Carlsen withdrawal after loss to Niemann

[Uri Blass]

I think that this is just another metric to look at. I always believed that there must be data or information in the games played if analyzed statistically en masse. Taking one game at a time is probably not useful unless the cheating is very primitive. Of course you have to weed out quick draws that are based on preparation and look at games that were actually played after book lines ... but maybe not! Maybe memorization of book lines in an unusualIy high fashion might also be unusual. I would not use that as the only metric but it would be something if there is one person that has a score that sticks out when compared to other players in the same category.

I am now pretty sure that it was Chess.com guys, after they bought out MC companies ... gave MC incriminating evidence of cheating by Hans and Dlugy. A few weeks before that Hans and MC seemed to be on amicable terms playing chess on the beach, but something changed after Chess.com bought out the MC group of companies.



On another note ... interesting that Bobby Fischer had the highest rating at 72% ... higher than MC and Kasparov! All way before engines were even a factor in chess preparation. That just shows what a genius he was!!!

I think that the % may be dependent on the style of play and not only on playing strength.
Even if Fischer had 72% and Carlsen 70% it does not mean Fischer played better and it is not clear to me how you calculate the percentage.

Even if you use the same engine than the question is how much time you give it.
Basically I would like an option to get deterministic results for every chess player in the future and in the past so I guess we need a fixed engine with a single core and fixed number of nodes for the analysis and even if we get it then it is not clear that a bigger percentage is always better because playing a different move that the engine does not choose does not mean worse move if there are some equal moves.

The question is not only what is the percentage of the correct moves but also about the quality of non correct moves.

[M ANSARI]

I might get this Chessbase software to check it out. I like the idea of using many different engines with analysis of each engine output. Today there are so many engines above 3200 ELO that it is impossible to check all possible engines. But it could be useful in a way to tune an engine to play aggressively or say more positionally ... without having it affect the ELO of the engine too much. I will say that if someone wants to cheat ... the database feature would be an extremely useful ... you could tune the database against different players and you would be able to navigate many lines that are super sharp and super critical and you would be able to do it instantly. That would be impossible to catch as you could just say it was home preparation. Some of the lines are so sharp that one single misstep and you are totally busted. Playing quickly would also put your opponent in time trouble ... thus much higher chances of a misstep due to time trouble. Another non engine and very useful tool would be EGTB's ... in many drawn positions it is extremely difficult for a human to play only moves under time pressure. If you look at Vincent Keymer's game against MC ... he was really doing well and holding the draw ... but then one little error and was lost.

[Chessqueen]

I guess all of Hans games will be analyzed now ... including OTB games. I hope nobody will accuse me of "gossip" but this french chess player seems to have some interesting data using a Chessbase feature which can check many different engines at once and give you a percentage of the moves that match engine moves. It seems Hans has 100% perfect play in several games! I don't know how other players relate to this but would be interesting to find out.

You can watch this here

Wow, just wow. Yet nothing will ever be good enough for some when it comes to this topic. SMH

As I mentioned in a previous post Carlsen accused GM Hans because He best him and Carlsen is in a mission to reach 2900, but it is so hard to reach 2900 when you have to face young talents that are rated much lower than you. As you can see Carlsen lost 4.7 ratings points https://2700chess.com/

[syzygy]

I believe that Niemann has cheated more - and more recently - than he has publicly admitted.

[M ANSARI]

Has anyone seen this game ... OMG !!! I will let the Chess speak for itself! This is against the youngest kid to ever get a GM title, the young american prodigy Mishra Abhimanyu. Poor kid had no idea what he was dealing with !!! Burying your Knight so that somehow your king can get one tempo ahead and win after 15 or more moves! This is just ridiculous.

I forgot how to post games but here is the pgn

[Event "CCCSA Fall GM"]
[Site "Charlotte, NC USA"]
[Date "2020.10.04"]
[EventDate "?"]
[Round "8.5"]
[Result "0-1"]
[White "Abhimanyu Mishra"]
[Black "Hans Moke Niemann"]
[ECO "C55"]
[WhiteElo "2342"]
[BlackElo "2465"]
[PlyCount "112"]

1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 4. d3 h6 5. c3 d6 6. O-O g5 7. Nbd2 g4 8. Ne1 h5 9. f4 exf4 10. Rxf4 Bh6 11. Rf1 Qe7 12. Nb3 Ne5 13. Bxh6 Rxh6 14. Nd2 Bd7 15. Bb3 g3 16. h3 Nfg4 17. Qe2 Nh2 18. Qe3 Rh7 19. Rf4 h4 20. d4 Ng6 21. Nd3 O-O-O 22. Re1 Re8 23. e5 Nxf4 24. Nxf4 c6 25. Nc4 d5 26. Nd6+ Qxd6 27. exd6 Rxe3 28. Rxe3 Rh6 29. Ne2 Re6 30. Rxe6 Bxe6 31. Bc2 Kd7 32. Nf4 Kxd6 33. b4 b5 34. a4 bxa4 35. Bxa4 Bf5 36. Bb3 Be6 37. Ba4 Bc8 38. Bd1 Ba6 39. Be2 Bxe2 40. Nxe2 Ke6 41. Nc1 Kf5 42. Nb3 Ke4 43. Na5 f5 44. Nxc6 f4 45. b5 f3 46. gxf3+ Nxf3+ 47. Kf1 Ng5 48. Ne5 Nxh3 49. Ng6 g2+ 50. Kxg2 Nf4+ 51. Nxf4 Kxf4 52. Kh3 Ke4 53. Kxh4 Kd3 54. Kg4 Kxc3 55. Kf5 Kxd4 56. Ke6 Kc5 0-1

[CornfedForever]

I believe that Niemann has cheated more - and more recently - than he has publicly admitted.

Nothing new there...he knows chess.com has more than Niemann has admitted to, because they have said so.

The real question has for sometime been: any evidence of OTB cheating? Frankly...it sounds like Magnus has none and is doubling down on the 'online cheating' idea. I can't say he 'wins' with that argument.

Hans' lawyer might well tell him to say, "You have my permission to speak freely about any OTB cheating evidence you have."

I think both sides are just trying to put each other 'in a box' at the moment.

[dkappe]

I guess all of Hans games will be analyzed now ... including OTB games. I hope nobody will accuse me of "gossip" but this french chess player seems to have some interesting data using a Chessbase feature which can check many different engines at once and give you a percentage of the moves that match engine moves. It seems Hans has 100% perfect play in several games! I don't know how other players relate to this but would be interesting to find out.

You can watch this here

One game with “computer like play” discussed in this video is this one:

[pgn]
[Event "CCCSA Fall GM"]
[Site "Charlotte, NC USA"]
[Date "2020.10.04"]
[EventDate "?"]
[Round "8.5"]
[Result "0-1"]
[White "Abhimanyu Mishra"]
[Black "Hans Moke Niemann"]
[ECO "C55"]
[WhiteElo "2342"]
[BlackElo "2465"]
[PlyCount "112"]

1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 4. d3 h6 5. c3 d6 6. O-O g5 7. Nbd2 g4 8. Ne1 h5 9. f4 exf4 10. Rxf4 Bh6 11. Rf1 Qe7 12. Nb3 Ne5 13. Bxh6 Rxh6 14. Nd2 Bd7 15. Bb3 g3 16. h3 Nfg4 17. Qe2 Nh2 18. Qe3 Rh7 19. Rf4 h4 20. d4 Ng6 21. Nd3 O-O-O 22. Re1 Re8 23. e5 Nxf4 24. Nxf4 c6 25. Nc4 d5 26. Nd6+ Qxd6 27. exd6 Rxe3 28. Rxe3 Rh6 29. Ne2 Re6 30. Rxe6 Bxe6 31. Bc2 Kd7 32. Nf4 Kxd6 33. b4 b5 34. a4 bxa4 35. Bxa4 Bf5 36. Bb3 Be6 37. Ba4 Bc8 38. Bd1 Ba6 39. Be2 Bxe2 40. Nxe2 Ke6 41. Nc1 Kf5 42. Nb3 Ke4 43. Na5 f5 44. Nxc6 f4 45. b5 f3 46. gxf3+ Nxf3+ 47. Kf1 Ng5 48. Ne5 Nxh3 49. Ng6 g2+ 50. Kxg2 Nf4+ 51. Nxf4 Kxf4 52. Kh3 Ke4 53. Kxh4 Kd3 54. Kg4 Kxc3 55. Kf5 Kxd4 56. Ke6 Kc5 0-1
[/pgn]

In particular, the move 17… Nh2 has just got to be computer cheating.

Since it’s so obvious, I figured I should be able to reproduce the analysis. My weapon of choice, SFDev from 9/18/2022 running at 20 ply multipv.

I looked at all the black positions from 17 ply onward. The columns are as follows:

PL = ply
LM = legal move count
GM = good move count, within 100 centipawns of the best move
DF = diff between the best move score and the chosen move score
BM = algebraic best move
CM = algebraic chosen move

Code: Select all

PL,LM,GM,DF,BM,CM
17,33,12,-29,f8h6,e5f4
19,32,6,-6,d8e7,f8h6
21,37,6,0,d8e7,d8e7
23,42,4,0,c6e5,c6e5
25,34,1,0,h8h6,h8h6
27,34,1,-102,g4g3,c8d7
29,36,1,0,g4g3,g4g3
31,39,1,0,f6g4,f6g4
33,42,1,0,g4h2,g4h2
35,44,1,0,h6h7,h6h7
37,42,1,0,h5h4,h5h4
39,41,1,0,e5g6,e5g6
41,41,2,0,e8c8,e8c8
43,40,1,-120,g6f4,d8e8
45,40,1,0,g6f4,g6f4
47,36,1,-111,h7h6,c7c6
49,33,2,-112,h7h6,d6d5
51,4,1,0,e7d6,e7d6
53,29,1,0,e8e3,e8e3
55,21,1,0,h7h6,h7h6
57,24,1,0,h6e6,h6e6
59,15,1,0,d7e6,d7e6
61,17,4,0,c8d7,c8d7
63,17,4,0,d7d6,d7d6
65,18,8,-76,d6d7,b7b5
67,17,2,0,b5a4,b5a4
69,16,4,-12,d6c7,e6f5
71,21,11,-2,d6d7,f5e6
73,16,4,-42,d6c7,e6c8
75,18,1,0,c8a6,c8a6
77,15,1,0,a6e2,a6e2
79,12,2,0,d6e6,d6e6
81,13,1,0,e6f5,e6f5
83,13,1,0,f5e4,f5e4
85,11,1,0,f7f5,f7f5
87,9,1,0,f5f4,f5f4
89,9,1,0,f4f3,f4f3
91,6,1,0,h2f3,h2f3
93,14,1,0,f3g5,f3g5
95,11,1,0,g5h3,g5h3
97,11,1,0,g3g2,g3g2
99,9,1,0,h3f4,h3f4
101,6,1,0,e4f4,e4f4
103,7,5,0,f4e3,f4e4
105,7,3,0,e4d3,e4d3
107,9,3,0,d3c3,d3c3
109,10,2,0,c3d4,c3d4
111,8,3,0,d4c4,d4c5

One thing you’ll note is that there are many positions with only one good move, especially in the “computer like” part around Nh2, and even here Niemann doesn’t always find the right move. (I know that when I think I’m about to lose or fritter away an advantage, I look a little bit harder for that “spectacular” move that saves the day.) The play of Niemann certainly didn’t match 100%. He had a very narrow path to follow and he could have lost the whole thing at times.

Certainly I could search a bit deeper here. If someone is able to reproduce the results in the video, please follow up.

[syzygy]

The real question has for sometime been: any evidence of OTB cheating? Frankly...it sounds like Magnus has none and is doubling down on the 'online cheating' idea. I can't say he 'wins' with that argument.

I think the real question is if Hans' rating is anywhere close to his chess ability.

If he is really around 2700, why would he cheat online (or OTB).

[Chessqueen]

The real question has for sometime been: any evidence of OTB cheating? Frankly...it sounds like Magnus has none and is doubling down on the 'online cheating' idea. I can't say he 'wins' with that argument.

I think the real question is if Hans' rating is anywhere close to his chess ability.

If he is really around 2700, why would he cheat online (or OTB).

According to Carlsen latest interview, he mentioned that GM Hans was coming with great moves and he wasn't really concentrating during the game, that is when he knew that he must be cheating ==>https://en.chessbase.com/post/carlsen-s ... on-niemann

NOTE: All that I have to say is that if a Patzer like me can memorize an analyzed game previously played by Carlsen, but analyzed up to 28 moves by stockfish or Dragon I really do NOT have to be fully concentrated, and there is nothing wrong by using your memory

[CornfedForever]

The real question has for sometime been: any evidence of OTB cheating? Frankly...it sounds like Magnus has none and is doubling down on the 'online cheating' idea. I can't say he 'wins' with that argument.

I think the real question is if Hans' rating is anywhere close to his chess ability.

If he is really around 2700, why would he cheat online (or OTB).

Good question. You might want to ask Parham Maghsoodloo. If I am not mistaken, I think he was actually over 2700 (blitz) and close to it in Standard play when he was caught cheating online.