVELOPMENT 260314 = 480 GAMES

Bench: 7682173 Timestamp: 1395813989

i7 980 3.33 Ghz.
6 real cores
Ponder: Off.
GUI: Fritz 12
Book: Perfect 2014c Sedat
No tablebases. No RTB used.
Hash 512
Relative Speed: 28.66
Knodes per second: 13759

Time Control= 4+0

Stockfish 260314 64 SSE4.2x - Houdini 4 x64_st_X6_CT0 22.0 - 18.0 +10/=24/-6 55.00%
Stockfish 260314 64 SSE4.2x - Komodo TCECr 64-bitx6 26.0 - 14.0 +16/=20/-4 65.00%
Stockfish 260314 64 SSE4.2x - Gull 2.8 beta x64_XP_LP 28.0 - 12.0 +18/=20/-2 70.00%

Time Control= 2+2

Stockfish 260314 64 SSE4.2x - Houdini 4 x64_st_X6_CT0 23.0 - 17.0 +13/=20/-7 57.50%
Stockfish 260314 64 SSE4.2x - Komodo TCECr 64-bitx6 22.5 - 17.5 +13/=19/-8 56.25%
Stockfish 260314 64 SSE4.2x - Gull 2.8 beta x64_XP_LP 21.5 - 18.5 +10/=23/-7 53.75%

Score using 6 cores: 143.0 – 97.0 = 59.58%
240 Games: http://www.mediafire.com/view/7c6dbf6ci ... 0Games.pgn

i7 975 3.33 Ghz.
4 real cores
Ponder: Off.
GUI: Fritz 12
Book: Perfect 2014c Sedat
No tablebases. No RTB used.
Hash 256
Relative Speed: 20.62
Knodes per second: 9.899

Time Control: 4+0

Stockfish 260314 64 SSE4.2x - Houdini 4 x64_StA_Ct0_X4 21.5 - 18.5 +11/=21/-8 53.75%
Stockfish 260314 64 SSE4.2x - Komodo TCECr 64-bitx4 24.5 - 15.5 +13/=23/-4 61.25%
Stockfish 260314 64 SSE4.2x - Gull 2.8 beta x64_XP_LP 28.0 - 12.0 +17/=22/-1 70.00%

Time Control: 2+2

201403SF260314_2+2 2014

Stockfish 260314 64 SSE4.2x - Houdini 4 x64_StA_Ct0_X4 22.5 - 17.5 +13/=19/-8 56.25%
Stockfish 260314 64 SSE4.2x - Komodo TCECr 64-bitx4 19.5 - 20.5 +10/=19/-11 48.75%
Stockfish 260314 64 SSE4.2x - Gull 2.8 beta x64_XP_LP 23.0 - 17.0 +12/=22/-6 57.50%

Score using 4 Cores= 139.0 – 101.0 = 57.92%
240 Games:
http://www.mediafire.com/view/e5tbge86v ... 4cores.pgn

Segmenting by Time Control:

Fixed TC = 150.0 – 90.0 = 62.50%
Incremental TC = 132.0 – 108.0 = 55.00%

Global Score= 282.0 – 198.0 = 58.75%

Against : Houdini 4.0 St. Ct0 (3227) = 55.62% ; Komodo TCECr (3181) = 57.81%, Gull2.8Beta (3141) = 62.81%

Average Estimated Elo Opponents = 3183
Estimated Elo Performance= 324
4


Error bars: +/- 23 EEP

Regards,

Tom.
Tomcass
Posts: 786
Joined: Sun Apr 16, 2006 9:09 pm

Re: Testing Stockfish 11-03-13. 480 Games.

Post by Tomcass »

TESTING STOCKFISH DEVELOPMENT 290314 = 480 GAMES

Bench: 7926803 Timestamp: 1396083902

i7 980 3.33 Ghz.
6 real cores
Ponder: Off.
GUI: Fritz 12
Book: Perfect 2014c Sedat
No tablebases. No RTB used.
Hash 512
Relative Speed: 28.66
Knodes per second: 13759

Time Control= 4+0

Stockfish 290314 64 SSE4.2DV - Houdini 4 x64_st_X6_CT0 20.5 - 19.5 +7/=27/-6 51.25%
Stockfish 290314 64 SSE4.2DV - Komodo TCECr 64-bitx6 29.0 - 11.0 +20/=18/-2 72.50%
Stockfish 290314 64 SSE4.2DV - Gull 2.8 beta x64_XP_LP 27.0 - 13.0 +19/=16/-5 67.50%

Time Control= 2+2

201403Stockfish_Dev_ 290314_2+2 2014

Stockfish 290314 64 SSE4.2DV - Houdini 4 x64_st_X6_CT0 24.0 - 16.0 +14/=20/-6 60.00%
Stockfish 290314 64 SSE4.2DV - Komodo TCECr 64-bitx6 27.0 - 13.0 +17/=20/-3 67.50%
Stockfish 290314 64 SSE4.2DV - Gull 2.8 beta x64_XP_LP 25.5 - 14.5 +14/=23/-3 63.75%

Score using 6 cores: 153.0 – 87.0 = 63.75%
240 Games: http://www.mediafire.com/view/bvgulu0g1 ... 4cores.pgn

i7 975 3.33 Ghz.
4 real cores
Ponder: Off.
GUI: Fritz 12
Book: Perfect 2014c Sedat
No tablebases. No RTB used.
Hash 256
Relative Speed: 20.62
Knodes per second: 9.899

Time Control: 4+0

Stockfish 290314 64 SSE4.2D - Houdini 4 x64_StA_Ct0_X4 18.0 - 22.0 +8/=20/-12 45.00%
Stockfish 290314 64 SSE4.2D - Komodo TCECr 64-bitx4 26.0 - 14.0 +15/=22/-3 65.00%
Stockfish 290314 64 SSE4.2D - Gull 2.8 beta x64_XP_LP 28.5 - 11.5 +20/=17/-3 71.25%

Time Control: 2+2

Stockfish 290314 64 SSE4.2D - Houdini 4 x64_StA_Ct0_X4 22.5 - 17.5 +12/=21/-7 56.25%
Stockfish 290314 64 SSE4.2D - Komodo TCECr 64-bitx4 22.0 - 18.0 +11/=22/-7 55.00%
Stockfish 290314 64 SSE4.2D - Gull 2.8 beta x64_XP_LP 29.0 - 11.0 +19/=20/-1 72.50%

Score using 4 Cores= 146.0 – 94.0 = 60.83%
240 Games:
http://www.mediafire.com/view/bvgulu0g1 ... 4cores.pgn

Segmenting by Time Control:

Fixed TC = 149.0 – 91.0 = 62.08%
Incremental TC = 150.0 – 90.0 = 62.50%

Global Score= 299.0 – 181.0 = 62.29%

Against : Houdini 4.0 St. Ct0 (3227) = 53.12% ; Komodo TCECr (3181) = 65.00%, Gull2.8Beta (3141) = 68.75%

Average Estimated Elo Opponents = 3183
Estimated Elo Performance= 3269


Error bars: +/- 23 EEP

This is a NEW BEST SCORE EVER in my tests, 8 EEP better than the previous ones (SF Ipman 290114 and SF Development 010314, both with 3261 EEP). Congratulations again to the Stockfish Team.

Regards,

Tom.
Tomcass
Posts: 786
Joined: Sun Apr 16, 2006 9:09 pm

Re: Testing Stockfish 11-03-13. 480 Games.

Post by Tomcass »

...woops!! :oops: I have noticed that file with games using 6 cores was wrong. The good one is:


Score using 6 cores: 153.0 – 87.0 = 63.75%
240 Games:
http://www.mediafire.com/view/lgiy09hgc ... 0Games.pgn

Regards,

Tom.
ouachita
Posts: 454
Joined: Tue Jan 15, 2013 4:33 pm
Location: Ritz-Carlton, NYC
Full name: Bobby Johnson

Re: Testing Stockfish 11-03-13. 480 Games.

Post by ouachita »

Tom,
Is Stockfish 290314 64 SSE4.2DV and Stockfish 290314 64 SSE4.2D the same version and bench? I assume "yes".
SIM, PhD, MBA, PE
Tomcass
Posts: 786
Joined: Sun Apr 16, 2006 9:09 pm

Re: Testing Stockfish 11-03-13. 480 Games.

Post by Tomcass »

ouachita wrote:Tom,
Is Stockfish 290314 64 SSE4.2DV and Stockfish 290314 64 SSE4.2D the same version and bench? I assume "yes".
Hi, Bobby!

Yes, it is the same version and bench. I included the final 'DV' -for Development- when I adapted the number of cores to my computers.

Regards,

Tom.
Tomcass
Posts: 786
Joined: Sun Apr 16, 2006 9:09 pm

Re: Testing Stockfish 11-03-13. 480 Games.

Post by Tomcass »

TESTING STOCKFISH ROCKWOOD COMPILE 170414 = 480 GAMES.

i7 980 3.33 Ghz.
6 real cores
Ponder: Off.
GUI: Fritz 12
Book: Perfect 2014c Sedat –Limit 5 moves-
No tablebases. No RTB used.
Large Pages allowed.
Hash 512
Relative Speed: 29.54
Knodes per second: 14177

Time Control= 4+0

201404StockfishRW_ 170414_4+0 2014

StockfishRW 140417 SSE4.2x6 - Houdini 4 x64_st_X6CT0 43.0 - 37.0 +19/=48/-13 53.75%
StockfishRW 140417 SSE4.2x6 - Komodo TCECr 64-bitx6 51.5 - 28.5 +31/=41/-8 64.38%
StockfishRW 140417 SSE4.2x6 - Gull 3 x64 50.0 - 30.0 +30/=40/-10 62.50%

Time Control= 2+2

StockfishRW 140417 SSE4.2x6 - Houdini 4 x64_st_X6CT0 40.0 - 40.0 +12/=56/-12 50.00%
StockfishRW 140417 SSE4.2x6 - Komodo TCECr 64-bit 48.5 - 31.5 +30/=37/-13 60.63%
StockfishRW 140417 SSE4.2x6 - Gull 3 x64 49.5 - 30.5 +27/=45/-8 61.88%

240 Games =
http://www.mediafire.com/view/7h909o20d ... 0Games.pgn


Global Score= 282.5 – 197.5 = 58.85%

Against : Houdini 4.0 St. Ct0 (3227) 51.87% ; Komodo TCECr (3181) 62.50% ; Gull 3.0 (3188)= 62.19%

Average Estimated Elo Opponents = 3199
Estimated Elo Performance= 3261


Regards,

Tom.
Tomcass
Posts: 786
Joined: Sun Apr 16, 2006 9:09 pm

Re: Testing Stockfish 11-03-13. 480 Games.

Post by Tomcass »

TESTING STOCKFISH DEVELOPMENT 020514 = 480 GAMES

Bench: 8678654 Timestamp: 1399021458


i7 980 3.33 Ghz.
6 real cores
Ponder: Off.
GUI: Fritz 14
Book: Perfect 2014c Sedat
No tablebases. No RTB used.
Hash 512
Relative Speed: 29.54
Knodes per second: 14.177

Time Control= 4+0

Stockfish 020514 64 SSE4.2D - Houdini 4 x64_st_X6_CT0 45.0 - 35.0 +20/=50/-10 56.25%
Stockfish 020514 64 SSE4.2D - Komodo TCECr 64-bitNOB 50.5 - 29.5 +28/=45/-7 63.13%
Stockfish 020514 64 SSE4.2D - Gull 3 x64 48.5 - 31.5 +24/=49/-7 60.63%

Time Control= 2+2

Stockfish 020514 64 SSE4.2D - Houdini 4 x64_st_X6 44.5 - 35.5 +21/=47/-12 55.63%
Stockfish 020514 64 SSE4.2D - Komodo TCECr 64-bitNOB 48.0 - 32.0 +28/=40/-12 60.00%
Stockfish 020514 64 SSE4.2D - Gull 3 x64 47.5 - 32.5 +21/=53/-6 59.38%

480 games=
http://www.mediafire.com/view/1fwrdweru ... 0Games.pgn

Global Score= 284.0 – 196.0 = 59.17%

Against : Houdini 4.0 St. Ct0 (3227) 55.94% ; Komodo TCECr (3181) 61.56% ; Gull 3.0 (3188)= 60.00%

Average Estimated Elo Opponents = 3199
Estimated Elo Performance= 3263


Only 6 Elo Points below the best scorer in my tests: Stockfish Dev. 290314. (3269) Within error bars. According to this test, SF Dev 020514 is 36 Elo points stronger than Houdini 4 Standard Contempt 0.

Error bars= +/-23 EEP

Regards,

Tom.
Tomcass
Posts: 786
Joined: Sun Apr 16, 2006 9:09 pm

Re: Testing Stockfish 11-03-13. 480 Games.

Post by Tomcass »

TESTING STOCKFISH IPMAN COMPILE 270414 = 480 GAMES.

i7 980 3.33 Ghz.
6 real cores
Ponder: Off.
GUI: Fritz 12
Book: Perfect 2014c – Sedat – Limit 5 moves
No tablebases. No RTB used.
Large Pages allowed.
Hash 512
Relative Speed: 29.54
Knodes per second: 14.177

Time Control = 4+0

SF 270414IPx 64 SSE4.2x6 - Houdini 4 x64_st_X6_CT0 42.0 - 38.0 +19/=46/-15 52.50%
SF 270414IPx 64 SSE4.2x6 - Komodo TCECr 64-bitNOB 54.0 - 26.0 +34/=40/-6 67.50%
SF 270414IPx 64 SSE4.2x6 - Gull 3 x64 50.0 - 30.0 +29/=42/-9 62.50%

Time Control= 2+2

SF 270414IPx 64 SSE4.2x6 - Houdini 4 x64_st_X6_CT0 45.5 - 34.5 +22/=47/-11 56.88%
SF 270414IPx 64 SSE4.2x6 - Komodo TCECr 64-bitNOB 53.5 - 26.5 +35/=37/-8 66.88%
SF 270414IPx 64 SSE4.2x6 - Gull 3 x64 51.0 - 29.0 +30/=42/-8 63.75%

480 Games =
http://www.mediafire.com/view/9995ynau5 ... 0Games.pgn

Global Score= 296.0 – 184.0 = 61.67%

Against : Houdini 4.0 St. Ct0 (3227)= 54.69% ; Komodo TCECr (3181) = 67.19% ; Gull 3.0(3188) = 63.12%

Average Estimated Elo Opponents = 3199
Estimated Elo Performance= 3281


This is a clear NEW BEST SCORE EVER in my tests, 12 Elo Points better than the previous record. (Stockfish Development 290314= 3269).

Congratulations to the Stockfish Team for your great work and to Ipman for this extremely strong compile, 54 EEP above the monster Houdini 4 Contempt 0.

Error bars= +/- 23 EEP

Regards,

Tom.
Tomcass
Posts: 786
Joined: Sun Apr 16, 2006 9:09 pm

Re: Testing Stockfish 11-03-13. 480 Games.

Post by Tomcass »

TESTING STOCKFISH ROCKWOOD COMPILE 110514 = 480 GAMES.

i7 980 3.33 Ghz.
6 real cores
Ponder: Off.
GUI: Fritz 12
Book: Perfect 2014c Sedat –Limit 5 moves-
No tablebases. No RTB used.
Large Pages allowed.
Hash 512
Relative Speed: 30.42
Knodes per second: 14602

Time Control= 4+0

StockfishRW 110514 SSE4.2_ - Houdini 4 x64_st_X6_CT0 23.5 - 16.5 +12/=23/-5 58.75%
StockfishRW 110514 SSE4.2_ - Komodo TCECr 64-bitNOB 22.5 - 17.5 +13/=19/-8 56.25%
StockfishRW 110514 SSE4.2_ - Gull 3 x64 XP 25.5 - 14.5 +14/=23/-3 63.75%

Time Control= 2+2

StockfishRW 110514 SSE4.2_ - Houdini 4 x64_st_X6_CT0 21.5 - 18.5 +9/=25/-6 53.75%
StockfishRW 110514 SSE4.2_ - Komodo TCECr 64-bitNOB 22.0 - 18.0 +11/=22/-7 55.00%
StockfishRW 110514 SSE4.2_ - Gull 3 x64 XP 26.5 - 13.5 +15/=23/-2 66.25%

240 Games =
http://www.mediafire.com/view/fme97uhn5 ... 0Games.pgn
Score using 6 Cores= 141.5 – 98.5 = 58.96%

i7 975 3.33 Ghz.
4 real cores
Ponder: Off.
GUI: Fritz 12
Book: Perfect 2014c Sedat – Limit 5 moves -
No tablebases. No RTB used.
Large Pages allowed
Hash 256
Relative Speed: 20.62
Knodes per second: 9.899

Time Control = 4+0

StockfishRW 110514 SSE4.2_ - Houdini 4 Pro x64_Ct0 22.5 - 17.5 +12/=21/-7 56.25%
StockfishRW 110514 SSE4.2_ - Komodo TCECr 64-bit_NOB 23.5 - 16.5 +9/=29/-2 58.75%
StockfishRW 110514 SSE4.2_ - Gull 3 x64 XP 23.0 - 17.0 +10/=26/-4 57.50%

Time Control= 2+2

StockfishRW 110514 SSE4.2_ - Houdini 4 Pro x64_Ct0 24.5 - 15.5 +12/=25/-3 61.25%
StockfishRW 110514 SSE4.2_ - Komodo TCECr 64-bit_NOB 23.0 - 17.0 +13/=20/-7 57.50%
StockfishRW 110514 SSE4.2_ - Gull 3 x64 XP 26.5 - 13.5 +17/=19/-4 66.25%
240 Games=
http://www.mediafire.com/view/u75o3w831 ... _games.pgn

Score using 4 Cores= 143.0 – 97.0 = 59.58%

Segmenting by Time Control:

Fixed TC = 140.5 – 99.5 = 58.54%
Incremental TC = 144.0 – 96.0 = 60.00%

Global Score= 284.5 – 195.5 = 59.27%

Against : Houdini 4.0 St. Ct0 (3227)= 57.50% ; Komodo TCECr (3181) = 56.87% ; Gull 3.0 XP (3199)= 63.44%

Average Estimated Elo Opponents = 3202
Estimated Elo Performance= 3267


Error bars= +/- 23 EEP

This is the best score so far for a SF Rockwood compile in my tests, only 14 Elo poinits below the top scorer SF 270414 Ipman. Please note the excellent performance of Komodo TCECr in this test.

Regards,

Tom.
Tomcass
Posts: 786
Joined: Sun Apr 16, 2006 9:09 pm

Re: Testing Stockfish 11-03-13. 480 Games.

Post by Tomcass »

TESTING STOCKFISH DEVELOPMENT 130514 = 480 GAMES

Bench: 8739659 Timestamp: 1400013448

i7 980 3.33 Ghz.
6 real cores
Ponder: Off.
GUI: Fritz 12
Book: Perfect 2012c
No tablebases. No RTB used.
Hash 512
Relative Speed: 28.66
Knodes per second: 13.759

Time Control= 4+0

Stockfish 130514 64 SSE4.2D - Houdini 4 x64_st_X6_CT0 23.0 - 17.0 +13/=20/-7 57.50%
Stockfish 130514 64 SSE4.2D - Komodo TCECr 64-bitNOB 24.5 - 15.5 +13/=23/-4 61.25%
Stockfish 130514 64 SSE4.2D - Gull 3 x64 XP 19.5 - 20.5 +7/=25/-8 48.75%

Time Control= 2+2

Stockfish 130514 64 SSE4.2D - Houdini 4 x64_st_X6_CT0 21.5 - 18.5 +8/=27/-5 53.75%
Stockfish 130514 64 SSE4.2D - Komodo TCECr 64-bitNOB 22.5 - 17.5 +11/=23/-6 56.25%
Stockfish 130514 64 SSE4.2D - Gull 3 x64 XP 24.5 - 15.5 +14/=21/-5 61.25%

Score using 6 cores: 135.5 – 104.5= 56.46%
240 Games:
http://www.mediafire.com/view/2bhflfca2 ... 0Games.pgn

i7 975 3.33 Ghz.
4 real cores
Ponder: Off.
GUI: Fritz 12
Book: Perfect 2012c
No tablebases. No RTB used.
Hash 256
Relative Speed: 20.62
Knodes per second: 9.899

Time Control: 4+0

Stockfish 130514 64 SSE4.2D - Houdini 4 Pro x64_Ct0 20.5 - 19.5 +10/=21/-9 51.25%
Stockfish 130514 64 SSE4.2D - Komodo TCECr 64-bitx4 24.0 - 16.0 +12/=24/-4 60.00%
Stockfish 130514 64 SSE4.2D - Gull 3 x64 XP 26.5 - 13.5 +14/=25/-1 66.25%

Time Control: 2+2

Stockfish 130514 64 SSE4.2D - Houdini 4 Pro x64_Ct0 25.5 - 14.5 +12/=27/-1 63.75%
Stockfish 130514 64 SSE4.2D - Komodo TCECr 64-bit_x4 28.0 - 12.0 +20/=16/-4 70.00%
Stockfish 130514 64 SSE4.2D - Gull 3 x64 XP 24.5 - 15.5 +12/=25/-3 61.25%

Score using 4 Cores=
240 Games: 149.0 – 91.0 = 62.08%
http://www.mediafire.com/view/m6gip5iy8 ... _games.pgn

Segmenting by Time Control:

Fixed TC = 138.0 – 102.0 = 57.50%
Incremental TC = 146.5 – 93.5= 61.04%

Global Score= 284.5 – 195.5 = 59.27%

Against : Houdini 4.0 St. Ct0 (3227) = 56.56% ; Komodo TCECr (3181) = 61.87%, Gull 3 XP (3199) = 59.37%

Average Estimated Elo Opponents = 3202
Estimated Elo Performance= 3267


Error bars= +/- 23 EEP

Regards,

Tom.