Anyone following the Chess drama?

[Ex-User (9086)]

For those who don't know there's been a big drama in the top level chess that I won't bother to describe. It includes over the board cheating accusations and lawsuits and it will take years to clear up probably.

I don't really care about who's right or wrong there, but I'm more interested in the principles. My point is that top level chess is kind of a solved game, no cheating is required.

As a side note cheating in live chess is quite easy. I see a few good schemes to cheat:
1. Version with no input device. Player has a signal receiver in their shoe. It's been shown that the official procedures don't detect hidden devices placed around armpits, or under players feet if said devices are on standby or are not receiving signals at the time of the check.
- Most top level games are broadcasted and observed publicly.
- One of the observers is working with the player or the player has a program that reads the publicly made moves and they input said moves to the chess engine which then chooses the right move and said move is transmitted to the player
- Players shoe-installed receiver buzzes a few times to inform about the best computer engine move

2. Input can be done with pressing the front or back of the feet. Chess moves aren't very complicated to describe so this doesn't need more than one or two buttons.


I want to say that cheating isn't needed. The guy currently accused of cheating may be really good at opening preparation and may have memorized so many moves and games that they simply play the game from memory. One of the accusations against him is that he used to cheat and another was that he plays some moves effortlessly. Making effortless moves is totally reasonable if the person just recalls a position from memory and relies on what they had studied using the chess engine before the game.


An average game of chess takes 40 moves.

Most people who don't play chess don't know that chess openings describe the computer-solved ideal 20+ moves at the beginning of each game and top players have most relevant 20+ move openings memorized. It doesn't take genius to have one player who memorizes 40+ moves instead and has an inherent advantage. Memorizing 40 moves takes a lot more than just 20, because of the variations and responses that branch after every move, but my point is-it's possible-and I don't see why it hasn't been done earlier.

After the opening, the middle game is said to be the only creative part of the game. That's where the game goes too deep for opening preparation, the variation on the board hasn't been repeated in other matches and there are enough pieces on the board that brute-force calculation does not give unambiguous answers.

Finally when the amount of chess pieces on the board drops to about 9 the game goes back to being completely solved. All endgames have been researched and most top players remember how to finish win or draw a game with 9 pieces on the board.

It has to be noted that top level players remember full games they played or studied in the past. So if at any point the game approaches the board state that they remember then they can just repeat moves from memory. Most players try to avoid repeating games that, as they remember, have ended with an unfavorable result.

There are specific theoretical lines and games that have been researched so extensively that they've been solved completely from beginning to end.

There are famous situations where one grandmaster player had the whole opening prepared and the other strong opponent did not memorize it as deeply and that was enough for said top player to completely destroy the other.


My conclusion is that at this point it's a contest of who can memorize more variations and repeat them accurately to the point of scoring sufficient advantage to close the game with their inferior real-time calculation.

 

[Hadoblado]

I followed it loosely when it first broke. There was also a poker cheating drama around the same time.

I don't believe the public (including myself) are adequately equipped to make the assessment as to whether someone is cheating. The knee-jerk response of many is that you can't possibly know, but for example, in speed-run cheaters can be identified with incredible accuracy if you know enough. While I don't expect the same level of provability in chess, there are almost certainly types of reads that most commentators are unqualified to interpret.

That said, an expert in chess isn't necessarily an expert in psychology or statistics. I don't think Magnus handled it well (he had no problem in previous tournaments when he beat the same guy who is known to have a history of cheating). He only made it an issue when he lost.

Either way, live competitions now have known vulnerabilities that need to be addressed if people are to trust results. This goes for chess, but also other games including esports, especially when there is an element of information deprivation. Even very small advantages compound over time, so relying on behavioural/statistical analysis will only invite more subtle cheating over time.

 

[EndogenousRebel]

Chess.com elo ratings are what most people are familiar with. Mid 2000s player like the alleged cheater vs capping 3000 players like Magnus rarely ever go to the towards the underdog.

In a real world elo measurements, which have lower sample sizes for starters, I'm not quite sure how those figures match up, so I agree that most don't know what the hell is going on. But there are plenty of pros and organizations that have analyzed the situation and agree that Niemans victories should be scrutinized.

As for the legal grounds the dude has to sue Carlson and others, does he really have any? Did he ever directly call him a cheater? I don't recall that ever happening at any point. This is mostly the media running away with a story after a recent surge in popularity in the game. I'm not sure Carlson can be held liable for that.

 

[dr froyd]

OP, your idea that there's no point in cheating is pretty absurd. Even from an engine point of view, chess isn't "solved". Nobody knows what the best 1st move is, because it's not computationally feasible to compute all variations. That's why you have various engines that rely on different strategies to find game-theoretically "good" moves. E.g. alpha zero that uses statistical learning to discard bad variations at a shallow depth of the tree so that it reduces no. of variations to compute.

many games go into well-known variations for 20 first moves, but that doesn't mean a player can choose alternative moves that takes them outside theoretical positions quickly - as magnus carlsen often does. Either way, the middle- and endgame will nevertheless remain and there's obviously a huge advantage in using an engine to cheat with.

when it comes to Niemann, i saw a statistical analysis of his past over-the-board games, and there were instances where he had 100% match against engine moves for entire games. Some of these games were 40+ moves, which makes it astronomically improbable that he wasn't relying on an engine. But, i didn't do the statistical analysis myself, so i don't draw any clear conclusions from that. It does look suss though, and i somehow doubt Carlsen would make these kinds of accusations without good reason.

 

[Rook]

https://lichess.org/@/chessbeads

i rebranded lol

 

[scorpiomover]

For those who don't know there's been a big drama in the top level chess that I won't bother to describe. It includes over the board cheating accusations and lawsuits and it will take years to clear up probably.

As a side note cheating in live chess is quite easy.

Then why not either change the rules to make cheating quite hard, or change the rules so that cheating is legal?

I don't really care about who's right or wrong there, but I'm more interested in the principles. My point is that top level chess is kind of a solved game, no cheating is required.

Then why ban cheating at all?

My conclusion is that at this point it's a contest of who can memorize more variations and repeat them accurately to the point of scoring sufficient advantage to close the game with their inferior real-time calculation.

Then why is chess considered a game of strategy, when it's really a game of memory?

 

[Ex-User (9086)]

OP, your idea that there's no point in cheating is pretty absurd.

Your ability to misrepresent what I'm saying is absurd.

I said cheating isn't needed, not pointless. People can benefit financially by cheating in chess. My point is people with great 10/10 memory and 9/10 skill can beat chess grandmasters with 5/10 memory and 10/10 skill.

The perfect game of chess is a draw. Most symmetrical openings lead to a draw and they are also the top engine moves for white and black pieces. Most top level chess games end in a draw.

To win a chess game one has to create asymmetry which in itself is sub-optimal in terms of engine evaluation, sacrifice pawns or pieces, take risks, have better move memory, better skill and hope that the opponent makes a mistake or preferably a few mistakes.

 

[dr froyd]

@Glaensaeth you wrote "top level chess is kind of a solved game, no cheating is required". I have no clue what your new distinction is supposed to clarify.

how do you know the perfect game is a draw? Nobody knows what the optimal 1st move is.

also i don't think engines prefer symmetrical positions in general. It highly depends on the particular engine. They all differ in their valuation functions and optimization methods, which leads to different styles. For example alpha zero is famous for prioritizing long-term mate opportunities over material, which makes its play highly asymmetrical.

 

[Ex-User (9086)]

@Glaensaeth you wrote "top level chess is kind of a solved game, no cheating is required". I have no clue what your new distinction is supposed to clarify.

how do you know the perfect game is a draw? Nobody knows what the optimal 1st move is.

also i don't think engines prefer symmetrical positions in general. It highly depends on the particular engine. They all differ in their valuation functions and optimization methods, which leads to different styles. For example alpha zero is famous for prioritizing long-term mate opportunities over material, which makes its play highly asymmetrical.

Your post is so basic. Kinda proof that you are not an Intuitive, kek.

I use godlike intuition granted to me as an iNtp, observe that no first move (as evaluated by every chess engine) by white improves the position beyond 0.5 which is half a pawn, and follow opinions of top chess grandmasters who say that a perfect game is a draw.

The game cannot be won without one or two pawn advantage or a major mistake on one side. The game is mostly solved, people who play the game casually find novelty in doing things that top players already know, keep arguing otherwise.

Chess search and move space is big, but the actual optimal move search space is quite small and repetitive.

I'm guessing all that just doesn't register for you. So truly sorry about it.

 

[dr froyd]

@Glaensaeth yeah, wow.. that's so genius it doesn't register with me lol. You're talking to someone who has actually researched and programmed game-theoretical solvers (even for cases with bigger search spaces than chess) – i should know how these things work.

you painted yourself into a corner of your own ignorance as per usual, so i already know how this will turn out