Tracking a player's progress

by Frederic Friedel
9/30/2022 – A few days ago Europe's biggest news portal, SPIEGEL Online, had YACS – yet another cheating story. It concentrated on a portrait of the accused Hans Niemann, and traced his chess development over the years. Did you know you can easily track the Elo progress in the ChessBase Players service? And also his best games, his best and his most difficult opponents, his favourite openings and most/least successful openings. If not, take a look at what you have been missing.

ChessBase 17 - Mega package ChessBase 17 - Mega package

ChessBase is a personal, stand-alone chess database that has become the standard throughout the world. Everyone uses ChessBase, from the World Champion to the amateur next door. It is the program of choice for anyone who loves the game and wants to know more about it.

More...

It features daily, in the mainstream media all over the world: stories about the cheating accusations raised by World Champion Magnus Carlsen against young upstart Hans Niemann. Europe's largest and most influential news magazine and portal has done multiple reports – the latest features a portrait of the accused. It is in German and behind a paywall. But we bring you excerpts, with kind permission. 

The career of Hans Moke Niemann 

But is his rise from the world's 1000 to the top 50 really that special? To cut a long story short: It is usually unusual. To understand that, you have to look at his story.

Hans Moke Niemann was born in San Francisco in 2003. His parents, the mother from Denmark, the father from Hawaii, worked as IT managers. For professional reason they the family emigrated to the Netherlands. Hans learned chess at the age of eight, comparatively late when compared to other grandmasters. He was then quickly discouraged when his chess teacher did not allow him to represent his school in the national championships, Niemann later wrote in an article for the magazine "Chess Life". He preferred to devote himself to cycling, racing dozens of times.

After the family returned to the USA, Hans, now nine years old, returned to chess. "I played a little over 100 tournaments in my first year back in the USA," he later recalled. He shot up the rankings, thanks in part to his mentor Maxim Dlugy, a Russian-born US grandmaster. "He was a traditional Soviet-style chess coach. He showered me with a lot of information and I soaked it up like a sponge," Niemann said in the "Chess Life" podcast. With him he raised his game to a higher level as a teenager, up to 2400 points in the world rankings. 

After a quick rise, Niemann's performance stagnated. On the one hand because of a lack of training, he says: his "curriculum" at times consisted of an American football game on the iPad, Netflix and a 15-minute look at chess books by Garri Kasparov. And on the other hand, the boring lessons at school had kept him away from chess. The turning point came for Niemann at the age of 16. He moved alone to New York City, went there thanks to a chess scholarship to a public school and earned money as a chess coach and Twitch streamer. But Niemann gave that up to pursue his dream: He wanted to become a professional chess player.

To achieve his goal, he rearranged his life. Getting up around 7 a.m., going swimming, then ten to twelve hours of chess training, eating on the side. He hardly had a social life outside of chess anymore, he said. Niemann learned and played a lot - and came to success with a courageous style. In October 2020, he was in the haze of the best 1000 players in the world with 2465 Elo points; today he is a grandmaster and is 49th in the world rankings with 2688 Elo points.

The rise

It is this rise of about 200 Elo points that some find suspicious. Yet it is not so extraordinary, according to former world champion Viswanathan Anand, among others. "Since there are a lot of promising young players who make dramatic leaps, he didn't particularly stand out for me," he told SPIEGEL.

Data surveys show that other grandmasters before Niemann made leaps from 2475 to 2675 Elo points even faster, such as super-grandmaster Anish Giri – or Magnus Carlsen himself. Niemann's performance is extraordinary compared to good chess players, but plausible in comparison with the top players.

Players.chessbase.com

We generated the curves above (click to enlarge) on the ChessBase Players Encyclopedia. There you can visualize the progress of any player in the FIDE list. 

 

So what does the rating progress of Hans Niemann look like, and how does it compare to other top talents in the current chess world? (Click ll images to enlarge)

Note that Hans Niemann and Arjun Erigaisi are both of the same age. Arjun is on an even steeper meteoric path and is now the second-highest ranked Indian grandmaster – and number eighteen in the world.

If you visit the ChessBase Players page, you will see that it has many additional functions. Let me show you some, using a dear friend from Holland as an example.

There are 20 pictures of Anish you can flip through. Here's one of the earliest, taken in 2010, when he was 15.

Moving your cursor over the progress curve will display the exact date when his ratings were achieved. For instance in April 2019 he reached 2797.

Now comes the interesting part. When we showed it to a top-ten grandmaster a few days ago his reaction was: "Hey, that is fantastically useful. I didn't know this function!" Which is why we are telling our readers about it here.

When you scroll down this is the first section you see:

It shows you how Anish has scored with white, black and in total. It also displays his favourite and least favourite opponent, and his most dramatic wins and losses. Scroll down further and you get the following:

"Oooh," said our top ten GM, "what happens when I click a link?"

You get a list of Giri games in the relevant opening, e.g. the C54 Gioco Piano, in which he scored - 46.5/76 = 61.18% against opposition rated an average 2714. You can load games from the list, and even start an engine to analyse. "Oooh", said our top player, whom I expect to climb even higher on the rating list now that he knows this feature. 

If you, too, want to make use of this utility, you can see all the functions described in the report: Online Player Encyclopedia - Players, profiles, and pictures (with video!)

 


Editor-in-Chief emeritus of the ChessBase News page. Studied Philosophy and Linguistics at the University of Hamburg and Oxford, graduating with a thesis on speech act theory and moral language. He started a university career but switched to science journalism, producing documentaries for German TV. In 1986 he co-founded ChessBase.
Discussion and Feedback Submit your feedback to the editors


Discuss

Rules for reader comments

 
 

Not registered yet? Register

lajosarpad lajosarpad 10/12/2022 02:31
@mc1483

Another point to consider is that there is a force of good that tries to prevent cheating, namely, the organizers of the tournament. If we say that someone was able to cheat at the Sinquefield tournament, for example, then we do not only accuse that person, but the organizers as well. If Niemann was indeed a cheater (a possibility I cannot exclude, of course) and would grab any opportunity he has to cheat, that does not automatically mean that he cheated against Carlsen, because it is quite possible that he did not even have the opportunity.
lajosarpad lajosarpad 10/12/2022 02:26
@mc1483

I'm confident-enough about the validity of my position to render the task of convincing others to my arguments, specifically the fact that any sequence of n moves has roughly (not absolutely exactly, I can explain why if needed) the same probability as the other concrete variations (I can explain this as well if needed) of the same length. We consider correlation with engine moves to be special and rightly so, because correlating to this specific pattern is suspicious and may motivate us to expect an investigation. But we cannot exclude the possibility of "correlation by chance", because whatever move sequence you play, that specific pattern is (roughly) equally unlikely as correlating to engine suggestions, so we would also have to exclude the possibility of the existence of the game you have just played due to the prior improbableness of it happening exactly as it happened, which disproves the assumption that the unlikely is impossible via reductio ad absurdum.

So, if we are consistent, then we recognize that low probability events happen all the time and we have no reason to exclude the possibility that a correlation to a pattern happened by chance, even if that pattern is viewed as something special/suspicious. If we allow someone to be punished based on such statistics, then that opens the door for a plethora of abuses. So I'm really really confident that I'm right on this one, but not because I happen to have some knowledge and know-how, because, to be honest, I catched myself being wrong even in the area of my expertise several times. But, on this one I do not see how would it turn out that excluding the unprobable special scenario is a good idea, as there are equally unprobable not special scenarios and we accept that they happened.
lajosarpad lajosarpad 10/12/2022 02:26
@mc1483 with all due respect, if I understand your current point, then what I have written is correct, but ignorant in the sense that there are much more to it. I think you are probably right in saying so, I have never studied the Italian mafia in-depth and, if the information I give about them seems to be superficial, the obvious reason is that it's probably is superficial. I'm quite interested to know more about the Italian mafia (I'm a curious person in general), but it is off-topic indeed here. It initially served the purpose to illustrate a point, according to which one may know something without a proof.

That's true. You do not need proof to know that you are in love, for instance. But in our discussion we are tackling the question of what we may consider as established knowledge. An institution can only condemn someone if it establishes that the person is the perpetrator of a certain kind. The institution needs to be very convincing when establishing something as a fact, especially if the act of considering something to be an established fact causes considerable damage to one or more individuals.

This is why we need proof before we establish that X has committed cheating in his game with Y. (or, more broadly, that X has cheated in a given period of time).

Some people here consider statistical evidence of correlation between a player's move and engine suggestions to be enough to condemn someone. I have researched conditional functional dependencies with a colleague a few years ago and written a program that automatically finds such patterns in any database and any data, we have also written a paper about this, that is also dealing with premature assumption and correlation biases. One may consider me to be an authority in this topic, but I kindly ask you not to accept my points just because I happen to be trained in this topic.
mc1483 mc1483 10/12/2022 11:14
@lajosarpad: yes, you misunderstood the whole thing. It's not your fault (nor it is mine), also what you have read after the "quick search" is correct but just covers the most blatant aspects of italian mafia. If I told you "The USA are a confederation of states", that would be correct, but you would reply "confederation is just one thing about the USA, there's a lot more to tell".
If you wish me to explain the whole thing, that I can do, but being long and totally OT, I think Frederic should approve it first (that's why I haven't done it until now). To "know and check", there are some books, but they are very specialistic and don't think they have been translated in other languages (just Sciascia's work, I think, but he was not a crime historian, just a clever storyteller).
lajosarpad lajosarpad 10/12/2022 01:09
@mc1483 You mix up the trust in authorities with the appeal to authorities at debates. The former is the optimistic attitude of assuming the best in case of authorities and it is a usual, even expected attitude of law-abiding citizen. The latter is the avoidance of arguing for one's point and, instead, pointing to an authority. I did not claim that I would know more than you about the Italian mafia and I did not distrust you. I merely refuse to be silenced for not being an Italian.

My understanding about the Italian mafia was that it's a body of organized crime, which claimed to serve a cause (Cosa Nostra, or Our Cause). According to my understanding they have a strong sense of honor and they view the group as a family.

A quick search yields "The Sicilian Mafia [...] is an Italian Mafia-terrorist-type[3] organized crime syndicate and criminal society originating in the region of Sicily and dating to at least the 19th century." and further down it claims they committed acts of terror.

When they killed someone - an act you yourself claimed they did -, then I really doubt that anyone likeminded to the killed one would not have a sense of fear and it is also difficult to imagine that the mafia did not know about this.

But, it is quite possible that I misunderstand the whole thing. However, instead of explaining why I am wrong, what the reality is instead, and, most importantly, how I can know it and how I can check this, you claimed that I cannot know it, you know it. I do not find this convincing. Not because I would not trust you, but, because the way of putting these arguments diametrically opposes many important epistemological principles.

https://yourlogicalfallacyis.com/appeal-to-authority
mc1483 mc1483 10/11/2022 12:55
@lajosarpad: if you like it or not, we trust some autorithy 99% of the times, because we have to - even if the autority we trust is indeed wrong every now and then. I'm eating some food, and I'm trusting the authority in charge of checking it's healthy (but some peope die of food poisoning). I'm driving a VW car, and I'm trustung VW the car won't suddendly have a major damage and let me be killed (but that happens). I got a diploma and trusted what I read in the books and what my teacher told me, otherwise I would not have got it (but books and teacher may be wrong, and sometimes indeed are).
So, we usually trust the authority, unless we have reason not to (such reasons may vary). Hungarian crime, there's no reason to think you're not trusty, and even if you weren't my life won't change for that. Chess.com, we know of their ties with Carlsen, also their timings are suspect, so there is reason not to trust them. That's why, about italian mafia, I expect you to trust me. You don't, your way to see things is too far from mine. End of discussion.
(not to mention that for us in Italy mafia is a very sensitive argument, but that I could discount)
lajosarpad lajosarpad 10/11/2022 11:20
@mc1483

"if this is the way you like to discuss, I simply won't discuss anymore."

It is up to you to decide whether you want to go on with this discussion (or any discussion) with me. If you consider me an unworty conversation party (I have reason to think so given your last comment), then I find that unfortunate, but I will simply accept and respect that you see it that way. Nevertheless, I wish you all the best and, if you decide in the future that we may have a conversation, even a debate, then I would not object. No hard feelings on my part :)
lajosarpad lajosarpad 10/11/2022 11:08
@mc1483 I think historians' work is pretty reliable. They can be wrong, of course, as human beings, but they are doing science and dig up (sometimes quite literally) interesting facts.

I surely did not mean disrespect and I am quite open to what you say and believe you, as I do not have any reason not to trust you. So, if you had a feeling that I do not believe you out of disrespect, then I would like to let you know that I deeply respect you, even though I disagree with you on several points. Like the following one:

"That's right: whatever you may tell me about hungarian crime, I will trust you."

This is the same kind of appeal to authority, but with reversed roles. Now I'm the authority, because I happen to be a Hungarian and you, a non-Hungarian just "cannot know". But, actually, your possible argument about Hungarian crime might be spot on even if you are not a Hungarian or you do not know anything about Hungarians. Invertly, a real authority of something can be absolutely wrong (think about the example of mistaken historians, that you have brought up, as an example of authorities being wrong).

As a matter of fact, if you would just say something about the mafia (or Hungarian crime, or anything), I would read it with interest and willingness to learn. But saying that people not experiencing the mafia cannot know them is not a good argument and that's not changed by the fact that I (fortunately) was not alive in the 60's yet and (unfortunately) did not have the chance to see your region. I certainly do not consider myself an expert of the Italian mafia in the sixties and I embrace any opportunity to learn, but when someone says "you cannot know it", then my answer is "no, thank you".
mc1483 mc1483 10/10/2022 11:50
@lajosarpad: "Then historians cannot understand the distant past because they didn't live back then"
That's right. Most of them cannot just because of that. From time to time archeological discoveries disprove what they believe to know.
"Mind you that the mafia is not a unique Italian thing. There is a Russian mafia, a Ukrainian mafia and so on"
You are wrong again. _Italian_ mafia _is_ a unique thing (especially it was until 30-40 years ago, now it's less unique). Other criminal organizations are called "mafia" because of some similarities and because the word "mafia" has become universal of a sort, and it's by now synonymous of "organized crime".
"In my region there are several groups of organized crime and you had no factual basis to assume that the phenomenon you speak about does not occur elsewhere from your region"
That's right: whatever you may tell me about hungarian crime, I will trust you. But you don't trust me when I tell you italian mafia is not what you think, no terror, no bullying and so on (at least until 30-40 years ago): if this is the way you like to discuss, I simply won't discuss anymore. Have a good week and "live long and prosper".
lajosarpad lajosarpad 10/10/2022 11:26
@mc1483

How is it so, that an extremely small p -> 0 probability is assumed to be impossible if it correlates with the sequence the engine suggests, but the same p -> 0 probability of another concrete sequence of moves is "normal" if it does not correlate with the engine?

There are two possibilities:

- either we exclude the possibility that a low-probability event occurs, in which case we will have to conclude that no chess game was played ever (despite the factual proof of the existence of many thousands of games) for the already played games had an extremely low probability of having the exact moves they ended up having
- or we accept that it's possible and then low probability does not amount to certainty and cannot be used as a proof to condemn someone
lajosarpad lajosarpad 10/10/2022 11:25
@mc1483

"And even if a non-zero probability still remains, we have to live with that."

So, would you accept your imprisonment or punishment for a crime you didn't commit based on the statistics of someone that underlines the high likelihood of you being the prepetrator of a crime? Would you accept that it is fair to stay in prison for 20 years or so for a crime that you did not commit, because you are suspicious based on some statistics?

In the case of a murder you have a body, which proves that the crime happened. In a cheating allegation of the kind we are discussing even the crime's very existence is not proven yet.

DNA sample is strongly suggests your involvement into a crime that provably happened. Comparison of your move with the engine is very different, the sample is non-biological, non-personal, but analytical. It is very much akin to the case when we see on the video footage Niemann pressing a button, in which case we would accept him being guilty of cheating and we would not start to argue that he did something else. That would be an unmistakable proof beyond reasonable doubt.

If you play a sequence of n precise moves, while the average number of viable moves is k, then you have a probability of 1 / k^n of making that sequence. But making any sequence of moves (correct or not) has the same probability as long as it has the same length. Yet, we see the correlation with engine moves to be a special case in comparison to other unprobable sequences. This is understandable so far. But why do we assume the execution of such a sequence to be impossible based on low probability if it coincides with the engine's choice, but we accept it as a normal sequence of moves if no such correlation occurs, even though the specific move sequence in the "normal" case has the same probability as the one that correlates with the engine.
lajosarpad lajosarpad 10/10/2022 11:12
@mc1483

"If you have not lived here for many years it's almost impossibile to understand its true ways."

Then historians cannot understand the distant past because they didn't live back then.

"But that's not important"

Then why did you open this off-topic in the first place and compare an underage girl to a notorious body of organized crime? The only reason I even responded to this was to be respectful, especially since you have made many valid points during our conversations, but I did not prefer to speak about the mafia when the topic is cheating in chess.

"The comparison I did was not meant to compare mafia crimes with chess cheating, but to let you know that sometimes there are things known for sure despite absence of proof."

You mix up the subjective feel of knowing something with the way legal bodies are to operate. I would not debate that you may know things without a proof, but institutions should not operate based on subjective feelings, but on objective facts, especially when they wreck the livelihood of an individual. Let me provide you an example. Let's consider the example of a person who visits 64 times a town and each time he visits the town, somebody is murdered. Is this suspicious? Yes, of course. Can we compute a very very small probability of this occurring without the person being the killer? Yes. Should we imprison this person based on this? No. Why? Because there is a chance that this was just a coincidence. So, if you are innocent, then you are guaranteed to have your individual rights being respected and you will not be persecuted based on statistical computations. Of course, the person in this example would be a suspect and the police would track all his movements to catch him if he is guilty. But institutions would not treat him as a guilty person until a proof is found.
lajosarpad lajosarpad 10/10/2022 11:08
@mc1483

I choose to not prematurely believe in any of these scenarios in this case, but I prefer institutions to refrain from treating him as a guilty person before his guilt is proven. Tournament organizers are of course free to choose who they invite.

"Also, you got the mafia wrong."

Another premature assumption. Mind you that the mafia is not a unique Italian thing. There is a Russian mafia, a Ukrainian mafia and so on, there are many copycat organizations and you took the liberty of making assumptions of what I may know, which is of course a fallacy, because, if I'm wrong, for whatever reason, then you can validly point out where and how I am wrong, instead of arguing why I cannot be right, which is an invalid approach to debates. This is precisely the mistaken judgement I especially would like to avoid when it comes to accusations of people. In my region there are several groups of organized crime and you had no factual basis to assume that the phenomenon you speak about does not occur elsewhere from your region. So, your way of arguing against the validity of any arguments I would make about the mafia based on what you perceived as my inability to understand it was wrong independently of whether I am right or wrong about the mafia.

I would not engage into arguing your inability to understand something properly, as I recognize that as an invalid argument which has the aim of excluding you from a conversation or a part of a conversation.
lajosarpad lajosarpad 10/10/2022 11:08
@mc1483

"But nobody in the world, except for you, would believe that, and I would be banned from competitions even if my method would not have been uncovered, even if I protested my innocence and cried all day."

Let me clarify: I do not believe anything in this context. It is the accusers who share a belief and want to convince others. I do not believe in the innocence or guilt of the accused by default. As a matter of fact, I could find the person very very suspicious, but I would still argue to not condemn him/her until a proof is found. The reason as of why I am arguing for the principle of "innocent until proven guilty" is that this is a way that prevents us from unjustly persecute innocent people. So let's separate the legal assumption of "innocent until proven guilty" from our own subjective opinions. For instance, in Niemann's case I do not believe that he cheated against Carlsen, but I do not believe that he did not cheat against Carlsen either.
lajosarpad lajosarpad 10/10/2022 10:52
@Jacob Woge

In correspondence chess, especially since the players are given plenty of time, it is of no wonder they will end up playing very precisely and it is difficult to distinguish a cheater from a very accurate player. In online tournaments, especially with quick time controls such a proof is impossible in practice if the person does not confess. So, I would not take online tournaments very seriously and I think FIDE should refrain from organizing online tournaments, leaving this to entities such as chess.com. Also, in case of online tournaments, since anti-cheating measures are difficult to introduce (but it's still possible, with webcams, screencasts, etc.) the entities which organize such tournaments are free to ban anyone they want for being suspicious. Notice that I did not debate chess.com's right to ban Niemann from its platform. When they had strong reasons to suspect his cheating and even he confessed, it was up to them to decide what measure they will take.

However, chess.com's recent ban, quickly following Carlsen's accusation is something I would not challenge in terms of whether chess.com has a right to ban him, yet, I find it very strange that even before they investigated the current scandal, they quickly ban Niemann.

"Without proof, there are two possible outcomes. Guilty, or not guilty. So, the odds must be fifty-fifty... "

The odds are of course very different from that and quite often. But I argue that we shouldn't condemn someone for a guilt we do not factually know that he/she committed. I'm all for condemning caught cheaters and to investigate suspicious cases. But I disagree with destroying people, if there is a non-zero chance of being wrong while we do so.
lajosarpad lajosarpad 10/10/2022 10:51
@Jacob Woge

"How far can credibility be stretched. May the OTB player be absolute beginner, and still IM strength on-line?"

Since without proof, judging only by the comparison to engine moves does not prove that the cheating actually happened, it just points out how likely it seems to be, I think that anti-cheating measures need to be ever improved in order to prevent/discourage cheating and suspicious cases need to be investigated by unbiased entities in a fair trial. Comparing engine moves with actual moves can strongly suggest, based on likelihood that such an investigation is needed, but in view of the (however unlikely, but existent) possibility that such a correlation can occur and in that case we would destroy an innocent person, I am convinced that unmistakable proof is needed before we condemn someone as a cheater.

In OTB chess it is much more difficult to cheat, there is video footage, there are potential witnesses (including the opponent and the arbiter), the cheating is manifesting in a very physical way (a device is somewhere to be found, or an accomplice), etc. So, anti-cheating measures need to evolve and do so quickly. But I view the possibility of baseless accusations to be a greater danger than the actual cheating.
mc1483 mc1483 10/10/2022 01:02
@lajosarpad: let's say I play some tournament, and _all_ the moves I play in every game are the best choices of some engine, i.e. Stockfish 10. And of course I win all games.
This is not proof I cheated, according to you, because - I quote your words - "there is a non-zero probability - however small it may be - that this might be happening by pure chance" (also, I may be a very strong player).
But nobody in the world, except for you, would believe that, and I would be banned from competitions even if my method would not have been uncovered, even if I protested my innocence and cried all day.

This is the situation regarding the little girl. Also, you got the mafia wrong. Movies got it wrong. If you have not lived here for many years it's almost impossibile to understand its true ways. This ways are even difficult to understand for people living in Italy, albeit outside Sicily. Any way, it's not a question of "frequent use of terror in order to bully the witnesses into submission" (this is the movie thing), it's something much more subtle and not easy to explain - nor this is the right place - that relies over NOT using terror or bullying witnesses. In fact, things started to change when mafia DID turn to terror for a series of reasons that, too, would be long to explain.

But that's not important. The comparison I did was not meant to compare mafia crimes with chess cheating, but to let you know that sometimes there are things known for sure despite absence of proof. And even if a non-zero probability still remains, we have to live with that. Courts find people guilty of some crime every day, and every time there is a non-zero probability (sometimes not even very small) they are wrong. Murderers won't confess, DNA may transfer from one person to another and so on. Still, murderers are found guilty every day. You may not like that and remind us of the "non-zero probability", but this is how the world works. And the girl indeed cheated- Niemann, I don't know.
Jacob woge Jacob woge 10/9/2022 11:36
“I understand that Niemann's play and behavior is very suspicious for you.”

Pertaining to the vs. Carlsen game of Sinquefield, it does not take much watching to realise Carlsen’s was an understatement.
Jacob woge Jacob woge 10/9/2022 11:18
There is a non-mob, more chess related example.

Back in the days, you would come across a correspondance chess (CC) player, rated 2400, i.e. close to IM strength. OTB, 1600. The CC player would beat the OTB version of himself 100 times out of 100, according to ELO.

No legal issues, a CC player could use whatever means he or she saw fit. But even if it had been forbidden - no proof.

Of course, CC chess is all but dead by now. There was a slow drift towards chess by email, then using a server - and this is perhaps the reason using a computer in on-line chess is not regarded as cheating to the same measure as using one in OTB chess. Didn’t used to be crime unless you are face-to-face. Of course, the clock, rather than a calendar, makes a difference.

The question is

How far can credibility be stretched. May the OTB player be absolute beginner, and still IM strength on-line?

Without proof, there are two possible outcomes. Guilty, or not guilty. So, the odds must be fifty-fifty...
lajosarpad lajosarpad 10/9/2022 10:32
@mc1483

"All this is no proof, but I really, really hope you won't say there are other explanations."

Actually there is at least another explanation, which is that there is a non-zero probability - however small it may be - that this might be happening by pure chance. Also, there might be other explanations that we may or may not think about. The explanation you believe to be correct will not be automatically validated even if we do not have any other explanation. If you say that something happened in a manner of X and I cannot provide a manner of Y as an alternative explanation, that does not prove that it must have happened in the X way.

It is very possible that the little girl you have spoken about was actually cheating. But treating an unproven accusation as factually true is premature in my opinion.
lajosarpad lajosarpad 10/9/2022 10:32
@mc1483

A non-tyranical approach to an accusation is to take into account the accusation and assume the innocence of the accused until (s)he is proven guilty. Instead, in the case of the little girl you assume her guilt because, according to you, she had a tendency to make moves that the engine agreed with. But this is not a way free societies operate. All the arguments that you have put together against the little girl are at best convincing that there is a reason to be suspicious about her activity and an investigation is needed. But you go beyond that and without actual proof you assume her guilt. https://www.law.cornell.edu/wex/presumption_of_innocence

This is very much akin to the law trials in Ceausescu's Romania that already assumed the guilt of the accused prior the presentation of proof and the accused was named "the enemy of the people" and imprisoned or executed.

If someone is accused of an unproven or unprovable guilt, then the accuser failed to prove his/her allegations and there is reason to consider the allegation slanderous. If the little girl happens to be innocent, then persecuting her is a really problematic thing to do, especially given the fact that she is/was of minor age.
lajosarpad lajosarpad 10/9/2022 10:31
@mc1483 I thought my earlier comment was clear.

"No, I as wrote before I think that a single move does not tell anything, and is evidence of nothing."

I said that 29... Nc4 is a strong argument against the cheating allegation and anyone aiming to analyze the question of Carlsen's allegation on that specific game should address this move. I did not claim that it is evidence of anything, so your answer was addressing something else than what I said.

"When the mafia killed someone - especially in the '60 and in the '70 - everybody knew who and why did it, but there was never proof."

I do not know the case of the girl you have mentioned. But I find it - with all due respect - awkward that you treat her as a guilty person before any proof is provided. If there was proof against her, then of course she is guilty. In the case of the mafia, which you compared this little girl with (which is not a very fortunate or valid comparison i.m.o.) the problem usually was not related to lack of proof, but to their frequent use of terror in order to bully the witnesses into submission (or else), the murders they committed were often committed in order to make sure they are not caught. Such a body of organized crime can be caught, if the law-enforcement officers are competent, not corrupt and if they are brave-enough to bear the consequences of doing their job properly, as a body of organized crime will sooner-or-later unwillingly leave some clues that might lead a prosecutor to themselves.

Here we are speaking about non-violent cheating allegations and not corrupt investigators, while the possible perpetrator is an individual, rather than a criminal organization and there is no fear involved. Saying that you know that the girl is "of course" a cheater without proof is a self-contradiction, since a proof is usually a prerequisite of knowing something, especially when we speak about accusations.
mc1483 mc1483 10/9/2022 07:49
@lajosarpad: "if someone claims to know that Niemann cheated, then a strong counter-argument is his imprecision/mistake with 29... Nc4 and this move needs to be taken into account by any serious attempts to convince others about Niemann cheating in his game with Carlsen"
No, I as wrote before I think that a single move does not tell anything, and is evidence of nothing. Had Niemann lost the game, then I would say that's evidence of not cheating. Similarly, his crushing victory against Yoo is indeed suspicious, but is no evidence of cheating (had other crushing victories followed, then it would be different).

"This is a self-contradiction. If there is no proof, then it is not evident that she was cheating and, invertly, if she was "of course" cheating, then we would have proof"

I think you're not familiar with the entity we call "mafia", in Italy (altough the word might be not unknown to you). When the mafia killed someone - especially in the '60 and in the '70 - everybody knew who and why did it, but there was never proof.
In this specific case, this is the ELO curve of the girl I wrote about: https://ibb.co/gdH2YJR (the blue line). I also could provide the game she won to the IM, if you like. All this is no proof, but I really, really hope you won't say there are other explanations.

I don't see Niemann's behaviour as suspicious (I don't like it, but it does not mean I find it suspicious), I just think Carlsen's opinion cannot be discounted (although I agree he has been very unwise) as he has lost many games before but never showed such an attitude nor there are other players he is not willing to face anymore. So every evidence - for both cases - is of great interest for me.
lajosarpad lajosarpad 10/9/2022 05:16
@mc1483

I understand that Niemann's play and behavior is very suspicious for you. I also see the reasons for your suspicion. I also grant you that there are elements that are strange/suspicious to me as well. But, according to my view, we should not say that Niemann (or the girl in your example) was "of course" cheating until this is proven. A possible counter-argument can be that if we take this attitude of "innocent until proven guilty", then some guilty people may get away with their crimes and that's a serious price we pay and I agree with this possible counter-argument. But, in return for that price, we do not condemn innocent people and therefore we will be morally superior in comparison to Soviet-styled communist states.
lajosarpad lajosarpad 10/9/2022 05:14
@mc1483

Carlsen - Niemann is a crucial point for the investigation. Because, if it turns out that Niemann was not cheating in that particular game, then Carlsen did not have any valid reason to start the scandal and tacitly suggesting that the organizers of the Sinquefield Cup didn't do their job properly and this is true even if, after due investigations we will find some examples when Niemann cheated in other cases and - horribile dictu - this is true even if Niemann was cheating against Carlsen indeed, because Carlsen did not have factual proof at the time of the accusation, he did not even witness the offence he claims Niemann committed.

So, even if we believe strongly that Niemann will cheat if he has the opportunity, it is quite possible that he had no such opportunity at the Sinquefield Cup. Carlsen's allegation suggests that Niemann cheated even there, in which case the organizers of the Sinquefield Cup were doing a poor job in protecting against the cheating.

In short: if we only wonder whether Niemann has cheated at OTB, then any evidence or proof including, but not limited to the Carlsen - Niemann game are interesting. But I'm much more interested about Carlsen's allegations, his reasons to harm the reputation of Niemann, which he did in a malicious, ill-perceived and premature way, as he did not have a proof at the time of the accusation and I'm also more interested in making sure that only proven cheating will be condemned, rather than perceived cheating. The region where I live was governed by Soviet-style communist dictatorships for decades and it was common in those regimes that some allegations were accepted as true based on loose, fabricated, or non-existent proof, with the verdict being written in advance and the accused was imprisoned or even executed for it. I do not like the fact that this pattern appears in the chess world, where unproven allegations are treated as being true in many cases.
lajosarpad lajosarpad 10/9/2022 05:09
@mc1483

"All this story told me it's no use focusing on one specific game"

It depends. In our case Carlsen strongly suggested that Niemann was cheating against him. So, this is a sample the accuser himself points to as something we should take a close look on. As a result, it is of extreme use to see whether we can find a proof about this alleged cheating. I should also say that Carlsen acted unwisely by making his allegations public, especially after he played with Niemann. He could have quietly continued playing the tournament and asking the organizers to watch out for Niemann, who, being unaware of the suspicions that arised against him would be easier to catch if he was cheating indeed. By withdrawing from the tournament Carlsen made it more difficult to catch Niemann if he was cheating and seems to be much more an act of damage control after a painful loss, than a genuine concern of a possible cheating incident and a truth-seeking attitude. This is why, along others, one of the reason, due to which I claim that Carlsen behaved badly in this context, as he:

- violated his contractual obligation
- encouraged by his actions all his fans and admirers not to respect their own agreements and to quit if a tournament does not go their way
- prematurely and baselessly accused a person, while, at the time the only thing he had against him was the suspicion of cheating
lajosarpad lajosarpad 10/9/2022 04:59
@mc1483

"Of course she was cheating, but nobody had proof."

This is a self-contradiction. If there is no proof, then it is not evident that she was cheating and, invertly, if she was "of course" cheating, then we would have proof. One may have a valid reason for being suspicious about her play, but factually claiming that she "of course" was cheating, but not having factual proof seems to be premature to me.

"In both cases - either HN is cheating or is not - the whole picture is difficult to see and won't likely be ever uncovered. "

1. If Niemann cheated and continues to cheat, then he will likely be caught sooner or later, given the fact that the more times one commits a crime, the more opportunities there are for mistakes and the more opportunities the rest of the world has to catch him
2. If Niemann cheated, but stopped cheating, then the only way of proving his past cheating is to find video evidence or other unmistakable proof of his cheating
3. If he didn't cheat but starts (or will start) to cheat, then he can be caught (see point 1.) and if so, in hindsight, we all will think that he was cheating even before he started cheating
4. If Niemann didn't cheat and doesn't start cheating, then he of course will not be caught doing an offence he is not doing

So, it is a good practice to double the effort of watching what Niemann does in order to catch him if he cheated.
lajosarpad lajosarpad 10/9/2022 04:57
@mc1483

As a result, it is quite possible that the engine regards two different moves to be yielding the very same result (two moves evaluated as winning in our particular example), while one of the two possibilities in reality is no longer winning, even though the reason for not winning is beyond the epistemological horizon of the engine. This is why choosing a significantly weaker continuation is a mistake even from an engine's point of view, with the exception (if we ignore the nuances outlined earlier) of fully calculated lines. So, if your engine did not see a forced mate following from 29... Nc4, then the engine does not know with full certainty whether 29... Nc4 is still winning, or it deteriorates the position to a draw. So 29... Nc4 is a mistake from the engine's point of view as well, even if we use your experiment's results as a premise in our thinking process.

However, one may argue that a cheater can afford temporarily getting out from the winning zone, in view of the fact that the other player will likely provide ample of opportunities to get back into the winning zone. So, if Niemann was cheating and he wanted to win, then he could afford himself to make a mistake with 29... Nc4, making the win more difficult, or even getting out from the winning zone (if an even deeper calculation would lead to a draw, or if Niemann did not run the engine long-enough to find the win), banking on the fact that his opponent will likely commit imprecisions or even blunders.

So, if someone claims to know that Niemann cheated, then a strong counter-argument is his imprecision/mistake with 29... Nc4 and this move needs to be taken into account by any serious attempts to convince others about Niemann cheating in his game with Carlsen. To be clear: 29... Nc4 does not disprove the allegation according to which Niemann cheated, but it's a strong counter-argument and a truth-seeking entity needs to take it into account while investigating the issue.
lajosarpad lajosarpad 10/9/2022 04:53
@mc1483

"I agree, but that's true from the human point of view. Nothing changes from an engine's point of view."

This statement would be mostly correct if chess was a fully calculated game, like tic-tac-toe, that is, the engine, from whatever position would know with full certainty the result of the game. In that case, the best variation(s) would be computed via the so-called minimax algorithm, that is, the engine would always choose in the fully calculated game tree the move that best suits the interests of the player who is to execute that move. In that case there is indeed no difference in the evaluation in theory. In practice, even in that scenario there would be differences. The longer the line, the more opportunities are for unknown bugs to creep in and the more time the engine needs to spend in executing the move, so even though, in theory, there is no difference in a fully calculated game between two moves with the same evalution (two winning moves, for example), in practice there are some nuance-level differences.

But, apart from tablebases, computed forced mates or stalemates, computed forced move repetitions of fully explored draws by 50-move rule, the engine does not know the evaluation of the position with full certainty.

Proof: we can find games between engines with different results (we can see engines winning with white, winning with black or drawing against each-other, which would be impossible if the engine would fully konw the perfect move from each position)
Jacob woge Jacob woge 10/8/2022 05:26
“...it's no use focusing on one specific game...”

The same applies to cheating in bridge.
mc1483 mc1483 10/8/2022 03:04
(continued from previous comment)
There were other interesting things:
- nobody has ever understood how the cheating was done. Games were not broadcasted, someone (myself included) thought she had "something" hidden in the glasses she always wore (with Niemann that would not be possibile). Someone suspected "something" was hidden inside the pen she used during the games. Heavy clothes (most victories occurred between december and april) and continuous use of facial masks likely helped to conceal some other "something". The fact she was a young, shy girl also must have helped.
- after the sudden drop her rating stabilized at 1500, and is now slowly rising (1544). One theory (also mine) is she played so much at high level that she learned a lot and became a better player. Another one is she changed the method, and is again cheating while avoiding a fast rise and all the resulting suspicions. A friend of mine (1800) recently played her, won and reckoned her to be "pitiful" (but I did not check the game).

All this story told me it's no use focusing on one specific game, checking for inaccuracies or mistakes. Ken Regan, Chess.com, Yosha and others, all are checking a lot of games because of that, with different tools and engines, and spend a lot of time doing just that. In both cases - either HN is cheating or is not - the whole picture is difficult to see and won't likely be ever uncovered.
mc1483 mc1483 10/8/2022 03:04
This is the story.

A couple of years ago we had a young girl (14) of about 1200-1250 ELO suddendly performing at 2000+ for a bunch of tournaments. When playing an Open she even defeated an IM whose ELO rating was 2300 at the time (he reacted just like Carlsen). After some _months_ she had peaked at 1837 ELO and had likely gained access to the World Youth Championship (she looked like one of the strongest females of that age in our country, maybe already the best one).
Of course she was cheating, but nobody had proof. But when she reached 1837 ELO she started to be subjected to serious scrutiny and eventually arbiters watched her more carefully. Then she lost a lot of games and her rating dropped quickly to 1514; also, she played less tournaments (I think in order to mask the sudden drop).
I studied some of her games. Moves were quite often the engines' best ones, let's say 80-90% of the times. The other times they were suboptimal, and in the end (when the game was already won) inaccuracies and even mistakes occurred. She even missed some easy checkmates, spending 10 o 20 moves where she could have spent 4 or 5. But in the end she won.
The method behind that, I think, it's quite obvious: she was guided by an accomplice who suggested, every now and then, suboptimal moves in order to raise less suspicions. In the end, when the game was impossibile to lose, she was left on her own.
(continues)
mc1483 mc1483 10/8/2022 01:23
@lajosarpad: "I think you will agree that a move that makes the win much harder to achieve rather than an alternative is a mistake.".
I agree, but that's true from the human point of view. Nothing changes from an engine's point of view. I could recount a story about a cheating scandal almost nobody knows, but I need some hours, as it's a bit complicated.
lajosarpad lajosarpad 10/8/2022 12:53
@mc1483 that's very interesting. My take was that after 29... Nc4 30. Bxc4 makes the draw more achievable than earlier. I may have been wrong with my engine-less analysis (I do not use engines for chess, as it is much more fun to have open questions, rather than almost certain assessments), maybe I have overestimated the strength of 30. Bxc4, but nevertheless, your engine-based experiment also yields that 29... Nc4 is an inaccuracy/mistake, even if it is not serious-enough to change the outcome. So, it was a-+, Niemann made a mistake, after which it was a -+ according to the engine.

"It looks like it's not a mistake after all, the game is still won"

I think you will agree that a move that makes the win much harder to achieve rather than an alternative is a mistake.

"How long did Niemann think on this move? "

I don't know, but I think it must have been significantly less than an hour.

@Matthias Ruf

"There, the body scanners detect not only metal but also silicone."

That could lead to some trouble at women's tournaments.
Matthias Ruf Matthias Ruf 10/8/2022 12:11
More control and spectators not welcome

As a result of the suspected fraud, chess could soon take place without permitted spectators. At the European Club Cup in the Europahaus in Mayrhofen, passing visitors were also not allowed in because of the approximately 700 participants. Smartphones had to be handed in at the checkroom and a body scan was carried out. A special thanks goes from my side to organizer Werner Crsnko, who let chess columnist Harald Keilhack from the Stuttgarter Zeitung pass through the barrier in an unannounced short trip.

At the US championship, which is taking place at the same time, the security measures have also been strengthened. There, the body scanners detect not only metal but also silicone. To the regret of chief organizer Tony Rich, only a few select spectators were received in the first 10 minutes at the venue of the well-known chess club in St. Louis. Precisely because of the new allegations by Chess.com about online cheating in over 100 games, GM Hans Niemann's presence games will now be broadcast with a 30-minute time delay.

In a summary for all chess fans, the Carlsen-Niemann scandal includes an unpleasant perspective of increased controls, which one already had to go through at airports.
mc1483 mc1483 10/7/2022 11:55
@lajosarpad: in the while I let Stockfish compute for a long time the move you pointed out as "allowing the draw", 29... Nc4 in the Carlsen-Niemann game. It looks like it's not a mistake after all, the game is still won, but the most peculiar thing is the time required to reach this understanding. At first, fxg4 is by far the best move, and Nc4 is indeed allowing the draw (at least white's position gets better). After 10 minutes or so the engine finds out Nc4 is a very good move, almost the same as fxg4, leading to victory. But after one hour the engine settles for fxg4 - crushing victory - even though Nc4 still is seen as good - victory, not crushing.
How long did Niemann think on this move?
Science22 Science22 10/7/2022 11:08
The fact is that Hans Niemann never told us the truth about anything, and that his back up team her makes things even worse. Liars and scumbacks of the worst kind

The good news is that we will never gani see Niemann play is a big tournament with private sponsor outside US. He is history
lajosarpad lajosarpad 10/7/2022 04:49
Unfortunately comments are not allowed at the first round article indeed. The game itself does not look suspicious to me. He played at super GM level and made inaccuracies. However, his behavior after the game was awkward to say the least, even to the point of being suspiciously unwise on his part.
mc1483 mc1483 10/6/2022 12:24
@lajosarpad: I, too, am inclined to respond to science22 abuses, such as the "reality" issue and so on. Then I realize that:
- I would waste precious time replying to his/her insults;
- he/she will insult me more;
- I won't be happier for that;
- other chessbase users would problably skip the whole debate and regard us all as trolls.
That's why I think that skipping is a better solution.

At this moment comments are not allowed to the live report; maybe, as you hope, they will be to the 1st round article (I think the live report would be better, as there are already too many reports with comments discussing the same issues).

edit: I see now the 1st round article is out, but comments are not allowed.
lajosarpad lajosarpad 10/6/2022 11:41
@Arzi

"It's sad that Carlsen can't admit the simple truth; he doesn't know, he believes he knows, without evidence. It does more damage to the chess world than Niemann's cheating two years ago."

Agreed.

@mc1483

Science22 behaves in a very abusive and irrational way with his/her comments. My approach to this is to respond, especially when he/she attacks me personally. I absolutely understand your point of view and from that position I would simply skip everything Science22 writes as well as any responses to him/her.

I'm happy to discuss Niemann's game, but I will need to analyze it and form an opinion first. Also, it would be great if we could discuss that game in comments to the USA Championship first round report if comments will be allowed at that article.