Carlsen withdrawal after loss to Niemann

[Uri Blass]

It's easier to cheat in OTB than it is to cheat in Bullet games online, that's fact.

No, it is easy to cheat in bullet games. You can simply let engines play. The hard thing is to make it play like a human so that an anti-cheat-software won't detect it. But this also should be doable: Take Leela with a human nn, and implement some human behaviour. Then try it with fake accounts until the detector don't recognize it.

Especially when compared to Vassily Ivanchuk's famous post-game analysis of an entire game completely from memory with no visual board to prompt him.

This is also the point where I am sure, he is cheating. Especially if you look at the video of his game against Carlsen where he doesn't even look at the board. His interviews show, that he cannot analyze without a board.

Obvioulsy it is hard to find sufficient evidence if someone is cheating OTB, and the technical progress will invite more and more to do it. The only chance I see is to ban every online cheater for lifetime from official tournaments, or use statistics like the STDACPL. No mercy.

Let me add: Until there is a solution for that, the only solution for real GM's is not to play against obvious cheaters.

I do not understand how you can practically cheat online in bullet with no convincing evidence against you and in a way that help you to increase your rating significantly unless you let a different human to play for you or to suggest you moves.

There can be evidence that is not based on the moves but based on the fact that you use some software that you are not allowed to use to play automatically(you have no time to look at engine output and copy the moves of the engine in bullet).

Note that I do not know if it is even possible to play automatically in games of humans and I think it should be easy to prevent it and a software can accept moves only if it detect moving the mouse.

Time management also can be an evidence against you because usually humans use premove in bullet and I think engines are not programmed to use premove.

I also expect times of the moves in buller to be also dependent on the length of the movement that you need to do with the mouse.
An extreme example is if you play Ra1-b1 and after premove of the opponent in the next move Qh8-a1
I expect in this case that Qh8-a1 is going to take you significantly more time if you are a human because you need to move the mouse from b1 to h8 and to a1.

[Alexander Schmidt]

An important line from the report: “…while Hans has had a record-setting and remarkable rise in rating and strength, in our view there is a lack of concrete statistical evidence that he cheated in his game with Magnus or in any other over-the-board (“OTB”)—i.e., in-person—games.”

You forgot to add: "OTB cheat detection has not, to date, been the primary focus of Chess.com."

[Uri Blass]

It's easier to cheat in OTB than it is to cheat in Bullet games online, that's fact.

No, it is easy to cheat in bullet games. You can simply let engines play. The hard thing is to make it play like a human so that an anti-cheat-software won't detect it. But this also should be doable: Take Leela with a human nn, and implement some human behaviour. Then try it with fake accounts until the detector don't recognize it.

Especially when compared to Vassily Ivanchuk's famous post-game analysis of an entire game completely from memory with no visual board to prompt him.

This is also the point where I am sure, he is cheating. Especially if you look at the video of his game against Carlsen where he doesn't even look at the board. His interviews show, that he cannot analyze without a board.

Obvioulsy it is hard to find sufficient evidence if someone is cheating OTB, and the technical progress will invite more and more to do it. The only chance I see is to ban every online cheater for lifetime from official tournaments, or use statistics like the STDACPL. No mercy.

Let me add: Until there is a solution for that, the only solution for real GM's is not to play against obvious cheaters.

I do not think that not to play OTB against obvious cheaters(when there is no rule that they are not allowed to play) is a good idea.

I think that it is easy to find sufficient evidence against a cheater that most people suspect that he is cheating simply by searching his body before or after the game and if he does not agree other to search his body(like Borislav Ivanov) then it is also a sufficient evidence against him.

I believe that Hans has only 2 choices if he does not quit chess and chess players continue to play against him.

1)Stop cheating
2)Continue cheating when there is going to be a sufficient evidence against him.

[pocpit]

How's Niemann able to play at 3000 rating level in Bullet games online, in the years 2020-2021?

Anyone who is a serious chess player online knows that it's practically impossible to cheat in Bullet games (1+0)

He's telecasted live in twitch simultaneously while playing bullet games even and still maintained that high a rating which is equivalent to fide of 2650+

It's easier to cheat in OTB than it is to cheat in Bullet games online, that's fact.

I've been told that there is software that does make it possible to cheat even at 1 + 0 bullet chess, though I would think that it would be pretty obvious to the chess.com cheat detection as it would have to be pretty much automatic. So I wouldn't say it's easier to cheat OTB than in online bullet chess, though it might be easier to catch the bullet cheater. Having a 3000 bullet rating on chess.com, if no cheating was involved, would mean that he is at least a strong grandmaster, but it doesn't tell us whether he did or didn't cheat in longer games.

I haven't tried it, but this software is supposed to allow you to cheat even on bullet games: https://chess-bot.com/

OTOH, I wonder if a player using a software like the above will move their mouse differently to a no cheating player. Actually I assume "mouse movement" is one of the parameters used by all these online cheat detectors.

[M ANSARI]

I would think that mouse movement can be easily made to move like a human.

[swami]

If you see Niemann's twitch telecast videos or on YouTube you would notice that he actually makes a lot of premoves. Preparing in advance of eventual recaptures, expected piece developments, carrying out piece manuevring according to a plan, and many other moves. Highly unlikely he used Chess Bots help, besides Chess.com is highly sophisticated in cheat detection in Bullet games. No way anyone can remain at the top for nearly 1500 days and counting.

[Uri Blass]

I would think that mouse movement can be easily made to move like a human.

I think that writing a software to move the mouse like a human including premoves is a lot of work and nobody is going to do it for cheaters.
There is not enough money in chess for it.

[CornfedForever]

As for Hans ... he is 19 and probably rated around 2500 ELO FIDE (real rating not his fake rating). So he should be around 2900 to 3000 ELO. There was a 1 0 bullet match against Nardosky a few days ago ... he did quite good ... but lost to Nardosky by around a 3 to 1 margin. Nardosky is a very strong bullet player but certainly not the best.

Naroditsky is actually rated #2 in Bullet on chess.com (behind Hikaru). Bullet is a totally different animal that 'real chess'. There is basically no calculation and about hand movement, pre-moves, and intuition (a place Aagaard says Hans excels...). There are some 25 players north of 3000 there. Nardo is #3 on chess.com in Bullet as well - just behind Hikaru and Carlsen. There a similar skill set is needed.

You know, your average active 2500 FIDE, from what I see, is perfectly capable of good after the game analysis, even without sight of the board. When I was at my best, I always struggled with 'good' post game analysis of games - without sight of the board - I had just played and some weaker players were clearly better at than than I...I probably won more often just because I understood the game a bit better.

Anyway...ducks aside, all you can continue to scapegoat (GM Smeardon's term/characterization - Perpetual Podcast interview...aimed at Carlsen) Hans for the OTB cheating sins of the world...not my bag.

[swami]

Bullet peak rating of FIDE Top 10 in in Chess.com

Ding Liren
3112

Wesley So
3206

Anish Giri
3206

Ian Nepomniattchi
3168

Fabiano Caruana
3112

Firouzja Alireja
3521

Hans Niemann
3227

One could say that based on Bullet alone, his performance seems to be even higher than many of the current top 10 players. Ofcourse, that cannot be compared with classical games but it does atleast make a case that Niemann has capability of playing over 2650.
There are few exceptions like Andrew Tang, who is thin, fit and agile hence able to play faster despite being rated sub 2550 in FIDE but the exception is not the rule.

[CornfedForever]

Bullet peak rating of FIDE Top 10 in in Chess.com

Ding Liren
3112

Wesley So
3206

Anish Giri
3206

Ian Nepomniattchi
3168

Fabiano Caruana
3112

Firouzja Alireja
3521

Hans Niemann
3227

One could say that based on Bullet alone, his performance seems to be even higher than many of the current top 10 players. Ofcourse, that cannot be compared with classical games but it does atleast make a case that Niemann has capability of playing over 2650.
There are few exceptions like Andrew Tang, who is thin, fit and agile hence able to play faster despite being rated sub 2550 in FIDE but the exception is not the rule.

And to clarify, FIDE itself does not rate bullet...in case anyone wonders.

But...

Firouzja Alireja
3521

Sounds suspect...even 'peak' itself may not be the best measure as it is not close to being a stable thing.