I wonder if a possible ‘path to redemption’ for Niemann is to admit he cheated, explain how, and then chess tournaments can implement measures to prevent it happening again. But I think it’s a highly unlikely outcome, because there is very little incentive for him to do this - since without such a confession, there can be no conclusive proof that he did cheat against Carlsen. Whether he did or not, I see several possible outcomes:
He did cheat that game (and in others, as per the chess.com analysis and his own admission), but doesn’t do so again. Then either a) he manages to improve his performance without cheating and becomes a bona fide super-GM or b) he rarely reaches that level of play again.
He didn’t cheat in that game, beating Magnus fair and square. Then either a) he plays well in other tournaments (I’m going to assume without cheating) and becomes a bona fide super-GM or b) he rarely reaches that level of play again.
How helpful would a simple binary code be? Buzz for “your opponent made a mistake and you have an opening” and nothing for no opening?
Apparently there are some games where chess.com allows their analysis bar. In low amateur games it’s pretty helpful since there might be a wild swing and for M3, which is something a 1000 might not see. I imagine the benefit would decrease as you go up and would need more direct instruction?
Even easier than that, since each of those characters only has 8 possibilities. So a mere three bits per character would suffice, or the equivalent of 1.5 standard characters for the whole move.
And of course you could get even more compact, if you wanted, though I’m not sure the added complication would be worthwhile: Each player has only 16 pieces, so you can specify a piece with 4 bits, and each piece has at most 27 different moves available at any given moment, so you can specify a move for a piece with 5 bits, for a mere 9 bits per move.
It would be incredibly helpful. Just the other day, I was playing a rapid game that ended up in a bishop ending that was a sort of fortress position. We each made a bunch of moves in the ending but no progress seemed possible and the game ended in a draw. Afterwards the computer analysis showed a big spike for one move in the ending - my opponent allowed a relatively simple tactic that neither of us noticed. During the game, if I had been notified at that moment, I would’ve found the win immediately. Obviously GM games would tend to have more complicated missed tactics, but the effect is the same.
Or, for a slightly more famous example, see game 6 of the second Carlsen-Anand world championship. Carlsen made a mistake with 26.Kd2, which would have given Anand a winning position had he noticed the tactic. But Anand continued with his normal play relatively quickly, and ended up losing the game. Had Anand been informed that there was a winning tactic in that position, he would’ve found it in just a few seconds.
I’ll let you correct me (since I’m no expert), but I’d guess that a few distinguishable buzz-types would be even more helpful. Something like “there’s a really good sacrifice here” or “you should play for a win, not a draw”.
It feels to me like trying to encode full moves into a signal is just too much to keep track of, at least without it becoming obvious. But a few high-level signals could give almost all the benefit if the player is already at a high level.
Sure, more information would be more useful. But you’re right that one or a few different signals about the state of the game is enough to give a strong player a huge advantage.
Magnus gave this interview (English subtitles) on the topic about a year ago. At about 5 minutes in:
But, had I started cheating in a clever manner, I am convinced that no one would notice. I would’ve just needed to cheat one or two times during the match, and I would not even need to be given moves, just the answer on which move was way better. Or, here there is the possibility of winning and you need to be more careful. That is all I would need in order to be almost invincible. Which does frighten me.
That’s a very interesting interview. It’s clear that Carlsen (and probably other strong players) have developed a sense for whether the other player is playing a move in line with their general style, or if it seems somehow out of the blue.
A more clever cheater could train a computer on their own games, producing moves with the same general style, but with no blunders, and some bias toward the better moves.
I didn’t see it mentioned in this thread yet: chess.com bid to acquire Play Magnus Group this August. There has been speculation that as a result of this deal Carlsen was aware of Niemann’s full history of cheating on chess.com.
I think this is a rare case where the truth ought to come out soon enough. Either Niemann can play at this level or not.
I’ve always been of the opinion that, once you show a propensity for cheating, that should be enough to kick you out. Which matches you cheated on is not so important.
There must be harsh penalties on the cheaters that get caught to discourage people for even attempting to cheat.
I would think even two instances of cheating in the past should have banned someone from play, at least, for some long duration. And this Man instances of cheating should be a lifetime ban, along with more if you can figure out how.
Heck, I would not trust this guy with anything ever in the rest of his life due to that much cheating. The idea is that someone who can so casually cheat clearly doesn’t see anything wrong with it, other than being caught. It would be like trusting a chronic liar.
That doesn’t mean I think finding out how he cheats would be a bad thing. For one, it would help with future detection. But I don’t think it should matter.
And as for the defamation situation: I don’t believe he would need to prove that he actually cheated. Just that it would not be unreasonable to conclude he had. Sure, the burden is on the accuser, but not to the point of having to prove the allegation entirely. Just that they weren’t recklessly disregarding the truth. If other experts would agree that it was suspicious, that would seem to have been sufficient.
That’s said, I think most people would prefer to avoid the need to prove that, and thus play it more cautiously. It makes sense that Carlson had more information than we were privy to.
Well, assuming he can’t maintain whatever cheating he’d been doing. Are they doing anything extra to make sure he doesn’t have some sort of cheating device on him?
Assuming he cheated against Carlson, the people in charge of detecting cheating apparently didn’t notice. So surely they need to step it up to detect him.
Thanks. The actual tournament website is behind, so it’s nice to see something current.
I watched the last 45 minutes, which doesn’t give the whole flavor of the game, but the complexities of the endgame were interesting. Also, I always enjoy Yasser Seirawan (or I have since the days of “Inside Chess”, anyway).
It seems that the current tournament is doing more: RF scanning in addition to the normal metal detectors, a cheating expert observing, and a longer delay on the live stream.
But my main assumption was the same as yours: it would be very difficult to cheat undetected for very long at that level of scrutiny.
The live stream has a 30 minute delay to combat cheating.
The Sinquefield Cup implemented a 15 minute stream delay after the kerfluffle with Carlsen. It was going out live before that, which absolutely blows my mind.
Personally, I don’t see how over the board events can possibly go forward without having stream delays. A 15 or 30 minute delay has no effect on viewers’ appreciation of the games and is short enough for people to avoid spoilers if they cared about that sort of thing but long enough to make it very difficult for players to take advantage of it.
?