Cheating suspicion at the Zadar Open in Croatia

[Dr.Wael Deeb]

I was sort of feeling sorry for this guy because he was accused of cheating, and although I am 99% sure he was cheating, 1% is still quite a bit to destroy someone's reputation. But then I read this today

http://www.chessbase.com/newsdetail.asp?newsid=8781

where he says:

How did you manage to beat the Croatian grandmasters?

I dont think there was anything special in my games at all. The Croatian GMs made some horrible mistakes during our games. Of course I practiced a lot with the computer, and after beating Rybka and Houdini by 10-0 each, i was absolutelly sure that no-one was gonna stop me winning. After all the arbiters were kind of polite and intelligent, you know, the other players weren't though....

He claims he beat both Rybka and Houdini 10-0 ... OK so now the 1% should be revised to probably about .00000000000000000000000000000000000000001% that he was NOT cheating.

This is a pure sample of stupidity from his side....

It looks like he hadn't a clue what the hell is he talking about
Dr.D

I glanced over it and was taken back while reading most of his comments.
Does he think everyone else are idiots? It's plain to all he is, along with being a cheat, a thief and a liar, a fool as well.

I think he'll be banned from future events.

The question that forces itself here Terry is how the hell did he do it

An advanced form of technology is involved here but what is it exactly
Dr.D

[Ajedrecista]

Hello:

I have used Chess Game Report 0.61 by Jack Welbourne in the game Kozul (0 - 1) Ivanov:

[pgn][Event "19th Open A"]
[Site "Zadar"]
[Date "2012.12.19"]
[EventDate "2012.12.16"]
[Round "6.2"]
[Result "0-1"]
[White "Zdenko Kozul"]
[Black "Borislav Ivanov"]
[ECO "E00"]
[WhiteElo "2638"]
[BlackElo "2227"]
[PlyCount "68"]

1. c4 c5 2. Nf3 Nc6 3. Nc3 g6 4. g3 Bg7 5. Bg2 e6 6. d3 Nge7 7. h4 h6 8. Bd2 d5 9. Qc1 b6 10. O-O Bb7 11. Rb1 d4 12. Na4 Qc7 13. a3 a5 14. Qc2 O-O 15. b4 axb4 16. axb4 Nxb4 17. Bxb4 cxb4 18. Rxb4 Nd5 19. Rbb1 Nc3 20. Nxc3 dxc3 21. e3 Ra3 22. Rb3 Rfa8 23. d4 Qxc4 24. Nd2 Qc7 25. Bxb7 Rxb3 26. Nxb3 Qxb7 27. Qxc3 Qd5 28. Qc2 Ra3 29. Rb1 Bf8 30. Nc1 b5 31. Nd3 Bd6 32. Nc5 Bxg3 33. fxg3 Rxe3 34. Kh2 Qf3 0-1[/pgn]

I used the following parameters:

Code: Select all

UCI engine: SF 2.3.1 w32.
2 threads and 256 MB/thread.
Analyze both sides.
Maximum depth: 12.
Skill Elo level: 2433 (I took the average rating of players: (2638 + 2227)/2 = 2432.5 ~ 2433).
First move to analyze: the 6th (I did not analyze the first five moves of this English Opening).

Here is the output:

Code: Select all

                                      CHESS GAME REPORT

Game scored at a 2433 ELO skill level and a max depth of 12.


                             ----------- WHITE ------------     ----------- BLACK ------------

GAME ELO                                 2355                               2618

AVERAGE ERROR FOR GAME       Expected    Actual    Rating       Expected    Actual    Rating
                               4,07       3,43      0,64          3,50       1,47      2,03

MOVE ERRORS                  Size       No.    %    Avg Err     Size       No.    %    Avg Err
                             None       13     45     0,00      None       18     62     0,00
                             Small       2      7     1,50      Small       3     10     1,28
                             Medium      5     17     4,16      Medium      5     17     3,17
                             Large       7     24     7,30      Large       3     10     7,63
                             Blunder     2      7    12,28      Blunder                  0,00

BEST MOVES                   Move        Exp   Act   Rating     Move        Exp   Act   Rating
                             22. Rb3      8,8   0,0    8,8      32..Bxg3    14,0   0,0   14,0
                             18. Rxb4     6,1   0,0    6,1      31..Bd6     10,0   0,0   10,0
                             16. axb4     5,9   0,0    5,9      11..d4       5,3   0,0    5,3
                             20. Nxc3     4,8   0,0    4,8      12..Qc7      4,1   0,0    4,1
                             10. O-O      3,8   0,0    3,8      29..Bf8      4,1   0,0    4,1

WORST MOVES                  Move        Exp   Act   Rating     Move        Exp   Act   Rating
                             11. Rb1      4,8  14,1   -9,2      10..Bb7      3,3   9,2   -5,9
                             30. Nc1      4,3  10,5   -6,2      22..Rfa8     2,6   4,8   -2,2
                             33. fxg3     3,3   8,9   -5,6      19..Nc3      3,4   5,5   -2,1
                             34. Kh2      0,9   6,4   -5,5      23..Qxc4     6,2   8,2   -2,1
                             25. Bxb7     4,7   8,2   -3,5      34..Qf3      0,1   0,0    0,1

MOST CHALLENGING POSITIONS   Move        Exp   Act   Rating     Move        Exp   Act   Rating
                             22. b1b3     8,8   0,0    8,8      32..d6g3    14,0   0,0   14,0
                             32. d3f4     6,8   8,0   -1,2      31..f8d6    10,0   0,0   10,0
                             21. c2b3     6,4   4,6    1,8      23..b7f3     6,2   8,2   -2,1
                             14. b1a1     6,3   7,0   -0,7      21..f8d8     5,8   1,8    4,0
                             18. b1b4     6,1   0,0    6,1      11..d5d4     5,3   0,0    5,3

  White       Pawn  WP   Exp   Act  Rating  Elo      Black       Pawn  WP   Exp   Act  Rating  Elo
  1. c4       0,00  50   0,0   0,0    0,0      0     1..c5       0,00  50   0,0   0,0    0,0      0
  2. Nf3      0,00  50   0,0   0,0    0,0      0     2..Nc6      0,00  50   0,0   0,0    0,0      0
  3. Nc3      0,00  50   0,0   0,0    0,0      0     3..g6       0,00  50   0,0   0,0    0,0      0
  4. g3       0,00  50   0,0   0,0    0,0      0     4..Bg7      0,00  50   0,0   0,0    0,0      0
  5. Bg2      0,00  50   0,0   0,0    0,0      0     5..e6      -0,36  41   0,0   0,0    0,0      0
  6. d3       0,20  55   4,9   4,1    0,8   2543     6..Nge7    -0,24  44   3,2   1,0    2,2   2618
  7. h4       0,24  56   3,8   0,0    3,8   2618     7..h6      -0,24  44   3,7   0,0    3,7   2618
  8. Bd2      0,20  55   1,6   1,0    0,6   2540     8..d5      -0,24  44   3,0   1,0    2,0   2618
  9. Qc1      0,08  52   3,8   4,1   -0,3   2354     9..b6      -0,16  46   3,4   2,0    1,4   2618
 10. O-O      0,16  54   3,8   0,0    3,8   2618    10..Bb7     -0,52  37   3,3   9,2   -5,9   1690
 11. Rb1     -0,04  49   4,8  14,1   -9,2    975    11..d4       0,04  51   5,3   0,0    5,3   2618
 12. Na4     -0,04  49   1,1   0,0    1,1   1744    12..Qc7      0,04  51   4,1   0,0    4,1   2618
 13. a3      -0,12  47   4,9   2,0    2,9   2618    13..a5       0,00  50   3,6   2,9    0,7   2524
 14. Qc2     -0,28  43   6,3   7,0   -0,7   1803    14..O-O      0,28  57   3,1   0,0    3,1   2618
 15. b4      -0,56  36   3,8   7,1   -3,3    289    15..axb4     0,56  64   3,1   0,0    3,1   1638
 16. axb4    -0,56  36   5,9   0,0    5,9   2618    16..Nxb4     0,56  64   1,1   0,0    1,1   1591
 17. Bxb4    -0,56  36   2,4   0,0    2,4    936    17..cxb4     0,56  64   0,9   0,0    0,9    662
 18. Rxb4    -0,56  36   6,1   0,0    6,1   2491    18..Nd5      0,56  64   0,8   0,0    0,8   1976
 19. Rbb1    -0,80  30   4,1   5,5   -1,4   1277    19..Nc3      0,56  64   3,4   5,5   -2,1   1869
 20. Nxc3    -0,56  36   4,8   0,0    4,8   2581    20..dxc3     0,56  64   0,2   0,0    0,2    184
 21. e3      -0,76  31   6,4   4,6    1,8   2577    21..Ra3      0,68  67   5,8   1,8    4,0   2618
 22. Rb3     -0,68  33   8,8   0,0    8,8   2618    22..Rfa8     0,48  62   2,6   4,8   -2,2   1737
 23. d4      -0,48  38   2,7   0,0    2,7   2618    23..Qxc4     0,16  54   6,2   8,2   -2,1   2126
 24. Nd2     -0,16  46   2,5   0,0    2,5   1746    24..Qc7      0,08  52   2,1   2,0    0,1   2414
 25. Bxb7    -0,40  40   4,7   8,2   -3,5   2159    25..Rxb3     0,40  60   2,4   0,0    2,4   1977
 26. Nxb3    -0,40  40   1,0   0,0    1,0   1460    26..Qxb7     0,40  60   0,5   0,0    0,5    391
 27. Qxc3    -0,40  40   3,2   0,0    3,2   1599    27..Qd5      0,24  56   4,7   4,1    0,6   2543
 28. Qc2     -0,36  41   5,1   3,1    2,0   2618    28..Ra3      0,36  59   3,0   0,0    3,0   2618
 29. Rb1     -0,36  41   3,5   0,0    3,5   2611    29..Bf8      0,36  59   4,1   0,0    4,1   2618
 30. Nc1     -0,80  30   4,3  10,5   -6,2    499    30..b5       0,80  70   3,7   0,0    3,7   1706
 31. Nd3     -1,05  26   2,9   4,9   -2,0   1969    31..Bd6      1,05  74  10,0   0,0   10,0   2618
 32. Nc5     -1,61  18   6,8   8,0   -1,2    331    32..Bxg3     1,61  82  14,0   0,0   14,0   2527
 33. fxg3    -2,86   9   3,3   8,9   -5,6   2105    33..Rxe3     2,86  91   0,1   0,0    0,1   1729
 34. Kh2     -6,54   2   0,9   6,4   -5,5     15    34..Qf3      6,54  98   0,1   0,0    0,1   1215

It looks like 2618 is a kind of upper bound because I see this value many times. WP columns mean 'winning probability' (in percentage).

If I take a look to 'MOVE ERRORS' section:

Code: Select all

  Accumulated error moves (out of 34 - 5 = 29):
                                    Kozul  Ivanov
None:                                13      18
None + small:                        15      21
None + small + medium:               20      26
None + small + medium + large:       27      29

It looks curious to me that a 2227 Elo player did not do a single blunder against a 2638 Elo player.

This tool must not be understand as definitive but only orientative.

Regards from Spain.

Ajedrecista.

[Lavir]

The question that forces itself here Terry is how the hell did he do it

An advanced form of technology is involved here but what is it exactly
Dr.D

I actually think that what he used to cheat was indeed very simple. He had a friend watching the broadcast in playchess/tourney site/whatever and telling him what he should play (and maybe even alternative moves) via a special tool inside his ear.

Very simple, but with searching you would never find that. The tool can be implanted so deep that only a doctor can find and remove the same.

There are two principal reasons why I think this was the case and something very simple as that was used:

1) When the broadcast was suspended Ivanov resumed to play as he played before. His style suddenly changed back to the semi-defensive/boring he had, and doing a lot of mistakes. When the broadcast resumed, he "magically" started again to play in attacking/chaotic style, deeply tactically and making no errors whatsoever.

2) When the time gets low, broadcast is an huge issue to rely on. The time lapse doesn't permit you to transmit moves in time. This is clearly seen in a very long drawish game where Ivanov, short on time, committed an horrible blunder that was very easy to spot even for a 1500 rated player. I think that he would himself spot it easily if he wasn't relying on the engine and not even analyzing. When the move didn't came because there was no time he had to move by himself, but since he didn't analyze the board he just moved without watching well and committed that horrible blunder. It can even be that he actually made that blunder voluntarily, as a sort of "evidence" that he was not cheating, but however this point is still valid, because you can see clearly that the last moves of that game (while being draw in any case) where somewhat subpar in confront to others, as it happens when engines have very little time to ponder, or they are forced to go on a line. BTW if he did the blunder voluntarily, he is an idiot, because nobody with his rating would make that blunder playing, there's no way, so, apart the suddenness change of style and other things, that particular blunder is one of the more glaring evidences that something was wrong.

[Dr.Wael Deeb]

The question that forces itself here Terry is how the hell did he do it

An advanced form of technology is involved here but what is it exactly
Dr.D

I actually think that what he used to cheat was indeed very simple. He had a friend watching the broadcast in playchess/tourney site/whatever and telling him what he should play (and maybe even alternative moves) via a special tool inside his ear.

Very simple, but with searching you would never find that. The tool can be implanted so deep that only a doctor can find and remove the same.

There are two principal reasons why I think this was the case and something very simple as that was used:

1) When the broadcast was suspended Ivanov resumed to play as he played before. His style suddenly changed back to the semi-defensive/boring he had, and doing a lot of mistakes. When the broadcast resumed, he "magically" started again to play in attacking/chaotic style, deeply tactically and making no errors whatsoever.

2) When the time gets low, broadcast is an huge issue to rely on. The time lapse doesn't permit you to transmit moves in time. This is clearly seen in a very long drawish game where Ivanov, short on time, committed an horrible blunder that was very easy to spot even for a 1500 rated player. I think that he would himself spot it easily if he wasn't relying on the engine and not even analyzing. When the move didn't came because there was no time he had to move by himself, but since he didn't analyze the board he just moved without watching well and committed that horrible blunder. It can even be that he actually made that blunder voluntarily, as a sort of "evidence" that he was not cheating, but however this point is still valid, because you can see clearly that the last moves of that game (while being draw in any case) where somewhat subpar in confront to others, as it happens when engines have very little time to ponder, or they are forced to go on a line. BTW if he did the blunder voluntarily, he is an idiot, because nobody with his rating would make that blunder playing, there's no way, so, apart the suddenness change of style and other things, that particular blunder is one of the more glaring evidences that something was wrong.

Thanks Fabio....

I extremely enjoyed your analysis of the situation....
Dr.D

[simonhue]

He's beating Houdini 10-0 and that's why he decided to copy Houdini's first lines in Zadar - doesn't fit

[styx]

Kenneth Regan, an associate professor of computer science at the University at Buffalo who is also an international master, has spent five years building mathematical models of how players at different levels of skill are expected to perform. The models are based on an analysis of tens of thousands of games.

By comparing tournament results against the models, he comes up with a score of how well players did, move by move, versus how well they were expected to do. He also compares the players’ moves with those of two top-selling computer chess programs.

Regan ran Ivanov’s performance through his system and found that he had the highest move performance score he had ever analyzed — better even than any individual effort by Magnus Carlsen, the player who just rose to the highest chess rating in history, at 2,861.

In an e-mail, Regan wrote that even assuming that Ivanov had a rating of about 2,700, which would make him among the best players in the world, statistically he played well above what a world-class player normally would and more like a computer.

Source: nytimes.com

[Lavir]

Btw Ivanov will play on the 35th Bulgarian International Open Chess Championship.
http://chess-results.com/tnr85628.aspx?lan=24

Valeri Lilov will be there too, so it will be interesting to see what happens and what Ivanov will do.

IMO whatever Ivanov will do will be bad for him. If he decides to continue cheating this time he will be caught because everybody is waiting for it; if he however doesn't cheat, people will immediately understand that something is wrong just by comparing the games. Many of those that still defend him (usually low rated amateurs that have not the experience to understand certain things) do because they think it's possible to increase your play so dramatically in even so short time, but if then he would return to play as badly as before (and with even that boring defending style he has in opposition to the attacking/inventive one he has shown), then even these people will understand that something is, indeed, not fitting.

The only thing that Ivanov could do to try to escape his written fate is to try to cheat in an intelligent way, but A) I don't think he is able to no matter what, B) it usually requires a knowledge of chess that IMO the guy doesn't have, C) also if he could indeed do it, by cheating in that gross way as he did before he actually prevented himself to cheat intelligently: he changed his style of play by following the engine style, so to be plausible he would have to cheat intelligently while at the same time simulating the attacking style he doesn't have; even for a GM this would be extremely difficult (if not impossible in any case) to do.

[JBNielsen]

Here are some analyses by the Dabbaba-tool: http://talkchess.com/forum/viewtopic.ph ... 32&t=45254

Ivanov has a brain like rybka4

[simonhue]

A new video by Valiery Lilov on Borislav Ivanov's last two tourneys:

http://www.youtube.com/watch?v=qhfCUdy2Tzk&hd=1

[Cubeman]

A new video by Valiery Lilov on Borislav Ivanov's last two tourneys:

http://www.youtube.com/watch?v=qhfCUdy2Tzk&hd=1

I wonder if Lilov will make a new video on Borislav from he latest tournament.
http://chess-results.com/tnr93301.aspx?art=1&rd=9&lan=1