Chess.com banned Hans after beating Magnus. Why?

So you've been stalking my page since day 1? You're a creepy old man, NervesofPoop. 

I am starting to think this is basically then end of chess. It may not be solved, but it may be solved "enough". 

Han's own admission in the Sinquefield Cup is that he played a computer move on move 10.  The whole cheating accusation is wether he played a computer move he learned before the match from a transposition of an early Carlsen game, or someone or something communicated it to him during the match.

How would you even detect cheating against someone whose major skill is the ability to absorb and regurgitate stockfish lines at a frightening rate? That isn't considered cheating, but it basically is.   

There's only so many good openings and only so many good moves from those openings, that someone who is insane at memorization could possibly be stockfish for 10 ~ 15 moves.  Someone wouldn't have to do full memorization, one would only have to know it good enough to repeat it on the board. This is especially possible for Carlsen, who seems to be playing the same opening over and over again. He even ended up being in the same opening for the match he resigned in.

Definitely turning chess into a game where who memorizes the dictionary more wins.

Players make mistakes in their games - that's part of being human. Even Carlsen.

Players who don't make mistakes are ... far less common.

This was Carlsen's recent victory over Keymer. Mistakes made by both sides. A back-and-forth battle, and an eventual winner. Accuracy percentages in the 80s.

That's, more or less, human chess at the GM level:

Carlsen's loss against Niemann, on the other hand, was very one-sided. Carlsen was on his heels for mostly the entire game.

I'm not saying that Niemann cheated, but it's inaccurate to say Carlsen simply "played poorly". More accurate would be: Carlsen played at a normal human level (around 90% accuracy).

But Niemann played at near 100% accuracy. Too strong to survive.

Either he received assistance at some point (the suspicion), or he played one of his best games ever (also within the realm of possibility).

Have they got to the erm bottom of this yet? 

They have not gotten to the erm bottom... as a matter of fact, I think we've gone straight back to the erm TOP! 

I won’t be asking to use his phone again ;o)

That's possible. But a few people have observed that Carlsen is currently playing some of the best chess of his career. I guess it's possible to be both near a performance peak and, at the same time, be burned out. 

I still think whatever is behind the recent kerfuffle is probably something that's not going to be very interesting. It's like being afraid of the dark. That noise from the other room wasn't a mass murderer breaking in and robbing the entire house. It was a rose that fell out of the vase and knocked over a wine glass. 

Probability of RemovedUsername333 being TonyaHarding/Melvin-whatever?  Pretty high .

This is not quite accurate.

A human cannot memorize all moves an engine provides in all lines.  Additionally, the notion that Hans has an amazing memory can be easily challenged by looking at his post-game analysis.  It would seem he is a reasonably strong player with perhaps an above average capacity to memorize, but nothing too special there.

The top guys generally do not memorize long variations in every opening they play.  Rather, they have a couple specific ideas they are working on at any given time and memorize those lines until they have a chance to play them.  Mostly, they memorize the ideas, not necessarily the moves (the notable exception is the Berlin endgame where several of the top players have memorized lines up to move 35+).  When you play a strategically complex opening, there are simply too many options for a human to memorize every possible move in every possible variation.

Hans is of the generation that learned chess with engines being omnipresent and recognised as stronger than human players. Perhaps it's not surprising that he has picked up 'computerish' ideas. Also, I hear he enjoys a blueberry yoghurt during OTB games.

Ok, let's start at the top.

Chessbrah
ONE rated games in the last year against any player over 2000. It was a cheater.
Only 27 total rated games against any player over 2000.
https://www.chess.com/games/archive/chessbrah?gameOwner=other_game&gameType=live&gameTypeslive%5B%5D=rapid&rated=rated&ratingFrom=2000&timeSort=desc

SeanWinshand
A lifetime accumulation of ZERO rated rapid games against any player over 2000.
https://www.chess.com/games/archive/seanwinshand?gameOwner=other_game&gameType=live&gameTypeslive%5B%5D=rapid&rated=rated&ratingFrom=2000&timeSort=desc

Polish_fighter3000
Three rated rapid games in the past year against any player over 2000.
Only 33 rated games lifetime.

https://www.chess.com/games/archive/polish_fighter3000?gameOwner=other_game&gameType=live&gameTypeslive%5B%5D=rapid&rated=rated&ratingFrom=2000&timeSort=desc

---

Chess.com rapid leaderboard is a joke. They pad the leaderboards like this to hide the cheating.

Wow, a handful of games a year against players rated many hundreds of points below them, and rating jumps of over 600 points in a single game.

I'm sure his 2900 rapid rating is statistically justified /s

Imagine if I had a 2000 blitz rating.

But I'd NEVER played anyone over 1100.

And I only played 1 rated game this year.

And I was on the leaderboard in spite of it all.

Now imagine someone using my account as evidence that blitz ratings are accurate and that strong players really do play blitz

---

Look at someone who maintains a ~2400 rapid rating (meaning they regularly play rated games).

Look back in their history about 1 year (allowing time for cheaters to be caught and banned).

Over 50% of their opponents will be banned. Sometimes close to 100%.

I know some GMs will play rapid against other GMs that they trust... these functionally private pools have nothing to do with rapid ratings overall.

I went through the list a while back, removing all players with a NM, FM, IM, or GM title.  Virtually everyone over ~2200 rapid that played consistently (5-10 games per week) was cheating in at least half their games (unless they were alt accounts of Hikaru and Magnus).  Hell, there are some people in that list who convert 99% of their advantages (no matter how small!).  Even Magnus and Hikaru do not do that.

But that gets into why the fair play algorithm is a joke ...

Yeah, I went though the top 100 non-titled rapid players at one point. IIRC about 30 of them were inactive and I thought about 30 of the remaining were cheating. Maybe that estimate was a little high. I didn't report them because I felt like if chess.com is so incompetent that they can't monitor that area of the site on their own then I'm not going to waste my time trying to help (I think they should be automatically checked).

I did report one guy because his account was so disgusting to me. Played perfectly 10, 20, 30 moves in a row (all moves after 4-5 seconds) then would play on his own the rest of the game and a blitz rating 800 points lower than rapid. In multiple games (where all the moves were made after 4 seconds) there was a 99% score. Checking now it's still open. At least it's inactive.

There are a few in the top 200 or so that annoy me because they are so obvious it isn't even funny.  2 I looked at a while back had their real name on their profile, claim they are coaches, with 2000+ FIDE ratings.  When you look up their profiles on FIDE's site, their actual OTB rating is 1400 (with recent games!)  With 2400+ chess.com online rapid ratings.  One even claimed he coaches people up to 2200 FIDE.  I am not sure what 2200 wants to be coached by a 1400, but okay ...

With a rapid rating of around 2000 I play several 30 minute games one or two days each week vs 1800-2200ish rated players. I am notified fairly regularly, probably once a week, about rating compensation for losses vs banned accounts and quite often if I check my opponents profile I see they are rated eg. 1950 rapid, 1280 blitz, 930 bullet and the like - I consider my own ratings in these categories to be more what one would expect to see. It could be that they can't handle the speed required for blitz/bullet, hardly ever play etc but I become less convinced when I am being outplayed and beaten by players with such ratings whilst also being 10-15 minutes down on the clock 😆

I would only look at that as a sign if they actually play blitz or bullet regularly, but yes, that is usually not a good sign.  Same with Puzzles and Puzzle Rush scores.  If the do puzzles regularly and have a ~1200 puzzle rating but never miss a tactic in a game, it is highly suspect.

I believe as well as the fact that there are players (often older ones) who are bad at blitz relative to rapid (but not absurdly bad), rapid is also relatively more popular with naive cheats.