Upgrade to Chess.com Premium!

Top 3 benchmarks

This won't make sense unless you know what a Top 3 analysis is. If you don't and want to, go to

http://www.chess.com/groups/forum/cheating-forum

 

OTB Games

{ Karpov, Anatoly (Games: 24) } Baguio 1978
{ Top 1 Match: 557/1051 ( 53.0% )  
{ Top 2 Match: 757/1051 ( 72.0% )  
{ Top 3 Match: 861/1051 ( 81.9% )  
{ Top 4 Match: 919/1051 ( 87.4% )  

{ Kortschnoj, Viktor (Games: 24) } Baguio 1978
{ Top 1 Match: 586/1045 ( 56.1% )  
{ Top 2 Match: 772/1045 ( 73.9% )  
{ Top 3 Match: 868/1045 ( 83.1% )  
{ Top 4 Match: 926/1045 ( 88.6% )

{ Fischer, Robert James (Games: 17) } Fischer-Spassky 1972
{ Top 1 Match: 356/609 ( 58.5% )
{ Top 2 Match: 465/609 ( 76.4% )
{ Top 3 Match: 502/609 ( 82.4% )
{ Top 4 Match: 535/609 ( 87.9% )

{ Spassky, Boris V (Games: 17) } Fischer-Spassky 1972
{ Top 1 Match: 345/605 ( 57.0% )
{ Top 2 Match: 444/605 ( 73.4% )
{ Top 3 Match: 495/605 ( 81.8% )
{ Top 4 Match: 519/605 ( 85.8% )

{ Karpov, Anatoly }  (Games 17) WC Seville 1985
{ Top 1 Match: 318/576  ( 55.21% )
{ Top 2 Match: 436/576  ( 75.69% )
{ Top 3 Match: 483/576  ( 83.85% )
{ Top 4 Match: 509/576  ( 88.37% )

{  Kasparov, Garry } (Games 17) WC Seville 1985
{ Top 1 Match: 311/583 ( 53.35% )
{ Top 2 Match: 411/583 ( 70.50% )
{ Top 3 Match: 477/583 ( 81.82% )
{ Top 4 Match: 503/583 ( 86.28% )

{ Anand, Viswanathan (Games: 10) } Morelia 2007
{ Top 1 Match: 163/274 ( 59.5% ) 
{ Top 2 Match: 203/274 ( 74.1% ) 
{ Top 3 Match: 228/274 ( 83.2% ) 
{ Top 4 Match: 240/274 ( 87.6% ) 

{ Carlsen, Magnus (Games: 9) } Morelia 2007
{ Top 1 Match: 186/325 ( 57.2% ) 
{ Top 2 Match: 245/325 ( 75.4% ) 
{ Top 3 Match: 279/325 ( 85.9% ) 
{ Top 4 Match: 293/325 ( 90.2% ) 

{ Carlsen, Magnus (Games: 9) } Nanjing 2009
{ Top 1 Match: 200/335 ( 59.7% ) 
{ Top 2 Match: 256/335 ( 76.4% ) 
{ Top 3 Match: 283/335 ( 84.5% ) 
{ Top 4 Match: 298/335 ( 89.0% ) 

(otb=5403 moves)
...................................................................
CC Games

{ Baumbach, Fritz dr (GER) (Games: 9) } ICCF 11th WC Finals
{ Top 1 Match: 180/283 ( 63.6% ) 
{ Top 2 Match: 227/283 ( 80.2% ) 
{ Top 3 Match: 248/283 ( 87.6% ) 
{ Top 4 Match: 255/283 ( 90.1% )

{ Nesis, Gennady Efimovich (Games: 12) } ICCF 11th WC Finals
{ Top 1 Match: 291/443 ( 65.7% ) 
{ Top 2 Match: 355/443 ( 80.1% ) 
{ Top 3 Match: 384/443 ( 86.7% ) 
{ Top 4 Match: 400/443 ( 90.3% ) 

{ Mikhailov, Aleksey Ivanovich (Games: 11) } ICCF 11th WC Finals
{ Top 1 Match: 221/397 ( 55.1% ) 
{ Top 2 Match: 309/397 ( 77.4% ) 
{ Top 3 Match: 339/397 ( 85.4% ) 
{ Top 4 Match: 351/397 ( 88.3% ) 

{ Palciauskas, Vytas Victor (USA) (Games: 11) } ICCF 10th WC Finals
{ Top 1 Match: 227/364 ( 62,4% ) 
{ Top 2 Match: 291/364 ( 80.0% ) 
{ Top 3 Match: 309/364 ( 84.9% ) 
{ Top 4 Match: 327/364 ( 89.8% )

{ Richardson, Keith Bevin (ENG) (Games: 10) } ICCF 10th WC Finals
{ Top 1 Match: 198/362 ( 54.7% ) 
{ Top 2 Match: 268/362 ( 74.0% ) 
{ Top 3 Match: 308/362 ( 85.1% ) 
{ Top 4 Match: 320/362 ( 88.4% )

{ Morgado, Juan Sebastian (ARG (Games: 10) } ICCF 10th WC Finals
{ Top 1 Match: 159/272 ( 58.5% ) 
{ Top 2 Match: 211/272 ( 77.6% ) 
{ Top 3 Match: 227/272 ( 83.5% ) 
{ Top 4 Match: 242/272 ( 89.0% )

{ Sanakoev, Grigory Konst'vich (Games: 13) } ICCF 10th WC Finals
{ Top 1 Match: 267/451 ( 59.2% ) 
{ Top 2 Match: 358/451 ( 79.4% ) 
{ Top 3 Match: 395/451 ( 87.6% ) 
{ Top 4 Match: 411/451 ( 91.1% ) 

{ Oim, Tonu O. (EST) (Games: 11) } ICCF 9th WC Finals
{ Top 1 Match: 236/396 ( 59.6% ) 
{ Top 2 Match: 313/396 ( 79.0% ) 
{ Top 3 Match: 349/396 ( 88.1% ) 
{ Top 4 Match: 363/396 ( 91.7% )

{ Baumbach, Fritz dr (GER) (Games: 12) } ICCF 9th WC Finals
{ Top 1 Match: 269/433 ( 62.1% ) 
{ Top 2 Match: 334/433 ( 77.1% ) 
{ Top 3 Match: 363/433 ( 83.8% ) 
{ Top 4 Match: 386/433 ( 89.2% )

{ Mikhailov, Aleksey Ivanovich (Games: 14) } ICCF 9th WC Finals
{ Top 1 Match: 331/573 ( 57.8% ) 
{ Top 2 Match: 437/573 ( 76.3% ) 
{ Top 3 Match: 476/573 ( 83.1% ) 
{ Top 4 Match: 503/573 ( 87.8% )

{ Sloth, Jorn (DEN) (Games: 7) } ICCF 8th WC Finals
{ Top 1 Match: 163/274 ( 59.5% ) 
{ Top 2 Match: 218/274 ( 79.6% ) 
{ Top 3 Match: 238/274 ( 86.8% ) 
{ Top 4 Match: 250/274 ( 91.2% ) 

{ Zagorovsky, Vladimir Pavlovich (Games: 11) } ICCF 8th WC Finals
{ Top 1 Match: 237/409 ( 58.0% ) 
{ Top 2 Match: 306/409 ( 74.8% ) 
{ Top 3 Match: 337/409 ( 82.4% ) 
{ Top 4 Match: 352/409 ( 86.1% ) 

{ Kosenkov, VsevolodTikhonovich (Games: 9) } ICCF 8th WC Finals
{ Top 1 Match: 157/296 ( 53.0% ) 
{ Top 2 Match: 203/296 ( 68.5% ) 
{ Top 3 Match: 226/296 ( 76.4% ) 
{ Top 4 Match: 237/296 ( 80.1% ) 

{ Estrin, Yakov Borisovich (Games: 10) } ICCF 7th WC Finals
{ Top 1 Match: 172/277 ( 62.1% ) 
{ Top 2 Match: 211/277 ( 76.2% ) 
{ Top 3 Match: 233/277 ( 84.1% ) 
{ Top 4 Match: 245/277 ( 88.5% )

{ Boey, Jozef Martin (BEL) (Games: 13) } ICCF 7th WC Finals
{ Top 1 Match: 276/450 ( 61.3% ) 
{ Top 2 Match: 355/450 ( 78.9% ) 
{ Top 3 Match: 388/450 ( 86.2% ) 
{ Top 4 Match: 409/450 ( 90.9% )

{ Estrin, Yakov Borisovich (Games: 10) } ICCF 6th WC Finals
{ Top 1 Match: 191/292 ( 65.4% ) 
{ Top 2 Match: 232/292 ( 79.5% ) 
{ Top 3 Match: 260/292 ( 89.0% ) 
{ Top 4 Match: 265/292 ( 90.8% )

{ Rittner, Horst Robert (GER) (Games: 8) } ICCF 6th WC Finals
{ Top 1 Match: 164/274 ( 59.9% ) 
{ Top 2 Match: 208/274 ( 75.9% ) 
{ Top 3 Match: 233/274 ( 85.0% ) 
{ Top 4 Match: 240/274 ( 87.6% )

{ Zagorovsky, Vladimir Pavlovich (Games: 10) } ICCF 6th WC Finals
{ Top 1 Match: 179/322 ( 55.6% ) 
{ Top 2 Match: 247/322 ( 76.7% ) 
{ Top 3 Match: 270/322 ( 83.9% ) 
{ Top 4 Match: 289/322 ( 89.8% )

{ Berliner, Hans Jack (USA) (Games: 10) } ICCF 5th WC Finals
{ Top 1 Match: 200/333 ( 60.0% ) 
{ Top 2 Match: 275/333 ( 82.6% ) 
{ Top 3 Match: 293/333 ( 88.0% ) 
{ Top 4 Match: 309/333 ( 92.8% )

{ Hybl, Jaroslav (CZE) (Games: 10) } ICCF 5th WC Finals
{ Top 1 Match: 200/341 ( 58.7% ) 
{ Top 2 Match: 263/341 ( 77.1% ) 
{ Top 3 Match: 294/341 ( 86.2% ) 
{ Top 4 Match: 310/341 ( 90.9% )

{ Husak, Karel (Games: 14) } ICCF 5th WC Finals
{ Top 1 Match: 251/466 ( 53.9% ) 
{ Top 2 Match: 333/466 ( 71.5% ) 
{ Top 3 Match: 371/466 ( 79.6% ) 
{ Top 4 Match: 402/466 ( 86.3% )

(CC=7711 moves)

(13114 moves)

........................................

CC averaged by year

ICCF 5th 34 games

1  651/1140  57.1
2  871/          76.4
3  958           84.0

ICCF 6th  28 games

534/888       60.1
687/             77.4
763/             85.9

ICCF 7th 23 games

448/727       61.6
566/             77.9
621/             85.4

ICCF 8th 27 games

557/979      56.9
727/            74.3
801/            81.8

ICCF 9th 37 games

836/1402     59.6
1084/           77.3
1185/           84.5

ICCF 10th 44 games

847/1445    58.6
1124/          77.8
1235/          85.5

ICCF 11th 32 games

692/1130    61.2
891/            78.9
971/            85.9

Comments

  • 3 years ago

    goldendog

    The hardware: Core2Duo 2.00GHz 1GB Ram

    The software: Batch Analyzer (BA) 45 sec move/ 16-30 ply min max/ 256MB hash/engine=Houdini 103a/ Windows XP 32 bit

    The Methodology: All benchmark games are rolled back to the time of play, using MegaBase 2010 and MegaCorr 4. A CC game is considered out of book if a game from the previous year no longer appears. For OTB games, the MegaBase/MegaCorr chronology is accepted even if the year coincides. When looking for book for an OTB game, a CC game from the same year is considered book.

    The benchmark games come from the World Championship OTB matches, very high quality modern Super GM tournaments, and the pre-computer era ICCF world championship finals. In other words, the best of the best.

    From those ICCF years I took the top 3 finishers--4 if there was a tie for 3rd-4th--and applied the databases to their games. If a top finisher's batch fell below 8 games after the 20+ non-DB move requirement, I discarded his results (with one exception when it was the winner of the tournament, and I included his results mainly for the sake of historical completeness). When a batch is so small, it tends to weigh insignificantly in establishing human thresholds.

    When the analysis from BA's book needs to be rectified with further analysis that BA has not done, I have done that analysis manually using Chessbase in multi-line analysis using Houdini 103a with 256 MB hash, and a stopwatch, mimicking the way BA ends analysis.

    All benchmark games must contain at least 20 non-database moves, the same requirements that will be applied to the games of the subject of a Top 3 investigation.

    All games of a batch must be against strong opposition, i.e. 2300+, or if this is impossible then 2200+ will be acceptable. The games must be chosen objectively, being the 20 most recent. More than 20 games is acceptable as well.

    When a game, either a benchmark game or a suspect's game enters tablebase territory, i.e. 6 pieces remaining, then the analysis is halted. The rationale is that any cheater can consult the online 6-man tablebases at that point, and including analysis of such positions becomes problematic in detecting engine use, as there is some likelihood that an engine is not being used then.

Back to Top

Post your reply:

Log In
Join
^ Ads keep Chess.com free. Upgrade to remove ads! ^
^ Ads keep Chess.com free. Upgrade to remove ads! ^