It's very remarkable that Carlsen is suspicious of Niemann's game against him in Sinquefield, since there's a clear consensus among other top chess players that there was absolutely nothing unusual about it (or at least about the moves played).
You missed a large part. Some of his moves were "somewhat suspect". However, he was interviewed after the game with Magnus and he really could not explain why he was making the moves he made. Even the interviewers were almost laughing as he gave his "analysis" for his own moves. He played off his top engine moves as just getting lucky, while at the same time stating he didn't make other moves because they would have weakened his position (when in fact it was the other way around), while also stating he made other moves to strengthen his position (when in fact it was weakening).
Nothing he said made sense. He is playing against the top players in the entire world, and he can't really describe his games. This is super genius territory, and yet he just claims his skills to mostly just be based on luck.
I agree, but as I tried to state, there is simply more to it. Giving bad interviews doesn't mean anything. Accidentally beating the world champion doesn't mean anything. Spending all night studying a rarely played chess line that just happens to be the exact line played the next day doesn't really mean anything either. Not really being able to analyze like a GM doesn't mean anything too.
But when you have all these factors happening during one game, statistically it is not probable.
And that one just happens to be a kid who was caught cheating online twice (unlike any of his other opponents) and was an unremarkable player until the age of 17 but has since attained 2700+ level (unlike any of the current young 2700+ players who all reached GM level before the age of 15).
You should see the sequence. He was totally unable to explain any lines he had in mind, stating some positions were « obviously winning » (where it was absolutely not obvious, and in fact the engine marked it as loosing), etc. A total disaster.
If you think about it: Magnus, is Magnus. He has an aura about him. People make blunders playing against him they wouldn't against others. This is known. Magnus is ALSO very good. But that "aura"... doesn't hurt him.
If for whatever reason, Hans saw far enough ahead, to not be worried... and Magnus hadn't, what does that say about Magnus?
He mentioned Hans wasn't nervous, in comparison to Magnus he had nothing to lose.
I won't defend his prior cheating. I will say: Prove it Magnus.
---
I'll draw a parallel to a game I have played at the national / international level. Bridge.
Bridge has had a TON of cheating scandals. People knew something was fishy. But they took the time, watched the videos, and figured out what happened.
Bridge cheating is a bit different, because you can find the smoking gun from video reviews, etc.
I wonder if "bridge supercomputers" as a cheating method has been tried. I assume the percentages on finesses working, etc, are easy enough for the experts to learn that they're not very worthwhile.
Well, remember, you don't have full information. Especially in an "all pass" auction from opponents.
Interestingly, the computers would to MUCH better on defense. Because as you bid, you speak about the distribution of your hand, and your partner does the same about theirs. (Even in negative inferences.)
And trust me: Good opponents will use that information, already.
So far, bridge has found the smoking guns because honestly: The cheaters have sucked at cheating.
If they bothered to actually encrypt their signals at all, they would have been suspected, but not caught.
---
To answer the question: Even today. Good players will know the answer to when to take which finesses. Where good, is probably around Life Master and a bit under.
When Robert Fischer achieved an unprecedented scores in the pretenders matches prior to winning the champion title, I believe that the biggest factor was not the quality of the play of Fischer, but the way he mentally unbalanced and broke his opponents, who absolutely did not give a performance they were capable of.
The unnerving, unbalancing effect Fischer had in the 1971 Candidate's tournament was absolutely due to his play. He put up immense resistance against lines prepared by teams of Soviet GMs (example here https://www.youtube.com/watch?v=X8-90cPf61M), which was the sort of thing that broke their nerve. He was not intimidating or manipulating his opponents outside of the board; if there were any outside-chess forces at play, it was the pressure from Soviet higher-ups embarrassed to lose to a lone American. Also, I don't recall that his opponents played worse than they would have been expected to against an opponent so much stronger than them — maybe after their fifth loss in a six game match, they'd become demoralized.
> He was not intimidating or manipulating his opponents outside of the board
They absolutely were blundering much more than they did vs other opponents, showing a far lower quality off play than usual. Also I believe you didn't read much about his championship match. Remember how the games were moved to a different room, and why?
Arjun blundered against Magnus far more than usual last weekend. That's what stronger players tend to do to you.
I thought you were talking about the Candidate's tournament. As for the championship match, it's not generally believed that Spassky severely underperformed anyway. He won several games and put up good fights in the draws and losses.
>I believe you didn't read much about his championship match
I am quite familiar with the surrounding circumstances, including the pre-match negotiations. I do not see how any of this "mentally unbalanced" or "broke" Spassky, unlike (arguably) Larsen and Taimonov. Furthermore, he's generally been commended for his behavior as a consummate professional and a "gentleman" in the circumstances.
I wonder if you are as familiar with chess (the game, professional play, and its history) as you lead on to be, posting all over this thread.
Well, I only visited the chess school for maybe 6 or 7 years in my childhood, and my peak lichess rapid rating was slightly below 2600, so I believe I'm definitely not the best expert on these matters.
In this case it seems most of the unusual behavior occurred after the match, where (I've seen claimed) Niemann gave obviously wrong reasons why he had played so well on such an uncommon line, and was also not able to explain why he made particular choices he did playing it.
Not only did he fail to explain his reasoning in any satisfactory way, but the suggestions he gave as responses to alternative lines from his opponent were outright losing which showed that he had a poor grasp of the position. This is extremely suspicious behaviour from a player who had just defeated the world champion while using the black pieces.
The consensus among the top GMs was that Hans’s postgame analysis was way below the level you would expect from a player of his rating, never mind a player near Carlsen’s rating (which is much higher)!
No it can't and it is damning evidence. When you wipe the world champion off the board as black (which he did), you need to be able to show you understand how the game progressed in a post-game analysis. Neimann's understanding of his own remarkable performance was seriously deficient.
Being unable to explain your own work is how a lot of academic cheating is confirmed. It doesn't matter if they don't know who you copied from, if you can't understand your own work right after you supposedly made it, they will fail you.
It's like when you get the correct answer on a math test, but then when the math teacher asks you to show your work you bumble around and can't reproduce the thought process required to arrive at the correct answer.
No, it’s more like someone come to see you after a four hours exam while you are tired and want to go home and ask you how you did the very tricky and somewhat open question 4 while suggesting different approaches and asking for your opinion.
The quality of players post-game interview varies widely. Some clearly don’t put much thoughts into them because they would rather go home. Niemann is in good company here. That was nothing particularly exceptional.
If you aced question 4, and most of the rest of the test also, yes, it would be easy to explain. Especially right after it happened, even if you're tired. If an answer is obviously correct to you, you should be able to explain how.
And if you don't want to and just would like to go home, fair enough, but then you really shouldn't give an explanation which is wrong!
It's a pattern with Niemann too, he's infamous for saying only "The chess speaks for itself" after beating Carlsen with black once before.
I think the parallel with academic cheating is accurate.
> If you aced question 4, and most of the rest of the test also, yes, it would be easy to explain. Especially right after it happened, even if you're tired. If an answer is obviously correct to you, you should be able to explain how.
You clearly have never been through a four hours math exam with open questions.
There is no obviously correct answer. There is the way you tackled the problem and the myriad of other ways you could have done which might be more or less obvious, easy or correct.
It’s the same with chess. There is the line you played, the line your opponent played, the other lines you could have played which you did or didn’t consider, same with your opponent. Some of them you considered seriously, other you didn’t. Plus all the things you missed but didn’t matter because your opponent didn’t go there.
Also, you seem to believe chess players are doing post-game interviews because they want to. It’s not the case. It’s a mandatory part of participating in the tournament. Most of them would decline them if they could.
And yes Niemann is infamous for hating post-game interviews and always giving poor answers which is why I’m surprised people actually base their argument on this.
I have, in fact, sat through many, many all days math exams. Not with open ended questions exactly as that wasn't the fashion in math at the time, but in other topics, sure. And yes, I was prepared to explain what I answered and why.
It's an exaggeration to say everyone hates post game discussion. Magnus used to dislike it somewhat, possibly - it was hilarious how Norwegian newspapers tried to turn him into a celebrity in those days because hey! Chess superstar! And it all fell flat because he was so unbelievably boring at the time, lol. He got more social confidence as he got older (and the media got better too, getting people who actually had a clue about chess to talk to him).
But these days, the young GMs are on twitch for heaven's sake. From a social and media standpoint, Niemann is perfectly competent, a lot more so than Magnus was at his age. It's explaining his play he avoids. So yes, it's suspicious. I'm equally surprised at why you would think people would just overlook this.
We can be reasonably sure that Fischer has not been cheating above and beyond unusual conduct. There hasn't been a way to meaningfully cheat. In our times however there are computers which can materially help, and there are technologies allowing someone to receive it, use of which is very difficult to detect without an unacceptably invasive search
Which are totally subjective. And they dont really make sense. He thought he was too relaxed? How would cheating OTB against Carlsen be relaxing? Carlsen is being really unprofessional here, even if he turns out to be correct. But the window has closed - we will never have evidence that he was cheating at this tourney.
And he wasn’t suspicious about his game when he beat Hans a few weeks earlier itself.
If Hans did go to all those extremes to cheat OTB it’s really surprising he would do so while playing black against Magnus Carlsen in an otherwise kind of pointless game.
It's a bit misleading to say that there was absolutely nothing unusual about the game. The Sinquefield game in question showed a very high correlation between engine optimal moves and the moves played by Niemann. His gameplay accuracy here is within the bounds of what the very top players can achieve in individual games, but high enough to raise some suspicion.
The next step is to place that mild suspicion in the context of both his history of admitted cheating, his unwillingness/inability to explain his remarkable moves post-game, and the additional context of many other games played in the last few years with _extraordinary_ accuracy. Now something that could be explained by just a very strong game appears very suspicious.
This is not true at all. For instance Nepominatchi commented on game saying Niemann's play was "more than impressive"[1]. Not commenting on the situation per se, merely your "clear consensus among other top chess players" comment.
You are hung up on the wrong thing. Parent writes: "[clear consensus that] there was absolutely nothing unusual about it", I quote one of the greats saying the game is "more than impressive", which, at least we should agree, is the exact opposite of claiming there is "nothing unusual about this game". Onus is on parent to come up with a list of grandmasters claiming the game was "usual", at least I provided one refutation but there are many.
Correct. They will not come up with a list of GMs saying Hans cheated. Many GMs have said that it doesn't seem like he did (at least during the game with Magnus).
Alireza, Ian, Magnus, Fabi, Wesley, and Levon have since made statements that imply they believe Hans cheated, or at least that they were suspicious of his play, as well as Yasser and Hikaru.
That's the majority of the players in the Sinquefield cup. Even Levon, who was initially skeptical, has since reversed his position.
As is tradition in chess, no one says "He cheated" they say things like "his moves were better than one would have expected" or "superhuman" or "I felt like I should trust my opponent over my calculation".
I watched them on stream. I don't do anything chess related on Reddit. You should perhaps actually engage with criticism and disagreement if you want to post here.
Some of them seem small enough that they won't trigger a metal detector. Currently they don't constantly scan the playing hall for wireless activity, which is what you'd need to detect this in use. I bet they start scanning for wireless transmissions soon, though.
I don’t think the “vibrating anal beads” theory is important because people actually expect it to be true in this instance. It is more about the chrisis if a mind sport where computers defeated humans soundly and througly.
The simple fact is that computers vastly outcompete human chess players. And not just big and expensive purpose built machines but the kind of computers everyone has access to.
Furthermore at the skill levels these players are you don’t even need constant handholding from a computer. A few hints at key moments would be enough to basically shift the balance in someones favour.
So if someone wants to cheat all they have to do is to receive a few bits of information from an accomplice. The question is not even if someone cheated in that particular game, but if cheating is possible.
We can imagine all kind of spy gadgetry one could use to communicate those few bits. People have two hangups with many of them: they can be found in a security screening, or they sound too sci-fy.
The vibrating anal beads combine three properties:
- they could transfer the few bits of information needed to tilt the game in favour of a cheat.
- they are not too far fetched. You can buy them right now commercially.
- they would be very hard to detect by security arrangements. It feels very unlikely that players would agree to the kind of invasive probing which would be necessary to detect one.
So it is not that people think that this particular player in this particular game actually used vibrating anal beads. It is more about the idea that someone could cheat at chess with covert communication methods.
The key takeaway is that if you have someone assisting you (entering the information into the computer) they only need a very simple way of sending a signal - which could be a "do something unexpected" or "this move is crucial". And you'd only need a time or two in a game to get the edge, assuming you're already skilled at the game.
It started out as a silly, obviously joking comment in the twitch chat of a GM chess streamer (Eric Hansen). Then he jokingly overreacted to it, and explained how it's possible, as an obvious joke. Then Elon tweeted about it as an obvious joke. But news websites don't care to make that distinction, and write about it.
2. Chess is full of VERY smart people. One of the most common ways to insult a smart person is to call into question their sexuality; hence why we have to have entire movements related to calling out anti-lgbtq+ statements like "that's so gay". [https://welcomingschools.org/resources/stop-thats-so-gay-ant...]
Anyway, combine those two things, and you get your answer. It's because the world hasn't really evolved at all in the last 30-40 years, outside of what we have been forced to do by law. It's easy and socially acceptable to call a man gay as an insult, so in a roundabout way, that's what we're getting with the anal beads talk.
I sort of laser focused on this last week when I heard this theory for the first time. It just struck me as so. . . odd. Why would that be a thing? That's what I came up with.
The contrast of the paranoid grievance you're responding to, with a sibling comment which does an excellent job of explaining "why anal beads", is remarkable.
Some other top chess players voiced strong suspicions regarding his post-match interview (including the interviewer), because he could not suggest basic lines in the post-match analysis, lines that even the interviewer could find without an engine (and for example Hikaru while watching the interview, being able to say the line instantly).