Engine insights on human and tournaments

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

Ferdy
Posts: 4856
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Engine insights on human and tournaments

Post by Ferdy »

Started a project to create a human rating system based on engine evaluations. For now here are some initial measures on how the rating will be built.

Couple of tournament samples:

Tatat Steel 2017:

Code: Select all

Player                   : WinMiss Dubious Mistake Blunder aveErr/Ga   Pos Games  Pts DangerG
Karjakin, Sergey         :       0       3       0       0     -2.09   344    13  7.0       1
So, Wesley               :       0       6       0       0     -1.65   325    13  9.0       1
Carlsen, Magnus          :       1       3       1       0     -1.70   459    13  8.0       1
Aronian, Levon           :       0       8       1       0     -2.19   367    13  7.5       2
Giri, Anish              :       1       6       2       0     -3.76   371    13  6.5       1
Eljanov, Pavel           :       1      12       2       0     -3.86   455    13  7.0       1
Wei, Yi                  :       0      10       4       0     -2.15   466    13  7.5       1
Nepomniachtchi, Ian      :       1      12       4       0     -3.18   392    13  5.0       2
Wojtaszek, Radoslaw      :       1      14       4       0     -3.21   471    13  6.0       3
Harikrishna, Pentala     :       0      10       5       0     -3.02   413    13  6.0       2
Adhiban, Baskaran        :       0       6       2       1     -3.80   474    13  7.5       3
Andreikin, Dmitry        :       1       8       3       1     -4.33   426    13  6.0       2
Rapport, Richard         :       1       7       4       1     -4.35   356    13  4.5       4
Van Wely, Loek           :       1      19       7       3     -6.93   429    13  3.5       6

TQ (Tournament Quality)  : -3.30
WijkaanZee 2000:

Code: Select all

Player                   : WinMiss Dubious Mistake Blunder aveErr/Ga   Pos Games  Pts DangerG
Leko, Peter              :       0       5       0       0     -1.60   295    13  8.0       0
Anand, Viswanathan       :       0       8       0       0     -1.15   306    13  8.0       0
Adams, Michael           :       0      23       0       0     -3.52   398    13  7.0       1
Van Wely, Loek           :       1       5       1       0     -4.54   334    13  4.0       3
Nikolic, Predrag         :       0       7       1       0     -4.00   374    13  6.0       4
Kasparov, Garry          :       0      10       1       0     -1.87   397    13  9.5       0
Morozevich, Alexander    :       1      12       1       0     -4.20   349    13  7.5       3
Kramnik, Vladimir        :       0       8       2       0     -1.42   292    13  8.0       0
Korchnoi, Victor         :       0       9       3       0     -3.67   363    13  5.0       2
Timman, Jan              :       0       9       4       0     -3.88   378    13  6.5       3
Short, Nigel             :       0       7       1       1     -2.98   302    13  5.5       3
Piket, Jeroen            :       0      17       2       2     -3.32   361    13  6.5       2
Polgar, Judit            :       0       9       4       2     -3.80   348    13  5.0       2
Lputian, Smbat           :       0      12       3       3     -6.95   429    13  4.5       4

TQ (Tournament Quality)  : -3.35
London Chess Classic 2016:

Code: Select all

Player                   : WinMiss Dubious Mistake Blunder aveErr/Ga   Pos Games  Pts DangerG
Kramnik, Vladimir        :       0       1       0       0     -0.76   272     9  5.0       0
So, Wesley               :       0       3       0       0     -0.87   218     9  6.0       0
Caruana, Fabiano         :       0       5       0       0     -2.49   256     9  5.5       0
Giri, Anish              :       0       4       1       0     -1.70   295     9  4.5       0
Aronian, Levon           :       0       0       2       0     -1.74   208     9  4.0       2
Nakamura, Hikaru         :       0       5       2       0     -3.68   239     9  5.0       0
Vachier-Lagrave, Maxime  :       0      12       2       0     -3.38   253     9  4.0       1
Anand, Viswanathan       :       0       4       4       0     -2.60   199     9  5.0       1
Topalov, Veselin         :       0      11       5       0     -7.89   257     9  2.0       3
Adams, Michael           :       0       9       0       1     -2.26   272     9  4.0       2

TQ (Tournament Quality)  : -2.74

Code: Select all

LEGEND:
WinMiss   - count of games where a player miss the winning move in one of the pos.
            Winning move: engine_movescore > 3.0 and
            player_movescore <= 3.0 pawn units and the player failed
            to win this game.
Dubious   - count of positions where a playable position
            turns into a difficult position.
            Playable position: engine_movescore >= -0.5
            Difficult position: -0.5 < player_movescore >= -1.0
Mistake   - count of positions where a playable position
            turns into a very difficult position.
            Very difficult position: -1.0 < player_movescore >= -3.0
Blunder   - count of positions where a playable position
            turns into losing position.
            Losing position: player_movescore < -3.0
aveErr/Ga - average error per game, in pawn unit.
            Measures the accuracy of a player, higher is better.
            error = player_movescore - enginemove_score
            total_error = pos x error
            ave_error = total_error/pos
            aveErr/Ga = ave_error/num_games
DangerG   - is the number of games where the player is losing.
            A game is losing when there is one position where
            engine_movescore < -3.0

NOTES:
1. Engine used in the analysis is Brainfish 300117 64 POPCNT.
2. Analysis time/pos is 10.0s.
3. Average analysis depth reached is 27.
4. When the player move error is below 6 pawns from a
   given position, error is set to -6.0 pawns.
5. Analysis of position starts at move 12 in every game.
6. When the score of the position in a game is already
   below -6 pawns, analysis of succeeding position in this game
   will be stopped and analysis will continue in the next game.
7. The table below is sorted by aveErr/Ga in descending order,
   then Dubious, then Mistake, then finally by Blunder
   in ascending orders.
8. Tournament quality is the average error per game of all players.
   TQ = total aveErr/Ga / num_players
   A high TQ means the games have less mistakes overall.
walter koroljow
Posts: 16
Joined: Fri Oct 15, 2010 11:06 pm
Location: Delaware, USA

Re: Engine insights on human and tournaments

Post by walter koroljow »

Take a look at this paper, "Understanding Distributions of Chess Performances", www.cse.buffalo.edu/faculty/regan/papers/pdf/RMH11.pdf

The author also looked at other variables, e.g., position complexity.
Ferdy
Posts: 4856
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: Engine insights on human and tournaments

Post by Ferdy »

walter koroljow wrote:Take a look at this paper, "Understanding Distributions of Chess Performances", www.cse.buffalo.edu/faculty/regan/papers/pdf/RMH11.pdf

The author also looked at other variables, e.g., position complexity.
Thanks for the info.
Ferdy
Posts: 4856
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: Engine insights on human and tournaments

Post by Ferdy »

Ferdy wrote:Started a project to create a human rating system based on engine evaluations. For now here are some initial measures on how the rating will be built.

Couple of tournament samples:

Tatat Steel 2017:

Code: Select all

Player                   : WinMiss Dubious Mistake Blunder aveErr/Ga   Pos Games  Pts DangerG
Karjakin, Sergey         :       0       3       0       0     -2.09   344    13  7.0       1
So, Wesley               :       0       6       0       0     -1.65   325    13  9.0       1
Carlsen, Magnus          :       1       3       1       0     -1.70   459    13  8.0       1
Aronian, Levon           :       0       8       1       0     -2.19   367    13  7.5       2
Giri, Anish              :       1       6       2       0     -3.76   371    13  6.5       1
Eljanov, Pavel           :       1      12       2       0     -3.86   455    13  7.0       1
Wei, Yi                  :       0      10       4       0     -2.15   466    13  7.5       1
Nepomniachtchi, Ian      :       1      12       4       0     -3.18   392    13  5.0       2
Wojtaszek, Radoslaw      :       1      14       4       0     -3.21   471    13  6.0       3
Harikrishna, Pentala     :       0      10       5       0     -3.02   413    13  6.0       2
Adhiban, Baskaran        :       0       6       2       1     -3.80   474    13  7.5       3
Andreikin, Dmitry        :       1       8       3       1     -4.33   426    13  6.0       2
Rapport, Richard         :       1       7       4       1     -4.35   356    13  4.5       4
Van Wely, Loek           :       1      19       7       3     -6.93   429    13  3.5       6

TQ (Tournament Quality)  : -3.30
WijkaanZee 2000:

Code: Select all

Player                   : WinMiss Dubious Mistake Blunder aveErr/Ga   Pos Games  Pts DangerG
Leko, Peter              :       0       5       0       0     -1.60   295    13  8.0       0
Anand, Viswanathan       :       0       8       0       0     -1.15   306    13  8.0       0
Adams, Michael           :       0      23       0       0     -3.52   398    13  7.0       1
Van Wely, Loek           :       1       5       1       0     -4.54   334    13  4.0       3
Nikolic, Predrag         :       0       7       1       0     -4.00   374    13  6.0       4
Kasparov, Garry          :       0      10       1       0     -1.87   397    13  9.5       0
Morozevich, Alexander    :       1      12       1       0     -4.20   349    13  7.5       3
Kramnik, Vladimir        :       0       8       2       0     -1.42   292    13  8.0       0
Korchnoi, Victor         :       0       9       3       0     -3.67   363    13  5.0       2
Timman, Jan              :       0       9       4       0     -3.88   378    13  6.5       3
Short, Nigel             :       0       7       1       1     -2.98   302    13  5.5       3
Piket, Jeroen            :       0      17       2       2     -3.32   361    13  6.5       2
Polgar, Judit            :       0       9       4       2     -3.80   348    13  5.0       2
Lputian, Smbat           :       0      12       3       3     -6.95   429    13  4.5       4

TQ (Tournament Quality)  : -3.35
London Chess Classic 2016:

Code: Select all

Player                   : WinMiss Dubious Mistake Blunder aveErr/Ga   Pos Games  Pts DangerG
Kramnik, Vladimir        :       0       1       0       0     -0.76   272     9  5.0       0
So, Wesley               :       0       3       0       0     -0.87   218     9  6.0       0
Caruana, Fabiano         :       0       5       0       0     -2.49   256     9  5.5       0
Giri, Anish              :       0       4       1       0     -1.70   295     9  4.5       0
Aronian, Levon           :       0       0       2       0     -1.74   208     9  4.0       2
Nakamura, Hikaru         :       0       5       2       0     -3.68   239     9  5.0       0
Vachier-Lagrave, Maxime  :       0      12       2       0     -3.38   253     9  4.0       1
Anand, Viswanathan       :       0       4       4       0     -2.60   199     9  5.0       1
Topalov, Veselin         :       0      11       5       0     -7.89   257     9  2.0       3
Adams, Michael           :       0       9       0       1     -2.26   272     9  4.0       2

TQ (Tournament Quality)  : -2.74

Code: Select all

LEGEND:
WinMiss   - count of games where a player miss the winning move in one of the pos.
            Winning move: engine_movescore > 3.0 and
            player_movescore <= 3.0 pawn units and the player failed
            to win this game.
Dubious   - count of positions where a playable position
            turns into a difficult position.
            Playable position: engine_movescore >= -0.5
            Difficult position: -0.5 < player_movescore >= -1.0
Mistake   - count of positions where a playable position
            turns into a very difficult position.
            Very difficult position: -1.0 < player_movescore >= -3.0
Blunder   - count of positions where a playable position
            turns into losing position.
            Losing position: player_movescore < -3.0
aveErr/Ga - average error per game, in pawn unit.
            Measures the accuracy of a player, higher is better.
            error = player_movescore - enginemove_score
            total_error = pos x error
            ave_error = total_error/pos
            aveErr/Ga = ave_error/num_games
DangerG   - is the number of games where the player is losing.
            A game is losing when there is one position where
            engine_movescore < -3.0

NOTES:
1. Engine used in the analysis is Brainfish 300117 64 POPCNT.
2. Analysis time/pos is 10.0s.
3. Average analysis depth reached is 27.
4. When the player move error is below 6 pawns from a
   given position, error is set to -6.0 pawns.
5. Analysis of position starts at move 12 in every game.
6. When the score of the position in a game is already
   below -6 pawns, analysis of succeeding position in this game
   will be stopped and analysis will continue in the next game.
7. The table below is sorted by aveErr/Ga in descending order,
   then Dubious, then Mistake, then finally by Blunder
   in ascending orders.
8. Tournament quality is the average error per game of all players.
   TQ = total aveErr/Ga / num_players
   A high TQ means the games have less mistakes overall.
Added Linares 1994, one of the best performance of Karpov in a tournament.

Linares 1994

Code: Select all

Player                   : WinMiss Dubious Mistake Blunder aveErr/Ga   Pos Games  Pts DangerG
Ivanchuk, Vassily        :       0       6       0       0     -4.62   354    13  6.0       3
Karpov, Anatoly          :       0       9       1       0     -2.20   391    13 11.0       0
Kramnik, Vladimir        :       0       7       2       0     -2.65   347    13  7.0       3
Anand, Viswanathan       :       0       8       2       0     -3.57   320    13  6.5       3
Lautier, Joel            :       1      10       2       0     -5.05   440    13  7.0       4
Illescas Cordoba, Miguel :       0      12       2       0     -5.44   345    13  4.5       6
Kamsky, Gata             :       2      11       3       0     -4.67   403    13  6.5       3
Topalov, Veselin         :       0       6       4       0     -4.59   395    13  6.5       4
Shirov, Alexei           :       0      13       4       0     -4.12   425    13  8.5       3
Polgar, Judit            :       0       6       8       0     -6.36   335    13  4.0       5
Kasparov, Gary           :       0       7       0       1     -3.27   356    13  8.5       2
Bareev, Evgeny           :       1       8       0       1     -4.44   416    13  7.5       1
Gelfand, Boris           :       0      10       5       1     -4.78   376    13  5.5       3
Beliavsky, Alexander G   :       0      10       7       1     -6.14   342    13  2.0       4

TQ (Tournament Quality)  : -4.42
London 2016 is best so far.
MikeGL
Posts: 1010
Joined: Thu Sep 01, 2011 2:49 pm

Re: Engine insights on human and tournaments

Post by MikeGL »

Ferdy wrote:Started a project to create a human rating system based on engine evaluations. For now here are some initial measures on how the rating will be built.

Couple of tournament samples:

Tatat Steel 2017:

Code: Select all

Player                   : WinMiss Dubious Mistake Blunder aveErr/Ga   Pos Games  Pts DangerG
Karjakin, Sergey         :       0       3       0       0     -2.09   344    13  7.0       1
So, Wesley               :       0       6       0       0     -1.65   325    13  9.0       1
Carlsen, Magnus          :       1       3       1       0     -1.70   459    13  8.0       1
Aronian, Levon           :       0       8       1       0     -2.19   367    13  7.5       2
Giri, Anish              :       1       6       2       0     -3.76   371    13  6.5       1
Eljanov, Pavel           :       1      12       2       0     -3.86   455    13  7.0       1
Wei, Yi                  :       0      10       4       0     -2.15   466    13  7.5       1
Nepomniachtchi, Ian      :       1      12       4       0     -3.18   392    13  5.0       2
Wojtaszek, Radoslaw      :       1      14       4       0     -3.21   471    13  6.0       3
Harikrishna, Pentala     :       0      10       5       0     -3.02   413    13  6.0       2
Adhiban, Baskaran        :       0       6       2       1     -3.80   474    13  7.5       3
Andreikin, Dmitry        :       1       8       3       1     -4.33   426    13  6.0       2
Rapport, Richard         :       1       7       4       1     -4.35   356    13  4.5       4
Van Wely, Loek           :       1      19       7       3     -6.93   429    13  3.5       6

TQ (Tournament Quality)  : -3.30
WijkaanZee 2000:

Code: Select all

Player                   : WinMiss Dubious Mistake Blunder aveErr/Ga   Pos Games  Pts DangerG
Leko, Peter              :       0       5       0       0     -1.60   295    13  8.0       0
Anand, Viswanathan       :       0       8       0       0     -1.15   306    13  8.0       0
Adams, Michael           :       0      23       0       0     -3.52   398    13  7.0       1
Van Wely, Loek           :       1       5       1       0     -4.54   334    13  4.0       3
Nikolic, Predrag         :       0       7       1       0     -4.00   374    13  6.0       4
Kasparov, Garry          :       0      10       1       0     -1.87   397    13  9.5       0
Morozevich, Alexander    :       1      12       1       0     -4.20   349    13  7.5       3
Kramnik, Vladimir        :       0       8       2       0     -1.42   292    13  8.0       0
Korchnoi, Victor         :       0       9       3       0     -3.67   363    13  5.0       2
Timman, Jan              :       0       9       4       0     -3.88   378    13  6.5       3
Short, Nigel             :       0       7       1       1     -2.98   302    13  5.5       3
Piket, Jeroen            :       0      17       2       2     -3.32   361    13  6.5       2
Polgar, Judit            :       0       9       4       2     -3.80   348    13  5.0       2
Lputian, Smbat           :       0      12       3       3     -6.95   429    13  4.5       4

TQ (Tournament Quality)  : -3.35
London Chess Classic 2016:

Code: Select all

Player                   : WinMiss Dubious Mistake Blunder aveErr/Ga   Pos Games  Pts DangerG
Kramnik, Vladimir        :       0       1       0       0     -0.76   272     9  5.0       0
So, Wesley               :       0       3       0       0     -0.87   218     9  6.0       0
Caruana, Fabiano         :       0       5       0       0     -2.49   256     9  5.5       0
Giri, Anish              :       0       4       1       0     -1.70   295     9  4.5       0
Aronian, Levon           :       0       0       2       0     -1.74   208     9  4.0       2
Nakamura, Hikaru         :       0       5       2       0     -3.68   239     9  5.0       0
Vachier-Lagrave, Maxime  :       0      12       2       0     -3.38   253     9  4.0       1
Anand, Viswanathan       :       0       4       4       0     -2.60   199     9  5.0       1
Topalov, Veselin         :       0      11       5       0     -7.89   257     9  2.0       3
Adams, Michael           :       0       9       0       1     -2.26   272     9  4.0       2

TQ (Tournament Quality)  : -2.74

Code: Select all

LEGEND:
WinMiss   - count of games where a player miss the winning move in one of the pos.
            Winning move: engine_movescore > 3.0 and
            player_movescore <= 3.0 pawn units and the player failed
            to win this game.
Dubious   - count of positions where a playable position
            turns into a difficult position.
            Playable position: engine_movescore >= -0.5
            Difficult position: -0.5 < player_movescore >= -1.0
Mistake   - count of positions where a playable position
            turns into a very difficult position.
            Very difficult position: -1.0 < player_movescore >= -3.0
Blunder   - count of positions where a playable position
            turns into losing position.
            Losing position: player_movescore < -3.0
aveErr/Ga - average error per game, in pawn unit.
            Measures the accuracy of a player, higher is better.
            error = player_movescore - enginemove_score
            total_error = pos x error
            ave_error = total_error/pos
            aveErr/Ga = ave_error/num_games
DangerG   - is the number of games where the player is losing.
            A game is losing when there is one position where
            engine_movescore < -3.0

NOTES:
1. Engine used in the analysis is Brainfish 300117 64 POPCNT.
2. Analysis time/pos is 10.0s.
3. Average analysis depth reached is 27.
4. When the player move error is below 6 pawns from a
   given position, error is set to -6.0 pawns.
5. Analysis of position starts at move 12 in every game.
6. When the score of the position in a game is already
   below -6 pawns, analysis of succeeding position in this game
   will be stopped and analysis will continue in the next game.
7. The table below is sorted by aveErr/Ga in descending order,
   then Dubious, then Mistake, then finally by Blunder
   in ascending orders.
8. Tournament quality is the average error per game of all players.
   TQ = total aveErr/Ga / num_players
   A high TQ means the games have less mistakes overall.
Thanks for this informative data. Looks like the real performance of a player in a specific tournament can be determined by AvErr/Ga column. The outlier is Wijk aan Zee 2000 (now called Tata Steel), because there are better performing GMs than Kasparov, but Kasparov won, really strange.

Would be nice to see some tournaments where Fischer played in order to determine the strength of Fischer compared to current GMs. Tournaments specifically where Fischer failed and won, like


Stockholm Interzonal (1962) -Fischer 1st place
http://www.chessgames.com/perl/chess.pl?tid=79444

Curacao Candidates (1962) -Fischer complained against Russian players for cheating
http://www.chessgames.com/perl/chess.pl?tid=80067



Is this program/script available for download?
Ferdy
Posts: 4856
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: Engine insights on human and tournaments

Post by Ferdy »

MikeGL wrote: Thanks for this informative data. Looks like the real performance of a player in a specific tournament can be determined by AvErr/Ga column.
That criteria seems to have a nice impact on the strength of a player indeed. Avoiding blunders and mistakes is also a good indication of a good player. WinMiss will measure the ability of a player to punish opponents that makes big mistakes. I have other measures that I am not showing, all these measures will be considered in the final rating calculation of a player including the win/loss/draw results.
The outlier is Wijk aan Zee 2000 (now called Tata Steel), because there are better performing GMs than Kasparov, but Kasparov won, really strange.
It is possible that the position is very complicated that opponents failed to see the correct continuation. Or perhaps due to time pressure. Right now I am evaluating all positions starting from move 12, I was thinking of including only up to move 40, but the strength of a player should also be measured by how one would handle the time as it is part of the game.
Would be nice to see some tournaments where Fischer played in order to determine the strength of Fischer compared to current GMs. Tournaments specifically where Fischer failed and won, like


Stockholm Interzonal (1962) -Fischer 1st place
http://www.chessgames.com/perl/chess.pl?tid=79444

Curacao Candidates (1962) -Fischer complained against Russian players for cheating
http://www.chessgames.com/perl/chess.pl?tid=80067
It is good that you mentioned this because I am searching tournaments where Fischer is one of the participants. I will search for those games and analyze it.
Is this program/script available for download?
Not at the moment, this is far from finished. At a time when I publish the initial rating list based from this system, I will share the script and the positon database it created.
Ferdy
Posts: 4856
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: Engine insights on human and tournaments

Post by Ferdy »

MikeGL wrote:Stockholm Interzonal (1962) -Fischer 1st place
http://www.chessgames.com/perl/chess.pl?tid=79444
Stockholm Interzonal 1962:

Code: Select all

Player                   : WinMiss Dubious Mistake Blunder aveErr/Ga   Pos Games  Pts DangerG
Petrosian, Tigran V      :       0       8       0       0     -1.55   404    22 15.0       1
Fischer, Robert James    :       0      11       1       0     -2.65   748    22 17.5       0
Gligoric, Svetozar       :       0      11       1       0     -3.21   695    22 13.5       2
Geller, Efim P           :       1      11       3       0     -2.85   640    22 15.0       0
Barcza, Gedeon           :       0      16       3       0     -3.66   761    22 11.0       6
Pomar Salamanca, Arturo  :       1      10       4       0     -3.33   617    22 12.0       3
Bilek, Istvan            :       0      19       4       0     -4.62   595    22 11.0       4
Yanofsky, Daniel Abraham :       1      33       4       0     -5.06   654    22  7.5       8
Kortschnoj, Viktor       :       1      16       5       0     -3.51   761    22 14.0       3
Bolbochan, Julio         :       0      22       5       0     -3.74   601    22 11.5       4
Cuellar Gacharna, Miguel :       0      28       5       0     -7.79   708    22  5.5      11
Teschner, Rudolf         :       0      15      13       0     -7.26   643    22  6.5      11
Bisguier, Arthur Bernard :       0      18       2       1     -4.28   572    22  9.5       7
Benko, Pal C             :       1      10       3       1     -4.24   759    22 13.5       1
Uhlmann, Wolfgang        :       1      10       3       1     -5.01   690    22 12.5       4
Portisch, Lajos          :       0      11       4       1     -4.18   636    22 12.5       4
Filip, Miroslav          :       0      12       4       1     -3.22   501    22 14.0       1
Olafsson, Fridrik        :       1      20       4       1     -3.38   590    22 12.0       3
Schweber, Samuel         :       0      18       5       1     -5.71   586    22  7.0      10
German, Eugenio Maciel   :       0      27       6       1     -5.49   649    22  7.0       8
Aaron, Manuel            :       1      23       7       1     -6.60   539    22  4.0       9
Stein, Leonid            :       0      19       4       2     -2.86   687    22 13.5       4
Bertok, Mario            :       1      20       7       2     -4.68   495    22  7.5       7

TQ (Tournament Quality)  : -4.30

Impressive stats from Fischer, zero WinMiss indicates that he would punish lousy moves from opponents. The average error looks good considering the long tournament with more games. He was never in danger of losing a game in those 22 games.

Petrosian is dangerous here with that kind of stats. Could be one reason why he won the championships in 1963 against Botvinnik.

Looks like the top players from the past can play toe to toe from our current top players :) despite the fact that current players have the best resources in our time.

Next is candidates 1962.
whereagles
Posts: 565
Joined: Thu Nov 13, 2014 12:03 pm

Re: Engine insights on human and tournaments

Post by whereagles »

Pretty cool topic. Keep up the good work 8-)
Ferdy
Posts: 4856
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: Engine insights on human and tournaments

Post by Ferdy »

Ferdy wrote:
MikeGL wrote:Stockholm Interzonal (1962) -Fischer 1st place
http://www.chessgames.com/perl/chess.pl?tid=79444
Stockholm Interzonal 1962:

Code: Select all

Player                   : WinMiss Dubious Mistake Blunder aveErr/Ga   Pos Games  Pts DangerG
Petrosian, Tigran V      :       0       8       0       0     -1.55   404    22 15.0       1
Fischer, Robert James    :       0      11       1       0     -2.65   748    22 17.5       0
Gligoric, Svetozar       :       0      11       1       0     -3.21   695    22 13.5       2
Geller, Efim P           :       1      11       3       0     -2.85   640    22 15.0       0
Barcza, Gedeon           :       0      16       3       0     -3.66   761    22 11.0       6
Pomar Salamanca, Arturo  :       1      10       4       0     -3.33   617    22 12.0       3
Bilek, Istvan            :       0      19       4       0     -4.62   595    22 11.0       4
Yanofsky, Daniel Abraham :       1      33       4       0     -5.06   654    22  7.5       8
Kortschnoj, Viktor       :       1      16       5       0     -3.51   761    22 14.0       3
Bolbochan, Julio         :       0      22       5       0     -3.74   601    22 11.5       4
Cuellar Gacharna, Miguel :       0      28       5       0     -7.79   708    22  5.5      11
Teschner, Rudolf         :       0      15      13       0     -7.26   643    22  6.5      11
Bisguier, Arthur Bernard :       0      18       2       1     -4.28   572    22  9.5       7
Benko, Pal C             :       1      10       3       1     -4.24   759    22 13.5       1
Uhlmann, Wolfgang        :       1      10       3       1     -5.01   690    22 12.5       4
Portisch, Lajos          :       0      11       4       1     -4.18   636    22 12.5       4
Filip, Miroslav          :       0      12       4       1     -3.22   501    22 14.0       1
Olafsson, Fridrik        :       1      20       4       1     -3.38   590    22 12.0       3
Schweber, Samuel         :       0      18       5       1     -5.71   586    22  7.0      10
German, Eugenio Maciel   :       0      27       6       1     -5.49   649    22  7.0       8
Aaron, Manuel            :       1      23       7       1     -6.60   539    22  4.0       9
Stein, Leonid            :       0      19       4       2     -2.86   687    22 13.5       4
Bertok, Mario            :       1      20       7       2     -4.68   495    22  7.5       7

TQ (Tournament Quality)  : -4.30

Candidates 1962:

Code: Select all

Player                   : WinMiss Dubious Mistake Blunder aveErr/Ga   Pos Games  Pts DangerG
Petrosian, Tigran V      :       0       5       0       0     -1.36   539    27 17.5       0
Geller, Efim P           :       0      18       5       0     -2.31   619    27 17.0       2
Fischer, Robert James    :       0      31       6       0     -3.57   853    27 14.0       5
Benko, Pal C             :       3      11       3       1     -4.01   842    27 12.0       4
Keres, Paul              :       0      20       4       1     -2.84   618    27 17.0       2
Tal, Mihail              :       0       8       6       1     -5.37   608    21  7.0       8
Kortschnoj, Viktor       :       0      14       6       1     -3.62   857    27 13.5       6
Filip, Miroslav          :       0      28       9       3     -5.40   645    27  7.0       7

TQ (Tournament Quality)  : -3.56
Petrosian rocks :!: Botvinnik is in danger.

Fischer is clearly out of form, probably discouraged/annoyed of what he had experienced as was written in Sports Illustrated by himself :(

Petrosian is consistent since the Interzonal.

Code: Select all

Results from file c_Candidates1962.pgn:

No. Name                   Win Draw Loss Unf.  Score Games       %
------------------------------------------------------------------
  1 Petrosian, Tigran V     +8  =19   -0   *0   17.5    27   64.8%
  2 Geller, Efim P          +8  =18   -1   *0   17.0    27   63.0%
  3 Keres, Paul             +9  =16   -2   *0   17.0    27   63.0%
  4 Fischer, Robert James   +8  =12   -7   *0   14.0    27   51.9%
  5 Kortschnoj, Viktor      +7  =13   -7   *0   13.5    27   50.0%
  6 Benko, Pal C            +6  =12   -9   *0   12.0    27   44.4%
  7 Filip, Miroslav         +2  =10  -15   *0    7.0    27   25.9%
  8 Tal, Mihail             +3   =8  -10   *0    7.0    21   33.3%

Total Games:     105
White Wins:       27 (25.7%)
Black Wins:       24 (22.9%)
Draws:            54 (51.4%)
MikeGL
Posts: 1010
Joined: Thu Sep 01, 2011 2:49 pm

Re: Engine insights on human and tournaments

Post by MikeGL »

Ferdy wrote:
Ferdy wrote:
MikeGL wrote:Stockholm Interzonal (1962) -Fischer 1st place
http://www.chessgames.com/perl/chess.pl?tid=79444

Code: Select all

Results from file c_Candidates1962.pgn:

No. Name                   Win Draw Loss Unf.  Score Games       %
------------------------------------------------------------------
  1 Petrosian, Tigran V     +8  =19   -0   *0   17.5    27   64.8%
  2 Geller, Efim P          +8  =18   -1   *0   17.0    27   63.0%
  3 Keres, Paul             +9  =16   -2   *0   17.0    27   63.0%
  4 Fischer, Robert James   +8  =12   -7   *0   14.0    27   51.9%
  5 Kortschnoj, Viktor      +7  =13   -7   *0   13.5    27   50.0%
  6 Benko, Pal C            +6  =12   -9   *0   12.0    27   44.4%
  7 Filip, Miroslav         +2  =10  -15   *0    7.0    27   25.9%
  8 Tal, Mihail             +3   =8  -10   *0    7.0    21   33.3%

Total Games:     105
White Wins:       27 (25.7%)
Black Wins:       24 (22.9%)
Draws:            54 (51.4%)

Then the conclusion of some analysts/journalists were correct, that Karpov is slightly better than Fischer when it comes to accuracy and if both their games against Spassky were calculated with regards to win/loss ratio. With your above data I think Karpov can even beat Fischer if Fischer is out of form. But very close to each other.


Linares 1994

Code: Select all

Karpov, Anatoly          :       0       9       1       0     -2.20   391    13 11.0       0 
Sotckholm Interzonal 1962

Code: Select all

Fischer, Robert James    :       0      11       1       0     -2.65   748    22 17.5       0 
Candidates 1962

Code: Select all

Fischer, Robert James    :       0      31       6       0     -3.57   853    27 14.0       5