Accusations of cheating in the highest tournament levels of chess

[Kyriakos]

Thanks to Arakhor for supporting the thread, this isn't about chess theory but hopefully discussion on what is going on in the global chess GM microcosm, which curiously has gained a lot of interest online - of course due to scandal

A brief summation (not particularly good, but gives the main events) can be found in this 9 minute video:

I will also provide a synopsis:
Hans Niemann is a 19year old chess player, who recently defeated the world champion, Magnus Carlsen. Immediately afterwards, Carlsen (implicitly, but very clearly) made the insinuation that Hans was using a chess engine in that game. Now, keep in mind that it was a game played not online, but at a venue, so cheating would require some ingenious method to bypass the checks (various theories appeared as to how it might have been done, but nothing concrete was established against Hans).
Hans had cheated in the past, but supposedly only when he was 15-16, and only in some online games which didn't have a monetary prize.
Fast forward a few days, and Magnus resigned the next time he played Hans - this was for another tournament, but this time online (no checks for cheating were possible).
In the current tournament, Hans made it all the way to the semi-final (defeating some great players on the way...), but lost to a chinese player (Liem Le). Magnus is also in the semi-final, and might proceed (it would have been very interesting to see what would have happened, if he had to play Hans in the final ).

In this brief interview with Magnus, you can see that he promised to speak openly (one has to assume he will try to back the accusations of cheating) about the affair, after the tournament is completed:

Now it has to be said that all this (possible cheating) is not just a serious issue but a game-breaking one, because in chess, for very many years now, no human player can defeat a high-end chess engine. Kasparov, the former chess champion, had famously said that computers no longer can be seen as something to beat, but a source of help for human players to better understand chess.

-Have you been following this chess mess? And if so, are you resigned to playing as black or white? Personally I find it funny and can't see how Magnus will manage to cement his accusations with proof - although the chess world would definitely throw Hans to the pit before disinheriting Magnus.

 

[Kaitzilla]

Magnus should take up Civ 4.

It is much better than chess, and so far no one has bothered to make a civ engine to play the game optimally.


Other than the Bongcloud attack, chess has had no innovations lately.

 

[Kyriakos]

Hard to have innovations in a system which can be brute-forced by a computer. Although (by now) the human contribution is always related to serendipity...
Still, it is quite curious that Magnus reacted as he did, and unless he can provide actual proof of official (prize) tournament cheating, it will harm his own reputation too.

 

[Rg339]

it’s been done before. It might not even necessarily be innovative. At another tournament, a player, GM Gaioz Nigalidze(sp?) claimed to have to use the restroom(looked a move up here and there on his phone). Chess at high levels is a gentleman’s game: security protocols you might assume to be present usually are not. It’s just assumed everybody is on the level

 

[Narz]

I know that Neimann kid he's a douche

He's also been banned from multiple online chess platforms for cheating

 

[Kyriakos]

Carlsen also came across as a bit of a douche now, though. Quitting on the second move and making it all one huge drama is hardly what the world champion should be doing

 

[EnglishEdward]

Let me think how this might work:

(a) The expert makes a move
(b) A watching accomplice types that into a computer
(c) The computer finds the best move
(d) Computer sends signal to cheater (internal electrodes?)
(e) Cheater plays a move.

That takes a finite time.

I'd look at the time intervals the alleged cheater takes to play their moves.

 

[Narz]

Yeah that's one way cheaters are caught. Unnatural move times

 

[Rg339]

Let me think how this might work:

(a) The expert makes a move
(b) A watching accomplice types that into a computer
(c) The computer finds the best move
(d) Computer sends signal to cheater (internal electrodes?)
(e) Cheater plays a move.

That takes a finite time.

I'd look at the time intervals the alleged cheater takes to play their moves.

Vaguely recall it being done like that in chess, before.

The Casino method. Unlike the movie, the chess world levies very light penalties. They don’t do a Sam Rothstein over signaling. No table saws. Just light suspensions.

Niemann’s a known cheater. Magnus is likely indignant because, well, you would be, if you’re an all time top-fiver going up against that guy. Carlsen has also been on a bit of a trip against the structural set up of chess tournaments; I assume this matchup plays into his already existing grievance with chess organizers.

 

[Lexicus]

Carlsen also came across as a bit of a douche now, though. Quitting on the second move and making it all one huge drama is hardly what the world champion should be doing

Just wait and let him explain himself after the tournament. Then judge.

 

[TheMeInTeam]

Yeah that's one way cheaters are caught. Unnatural move times

much more common online than otb though. it is also harder to catch gms, because unlike random online cheaters they don't need to refer to their engine for 4-5 seconds (like every other move) to determine whether they should recapture the enemy's queen in an early queen trade when it's the only legal move or other obvious tells. gm can probably get 1 or 2 big hints across an entire game when a move isn't relatively obvious and it will benefit them greatly vs other gm.

there are a few things that bother me about this case:

• even if someone has cheated previously, presumably if allowed to play in this tournament at all, the past conduct has been forgiven for one reason or another.
• thus while our priors make us more inclined that niemann would cheat than other gms at the tournament, we still need evidence that niemann cheated. if that wasn't required to act on suspicions of him, he shouldn't be playing in tournaments again yet.
• the way carlson handled this was extremely unprofessional (quitting round robin tournament outright without presenting sound evidence, which has impact on both sponsors and other players who are innocent)
• the longer this goes w/o presenting evidence, the worse it looks for carlson.
• niemann will understandably endure suspicion for quite some time regardless. but nobody should be acting on that suspicion w/o observations consistent with cheating.

Carlsen also came across as a bit of a douche now, though. Quitting on the second move and making it all one huge drama is hardly what the world champion should be doing

maybe he's in some kind of bobby fisher esque phase. he didn't even feel like defending his world champion title, so top 2 from candidates will be playing for it instead. he won't be the world champion for much longer.

Hard to have innovations in a system which can be brute-forced by a computer.

i don't think there are many games left where a well-trained machine learning algorithm can't already outperform a human easily. it's true that nobody has bothered in many games, but bots have become ridiculous in a few cases (beating best sc player, getting very near unbeatable in rocket league, unbeatable in go etc). i'd guess games like dominions would take the most compute, but even there i'm not convinced it's beyond ML to play at least the scripting and empire building decisions better than humans.

 

[Kyriakos]

Tournament nears an end - so the promised explanation by Magnus draws nearer.

It may end the next time they play, since he won the first batch.

If Hans had played those engine-like movies Magnus does in the video, the tone may have been very different

 

[Birdjaguar]

i don't think there are many games left where a well-trained machine learning algorithm can't already outperform a human easily.

Civ I-VI

 

[amadeus]

Civ I-VI

King, we have found oil in our lands!

Excellent! Begin construction of a farm immediately!

 

[Kyriakos]

Ok, doesn't sound that promising tbh. Let's see. Supposedly (as he says in the video) he will make a first statement today or later this week.

 

[TheMeInTeam]

Civ I-VI

it hasn't been done yet to my knowledge, but i don't believe for a second that trained ML AI couldn't outperform humans in civ games.

 

[Kyriakos]

it hasn't been done yet to my knowledge, but i don't believe for a second that trained ML AI couldn't outperform humans in civ games.

At least it shouldn't be like the current civ "AI", which simply gets bonuses and thus "cheats"

 

[stinkubus]

It's already been mentioned ITT, but I have no idea why competitive chess persists when the game has, for all intents and purposes, been "solved". What can a human player do at this point that would be novel or interesting?

 

[Kyriakos]

It's already been mentioned ITT, but I have no idea why competitive chess persists when the game has, for all intents and purposes, been "solved". What can a human player do at this point that would be novel or interesting?

Pretty much nothing. They can only offer their own analysis for moves that mimic the chess engine's moves (which, unlike the chess engine's, only compute 4-5 moves ahead, in the case of both Hans and Magnus in the last two tournaments).
No human player could even win a single game against the top chess engines, afaik.

I suppose this is another reason why the current chess GM crowd is filled with clowns/characters; it wasn't this way before computers obliterated everyone.

 

[Narz]

It's already been mentioned ITT, but I have no idea why competitive chess persists when the game has, for all intents and purposes, been "solved". What can a human player do at this point that would be novel or interesting?

Chess will never be solved.

Computers will come to dominate more and more human endeavors including conversation, but we'll still want to talk.