Happy holidays, dang! Thank you for all the effort you pour into making HN as useful and vibrant a forum as it can be. You help set the bar for productive online discourse, and that means something and is more and more important in the world.
The report outlines the rationale in some detail. The removal of Hans' ban was controversial among chess.com's Fairplay staff. In recent months, suspicions grew about Hans' tournament performance among both elite chess players and chess.com staff. Simultaneously, chess.com is readying to host its largest ever (by cash prize pool) tournament, and Hans recently qualified for it. Then the Sinquefield incident happened, resulting in both Hans' play specifically, and the potential for cheating in chess tournaments generally, being called into question with a great deal of attention.
All of this led chess.com Fairplay staff to request that leadership re-review Hans' account, with updated analysis they provided, and to reconsider the removal of Hans' ban.
> Where does this report state that the "removal of Hans' ban was controversial among chess.com's Fairplay staff?"
Page 58, from the text of a letter chess.com sent privately to Hans, just after they banned him:
"When I received your confession back on August 12th of 2020, in light of your age, I allowed you to create a new account with no fair play markings to continue to stream chess... For my team, however, there always remained serious concerns about how rampant your cheating was in prize events. In finalizing the field for the upcoming CGC, and based on a growing concern regarding ensuring fair play in Chess.com’s first million dollar prize event, my team did a deep review of your past history, and encouraged me to rethink my position of letting you continue to play in prize events on Chess.com. I ultimately made the decision that too much was at stake given our ongoing suspicions and past violations." [Emphasis added, 1]
This is an example of where such internal controversy is discussed.
[1]Full text of this section of the letter:
"Moving on to my second point, I want to address both the reasons and timing for freezing your account and rescinding your CGC invite. For my team, however, there always remained serious concerns about how rampant your cheating was in prize events.
When I received your confession back on August 12th of 2020, in light of your age, I allowed you to create a new account with no fair play markings to continue to stream chess. You’ll remember that I worked hard to both advise you on this process and to protect you as much as I could. I would do that again for you or any young player I deemed to have lost their way and wanted to choose a better path forward.
For my team, however, there always remained serious concerns about how rampant your cheating was in prize events. As you know, we’ve closed the accounts of hundreds of titled players (including 4 of the top 100 Grandmasters who have confessed to cheating), and we carefully monitor and help all of them as they rehabilitate into participating in our events. You and I had many subsequent discussions in our Slack DMs where we openly cooperated on the right way for you to rebuild your reputation.
In finalizing the field for the upcoming CGC, and based on a growing concern regarding ensuring fair play in Chess.com’s first million dollar prize event, my team did a deep review of your past history, and encouraged me to rethink my position of letting you continue to play in prize events on Chess.com. I ultimately made the decision that too much was at stake given our ongoing suspicions and past violations.
Considering the above, we made this decision to close your account privately and uninvite you from the CGC. I regret the timing, but the timing between the Sinquefield Cup and the CGC required me to move quickly to replace your spot. I believe I acted in the best interest of the game and all participants to
reconsider our invitation with so much at stake.
I’m going to bring my letter to a close with an offer to have a call. If you are willing to correct the false statements you made about having never cheated when it mattered (now that you have said these untruths publicly), acknowledge the full breadth of the above violations, and cooperate with us to
compete under strict Fair Play measures, Chess.com would be happy to consider bringing you back to
our events. In fact, I think it would be a wonderful redemption story for the full truth to come out, for the chess world to see this and acknowledge your talent regardless of your past, and give the community what they deserve: The truth."
The full report is worth reading, if you're interested. It reveals a great deal of additional information, not just about Hans' history of cheating and his misrepresentations, but of wider spread cheating (or suspected cheating) by high level chess elites online.
I've now gone through the full report and note that the vast bulk of it has little to do with the accusations themselves (or evidence) but ELI5 type stuff about the context including analysis, and - as you say - indications of a wider problem.
That's fine, but gives the surface impression of being heavyweight while the important stuff is only a few pages and adds little that is new and nothing that delivers a knockout blow that justifies the onslaught against Niemann led by Carlsen's antics that kicked the whole thing off.
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 to cheating on chess.com. The fact that online cheating is so widespread even among top chess players is certainly news to many, including me. Perhaps it is a good thing that this scandal is highlighting the issue, and given how widespread cheating may be, perhaps chess tournaments both online and physical need to take cheating much more seriously than they apparently have been.
There is also an interesting analysis of Hans' rating improvement history, his over the board tournament performance and key game analysis, and a rundown of key moments in his game against Carlsen in the Sinquefield cup. Each raises concerns.
Chess.com's report also makes it clear that Niemann lied outright about his history of cheating in post-Sinquefield interviews, as he admits in communications with chess.com Fairplay staff to much broader cheating.
All in all, the report raises many concerns and it seems reasonable for the chess community to demand much higher standards of cheat prevention and detection across competitive venues. How long might cheating issues have gone on merely rumored vs fully investigated or acted upon, had this intrigue not developed due to Carlsen's withdrawal from Sinquefield '22?
[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 their ratings progress over time carefully, so as to stay in acceptable ranges of engine move correlation, rate of improvement, etc.
> [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 their ratings progress over time carefully, so as to stay in acceptable ranges of engine move correlation, rate of improvement, etc.
Important point I'd say. I really can't shake my personal feeling that chess as a sport is just simply dead, especially as an online e-sport. Especially in combination with the possibility that there's plenty more cheating that they're not catching/detecting.
Taking this report along with Hans's admission to only cheating twice, for example, and accepting Chess.com's assessment as accurate, it would seem that Hans's mistake was to confess only to the instances of cheating that he thought had been caught. Which indicates the mentality and experience involved where the actual game is not getting caught and many, just as Hans was before beating Magnus, are playing it successfully.
> I really can't shake my personal feeling that chess as a sport is just simply dead, especially as an online e-sport. Especially in combination with the possibility that there's plenty more cheating that they're not catching/detecting.
Online chess is bigger than ever, and some kind of botting/cheating is possible for virtually every e-sport (indeed I'd say many are easier than chess). This is a major scandal that should have serious consequences, but there's no reason for it to be the end of online chess.
>>Online chess is bigger than ever, and some kind of botting/cheating is possible for virtually every e-sport (indeed I'd say many are easier than chess).
I think it's the other way, chess is about as easy to cheat as it gets. With chess you can have a perfectly usable cheating system that isn't on the same computer at all - computer A is clean, and is where you actually play on chess.com, computer B is where you enter the moves into the chess engine. That doesn't work for something like CS:GO or League of Legends or whatever, cheating or botting needs to be in real-time and on the same computer.
Yeah chess is the perfect type of game for cheating - it has a tiny (relatively) gamestate, perfect information, and it's turn-based. At the level of chess player we're talking about it's sorta trivial for them to almost immediately memorize a boardstate.
> At the level of chess player we're talking about it's sorta trivial for them to almost immediately memorize a boardstate.
And the rest! I'm pretty crap (hovering 750-850 on chess.com, idk if I play enough for it to be accurate though, probably lower) but I was still thoroughly impressed when I got thrashed by someone (~1200) playing the entire game in their head. As in I had a local game running on my phone, called out my moves, entered his, and he never saw the board, just held it in his head. Thrashed me.
(And there was background conversation going he'd occasionally chime into, ordering pizzas, etc. Ridiculous.)
A 1200 playing full games in their head sends up some warning flags to me?! I'm over 1500 on Lichess (yes, ELO calculation is not the same, but likely within 300 points at that level?!) and can't even dream of doing that. Did your friend cheat? ;)
I don't see how he could (or why he would) have, in the same room, eyes often closed, yelling at people to shut up so he could concentrate; occasionally checking the board state with me (always bang on) - but as I replied to sibling, it's not necessary to have a great memory like that is it? Of course, picturing/tracking the current state of the board has a lot of overlap with thinking x moves ahead though I suppose.
1200 rapid if that makes a difference? Also, I suppose it's.. certainly not orthogonal, but it's not dead parallel either - can have a good/bad memory and play well/not well.
Depends. Some fish tournaments, like bass, are typically catch and release after weighing. This makes it harder to examine the fish if the goal is to release. Others like kingfish are catch and kill, though people cheat by other means like catching a fish earlier in the week.
The big money tournaments like bill fish, where millions of dollars can change hands, are usually catch and release on the spot but entrants have a judge on board to measure any fish caught.
I've done competitive king fishing in the past, and cheating or suspicion of cheating has been around for decades.
From the article they describe many ways they can cheat in fishing (e.g. putting ice in that then melts) that aren't so easy to catch.
If there is a way to cheat, someone will inevitably give it a go with enough incentives. Cheating is the meta-sport and I would think no sport could ever be completely free of it.
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, but still not going to take too long before those exist.
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.
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.
It is not hard to run cheats for basically any game that uses video on another machine. Just a little more expensive (i.e. it is harder for someone in high school, but not anyone over 20).
Splitting the video stream to a second computer to run real-time video analysis and then send back commands through an emulated mouse may be doable, but it's certainly harder for anyone to do than copying chess moves on your phone.
Certainly! But it does show the amount of effort people will go to cheat - though lots of the "online shooter" cheating is people selling cheating tools/kits to idiots who want to win (basically turning games into pay to win).
The number of people who can build that kind of system from scratch is small.
I don't have details, but they're originally were designed for specific games - but the mechanisms can be adopted for others. The goal is to get a slight advantage on headshots, I believe.
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.
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 the video stream into something the model can process quickly enough.
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”.
That's a generic AI. I would be very surprised you could get it to work accurately for games and hyper specific the heads of (possibly non-human)models. The box you see in the marketing material is completely useless for this use case. Besides how do you get the data to your standard second PC, scale the input the USB thing expects. Send it over USB and back, and then you have to send it to a custom mouse that receives the click command over bluetooth? I don't think 7ms just for the inference step is going to cut it.
A weekend project were you have to design an ultra low-latency software + hardware stack, a custom mouse and train a custom AI. I would be incredible if anyone manages this in a weekend.
If you want absolutely olympic response times (110-120ms), it is a little tight but nowhere near impossible. 7ms as the other poster says is absolutely insane. Most videogames will already trigger on consistent peak human response times, so you'll want to trim down your bot to go to a more average person (160ms, maybe a bit higher) to fly under all the heuristics they might (but probably don't) use. Plenty of time to compute things.
7ms is enough. But that is one in a chain of latencies. 110ms is way too long if you want your trigger to be effective. In 110ms 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.
You are completely misunderstanding the goal of such a cheat. It's a "triggerbot" as in it fires the mouse when an enemy head enters the crosshair. The classical use case is that the user camps at an edge and waits for the enemy to peek. The user has his crosshair at the correct offset from the edge so that when the enemy peeks he could hit them with a human reaction time.
However a human sees the enemy BEFORE it enters the crosshair and estimates and corrects when the enemy will enter the crosshair. A trigger bot measure exactly WHEN the head is in the crosshair, it has no predictive power of enemy dynamics. This totally changes the latency game. A 150ms trigger latency from the time a head is in the crosshair basically means you shoot when either the crosshair or enemy has moved significant amounts. This also means that you can very obviously cheat with a trigger bot if you use the trigger badly, the human needs to "hide" the trigger latency to make it appear human. You can't compare it to human reaction time at all, a trigger needs to be about an order of magnitude faster to be useable.
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.
Leading the targets makes this MUCH harder from just an over the counter object detection AI. In fact making that and making it look human so it's not trivially detectable would blow this from a weekend project you can do in a few hours(Which I already doubt.) To weeks/months/years long project since you need it be really humanlike... Even now cutting edge AI can be spotted by humans.
I was thinking a decent proof-of-concept. All you'd really need is a generic object detector and fake mouse.
Latency _could_ make it a lot more difficult, but beating a human I don't expect would be hard.
You don't need to do hundreds of fps. Just grab the newest frame, process it, repeat. Missing frames at most increases effective latency or means your cheat isn't 100% effective if you miss a head. It's a sliding scale of improvements, not a deal breaker.
You also don't even necessarily have to process the whole frame. Just the bit actually _at_ the crosshairs is probably going to be enough for a crappy version.
And a fake mouse is just usb-hid, usb gadget whatever search terms, not like you'd have to break any new ground there.
I consider it a "weekend project" because it's just throwing together a couple of existing libraries in a fairly standard way. Like most things, cleaning it up enough to be perfect could/would take far longer.
You don't need 'hundreds' fps - you can do it w/ 10sec and get better reaction time than humans (reaction under 100ms in track events is consider foul). It should not be so difficult with a separate GPU and a capture card.
Serial mouse/keyboard inputs are trivial as well.
How is Serial mouse/keyboard trivial? You'd need to make a custom mouse that receives data from your second PC. You'd probably want it to be wireless else you will have 2 cables coming out of it. And no just having a second fake mouse is not fair. That's easily detectable by software and would definitely raise a lot of red flags.
This is how: take ESP32 - it has bluetooth, wifi and serial port all integrated and it costs couple of dollars. It's small and it runs on 3.3V, it has a voltage regulator that it allows to connect to 5V.
I can do the code and place the esp32 in a mouse on a Saturday (my C is always rusty, not using it professionally). And I am a hobbyist at best. So the original serial mouse/keyboard are connected to ESP32 that normally proxies the signal of the hand movement to the PC.
If I press a pedal (w/ foot) the second computer would receive a signal, calculate a human alike trajectory to move the mouse and send it to the ES32. The latter will execute it along with the left click to shoot.
Like I said - trivial. The same can be done with the USB port as well, and it's not harder. Just that for PS/2 I have the tools laying around.
It is trivial. All you need is a device that can work as a USB gadget that you can plug your keyboard into. I could do it with my phone, many Raspberry Pi-like SBCs, or even Arduino-like boards...
Using computer vision on the video stream (eg via a capture card) and then sending valid mouse and keyboard inputs most definitely has not been around for decades. It's a mechanism to cheat that has zero binaries running on the host computer.
You can use (the data of) statistical analysis to train a deep neural network to avoid detection. That may also help avoiding chess detection cheaters, so the moves are not perfect anymore.
You can point a camera on the first computer screen to capture and analyze GO board and second computer will provide suggestions. No need to memorize anything.
Or if the system allows live spectators, just log another machine in as one of those and scrape the data that way, possibly more reliable than adding machine vision into the mix. That won't work for contests that have a broadcast delay though.
Totally fair ... and it would make a lot of sense that this event brings about some sort of cultural cleansing or realignment. I guess I really just meant "dead to me."
That being said, if it turns out that online chess continues to do well, but there's still plenty of cheating going on, I don't think it'd be accurate to say online chess is alive because it wouldn't be chess but something different, for better or worse. And whether such an outcome occurs might be what's really at stake here.
Cheating is common in e-sports, but it is rare in high-level competitive e-sports. Sure, you may be a 'gaming chair enthusiast', using a cheat when in your online Apex or Forntite pub games, but you can't bring it with you to a tournament.
Starcraft is about the only e-sport where I can see tournament cheating providing a very large advantage for a single bit of information - a player who can be notified that he is being all-inned would have a huge advantage in tournament play.
Dota2 could have a similar situation. Enemy is doing Roshan. Obviously not as significant as your example since teams play around the control of that area and have very good gamesenses. But sometimes that few extra seconds or confidence to initiate would make a gamechanging difference.
It would also be a lot harder to detect since just having a hunch is a perfectly valid reason to jump into the fog.
I'm sensing a market in dedicated hardware here. For players who spend upwards of 1000 hours on a single game it could be worth it to have a locked down machine for certain contexts where you have to prove your bona fides.
I guess I’m not as cynical. I picked up chess again when I was deployed (I was on the chess team in high school) and have been closely following this drama, and my sense is that perhaps we all admit that online chess is dead, but a return to analog chess is possible and, perhaps, preferable. I know this begs the question of defeating cheats in irl play, but shouldn’t that be an easier problem to solve?
I would be surprised of over-the-board cheating was viable in the long term. Seems like it should be relatively easy to stamp out provided there's a will to do so.
I've seen on the last thread that sometimes for cheating one bit of data is enough, something like a signal at some point where you could make a decisive move. I wonder, then, how you could send stealthily to someone a bit of data. Something like an implant under the skin giving a slight shock or vibration or heat. The answer to that would be to block signals from entering the space where the game is played, but then maybe you could implant a whole device in yourself to do that. I'm not sure how you could give it the data of the current game though.
> I really can't shake my personal feeling that chess as a sport is just simply dead, especially as an online e-sport.
This sent me musing whether we are actually witnessing the emergence of a new type of gamesmanship. If we don't end up killing each other in another global war (sadly there won't be subsequent volumes about "the war no one wanted"), then having men and machines team up in competitive environments is a given. (This is already happening, no doubt, in military settings.)
Maybe new games can be devised, or existing games modified, that can't default to games that are machine vs machine with the human teammates reduced to secretaries in the game. Games designed with AI teammate already in mind. The human role can't be just physical. I wonder what that would be like.
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.
One of the most interesting (but inconclusive) points they found was how Hans' evaluated strength dropped after they introduced the 15 minute broadcast delay.
Except, introducing the 15 minute broadcast delay just happens to coincide with Magnus scandalous withdrawal, Hans being especially carefully searched and being stared at by everyone (that is, if we trust how it was evaluated over a couple of games in the first place).
So, again, even though some statement from chess.com was expected, this one probably does more harm than good. Heavy implications, winking and chuckling, but nothing that would allow one to close the case (quite naturally). All of this historical progress evaluation was done by people a month ago already, and did achieve as much as this one. This one is "official" though, so pours another barrel of fuel into the fire.
That said, I wonder why 15 minute (or even more) delay isn't standard in events like that. Seems like the least you could do, given how questionable chess in 2022 is in the first place.
I do think all of the implications on Hans are well deserved, and the deep exploration of cheating within the chess world is well placed. And after looking through the evidence myself, I do believe that Hans is cheating in over-the-board games. I don't think any of the analysis presented on YouTube have yet met a sufficient standard of quality (statistical mistakes everywhere), but I do have faith that someone will come forward with an analysis that both doesn't make any mistakes and also firmly implicates Hans is a cheater. Time will tell.
I also agree that Hans material drop in performance after the broadcast delay could be reasonably attributed to:
1. he's 19
2. the best chess player in the world said he's a dirty cheater
Such circumstances would throw off many prodigal 19 year old's, I'm sure his emotions got in the way of his playing. (I also believe he was cheating here, but I don't think the drop in his performance meets any standard for acceptable evidence)
I didn't look at the chess.com thing since the historic progress is eclipsed by a streak of 5 important games where he simply scores 100%. Everything else is irrelevant. GM's usually have zero games at that level. Dropping to his usual level after that doesn't average it out to human level. lol
Someone should organize that running joke: naked chess tournament.
There's a video of Hikaru Nakamura looking at some of the games identified by this report, and he was admitting that it wasn't immediately obvious to him: the games looked like good games (although I didn't watch the whole video).
So such a test of the system almost feels necessary for the chess world to see if they really can tell the difference between high level classical OTB play and cheating. It would be interesting indeed if ever the consensus was that a fair player was the cheater and the actual cheater just looked like "good play"!
Another player decreased by 10 vs 15 for hans. Other players increased by a similar amount to hans. It could easily be noise. Hans under the most public scrutiny and criticism of his life, so you would expect his play to suffer.
I kind of suspect magnus misplayed because he knew that Hans cheated so much online.
Does engine correlation actually prove anything though? Some of the 'statistical analysis' that has been posted on twitter regarding it in the last week has been against hundreds of engines, so 'engine correlation' seems to mean "the move made matched against at least one engine that would have made that move" I think?
1) You can look at the "strength" of individual moves. Someone who plays at 2000-level normally but magically coughs up 2600-level moves when in trouble is probably cheating (watch some of the live chess streamers--you'll regularly see this in real time). Computers are quite good at estimating the strength of a move after the fact.
2) Quite often there are certain "play lines" that computers will play that humans simply can't find over the board.
For example, a computer can take a defensive "play line" that is littered with traps with only a single non-losing path for 30+ moves and work it out really quickly (there is only one non-losing path to take so it prunes the search space mega fast) and play it perfectly. A human playing such a line is almost always cheating--humans simply can't run those kinds of lines in real time.
If you look at computers analyzing even the highest end games, you see the humans making quite a few mistakes that the computers will spot and take advantage of immediately. Someone who walks down these kinds of paths regularly is a statistical anomaly.
That having been said, given the current crop of computer-trained chess kids, it IS possible that we'll grow a prodigy that can run those kinds of lines. However, it doesn't seem like that person exists, yet.
I'm haunted by the possibility that humans might (at least half) catch up, too. When I look at how AI beats humans, I can't help thinking that AI shows us that human narcissism holds humans back. We don't want to look stupid, or mediocre - we don't want to make moves that are hard to explain the value of clearly.
In Go, we can't make ourselves spread our moves around the board as much as we should, we tend not to choose a maybe good move elsewhere over a clearly powerful move where the board is developed, for example.
Maybe there's a pattern to the moves AI chooses that is also a pattern humans can see without running every line; we're just reluctant to choose moves that we can't clearly justify in the shorter run.
He's just referring to the fact that after game is over you can let your program stew on any given move for a weekend or more before reporting back on how strong it was or wasn't.
But standard programs like Stockfish won’t tell you how strong a move was. They’ll just tell you how much it changes the evaluation.
E.g. if you initiate a queen trade in a straightforward position, on the next move I have to take back my queen; any other move will show a gigantic evaluation drop by the engine. But that doesn’t mean it’s a particular strong move — even an absolute beginner will play it. Thus it’s of no value for determining whether the person who played it is cheating.
It’s entirely possible that chess.com has access to more sophisticated software that can estimate the strength of players (they sort of allude to this with their “strength score” metric) but AFAIK it’s not publicly available and not clear how it works, or whether it can evaluate individual moves as opposed to the game as a whole.
Am not a statistician, but at least in an online analysis I saw, seems like correlation can effectively identify players who are playing too much like a computer. Because they don't just run correlations on Niemann, but on all the top players, and do comparisons (and for certain long stretches of tournaments, Niemann's is playing way, way above how anyone else has ever played).
This is video explains it pretty well, and seems like a very compelling argument (at least to me): https://www.youtube.com/watch?v=qjtbXxA8Fcc - and just know that the woman talking is a bit hard to understand because of her accent.
... oh, and to address your point about 100's of engines, my first thought was that are only a handful that everyone uses (Stockfish?) (and also, just guessing, but I get the impression that most top engines recommend similar moves, but again, just a guess!).
Ah, have to admit, this is a very good counter argument, calling chessbase's methods in to question. And it's surprising that chessbase does not always do the same analysis for each game, that its nodes aren't setup with the same set of chess engines (although, again, maybe most top chess engines suggest similar moves??). Hmm...
... also, not sure I agree with his opinion that for each move that is analyzed, LetsCheck will return 100% if any engine returns 100% (and there could be multiple computers that were used, each with different engines). The point of the analysis is to determine if a player is playing like a computer, and the user may himself have multiple chess engines open in order to confuse the cheat detection. But again, am not an expert at chess engines or statistics, so am not sure what effects checking multiple engines has...
... also, he says that "Ken Regan's scientifically valid method has exonerated Hans by saying his results do not show any statistically valid evidence of cheating." This is very confusing, because Chess.com post has basically said the opposite (maybe Ken Regan's analysis is referring to a different subset of games?). Guess nothing is definitive. But at this point, I still lean towards Hans cheating (on top of this analysis, there is also a lot of circumstantial things he did that to me indicate he might have cheated, which is too long a topic to go into).
Chess.com explicitly states in the report that that sort of methodology does “not meet our standard” for cheating detection. If they don’t feel comfortable using it, I certainly don’t.
> Does engine correlation actually prove anything though?
It probably doesn't, and for many reasons, both because the more engines you add the greater the chance of falsely accusing someone (so an analysis that features hundreds engines is probably worthless), but worse than that, you can manipulate the result of the analysis through the selection of engines
I don't think engine correlation necessarily proves anything, on its own. It's worth remembering, though, that chess.com's report a) presents more than merely raw engine correlation, and b) its correlations do not seem to match against hundreds of engines.
But even with all the evidence presented, "proof" is a tricky thing. To what standard would we be trying to prove a claim?
Does this report prove beyond all reasonable doubt that Niemann cheated? I'd say no, but others may disagree.
How about to a preponderance of evidence? Perhaps. But even that is hard to say when no one has yet presented a rigorous defense or set of counterpoints.
In any case, my post wasn't meant to say that Niemann cheated per se. I have no idea, and chess.com themselves may not be able to actually prove whether he did. But I found the report interesting, even beyond the current issue surrounding Niemann and speaking to potential cheating in high-level chess more broadly, and if you re-read my post, I tried not to state anything definitive about whether Hans actually cheated or not.
Not anymore than elevated testosterone levels are "proof" of performance enhancing drugs. Engine correlation is a marker and when combined with other markers, can be meaningful.
What it mostly shows is that Hans move strength is unnatural.
Further evidence to support this is that he often plays bad moves. That is, moves that are considered blunders, with a high frequency. This is either an attempt to cover up the engine moves or representative of his actual capability. For instance, the report mentions that in a post-game analysis he suggested a move that would be an obvious blunder. When the interviewer pointed this out, Hans wasn't fully convinced until he was shown the engine analysis. So he also is showing a habit of deferring to what the engine suggests.
> The fact that online cheating is so widespread even among top chess players is certainly news to many, including me.
A few weeks ago my daughter very proudly announced that she managed to "outsmart" (=hack) her online test app used by her class. I was shocked and asked why she would do that, she's smart enough to get an A without that. She seemed obviously puzzled by my question and my lack of enthusiasm about her "achievement" she was so proud of (it involved some JS modifications). I guess is it's just another kind of thrill.
> A few weeks ago my daughter very proudly announced that she managed to "outsmart" (=hack) her online test app used by her class. I was shocked and asked why she would do that, she's smart enough to get an A without that. She seemed obviously puzzled by my question and my lack of enthusiasm about her "achievement" she was so proud of (it involved some JS modifications).
I wonder if maybe it was the chance to solve a different challenge than the one before her (the test), which if she could already get an A maybe wasn't the right challenge?
She should be proud in the sense that real world problem solving and valuable technical skills is a lot more important than any school test. That said she will have to learn not to let arrogance and risk-taking be her downfall.
> two dozen chess Grandmasters who have admitted to cheating on chess.com. The fact that online cheating is so widespread
chess.com globally has more than 93M members. There is cheating, but a dozen admissions can't be accurately described as "so widespread." Research has shown fewer than 0.02% cheat. While this is mildly shocking, your statement based on 12 grandmaster admissions is a sweeping generalization.
This is how many grandmasters admitted cheating, not how many players globally have admitted it.
As of 2021 there were 1,315 active grandmasters. [0]
24/1,315 = ~1.8% of grandmasters admit cheating (give or take a few depending on how many play online chess or are still active). In my opinion, that is a serious problem.
Even with my careless error, the statement in question is still a sweeping generalization.
> As of 2021
This is a tough number to track down, and you've limited the count by "active," in 2021, but the number is rapidly increasing. According to the FIFA Database as of a few moments ago, there are 1771 chess grandmasters.[1]
24/1771 < 1.4%
> In my opinion, that is a serious problem.
Even assuming 1.8% of chess grandmasters are cheaters, this means that 98.2% of them are not. If you tested a 98.2% of a perfect score on a test, would you really think your grade was a serious problem? If you had the chance to retake the test for a replacement score, either better or worse, would you?
> That's how many grandmasters there are total (1771). I was quoting the number active, hence the hedge of give or take current active / cheating.
The difference between 1.8% and 1.4% is negligible and is only in regards to the limited population of chess grandmasters which could not be a valid sample representation of 93M chess.com members.
> Yes, I think cheating more than a fraction of a percent as you originally posited is a detriment to competition.
Your answer is apparently in reply to some question that was not asked. On the contrary, my claim was that the argument that cheating was "widespread," by extrapolating a mere two dozen cheaters among 93M, is fallacious reasoning, specifically a sweeping generalization, and also that
>>> Research has shown fewer than 0.02% cheat.
which is not a postulation but a published fact.[1] Being that two hundredths of a percent may be described as a tiny fraction of a percent rather than more than a fraction of a percent, you can clearly see in this case, by your own scrutiny and straw man, cheating is not a detriment to competition.
The one thing that baffles me with SEO is how it's just guesswork yet is a massive business. It's like promising to someone that you'll get their name listed sooner in the phonebook without any control over the phonebook itself, and then that person pays you to do it.
It's no different, since we have no control over Google's ranking mechanism and they won't explicitly tell you what the algorithm is (and change the rules daily), so it's just guesswork.
Baffling an entire colossal "industry" is built from guesswork.
They say they estimate that less than 0.14% of players on chess.com ever cheat, so it may be less than everyone assumes just due to this current drama.
They also claim 100 million subscribers, so that's 140,000 cheats. Since if you're cheating you probably do quite well (I pity the engine that could lose to me!), that's likely most of the high-rated players on their site.
And actually this tallies quite well with my experience of amateur physical sport, where at the lower levels everyone's honest and sporting and it's all great fun, but as soon as you get to the point where people are putting their heart and soul into the game, everyone's cheating, everyone knows everyone else is cheating, and anyone who's not cheating might as well give up and go home, as far as winning things is concerned.
Once you put money on it, and you're playing against people you're never likely to meet in real life, Jesus I can't imagine what it's like.
This number seems wildly low to me. I've been playing on chess.com for 2 years and have run into ~30 cheaters. I know they were cheating because chess.com told me they were. I've suspected a handful more of cheating who may not be as they were never banned. Either way this number doesn't seem correct.
> I find the decsription "lied outright" unfitting
Hmm.. if time allows in the next hour or so, I will find his exact quotes from the interview and will definitely edit this out if it is unfair. But my recollection is that Hans stated explicitly that he had never cheated in a serious tournament since he was 12, that he only cheated thereafter in "unrated" games, that he only cheated twice overall, and that he hadn't cheated at all in the ~2.5 years (since he was 16) preceding his game against Carlsen in Sinquefield this year. From the report, chess.com is saying any or all of those statements are outright lies.
But it's good to call this out and it sucks to unfairly malign anyone so I'll follow-up if other obligations allow me to scrub through the post-game interviews videos quickly.
> Expecting a perfect verbal 72 page report from him under these circumstances fells pretty cruel.
To be equally fair, 'a perfect verbal 72 page report from him' is not at all what I stated or implied as a standard, and saying so is a strawman.
I finally got a chance to look this up, and I think saying that Hans "lied outright" is fair. (Obviously on the assumption that you believe chess.com's analysis of his play on the site, and their claims of Hans' admissions to cheating there.)
I listened to the entirety of Hans' interview again, and also read a transcript thereof. But I could've saved some time because the chess.com report itself quotes (and paraphrases, separately) Hans' statements about his play there, and calls Hans' statements false outright.[1,2,3]
Here is a selection of Hans' direct quotes from the post-Sinquefield interview which are outright falsehoods according to chess.com:[4]
"I have never, ever in my life cheated in an over the board game, or in an online tournament. They were in unrated games."
"Other than when I was 12 years old, I have never, ever, ever--and I would never do that, that is the worst thing I could ever do--cheat in a tournament with prize money."
"Never when I was streaming did I cheat."
"I did this when I was 12 years old. And then when I, and then the second, the other times I did it, it was not even in an over the board tournament, it was not even a prize money online tournament. It was in absolutely random games."
According not only to chess.com's analysis and the evidence they present, but apparently to Niemann's own admissions to them, each of these statements is outright false on their own. (And if they are false, then together they also grossly misrepresent the overall picture of his behavior.) It appears Hans did in fact cheat in rated games, and in cash prize games, as well as cheating while streaming, and while playing against highly rated players in "real" games.
1: "Consistent with the letter we sent Hans privately on September 8, 2022, we are prepared to show within this report that he, in fact, appears to have cheated against multiple opponents in Chess.com prize events
(beyond the Titled Tuesday event that Hans admitted to having cheated in when he was 12), Speed Chess Championship Qualifiers, and the PRO Chess League. We also have evidence that he appears to have
cheated in sets of rated games on Chess.com against highly-rated, well-known figures in the chess
community, some of which he streamed online. These findings contradict Hans’ public statements.
In particular, in interviews given during the 2022 Sinquefield Cup, Hans made several comments to the
press about alleged instances of prior cheating
:
• “Other than when I was 12 years old, I have never, ever, ever – and I would never do that, that is
the worst thing that I could ever do – cheat in a tournament with prize money.”
• “Never when I was streaming did I cheat.”
• “Keep in mind I was 16 years old, I never wanted to hurt anyone, these were random games. I
would never – could even fathom d
doing it – in a real game.”" -Page 4 of the chess.com report on Neimann.
2:"If you are willing to correct the false statements you made about having never cheated when it mattered (now that you have said these untruths publicly), acknowledge the full breadth of the above violations, and cooperate with us to compete under strict Fair Play measures, Chess.com would be happy to consider bringing you back to our events. In fact, I think it would be a wonderful redemption story for the full truth to come out, for the chess world to see this and acknowledge your talent regardless of your past, and give the community what they deserve: The truth." -Page 58 of the chess.com report on Neimann.
3: "In your
interview you mentioned (paraphrased) that you “cheated when you were 12” and then “later when
you were 16 in an unrated game”. This directly contradicts our statistical evidence, as well as the conversation you and I had in our private call when you confessed to cheating, and there is written
evidence from you that substantially corroborates this. You also contradicted your own statement that
you had only cheated in unrated games in the interview by later stating that you did it to gain rating
points, which obviously indicates cheating in rated games." -Page 57 of the chess.com report on Neimann.
Hi, welcome to HN. Speaking for all of us long-term users, I think we would very much enjoy that 72-page report that you've offered to compile. If you post a submission linking to it, let us know. (Especially interesting would be a meta-analysis of the types of lying and so on!)
It's a good point, I do also think that people (especially on social media) are too quick these days to jump all over any minor mistake a person makes. But (in my humble opinion), if the report is correct, then Niemann went waay too far, beyond a little lying and cheating to possibly raising himself at the highest levels of chess (and winning actual money). Yeah, as you probably saw, it says he cheated online an estimated 100 times, some for actual prize events (amongst other implications).
Yeah, personally, I do believe the report as it does seem to match a lot of other online analysis of his game history (this isn't the greatest video, but it is compelling: https://www.youtube.com/watch?v=qjtbXxA8Fcc).
And sure, it does seem cruel to call him out in an interview, but so is cheating to get to the top. Yeah, someone who cheated on this scale has robbed those players of the recognition and prize money they worked so hard to deserve. Honest players who have probably devoted way more time than Niemann had (the two years of study you mentioned is impressive, but maybe others have studied like this for decades??)
> He seemed a bit on the spectrum and oversharing, literally hyperventilating and not talking in coherent sentences.
Yeah, not really. He's 19 and is a non-native English speaker. He understands that he's under siege and has to show some vulnerability to gain trust in listeners. IMO, the way he spoke was completely in accordance with all of this, while also remaining coherent.
He’s born in San Francisco, very classic white American native. While he did spend ages 7-10 in the Netherlands, he went to high school in Connecticut and New York. His uncharacteristic accent is from spending the last few years studying chess in Europe.
I can't speak for SV, but I think a 72-page report from every confessed past fraudster that's engaged in what seems to be fraud-adjacent activity would do wonders in clearing the crap out of the crypto scam space.
It's chess, it's not heart surgery exams. The cheating sucks, but anti-cheat measures are infinitely worse. There is no better way to suck the fun out of something than demanding you submit to a patdown and metal scan or a rootkit before you can participate.
It's just soccer, bicycle racing, weight lifting, sprinting, marathon running, ..., no need to submit to a urine and blood test, if you want to compete in the olympics. That sucks the fun out of it. /s
With respect, in those examples the extreme amount of money to be made with the first place motivates the cheating. The anti-cheat for such competitions only barely works, as top competitors do as many (possibly dangerous) things they can to come out on top. As long as they are undetectable. We should approach a disincentive to cheating in a different way, so it is just a few lemons as it is in chess.
>The fact that online cheating is so widespread even among top chess players is certainly news to many, including me.
Not to me. When the stakes of winning are so damn valuable, sometimes literally so, it would be patently stupid to not cheat and increase your odds of winning.
Remember: Cheating is only a problem if you're caught. If you're never caught, cheating is completely legitimate.
These aren't games of pleasure, played for the sake of playing chess. These are games played for the sake of winning. Prize money, ELO, fame and renown, etc. When the only thing that matters is winning, you gain nothing by playing honestly.
If your attitude about chess competition was applied toward you in all of your routine irl activities, you would be extremely unhappy.
The cop does better in his job by meeting his quota of speeding tickets, so it objectively makes sense that he would issue you a speeding ticket every day with no evidence and over your strenuous objections... except you wouldn't object because the cop is working the system rationally. Yeah, right. Anyway, the lawyer you hired to work through this issue in court would of course inflate the number of hours he worked on your case, because why wouldn't he, he'd be an idiot not to!
Can't wait for you to apply for you YC investment, you make a great business partner, so rational and all.
He didn't say it's a good thing, nor that he approves, just that it is unsurprising that people (attempt to, and sometimes) cheat because cheating is a rational response.
It's just a different way of saying that when the measure of something becomes the target, then that measurement is gamed.
The sad fact of reality is honest people finish dead last in the race known as life. Anyone who succeeds in life has cheated in some form or another, unless their life was a miraculous chain of stars aligning one after another.
Human society and life isn't one of cooperation and harmony, it's a competition and we're all striving to one up each other to varying degrees. If that involves literally kicking another guy off the ladder, so be it. If a cop is feeling particularly irate and/or is hard pressed to satisfy a quota with frivolous speeding tickets, so be it. Contractors inflating their hours is par for the course.
Life ain't all sunshine and flowers, it's also dirty and ruthless. Nobody's a saint in this plane of existence.
I'm just happy to say that most people that I call my friends don't share your vision of what it means to have a life well-lived, or how they define success, or have lived with this attitude of dirt and ruthlessness being necessary.
This is sociopathic rationalization for poor behavior. It isn’t true. If you are a compulsive liar and cheater, perhaps you will justify your immoral behavior by deluding yourself into believing that everyone else is doing it too- but it is not so. We are not all behaving this way.
You can claim "Well, we are/I am honest!" all you want, but the reality is nobody is a saint. Nobody. Full stop. If someone says they're a saint, they are lying out their ass. We have all cheated at something at some point in our lives, whether we got caught or not.
You also fail to understand I'm not passing judgment on such behaviour. Rather, I'm saying we need to take into account the ruthless nature of reality if we don't want to get screwed over ourselves.
Calling evidence of Neiman cheating a surprise is an honest and naive understanding of reality, one which leaves you wide open to cheating thrown your way that you will never realize because you never expect it.
Nobody is a saint, much less when achieving victory is so heavily incentivized as in professional games.
> You can claim "Well, we are/I am honest!" all you want, but the reality is nobody is a saint. Nobody. Full stop. If someone says they're a saint, they are lying out their ass. We have all cheated at something at some point in our lives, whether we got caught or not.
That's a total non-sequitur. Perhaps no-one is perfect (though frankly I doubt it), but that doesn't mean that cheating more means more success than cheating less.
Cheats definitely exist. But many of them end up doing worse than people who live more honestly, as we see here.
Cheating might work if you have nothing to lose. Example: you are supposed to be present in an online lecture as a formality, there is no interaction and nothing to gain by being in it. Sure, people are likely to join, mute and do other things.
However, for serious real life situations where stakes are high, risk of ruin is too high. This is not rational. In fact, those stars have to align for the cheater, not the honest ones.
(I still don't think Hans cheated OTB and chess.com is being ridiculous)
A game is defined by its very specific set of rules. This can easily be seen in chess, which has a huge variety of variants. Or how one, minor rule change in baseball and you've got people wondering if the Hall of Fame even make sense anymore.
Once a set of rules are defined, competitions and tournaments within those rules are meant to determine the best people in that game.
When a player cheats, and rationalizes it as "only matters if you get caught", that means the cheater is playing by different sets of rules from his opponent. They are playing literally different games. If the two opponents are playing different games, it completely undermines the purpose of the competition.
Here's the thing though: How do you know a player cheated? Because you caught him. So what about all the cheaters who are never caught? You can't say they're cheaters.
Therefore, if a player cheats but is never caught, he isn't a cheater. If a player isn't a cheater, whatever he did is legitimate and it isn't a problem.
Therefore: Cheating is only a problem if you're caught.
Put another way, the rules only tell you there will be consequences if you are caught breaking them. The rules do not and cannot police the act of cheating itself, only the results of such actions and only if the act of cheating becomes known.
So the optimal way to play a game to win is always to play skillfully /and also/ cheat without getting caught. Not cheating is not optimal to winning. It is absolutely a high risk, high return course of action, but that's the nature of optimizing.
If all you care about is the prize money, then there are easier ways to rob people, e.g. at gunpoint on their way out of the competition venue. The Rules say "If Person X does Y, then Person X gets Z". You want to rationalize interrupting that process to take Z for yourself. Y could be "their job" and Z could be "their salary". I suppose, in your terms, highway robbery is also a "high risk, high return course of action".
But entering into the competition under the guise of legitimate play means you also care about the prestige of being declared the winner. But you haven't won the game. You've played a completely different game. You can claim to have played the same game, but that would be a lie. You could just as easily redefine "narcissistic sociopath" in your own head to exclude adhering to The Rules. But it'd still be a lie.
I see in your other comments in this thread that you claim "most people finish dead last in life", intimating that they are fools for not cheating. But that's the thing: if everyone cheated, it would be a complete breakdown of society. Rules, laws, and systems to enforce them, are how we can leave our houses in the morning without wearing bulletproof armor from head to toe.
But what do I know? I only have a happy life with my wife and two kids. I'm a loser. So go! Cheat at science and get that Nobel Prize money! Cheat at keeping your restaurant clean and pocket the savings on soap! Never mind the wasted time and resources and lives ruined. You're a winner.
I cannot agree. First, your position comes with an assumption that one feels okay when cheating. Many people simply do not, for personal integrity, anxiety and other reasons.
Also, if you expect a game to bring profits in a long run, risk assessment may become too complex, if you can be caught retrospectively (like that top-score trackmania guy for example). Some people understand it from the beginning and decide to not put themselves into an unstable trap for life, even if cheating is compatible with their values.
You’re right that games (businesses, situations) themselves may have a long-term model which includes cheating as an optimal solution. But assuming that people playing it are all cheating sociopaths is incorrect. Tbh, your persistence all over this subthread is a sign that you have some sort of a close personal relationship with this topic.
Legitimate would mean condoning cheating. Which then means everyone should be allowed to use engines, which then would mean engine vs engine, which we already have.
I commented on this already in another post, but I'll reiterate here: How do you know a player is a cheater?
If you can't prove a player cheated, that is if you never catch the cheater, then whatever the player did is legitimate. It doesn't matter if the player in fact cheated and got away with it, it's all legitimate if he wasn't caught.
The rules do not and cannot police the act of cheating itself, only the results and only if the cheater is caught.
I'm arguing that it is in one's interest to not hold naive ideals about the world, about reality. Reality is a dirty, ruthless place. If you don't conduct yourself being aware that other people are going to take advantage of you, or even cheat on you, then you will end up getting screwed because you cannot defend against attacks you aren't aware of nor expect.
If you made your statement that "the world is a dirty place" I think most would agree with you. The point almost every reply is taking issue with is when you say "cheating is legitimate".
The replies, including my own, are saying "cheating is not acceptable". By originally saying "cheating is legitimate" you are saying "cheating is acceptable", but in your follow up arguments it's "cheating moves that are not caught are considered legitimate", which is the definition of legitimate and thusly there's no substance.
You seem to define "legitimate" as depending on society's perception. As another thought experiment, what if someone didn't cheat, but society erroneously came to the conclusion that the person did cheat. Is what that player did legitimate?
I guess it comes down to: does objective truth exist, or is everything relative to society's beliefs?
>You seem to define "legitimate" as depending on society's perception.
For the purposes of this discussion, I define "legitimate" as whether anyone wants to cry foul on a player's performance. If a player cheats but doesn't get caught, it's legitimate because nobody can cry foul and make it into a problem.
>As another thought experiment, what if someone didn't cheat, but society erroneously came to the conclusion that the person did cheat. Is what that player did legitimate?
Nope.
>I guess it comes down to: does objective truth exist, or is everything relative to society's beliefs?
If we were to drill down to it, it can be argued humans can never truly be objective. Take Neiman for example; so far we haven't seen any evidence (that I'm aware of) that Neiman did cheat in his game with Magnus without any doubt, but it appears that society at large has decided that he was a cheater in that game.
This and the aforementioned thought experiment actually also come back to one of my original arguments: That not cheating is stupid. When you can be judged a cheater despite being honest, or judged honest despite being a cheater, it is much more beneficial to just cheat and attempt to reap the rewards of cheating. Being honest gains you nothing while burdening the risk of losing everything, while cheating potentially gains you more at the risk of losing everything.
Also, I hope you don't misunderstand any of this as me passing judgment one way or another on this sort of behaviour. Personally, I would prefer no cheating, but there are my own thoughts and then there is reality, and reality is as filled with sunshine and flowers as it is dirty and ruthless. I see all this as a great opportunity to appreciate reality for what it truly is, so we hopefully don't end up on the short end of a stick insofar as what's within our powers to affect.
Hans cheated in over 100 online chess matches, and now he's stuck in this giant controversy. If he didn't cheat in those matches, none of this controversy would have happened, and he would be judged based on his skill.
I would argue his failing wasn't that he cheated, but that he got caught by cheating too obviously against someone who wasn't going to take it lying down.
The people who succeed cheat only as far as to reap the rewards while avoiding the risks thereof, they pick the time and place to cheat and not cheat. Of course, we don't know as third-parties whether any such successful individuals are in fact cheaters or not since nobody caught them.
To put it another way, if Neiman cheated in over 100 online chess matches and nobody caught him, we wouldn't know he's a cheater and we would judge him based off his apparent "skill". Thus, cheating is only a problem if you're caught.
I would say there are 4 main motivations to not cheat:
1. Getting caught and problems arising from that
As you stated, there is a non-zero chance that someone who cheated will get away with it. And there is a non-zero chance that someone who didn't cheat will be wrongly widely accused of cheating. But someone who cheats has a much much higher chance of being widely accused of cheating that someone who doesn't. Cheating purely increases your chance of being widely accused of cheating.
You as a single individual cannot know all the various analyses that will be done on your games in the future (whether analysis of your moves vs engine moves or analysis of video or data recordings from websites you're on). You cannot foresee all the different possible ways you might be caught. The better you get, the more analysis will be done on your past games. Every past cheating instance is a liability for the future.
And if you rely on the cheating to get to a high level, you'll be forced to continue doing it to retain that level. You can't foresee the various anti-cheat mechanisms that will be put in place in the future that will catch your future cheating.
Once you're caught, it's basically disaster for you. Even when you don't cheat in the future, people will still suspect that you're cheating.
2. A feeling of accomplishment
People are often motivated by a feeling of accomplishment. If you're cheating to get ahead, you're not accomplishing as much, so you won't have as much of a feeling of accomplishment. If you use a cheat code to beat a videogame you don't feel much accomplishment. You won't feel like the best person in the world at chess if you only won due to a computer feeding you moves.
3. Enjoyment of a better world
You said you'd personally prefer a world without cheating. By not cheating you would be helping create that world. By cheating you would be helping to create a cheatful world. You're going to have a hard time advocating for a cheat-free world if you yourself are cheating due to the conflict of interest. E.g. you won't be able to advocate for installation of anti-cheat protections because if they get installed, they'll catch you. Same with advocating for stronger statistical analyses.
If you cheat, you're disrespecting the people around you, both your competitors and the audience. When you're consistently disrespecting everyone around you, I would expect that to have a negative impact on your ability to make friends with them. It could negatively affect your mood and your outlook on life. Overall I expect your life to be less enjoyable.
4. Morality
Is there objective morality? Do we have a moral obligation to respect other people? What is the meaning of life?
I say yes there is objective morality. Yes we have a moral obligation to respect other people, it's one of the components of the purpose of our lives.
I don't agree with blanket statements like that but I think there's a kernel of truth in your mindset that many people are oblivious to.
Chess, as a competitive game with meaningful stakes, is obsolete. It became obsolete when the first computer outplayed the reigning human champion, and it's been gradually becoming more obvious. Recent events have started to demonstrate precisely why: people are becoming better at cheating at chess way faster than they're getting better at playing chess.
All of those things you can get by cheating. But cheating is never winning. The whole point of a game or sport is that it ceases to be a game when you go outside the rules.
Yes, many can square this circle in their own heads and convince themselves that anything goes. Some, like Lance Armstrong, will enrich themselves enormously. But none of them will have “won”.
The question is one of incentives and motivations.
Am I playing chess because I love chess? To have fun? In that case, cheating is counter to my goals (unless I enjoy the act of cheating itelf, but that's getting besides the point).
On the other hand, is chess simply a means to an end? Am I playing chess because I want to win something? If so, cheating is but one of many possible decisions I can undertake to increase my odds of winning.
Where professional games are concerned, where money and fame are at stake if you lose, cheating is downright inevitable and certainly not something to be surprised about.
I don't disagree at all. But there are people who just want to play the game by the rules. There are people who won't cheat for moral reasons. There are people who lack competence to cheat. There are people who are risk averse. Cheating should be expected, but you were implying every single person has more incentive to cheat than not, which just isn't true. You could even argue most people are in fact cheating and I might even agree with you, but without evidence it's just slandering, something even the world champion should be reprehended on.
>but you were implying everyone has more incentive to cheat than not,
I am implying that because that's inevitable in a situation where winning is so much more important than simply playing. Again, these aren't games of pleasure, games "for fun". These are games with valuable stakes on the table.
Even Magnus himself can be argued as playing to win in order to eventually obtain the famed 2800 ELO or whatever the magic threshold number is. If evidence comes out that Magnus cheats, it would be a shame but I wouldn't be surprised either.
In situations where the ends justify the means, the means will include cheating.
A key fact to understand in thinking about cheating in over the board chess: a strong player can defeat a much stronger opponent with just 1-3 hints per game indicating the strongest move. For example, most chess experts agree that a ~2600 rated player with 2-3 hints at key moments per game would be expected to beat a ~2800 rated player. Many people might assume that a cheater needs guidance every move, thereby requiring a potentially more obvious cheating mechanism. That is not the case.
Also, clever cheating devices have been found in over the board chess competitions. So, this is possible. Moreover, one needn't carry a device on themselves. A cheater may have accomplices providing hints, if they carry a device.
It will be interesting to see how chess tournaments, as well as FIDE, chess.com, and other major chess institutions react to this situation. The potential for cheating has now been brought to the absolute forefront of chess discussion. And Carlsen's actions have been questioned by FIDE in recent interviews, with FIDE staff condemning "vigilantism" of a kind.
Some set of resolutions seems necessary--perhaps standards for security in major chess tournaments, perhaps an alliance to share cheating or reliability data amongst major chess operations, perhaps a standard term in major chess tournament agreements that no previously identified cheaters (online or otherwise) will be allowed to play, and perhaps sanctions in some form against Carlsen (or Niemann, if concrete evidence against him emerges).
> Also, clever cheating devices have been found in over the board chess competitions
The most convincing candidate for such a device I've seen is an Illuminati Thumper Pro hidden in Hans's shoe: https://illuminati-magic.com/products/thumper. If you watch the footage of him getting scanned before his match with Alireza (and, crucially, before Magnus announced he was dropping), there are a couple of subtle things that are consistent with this theory: https://www.youtube.com/watch?v=PIulWkTHuu0:
1. He swallows (seemingly nervously) at 4 seconds shortly before his left shoe is scanned.
2. As the left shoe is being scanned, there are 2 beeping noises that at least to me, sound like they are coming from the wand, but are seemingly ignored by the wander. The same beeps do not repeat when his right shoe is scanned (that is, it's not just metal parts built into the shoes themselves). Two caveats to this part: First, I've heard differing opinions on whether the thumper will trigger metal detectors. Second, it's possible (even if unlikely IMO) that those beeps are not from that wand and it's just a coincidence that another wand or object beeped - since we can't see the wand in frame.
3. At 1:17 he starts nervously fidgeting with the credit card as the RF scanner gets close to his left foot and noticeably slows down when the scanner switches from his left foot to the right foot, and appears to stop completely as soon as the scanner is moving up away from his right foot. The RF scanner, to my understanding, would only detect devices that are actively transmitting which the thumper shouldn't need to do at all if Hans were using purely to receive engine moves/hints during the but the fact that it theoretically could transmit would explain why he'd be nervous about getting scanned anyway.
Of course none of these observations are proof but they sure look suspicious to me.
You don't need something that transmits if you're searching for bug-like devices or any general integrated circuits with a nonlinear junction detector:
I am very far removed from anything related to Chess, but if they want to get serious about this they should hire people who specialize in the federal-contracting adjacent field of TSCM (technical surveillance countermeasures).
I also think that people putting a lot of focus into shoes or other clothing articles underestimate the motivation and capability of people to use the traditional "prison wallet" method of concealing things.
The wiki said that it doesn't work against shielded electronics, however who knows how accurate that actually is. I really enjoyed the anecdote about the US embassy in Moscow having diodes embedded in the cement throughout to make finding actual bugs much harder.
With that nest of hair I wonder maybe he disassembles the device and puts small parts in various locations, the most detectable in the shoe or "the pocket".
Someone familiar with slight of hand could comfortably scratch here and there while dropping pieces in a 'build' pocket.
You've seen the film The Man with the Golden Gun ? A cufflink here, a pen there, that pack of chewing gum... assembled together could be a cheating device.
I love it. The idea that someone would undergo training as a magician in order to cheat at chess is just hilarious! It’s not totally absurd though, given the history of cheating (as well as espionage) in sport.
The tricks people have pulled to cheat in baseball and (NFL) football are similarly amusing!
Doesn’t seem absurd at all to me. There might not be as much money in chess cheating as other scams but someone could be motivated to just become known as one of the best chess players.
>I also think that people putting a lot of focus into shoes or other clothing articles underestimate the motivation and capability of people to use the traditional "prison wallet" method of concealing things.
I felt silly for even thinking this, but seeing as you've mentioned it. It would be so hilarious if true considering he has offered to play naked[1] to prove his innocence!
You miss the vitally important point: they weren't needed in the game in question, so we don't even get to the point where fanciful theorising is relevant.
Magnus didn't play particularly well and Hans played ok. This was not an example of a superhuman intelligence passing hints to overcome Magnus at his best.
Understated part of Magnus’ play is that he may have been playing a worse line that should have pushed Hans out of theory, but apparently didn’t.
I don’t totally buy Hans didn’t prepare the weird line, but it’s worth calling out; it’s at least marginally possible that Magnus out himself in an unwinnable position on purpose, but couldn’t convert it.
> Magnus didn't play particularly well and Hans played ok.
Everyone says this, but do you really know? Those statements are after-the-fact observations of engine evaluations. They don't speak to the amount of mental work that Hans would have had to put in to play optimally (or 'ok' as you say) in those positions.
You might find yourself making the same remarks when looking at the post-game analysis of any top player against an engine. Everyone crumples eventually against perfect play.
Yah, even to club level players, Magnus played a bad game. Hans had nowhere close to perfect play ( I think the stock fish analysis says it’s close to 70% best moves, which is equivalent to Hans rating ). As a club level player, it blows my mind how many people are siding with Magnus.
: edit, typo
> They don't speak to the amount of mental work that Hans would have had to put in to play optimally (or 'ok' as you say) in those positions.
Yes, they do. When Magnus makes poor choice - not giving himself an advantage or playing moves giving black an advantage - it makes it easier for black. That’s the whole point.
No, that is not how that works. An engine evaluation saying that a position is better for black does not mean that it is easier for a human to play that position. Easy to play and winning for an engine are orthogonal concepts.
Putting your opponent in positions that are better according to the engine but only with engine-like perfect play is a strategy at the highest levels of chess. Because the move is objectively worse, it won't be played, because it's not played, your opponent won't know it, because they don't know it, you'll play it better, then you win.
Even better, in that video he has a pack of gum that sets off the sensor, the security guard takes it, finishes the scan, and then gives it back! Obviously not proof of cheating but how hard would it have been to hide a device in that pack of gum?
is there any thinking on how many bits of information do you need to cheat, and how many can be communicated via thumper?
e.g. is the bit of information "move the knight" aka theres only about 4 bits of info, or is it "move the knight to E6" which is a good deal more bits, that could be lossy/error prone.
just on the surface of it, i dont see how this thing could give enough info but i suppose with a loooot of training you could improve the info transfer rate?
Here's a relevant quote from Magnus regarding cheating:
"I would just needed to cheat one or two times during a match, and I would not even need to be given moves, just the answer on which move was way better, or here there is a possibility of winning and here you need to be more careful. That is all I would need in order to be almost invincible."
Even just 1 bit - an indication to be careful - would be enough to boost the strength of a GM. An accomplice coughing in the background to let you know there's something to watch for. For a strong player - and there's no doubt that Niemann is a strong player, the question is just how strong - that's all they need to avoid making mistakes. GMs can solve insanely hard puzzles, because they know it's a puzzle and has a specific solution. Same thing with 1 bit of info.
Of course, realistically they could simply use Morse code instead of "bits" and transmit two squares (just 4 Morse "letters").
yes but against magnus, who is supposedly two levels above Hans, this is not just a one move cheat, he'd have to cheat + have a continuous absence of mistakes, which is an awful lot of information to transmit.
i dont have a horse in this race i just like thinking about things in terms of information theory since this is a remarkable applied case
another way to decide this - have them play blitz (where the moves are way too fast for info transmission to happen), and see if the skill level scales accordingly?
have them play blitz (where the moves are way too fast for info transmission to happen), and see if the skill level scales accordingly?
Not a fair contest. There are plenty of top classical chess players who are weaker in blitz and vice versa. It’s a different skillset. Classical is all about preparation for the opening followed by some deep thinking in the midgame. Blitz is all about pattern recognition and the ability to simplify down to an ending where you can blitz out the exact solution from a database.
It's true that some people are worse at blitz, but if Niemann's OTB blitz rating is as good as his OTB rating at slower time controls, that's evidence against him cheating ... at least in ways that materially affect his rating. I guess it would be possible to cheat with an engine tuned to your actual rating just to make it less stressful.
Well, even endgames with as few as 5 pieces on the board are beyond what a human can solve with memorization. I don't disagree that blitz is mostly pattern recognition and rapid tactical thinking, but that applies all the way down to the end.
magnus completely destroyed hans in two games, as black. I think the ease with which magnus took hans apart in these beach games, presumably added to his suspicion when hans played so much better in the Sinquefield cup.
Eh casual play has so many factors. I wouldn't put much weight on how badly a gm loses on the beach. For example, drugs could have been involved.
I do think Hans is cheating, but I think the proof will lie in statistical analysis of his games and demonstrating that he has an unusual (>3200 rated) propensity to clutch out specific moves. I think everyone suspects at this point that if Hans is cheating, its only a handful of moves per game.
>this is not just a one move cheat, he'd have to cheat + have a continuous absence of mistakes
Blunders are exactly what a device like the one described would seriously help with. If the buzz means both "there is an only move here and it's not immediately obvious" and "at least one of the natural moves here is a blunder or very inaccurate" then you need to just send a buzz and you've probably cut inaccuracies significantly. That said, a very simple communication device like this is probably badly hurt by a 15 minute delay.
> at least one of the natural moves here is a blunder
Interesting, I'm not sure if a computer has the ability to recognize something as a "natural move but also a blunder." It would require a very human-like way of thinking about moves, which computers don't generally have.
Anyone 1500+ can recognize the natural move--that's what makes it natural.
Probably the easiest case of "natural move but blunder" is anything that is a top 3 engine move when looking 3-5 moves deep, but losing significantly on deeper evaluation.
Also, this sort of categorization is at the heart of how chess puzzle collections are automatically assembled. A good chess puzzle contains an unnatural move that wins--the exact opposite of the natural but blunder. Chess sites scan their online play databases for these all the time, and serve them up as puzzles.
> he'd have to cheat + have a continuous absence of mistakes, which is an awful lot of information to transmit.
Perhaps a top-level player can jump to a higher level if they can stop worrying about coming up with brilliances in the macro strategy, and instead focus entirely on making their micro-level play spotless.
This is the opposite of what a computer-assisted player would do. Computer chess engines (generally speaking, somewhat less true of the latest generation) are not great at high level strategy but will never miss a tactic (micro level play).
ELO gives you a statistical evaluation of how likely victory is for one of the player. Hans rating means he has non insignificant chance of winning against Magnus.
Hans can win without cheating as this last game proves. There is not a shred of evidence against him after all.
Sounds like not just the players need to be checked for cheating devices, but the audience too. Or the players have to play in a faraday cage without an audience.
It's not going to make the game more fun, but it's probably necessary.
From what I understand you only need one bit. The assistance doesn't need to be "move piece P to square S", but "this position is critical, if you spend extra time exploring here you will find a winning move".
As these players are on timers there is a race against a clock. So if you know where to focus your time/effort you can easily gain an advantage.
Just "e6" would likely be enough context in most situations. Sometimes only one piece can legally move there. Sometimes it's obvious which piece should move there once the position is pointed out.
I watched the livestreams of Andrea Botez's games in Vegas. Something I was really struck by is even though she's not a GM strength player, her mental board visualization skills are way up there. She did post game debriefs on the streams where she went rapid fire through hundreds of variations with her cohost just verbally. She'd go through them faster than the host could click to show chat at times.
Now imagine what people like the Super GMs are capable of.
They put a number of mid-game positions on the board, and Magnus was able to guess the players, tournament, game number, who won, what the next few moves were. Who was playing on the table next to him. What their moves were.
Yeah. This is why I think those people are full of shit.
First, those people should look at his YouTube. Obviously he's capable of analyzing games. To think he'd be incapable of analyzing a game he just played? What? It makes no sense.
To think that someone, even if they were cheating in every game, was not a 2600ish rated player and to perform like Hans is just ridiculous. Every 2600 player can out analyze anyone who is a 2100 player (Botez).
You don't have to imagine, check out any post match press conference or when players discuss over the board after the game.
They remember all the variations they consider, and they've considered most of the variations their opponents have calculated, so the variations aren't new branches, they're just pointers to spots in the game tree both players have in their heads.
At their level it's pretty much known what location / piece they're thinking about. For key moments, it may be enough to transmit the piece name only. And have some follow up for destination if they really need them.
Not many at all. For instance it takes a maximum of 6 bits to encode a given destination square on the board. This is probably sufficient, or very close to it.
As opposed to chess, magic is about lying and cheating. Everybody knows it and everybody is fine with it in magic, as you can see here in Dary's legendary ambitious card routine:
Nonetheless there's a rough ranking of "acceptable lies", and a thumper - anything with a secretly complicit audience member, really - ranks near the bottom of that.
> Many people might assume that a cheater needs guidance every move, thereby requiring a potentially more obvious cheating mechanism. That is not the case.
IMO, this reasoning potentially implicates every high level player. If it's possible that two hints can account for the difference between 2600 and 2800, and a 19 year old kid under heavy scrutiny can exploit this weakness without being detected, it seems assured that other more experienced players are also exploiting this.
A bunch of grandmasters have now talked about the psychological aspect of even just wondering if your opponent could possibly by cheating, and second guessing if a bad looking move by your opponent might actually be a brilliant engine line.
It seems that might even be enough for Hans as a 2675 rated player to get an edge against 2800+ player without even actually cheating
IIRC Kasparov was psychologically defeated by Deep Blue with the exact opposite of this play.
After playing what seemed at the time like 'computer-type chess' - relentlessly accurate goal-seeking strategy, Deep Blue started to play far less obvious and riskier moves. Kasparov's prejudice that a computer couldn't play like that led him to believe that Bobby Fisher was hiding inside the machine with an oxygen tank and a sandwich.
I thought his primary complaint at the time was that they were reprogramming in the evenings in response to the day's games, providing a lot of grandmaster human input during the tournament. I could be wrong there.
I watched his later matches against Deep Junior, around 2004 (?) in New York City. Match was tied, in the final game Junior made a mid-game move that was surprising to everyone in the analysis room. They were using a different software to analyze the potential lines and not finding the advantage for DJ. Yasser Seirawan and Maurice Ashley couldn't 100% agree that it was a bad move, but they said from what they can see it looked like a mistake by Deep Junior. Kasparov to a lot of time to ponder, and they accepted an exchange that would lead to a draw.
It was a very psychological moment in that era when machines were not clearly superior to the best humans.
I've definitely experienced that, playing mahjong against a guy who was behaving oddly in the European championships. But all you can do is have strict referreeing and stricter penalties for anyone who is caught cheating even once (which seems to be missing in the chess world given the player in question's record). You certainly can't try to retroactively impose a vigilante penalty that FIDE haven't.
To some extent yes, but humans tend to make moves that follow some kind of reasoning or logic. When you play a strong opponent, it's possible that you won't see the next move that they play, but once they have played it, you can deduce the logic they used in order to make it. An engine move on the other hand can easily reject the standard strategies and can appear highly irrational. When you see this kind of move, it becomes easier to suspect cheating.
yes but the computer is just so much more powerful, as well as it makes seemingly weird moves more often, as well as it can get itself out of a "bad" line if pushed that way much easier.
That doesn't scale down the skill level. At top level of just about any thing the difference between player is decided by few mistakes (by that I define less than optimal action).
If average player does 100 mistakes per match fixing 4 of them won't matter. But if great player makes 6, fixing even single one can be deciding
Software running in a smartphone can play deep games against each other.
Chess engines aren't like car engines are to sprinting. They are more akin to text-to-image AIs but as though every single picture they produce is better than what any artist ever could produce.
Part of that is because Chess is easily defined (the win conditions are comparatively simple).
I'm rambling now and I think that's enough wall of text for a hot take.
Machines can throw objects faster, further, and more accurate than humans, but field sports are still popular. It's interesting not because the object goes far, but because it's being done by a human.
> Chess is a game of symbol manipulation. It isn't played in the real world and its rules don't require human bodies the way field sports do.
I didn't mean "humanoid" as in "C3PO sitting across from the player at a table", I meant humanoid thought processes.
As far as I know, none of the chess engines are humanoid in the way they determine the next best move.
For one, they are all using far, far more instant-recall capacity than any human, ever.
> its rules don't require human bodies the way field sports do.
Which rule in football, tennis, american football, baseball, basketball or hockey requires humanoid players?
They may preclude robots as players, but that's a post-hoc fallacy - "they require all players to be humans, so therefore robots cannot replace humans like in Chess".
> I didn't mean "humanoid" as in "C3PO sitting across from the player at a table", I meant humanoid thought processes. As far as I know, none of the chess engines are humanoid in the way they determine the next best move.
Sure, and if you had android basketball players or soccer players, they would probably play the game differently as well.
> Which rule in football, tennis, american football, baseball, basketball or hockey requires humanoid players?
The totality of the rules put together tend to require that. For instance, in the NFL, whether or not a player carrying the ball is considered "down" depends upon their elbow or knee touching the ground, which implies that they need to have elbows and knees. Whether or not a catch is considered in bounds depends upon both feet touching the ground in bounds, which implies that they need to have feet. And so forth. Soccer has specific rules about the hands, feet, and head, while basketball's rules around dribbling specify hands and footsteps.
Obviously you could build a robot that looked like a dalek with a pneumatic cannon and design the robot to shoot a ball, and that robot would probably be better than human players at shooting baskets or completing football passes, but it wouldn't actually be able to play the full game according to the same rules that apply to human players.
From what I've read on this exact topic w.r.t. robotics, there are a few select places you could probably replace humans with a robot using current technology, and achieve almost perfect results. Kickers in American football, free throws in basketball are two such examples.
Kickers have to sometimes tackle, run fakes, kick onsides, deal with mishandled snaps or holds, and adjust for the conditions, like wind or how the opponent is trying to bother the kick. Plus the kick is from different parts of the field, unlike a free throw. But it doesn't matter, since someone taking a free throw has to be already in the game playing. There's no designated FT shooter.
At some point, being able to score 3 points from unlimited range just breaks the game of football to the degree that an increased risk of mishandling a bad snap/hold doesn't really matter, and you won't ever need to kick onside, either.
I don’t really think you can compare those though; robots are crap and easily beaten at most physical sports. I mean, as far as I can find, there is not even a bipedal robot that outruns humans (there is the cheetah robot but it has 4 legs ; that’s like having a dog or, you know, cheetah compete against Usain).
But yes, people will continue to play chess, go and spear throwing because it doesn’t matter if something non human is better.
It’s not required, we just don’t allow anything else because it would no longer be interesting at all. That’s the point. Same why we don’t watch grandmasters play computers; they lose. But humans vs humans is still a good watch.
Are they? The throwing sports? How many people do you know who regularly follow shotputting or javelin (outside of possibly the olympics). How much do the top 20 javelin throwers in the world earn in sponsorship and prize money and how does that compare to other actually popular sports.
I have no doubt that chess will remain at least as popular as javelin or shot-put for the foreseable future. I'm just not sure that counts as 'popular'.
what if Hans is not human, but an advance AI cyborg? maybe he's not communicating with a chess engine, he IS the chess engine. Maybe his name is an anagram for the word Enigma or something.
In a similar vein, Tool Assisted Speedruns for video games can always outperform human players, but TAS streams and youtube videos don't get nearly as many viewers as real humans running. And human speedrunners caught cheating and using tools etc. have drastic hits to their popularity after they're exposed.
I'm actually on the whole more interested in a TAS. I would often watch a big fraction of the TAS block at a GDQ, and only RTA for niche cases where I especially care about that game.
However, TAS is very unlikely Computer Chess because the TAS is actually a composite of a vast number of individual human player inputs - assembled by in effecting rewinding and continuing the game over and over. The TAS is not a machine beating the game, but the effect if humans played the game as well as they know how. That's why they have "sync" problems during a GDQ, the playback device has no idea how to play, it's just robotically carrying out actions.
Yeah I know what you mean but human chess is much more exciting because you can see themes appear in a player's strategy. You can sort of follow what they're thinking. Computer chess is next level precisely because the computer has decided that is the best move, but it's not part of some kind of narrative, it's just based on the engine's analysis.
Isn't that way more beautiful? Pure brilliance and perfection. There is no emotion in the machine making dumb mistakes. There is no fear in the machine questioning his move. Is there some kind of video of a GM analyzing an engines moves step by step? Can they even understand every move without seeing the whole picture?
Yeah, in a way it's more beautiful I think, but only in some abstract sense because it's still above everyone's heads. What's missing to make it interesting to watch is that narrative I was talking about, and that's a distinctly human feature of gameplay. I don't know if there's a video of GM's analyzing engine's moves, mostly because I think that's literally what the engine is for. And I don't think they can understand every move, which just goes to show how strong engines are. It's like alien tech strong.
> Chess engines aren't like car engines are to sprinting.
I think this is questionable. While we can understand the physical limitations of a human compared to an engine, we tend to alleviate the intellectual limitations. Just like an engine can deliver far more power than a human could ever do regardless of their training, a computer performs far more chess move computations than a human ever could, regardless of their training. It's just that our brain is biased toward alleviating computational cost, because we implicitly think "in the end, a human could as well play the same moves as a computer"
I do agree however on the premise that chess is a zombie sport, but I think it has more to do with the ease of cheating. If you consider cycling for example, there has also been cases of cheating with an electric engine inside the bike, and new cheating methods are likely to be developed faster detection procedures. And in this case "Bike engines are like car engines are to sprinting"
In fighting games, most AIs are discredited and stupid because they have no reaction time. I don't know of any that name in a nondeterministic 10-15f of reaction time. It really complicated things.
That's fairly counterintuitive to me. The easiest avenue for an AI to gain advantage over a human is reaction time and accuracy. E.g. an AI that reacts in microseconds will never be beaten on pure reaction by a human given that there's a floor of some 50-100msec for a human to be physically able to react to stimulus.
E.g. I remember early 3rd party Starcraft AIs would beat humans just by micromanaging certain nimble flying units.
> (or Niemann, if concrete evidence against him emerges)
This window seems closed though. Carlsen seems to have no evidence. Where else could the evidence come from? All we have is character attacks. Even if justified, they can't prove that he cheated.
All we know for sure is that Carlsen accused him of cheating with no evidence.
All this talk of "Carlsen accused him of cheating with no evidence" reminds me of the blowback against some athletes in the 70s and 80s who accused rivals of taking PEDs "with no evidence".
Sometimes the evidence of someone doing monstrously better than can be expected by their history is sufficient IMO. I mean, look at this article about swimmer Shirley Babashoff [1], dubbed "Surly Shirley" at the time by the media, for suggesting the East German women were on PEDs in the 70s. Nowadays we look back on those images of the East German women, looking more manly than any dude I've ever seen, and wonder how we considered with a straight face that they weren't on a boatload of drugs. Similarly, it completely baffles me how any sane person can think that Flo Jo wasn't on PEDs in the runup to the 1988 Olympics - her 100m dash record still stands today.
I'm not saying Carlsen went about it in the right way, because now Niemann is basically in an indefensible position, but I'm also not willing to quickly dismiss it because Carlsen has "no evidence".
Usain Bolt is another interesting case. Commentators have made a pretty compelling case based on the circumstantial evidence that he was doping, but due to weaknesses in Jamaica's anti-doping program it's likely impossible to prove one way or the other.
A huge number of (male) athletes across all sports are doping now. Testosterone deficiency is a medical condition, and taking “testosterone replacement therapy” is explicitly allowed by many sports. Even if it’s not allowed, it’s not really possible to robustly test for, because in some forms at least it’s simply increasing the levels of the same hormones that your body is naturally producing. Same goes for human growth hormone. Just take enough to put your levels at a very high, but plausibly natural level and you essentially can’t fail a drug test.
It’s only called “testosterone replacement therapy” when it’s prescribed by a doctor btw. Taking exactly the same course of medication in any other circumstances is called “taking steroids”.
Is trivially Easy to test for TRT. Taking testosterone shuts off your body’s LH and FSH production and those can be tested. These can be biologically mimicked with hCG and HMG but those don’t chemically show up on LH and FSH tests.
LH ratios aren’t a robust test, aren’t accepted by any governing body alone as far as I’m aware, and are only a very weak indicator if you’re only testing urine. Testosterone to epitestosterone ratio is still the gold standard, and to fail that test you need a ratio about 3x above normal. The efficacy of these tests are also highly impacted by when the PEDs were last taken, and that’s true for all classes of PEDs, not just the bio identical ones.
There are also faaaaaar more PEDs than the labs test for. The state of PED testing is completely unreliable, and you really have to be an idiot to get caught. But that’s only if your PEDs are actually prohibited, which they often aren’t.
FIFA call it a “therapeutic use exemption”, which basically anybody can get with a note from a doctor. A lot of other sports have similar exemption processes. I know the NFL does too.
> FIFA call it a “therapeutic use exemption”, which basically anybody can get with a note from a doctor.
In my quick research on this, the TUE process appeared to be quite strict, and definitely not an easy "you just need a note from a doctor" kind of thing. For example, in the NFL, the only example I could find where they'd give a testosterone exemption is if you had testicular cancer and had removal of 1 or both testes, which seems reasonable. I also found examples where both the NFL and FIFA had recently suspended players for testosterone use.
If you have any other examples or sources that counter this info, I'd be very interested to see it.
You can’t find examples because the treatments are part of the players private medical records and are not published. But it is as simple as getting a doctor to say that therapeutic use is required.
What hard evidence—exactly—do you expect Carlsen to be able to produce? Alternatively, imagine anyone in a similar position. What hard evidence can anyone produce in situations such as this?
I don't expect him to produce evidence, but I expect him to say more than "I suspect he cheated"
If he saw something unusual, like "Hans was messing with his shoe" or "I heard several vibrations coming from Hans during the game" etc.. that would be at least something.
Magnus has produced what he can given the situation and has staked something of extreme personal value—his legendary near-2900 ELO—on it with his move-1 resignation.
If he'd heard the guy's damn shoe buzzing he would have insisted on a search.
I believe you could have an engine look at the historical games of a player and identify the "strength" of each move. How strong (in terms of elo) does a player have to be before they find a certain move? How often do the top players find moves that greatly exceed their own elo? Does Hans find top moves more frequently than his opponents?
The challenge with this appoach of course is identifying a players strengths and adjusting for their preparation. Making 20 top engine moves in a row is not odd if both players studied that exact line before the match.
What's odd is making 20 top moves in a row on a bizzare line that nobody has ever played before that Magnus specifically prepared because he knew it was unusual (and engine disadvantaged) and unlikely to be in anyone's prep.
Niemann have admitted cheating before when playing online, so Carlsen is just not making this up about any random player. There is a history of cheating.
There is no smoking gun, but there is a lot of smoke. The ease with which Niemann pulled out of the hand a couple of brilliant moves, without spending too much time thinking about it, on an unusual line, is highly suspicious.
That probably would have been even worse of a shit-storm. You think you are rational and will ask good questions, but if you are not a trained journalist, and you haven’t prepared, you will most likely only ask really dumb question that do nothing except cause more drama.
I expect people to hold off the public allegations if they can't prove anything.
I could name specific players who I'm pretty sure were cheating in my own game. I've sometimes had a quiet word with a ref and asked them to watch a particular player closely. I've occasionally had a louder word with a ref and asked them to enforce the rules that are in place to make cheating harder. But you can't pull something like this based off of nothing but your own feelings.
I don't know much about chess, but it seems like Niemann now has to either maintain his performance in Chess without cheating, or cheat to maintain it if he can't without - in which case he could still be caught.
That's not true. Both he and Chess.com both say that they have evidence to the extent of Hans' cheating. Both have asked Neimann for the ability to speak freely without threat of libel.
Apparently this (the idea that telling the truth is necessarily not libelous) is untrue in some jurisdictions? Or perhaps they fear that they are not 100% correct?
Not quite. There is still a possibility that Niemann will admit cheating. If he actually cheated there may be a time—in years or decades rather then months—that he fills with remorse and admits it (hopefully with a detailed description on how he did it so we can verify). If however he didn’t cheat, we will probably never actually know the truth.
Very interesting. I don’t really understand chess beyond the basics but when I think of sports the difference between good and great really seems to be, in baseball just a hit or two per week, in American football a running back who has the vision to cut decisively a second or less before another running would.
When you see it consistently the difference seems enormous, but the math … is surprisingly tiny.
“or less” is doing a lot of work in that first paragraph. Running plays typically only last like two seconds, maybe four if they get to the second level (linebackers).
Also, I agree the “math is tiny“ but the talent, work ethic, luck, etc. required to separate oneself from the good to be great is *enormous*.
You're right: the hint doesn't even have to be a move. It could also be an evaluation "it's better for white", or even: "there is a winning combination" which might be enough to get them to focus on finding it.
As an electrical engineer currently in the process of getting a dental implant, I would say it's definitely doable. But it would present a pretty serious packaging challenge, particularly the power supply.
[UPDATE] Turns out you can get dental implant hardware on eBay:
In 1945 the Soviet Union gifted a wall hung artwork to the USA embassy which passively transmitted audio signal from the room with no power supply or active electronics.
No internal power supply. Very important distinction.
"The device, a passive cavity resonator, became active only when a radio signal of the correct frequency was sent to the device from an external transmitter."
But the Thing was a transmitter. A receiver could be entirely passive and very small. It could easily fit in a dental implant. (There have been reports of people hearing AM radio broadcasts through their fillings.) But the risk of detection would be very high.
It's a pretty straightforward calculation, which I don't feel like doing right now. But I think Stockfish is going to be pretty power-hungry at the grandmaster level.
I have some smart light switches in my house that don’t contain batteries and aren’t wired to mains. They’re powered purely by the force of pressing the switch. A bite-powered receiver might not even need batteries. Okay, maybe I’m getting too crazy?
You don’t need a faraday cage if you suspect he gets the moves from outside. Just put a 20 minute delay to the video feed and don’t allow random people in the room.
The St. Louis chess club provided more checking / scrutiny than any other OTB tournament I have seen. How would you improve their process? Honestly would love to hear.
Daniel Naroditsky[1] said at the St Louis chess club specifically it would be pretty easy to cheat OTB no matter what searches they do. He said there is a balcony which the players have access to when they walk away from the board which has a clear view of the car park and you could have an assistant signal from there at crucial moments. He also cited a case of a player who was definitely cheating OTB[2] and was never caught and where no mechanism was ever found.
It really seems to me that everyone is overthinking this. Hans has admitted to cheating and just says he didn't cheat here. The idea we should give him the benefit of the doubt seems really odd. It doesn't really matter to me whether he cheated in this specific tournament - he shouldn't have a place in the chess world.
[1] https://www.youtube.com/watch?v=eJVzSXsZ10I&t=3291s
[2] in the sense that every move for an extended period covering multiple tournaments was exactly the top engine move - way more accurate than any human ever
Word, I hear ya, a cheater is always gonna be a cheater. … But, honestly, we are going through a strange transition in chess. I’m not very familiar with online chess, so I take that sort of cheating lightly. Was surprised that online games can effect OTB ratings. I’m guessing FIDE, USCF , and other orgs will start separating online vs. OTB ratings as a result of this friction.
Chess tournaments could use the services of already existing casino anti cheating experts. But I imagine that would be very expensive and not feasible for most events.
I can't think of any effective way to curb online cheating in chess. Ultimately, online chess with money prizes shouldn't really exist.
Chess is not a game where there is always one right move, but where positions are threatened/protected to a greater/lesser degree, and the _value_ of those threats/protection can only be realized much further along in the game.
The best players in the world, these 2700+ rated players have:
1. Played many many positions many times over and have incredible recall of those general positions.
2. Know how to analyze a position/state of the game at any given moment and have a better "feel" for who is advantaged and/or where the greatest strengths/weaknesses of black/white are located.
However, none of them have the power of chess engines, which analyze singular moves (or poll a db) for the hundreds/thousands of possible outcomes 1, 2, 3, ..., n moves ahead (this is why the best engines are strictly better than humans at this point), so unless a player has both played and committed to memory the exact line being played in a game, the best they tend to get to is "having a feeling" about the state of the game (please forgive my oversimplification here, chess fans)
Now if a 2600 rated player - someone who's still easily in the top 1% of chess players and incredibly capable player of the game - were to be playing a game against a 2800 rated opponent, but had a computer tell them "Hey, this one move is critical" without being told the exact move, they would almost _certainly_ become heavily favored to win. That "feeling" about a position is now irrelevant. There are only a few pieces that will be likely moved on any given turn, and now you can narrow down your own analysis to what is different about moving any one of them in particular because you've been given advanced warning that the most-likely-to-be-played moves will result in wildly different consequences n moves later.
These are hours-long games. Taking 15-20 minutes on a turn is not unheard of, and doing so on a turn that is proven-critical can make all the difference.
You don't know there are exactly 3 bugs to start with. While in chess, esp if you're experienced, you know what your possibilities are. Just need to pick one.
chess experts do agree on that, they've all been saying it on various youtube channels, i was even irritated when I saw it written here cuz I was like "you're just repeating the same stuff I've been hearing"
200 points difference means a 25% chance to win, so I doubt just 1 hint is enough to bridge that gap consistently. Many high level games are won by grinding out a small advantage. I'll take a 2700 with 3 hints against a 2800 though.
The "200 points difference means 25% chance to win" breaks down at the highest levels. It works fine near the middle of the bell curve -- i.e. 800-2000 Elo -- but once you get to 2200 Elo you are talking about the >99th percentile. For example, I don't know of a single 2400 player who can score an average of 0.25 against 2600 players.
Even at my own mediocre level of 1800, I definitely do not score 0.25 against 2000 rated players. More like 0.1 if I'm feeling sharp.
As an example of this, until the Niemann game Magnus had a 53-game win^H^H^Hunbeaten streak. Prior to this he had a 125-game unbeaten streak. Many (most?) of these games were played against competitors within 200 Elo. Many of these were played against the 10 next-best chess players in the world.
The back of the envelope percentage calculation absolutely does not apply at this level of chess. In reality if Niemann were to play Magnus in 100 games, he would be exceedingly lucky to win one game.
I disagree with the second paragraph but not enough to get into a public debate about it. But it is worth pointing out that Carlsen's 53-game streak was a non-loss streak, not a win streak. Many of those games were draws.
First, I said he would be lucky to have a 1 in 100 chance. Second, absolutely nobody is saying that's the only reason to be suspicious of this game. Regardless of whether or not you believe Niemann cheated, if you think the fact that he won is the only claim in this accusation you simply aren't paying attention.
>I don't know of a single 2400 player who can score an average of 0.25 against 2600 players
I mean, you can look at the stats. They play all the time and while it becomes less accurate at the highest ratings (more so at the 2800+ level), 2400 vs 2600 does still result in something in the general range of 0.25. However, if it's 0.1 (like in your example) then my point is even stronger since it would be even harder to turn that into a win consistently with just 1 hint.
>Even at my own mediocre level of 1800, I definitely do not score 0.25 against 2000 rated players.
If you are noticing that at your level, it is probably either selective memory or specific to your play as ELO-estimated winning chances hold up well enough at 1800-2000.
The Elo system is calibrated so that that the expected value from playing a player 200 points stronger than you is 0.24. This is true independent of the strength of the players. If you are scoring 0.10 against players 200 points stronger than you (that would mean, for example, 1 draw and 4 losses over 5 games) but maintaining a stable rating, then you must be crushing players that are weaker than you and/or doing very well against players at your level.
(FWIW, I am 2000 USCF and an expected value of 0.24 vs a 2200 and 0.76 vs an 1800 feels quite reasonable to me.)
Elo assumes that performance across n^2 players fits the logistic curve model of n players. There is no reason to believe that assumption is remotely accurate and that Elo would ever stabilize. Players often avoid playing lower rated players, for this reason.
I don't think people are saying that it cannot happen, just that you need to prove it instead of hurling empty accusations, especially when it can destroy someone's career. I personally think it sets a bad precedent if every top player immediately starts crying "cheating!" when beaten by a lower ranked one.
Surely he means just that; i.e. Hans Niemann is known to have cheated in past games (these were online games and it must be noted that Niemann maintains that he has never cheated before or since the and never in an "over the board" tournament).
Does the hypothesis in question seem to hold for, say, the world's top 10 languages (by, say, # of living fluent speakers globally)? Top 30?
I'm genuinely curious. And also trying to discern whether your point is that many-- but not necessarily _all_-- current languages seem to conform to this hypothesis, or that actually very few seem to.
This is known as Universal 20 of Greenberg 1963. Not so universal, but you have to get down to languages way out of the top 30 like Chechen to find one that don't obey the typical rule. Of 576 languages catalogued in Dryer 2018, 113 are English-like, 182 are English-reversed, and the rest somewhat varied.
Mathematics often deal in abstractions, and "God" may be used as a linguistic abstraction for ideal or perfect knowledge. It's a shorthand in writing casually about the most elegant mathematical insights. Its use is also a tradition via Erdos with "The Book", which Aaronson and many mathematicians pay homage to.
Right. AFAIK there is no universal Math deity like there is Caissa for chess that can serve as a representation of "perfect math", God is the next best thing.
If it's any consolation, Erdős himself said in one famous lecture that the "God" part is irrelevant: "You don't have to believe in God, but you should believe in The Book."
I see many comments here asserting that RCV is "fundamentally sound" and that there is proof that it satisfies at least strategyproofness (in the sense of Cooperative Game Theory and Social Choice Theory, s.t. revealing/voting for true preferences is at least weakly dominant).
But can anyone provide backup for these claims?
I'm far from an expert, but I'm curious to reconcile these claims vs Arrow's Theorem and Gibbard's Theorem. My (potentially flawed) understanding as well is that RCV leads to a greater propensity for "extreme" (in the strict sense of being the top choice of a small minority of voters) candidates to win.
FPTP makes the mistake of assuming that the person with the most votes is the best.
Likewise, the biggest flaw of RCV (to me) is that it makes the opposite mistake: it assumes the person with the least votes in each round is the worst. This isn't necessarily the case.
It's still much better than our current system because it collects more data-points from voters about their preferences, and uses that data to make more comparisons between different batches of candidates.
But there are simpler systems out there with even greater voter satisfaction efficiency, because they don't make the same flawed assumptions about voter preferences. In approval voting, you either approve of a candidate or you don't, and you vote as such.
Election science wonks like to talk about theorems and satisfaction of obscure criteria, but I think they often undervalue the two things that matter most: voter satisfaction and simplicity. "Simplicity" meaning: an optimal strategy that's simple-to-explain, an easy to understand outcome, low cost of implementation, etc.
Most of us here are software folks, and we understand that designing a great system is the easy part; the hard parts are implementation and change management. Voting systems should be thought of the same way.
I'm glad you brought this up, because ~RCV~IRV does not actually satisfy these conditions (it is a personal point of frustration how Fair Vote is misleading, even by using the term RCV). IRV doesn't pass the favorite betrayal criteria. IRV also causes a weaker definition of what it means to be a spoiler[1]. Specifically you should look at monotonicity[2].
I'd encourage you to dig a bit into electionscience.org (they have some good YouTube videos as well) as they do a deeper dive on some of these topics. There's a reason Arrow himself was a fan or cardinal systems. The other part is that we need to consider factors other than VSE and spoilers. A lot of people get hyper focused on VSE (typically Condorcet supporters) and while it is an important factor 0.5% isn't a big deal (really a few percent isn't). Also you might want to check out our (HN) resident voting expert's[3] comments, because he links to a lot more information. I'd call myself a hobbyist where Clay is an expert.
RCV is basically just a very quick runoff election; it has the same propensity for extremeness that a runoff does.[1] That said, I think it is very likely that more people will vote for third-party candidates (TPC) under ranked-choice.
You could make the case that TPC will allow for the representation of a wider spectrum of views, or alternatively that TPC are likely to be extremists. Both may be true.
I would be sensitive to the "TPC under RCV are more likely to be extremists" argument were it not for the fact that FPTP has already allowed profound extremism in one of our two parties, and since FPTP is so anti-competitive it means that it's extremely difficult to combat. I'll take a system that makes parties more competitive, which non-FPTP systems all do.
I'm not sure what your definition of 'extremist' is. If either major party is getting close to 50% of the votes, it must be quite mainstream. I don't have to like it, and I may in fact find it abhorrent, but that has nothing to do with extremism.
A district tilts heavily toward one of the two main parties -- it is a "safe" district for that party. The candidate nominated by that party in the primary election is nearly certain to win the general election. If the candidate is highly extremist / flawed, they might lose, but party affiliation is sufficiently strong nowadays (aka the electorate is sufficiently polarized) that a candidate can be pretty far to one extreme -- farther out than the bulk of the electorate -- and still win.
Meanwhile, there is a tendency for the more centrist voters in both parties to skip the primary. Thus, the candidate who survives the primary is often relatively extreme.
The result is that the victor of the general election is often to the extreme side of not only the electorate as a whole, but the membership of their party.
To oversimplify, imagine that political views fall on a one-dimensional spectrum ranging from 0 to 1, and the electorate consists of:
- 40% at 0.4 (center-left)
- 30% at 0.6 (center-right)
- 30% at 0.8 (heavy right)
In the primary, center-right voters are under-represented, and a candidate at or beyond 0.8 has an excellent chance of being nominated. Then in the general election, at least 2/3 of the center-right voters are likely to swing toward that candidate (because of polarization / strong party affiliation).
Agreed, the spectrum ultimately is arbitrary. I was thinking in terms of a spectrum normalized to the country as a whole. In a particular district, you often have a breakdown that is off-center relative to the overall nation (of course that can be in either direction).
And in any case, the key point is that the primary process can lead to a candidate taking office who is well off-center even within their district.
Amusingly, upthread you and I are talking about whether encouraging more people to vote leads to better outcomes, and this is an example where the answer is yes: a candidate can be both an extremist and receive 50% of the vote if, as in is true in the US, less than half the eligible electorate actually votes. The current president was elected on the backs of 25% of eligible voters; does that make him an extremist candidate even by the standards of the US electorate? If we had more data on account of more Americans asserting their preferences, then we could actually determine that.
I think that's a sign that each of us has nuanced views!
'Extremist' is a bit of a slippery fish; I can't imagine calling something with even 10% support 'extremist'. If I had to draw a line, it would probably be that the sum of all 'extreme position support' would have to be less than 5% for a given subject. That is not to say that positions with >5% support are necessarily correct or conscionable.
Well it's similar to a traditional runoff when there are two major candidates, but it could lead to different results when support is split more evenly among three or more candidates.
RCV gives you the same results as a runoff if the voters have ordinal preferences that they would have exercised in a runoff. There are a number of reasons why either of these may be wrong, but it seems to be a fairly reasonable set of assumptions.
IMO it also lets you just count the interest, regardless of outcome. Now you could see Nader (or Bernie, or the libertarian candidate) get 5, 10% of 1st choices and know it's legit, even if it doesn't change the outcome of the election.
This is why it's opposed by both faces of the status quo party. Once a third party gets to 10% for a given office, there won't be much stopping it from getting to 40% in the next election for that office.
I can't speak for any theorems so take this with a grain of salt, but if the candidate who was the top choice of a small minority manages to win that would mean that they are someone that a majority would settle for, which I would think is a more optimal outcome than having a highly polarized winner who half love and half hate when the goal of an elected official should be to represent as many of their constituents as possible.
There are properties that different voting systems have. Not all voting systems can fulfill all of these properties. In many cases, it is a trade-off of some having some properties but not others and people deciding what they care about more. For instance, RCV satisfies the majority criterion, meaning that if a candidate receives more than 50% of the vote, then they are the winner. However, it does not satisfy the Condorcet criterion, which is that if there is a person who wins in a pairwise matchup against every other candidate, then that person is the winner.
RCV is also sensitive to different tactical strategies. You can rank a weak candidate higher in the hopes that they do well in one part of the run-off and then worse later. You may also have a preference for voting a less preferred candidate first for other reasons.
I wouldn't say it's perfect. It has some nice properties. Other systems may have nice properties too. The useful question to me is if you compare the current system vs. RCV, would it elect of those who more closely match the preferences of people they represent. I would say yes to that, but I also think Condorcet methods would do an even better job.
Have you seen any research on which parties or groups might benefit most from Condorcet methods? It would seem to me that it would benefit smaller parties who share something in common with a couple of major parties (such as the Libertarian party in the USA).
The more typical definition of an extreme candidate is not "this person is not my top pick", but rather "I would not vote for this person at all". Candidates under RCV win when they get 50% of the vote, so if you're so extreme that only a small minority bothers to rank you then you're not going to find yourself elected on behalf of a minority. This is in contrast to FPTP, where any extremist can win with any miniscule percent of the vote if there's enough vote-splitting going on.
I'd just first like to say that I love reading your work. I'm always delighted when I see an update on Shtetl-Optimized and I admire how you are both simultaneously rigorous and funny in your papers and posts. Your PnP survey stands out in my mind as a truly fun and insightful read in particular.
Two questions:
1. Any meaningful updates you'd make to the PnP survey today?
2. As a total aside, I'm curious for your thoughts on blockchains. Not specifically proof of work as a BFT system per se, but more broadly curious to what degree you think "trustless" transactions and data processing might or might not be transformative.
Apologies in advance if you've recently provided thoughts on either of these topics and I've sadly missed them!
1. The P vs. NP survey is only 2 years old, so all the edits I'd make to it now would be rather minor ones: for example, including some more recent circuit lower bounds of Ryan Williams and others, some more no-go results for Geometric Complexity Theory, and Raz and Tal's BQP vs. PH breakthrough (which required a new circuit lower bound, though not of a kind that can evade the natural proofs barrier).
Thanks for the reply! And apologies I didn't realize the autocorrect on my phone turned "PvNP" to "PnP". I'll read up on the recent works you mentioned here. Appreciate the link to your comment on Blockchain technology as well.
If you still have time to answer questions.. is there any recent work in the cryptographic or blockchain space you see as standout? I haven't referred recently to any citations or updates, but found Ben-Sasson, Bentov, Horesh and Riabzev's recent work (https://eprint.iacr.org/2018/046) on ZK-STARKs intriguing.
