The Full Chess Cheating Report of Hans Neiman • Hacker News Board

Re: The Full Chess Cheating Report of Hans Neiman

#431

Post by ninethirty » Wed, Oct 05, 2022, 4:05 PM UTC

maegul wrote: post #123

erdevs wrote: post #27

This report provides a detailed background of Hans' potential cheating, and detailed breakdowns of certain aspects of chess.com's cheat-detection methodology, including previously unknown (or little known) methods such as window focus change event monitoring and post-focus-change move analysis.[1] The report also reveals Niemann's engine move correlations alongside over two dozen chess Grandmasters who have admitted…

> [1]Tangentially, this induces an obvious concern about cheat and cheat-detection arms races. A clever cheater might scrutinize this report and refine their cheating plan. For example, they might recognize the need to use a second device (such as a phone) to cheat. They might use the data corpus presented in this report to establish limits on how often they use chess engine moves per game, and they might manage thei…

Makes me wonder how many people are running game-theory-optimal poker bots and raking in fortunes.

Re: The Full Chess Cheating Report of Hans Neiman

#432

Post by msbarnett » Wed, Oct 05, 2022, 4:05 PM UTC

gbear605 wrote: post #364

msbarnett wrote: post #59

Earlier quoted context omitted.

> They didn’t find any instances of cheating after August 2020 They banned him after August 2020, and only reinstated him after a full confession the next year. The fact that, after having being caught red handed in online for-prize tournaments and knowing he’s being monitored very closely online going forward, he suddenly starts rising at a world-historic pace (at an age this has never before happened at) at OTB tou…

They reinstated him January of the following year, only five months later.

He was (and still is) banned from prize money tournaments even after that partial re-instatement.

Regardless, there's obviously a huge difference between "no evidence he cheated after August 2020" (wow, guess his heart grew 3 sizes that month) and "no evidence he cheated after a 5 month ban which required him to beg to get his account back, made it very clear to him that they were watching him like a hawk, and they wouldn't let him play in tournaments even after that" (not exactly indicative of an unprompted improvement in his morals).

Re: The Full Chess Cheating Report of Hans Neiman

#433

Post by rrrrrrrrrrrryan » Wed, Oct 05, 2022, 4:19 PM UTC

ninethirty wrote: post #431

maegul wrote: post #123

Earlier quoted context omitted.

> [1]Tangentially, this induces an obvious concern about cheat and cheat-detection arms races. A clever cheater might scrutinize this report and refine their cheating plan. For example, they might recognize the need to use a second device (such as a phone) to cheat. They might use the data corpus presented in this report to establish limits on how often they use chess engine moves per game, and they might manage thei…

Makes me wonder how many people are running game-theory-optimal poker bots and raking in fortunes.

I don't know if a game-theory optimal poker bot would work well - You'd really want a battle-tested psychologically optimal bot that can convincingly play like an amateur, except when it really counts, to give your victims a false sense of confidence. A hustler-bot, essentially.

I'm sure it'd be highly illegal.

Re: The Full Chess Cheating Report of Hans Neiman

#434

Post by pedrosorio » Wed, Oct 05, 2022, 4:22 PM UTC

treis wrote: post #412

pedrosorio wrote: post #239

Earlier quoted context omitted.

> Anand, a former world champion, became a GM at age 19 At the age of 18: https://en.wikipedia.org/wiki/Viswanathan_Anand#Early_chess_... Also, according 2700chess.com he was #96 in the world with a rating of 2520 on January 1988. That's right after turning 18. Growing up in India (before the chess boom). Before chess engines existed. 6 months later he was #49 in the world. https://2700chess.com/players/anand A year…

This is all pointless comparisons though. Hans made his improvements during Covid times and The Queen's Gambit boom in online chess. Even if it's an extraordinary thing to happen we were in extraordinary times.

> Hans made his improvements during Covid times

None of the similarly rated players at the same age did anything similar during "Covid times and The Queens' Gambit boom".

Watch the videos I linked at the end of my comment and tell me how covid times contribute to that anomaly.

Re: The Full Chess Cheating Report of Hans Neiman

#435

Post by treis » Wed, Oct 05, 2022, 4:33 PM UTC

pedrosorio wrote: post #434

treis wrote: post #412

Earlier quoted context omitted.

This is all pointless comparisons though. Hans made his improvements during Covid times and The Queen's Gambit boom in online chess. Even if it's an extraordinary thing to happen we were in extraordinary times.

> Hans made his improvements during Covid times None of the similarly rated players at the same age did anything similar during "Covid times and The Queens' Gambit boom". Watch the videos I linked at the end of my comment and tell me how covid times contribute to that anomaly.

Unless they make a concrete accusation of cheating I'm not really interested. These circumstantial accusations all boil down to "something unusual happened" but the world is full of unusual things happening. Nobody, AFAICT, has come out and said "Hans cheated in X OTB event. He did so by getting Y information by Z method".

IMHO everyone needs to either put up hard evidence or shut up.

Re: The Full Chess Cheating Report of Hans Neiman

#436

Post by WhitneyLand » Wed, Oct 05, 2022, 4:45 PM UTC

rippercushions wrote: post #262

rowanG077 wrote: post #254

Earlier quoted context omitted.

I doubt that since a weekend project. You need to analyze a video stream of hundreds fps in real time with very low latency requirments. You'd probably need an FPGA to do that. Custom training an AI to only trigger on heads.

No, it's trivial these days, and this very task is one of the introductory tutorials to Google's Coral AI Tensorflow boards: https://coral.ai/models/object-detection/ Which you can buy as a USB stick for $59: https://coral.ai/products/accelerator/ Sure, it's a little more complex if you want to train on specific heads as opposed to a generic model, but not hugely so. The biggest challenge would likely be downsampling…

When you say it’s a weekend project for a college student it suggests it’s mundane or easy for most professional developers within a weekend.

This sounds exactly like the old “Twitter would be easy to implement give me a month” naivety.

Development is hard, even small projects aren’t usually finished in a weekend, let alone something there’s no reference implementation for and that we’re merely speculating about.

If such a project was built in a weekend to work reliably enough others could use it successfully in gameplay, it would be highly impressive. Maybe we’d call it an extraordinary effort, if for no other reason “because software”.

Re: The Full Chess Cheating Report of Hans Neiman

#437

Post by kadoban » Wed, Oct 05, 2022, 4:56 PM UTC

patmcc wrote: post #192

kadoban wrote: post #180

Earlier quoted context omitted.

For shooters, bots or assists that only use vision and normal input (special mouse/keyboard) aren't hard to imagine and I'm positive already exist. I know specifically I read a while back about mice that do recoil correction to improve aim, and things like "click when something that looks like a head is in the crosshairs" would be a weekend project at most for a bored college student. Strategy games may be harder, bu…

It's not hard to make cheating tools, it's hard to run them on the same hardware as the game and not get caught. Chess cheating tools don't need to be on the same hardware, so they're not detectable in the same ways.

My point was that video game cheats _don't_ need to be on the same hardware.

Re: The Full Chess Cheating Report of Hans Neiman

#438

Post by kadoban » Wed, Oct 05, 2022, 4:58 PM UTC

perth wrote: post #271

kadoban wrote: post #180

Earlier quoted context omitted.

For shooters, bots or assists that only use vision and normal input (special mouse/keyboard) aren't hard to imagine and I'm positive already exist. I know specifically I read a while back about mice that do recoil correction to improve aim, and things like "click when something that looks like a head is in the crosshairs" would be a weekend project at most for a bored college student. Strategy games may be harder, bu…

Funny thing, there are already YOLOv3 derived aimbots, and a lot of anti cheat now scans pci-e cards to try to detect them. They intercept video input and act as mouse and keyboard devices, acting almost completely independent of the host computer.

Yeah that's not far at all from just running as a separate machine, either grabbing the hdmi output or going full webcam pointed at the screen.

Re: The Full Chess Cheating Report of Hans Neiman

#439

Post by alephxyz » Wed, Oct 05, 2022, 4:58 PM UTC

cryptonector wrote: post #400

So chess.com says Niemann cheated online. Well, I'll buy that, why not, it's so easy to cheat at chess online that maybe we should assume most online players do. Oh? What? And GMs admit to cheating online? Why then is Niemann's possible/likely online cheating news and not the other GMs'?! Fine, but let's say he's cheating online and that it's more a scandal than those GMs' cheating, and that it being more of a scanda…

Niemann himself admitted he cheated online on an interview.

Re: The Full Chess Cheating Report of Hans Neiman

#440

Post by barrkel » Wed, Oct 05, 2022, 5:00 PM UTC

rowanG077 wrote: post #342

barrkel wrote: post #270

Earlier quoted context omitted.

7ms isn't going to be the bar. Human reaction times are in the 150+ms range, you have 20x your budget. I agree it's still tough though, most I/O paths on generic hardware will have buffers at every step that you'll have to fight to eliminate.

7ms is enough. But that is one in a chain of latencies. 150ms is way too long if you want your trigger to be effective. In 150ms the enemy has moved out of your crosshair. Humans deal with that with dynamic adjustment but the AI will just click the mouse. I'd think you'd need to hit less then 25ms end to end and even that might not be enough for far away targets.

The AI would need to lead the target under any reasonable implementation. Just clicking the mouse when crosshairs are over a target would scarcely deserve the name AI.