Advertisement
Boards Golf Society are looking for new members for 2022...read about the society and their planned outings here!
How to add spoiler tags, edit posts, add images etc. How to - a user's guide to the new version of Boards

Bunratty 2021 On-Line Event

2»

Comments



  • The player I'll assume you're talking about did not get kicked out of the event from what I saw during the event and what I heard in the aftermath. They scored 5 wins from 5 and then left the event. Certainly was no "banned" banner on their page at any time that I saw. There was a second player who was removed for breach of fair play rules.
    I'm aware his account wasn't closed until later but is it possible the organizers asked him to not play any further games? Or maybe it was a sudden pang of guilt after all.
    Anyone can point out an IM can play a bad game - however, that IM said that they played miserably and didn't suspect anything really. This is a similar story to a recent Irish OTB board cheating incident where some players didn't suspect anything because they found they played miserably; but others did and they were proved correct.
    It's not meant as a slight against any player but merely the pervading attitude that a 1700 player catching an IM or a talented junior napping must be cheating. Then you get the pompousness of the Guardian article portraying it as some sort of masterpiece only Tal could manage without an engine, when in reality any competent 1700 player, known or unknown, would likely do the same. The Bulmaga game probably wouldn't even be considered in any cheat detection analysis because a winning position was achieved after a few opening moves.
    The chess.com accuracy score is kind of a nonsense, but probably is telling as to the system they use for detection. Arbiters are specifically trained in online event anti-cheating measures- nowadays - I should know having done the test before Christmas. The main anti-cheating measure for online events is the score created by Prof Ken Regan because most of the time you won't be in the physical venue with the player and even cameras can only tell so much. If you're scoring something in the .7 range of performance, you're either playing short draws, having a cracking event and/or are to be monitored. Certainly nothing black and white there. If you score above .8, you are using assistance. If you score a .9 or above, you are stockfish.
    The chess.com accuracy score is certainly evidence, I'm aware of that. It is engine based after all, but it's not a simple exhibit of how often the top engine move was played and I just wanted to clarify that in case some users are confused. I don't know enough about either the chess.com system or the Reagan method to say how similar they are but Danny Rensch does say in the link I posted above that the accuracy score has nothing to do with their actual detection method.
    I think most of the main arbiters will preach to you that these scores will have to be believed eventually if we are to have any online opens in the future with any degree of confidence; and that these scores should be actioned upon and that terms and conditions should point to these in advance of you playing an event (the 4NCL online does this). These will not catch you for spotting one tactic legally or illegally but, over the course of an event, they will catch you for 3000 standard play.

    "Using accuracy figures in isolation as a smoking gun is extremely dangerous." I mean, this is kind of true for the chess.com metric but also the entire problem with online chess - there are no better way to catch an online cheat than using Prof Regan's metrics. There is also precisely zero other ways of telling someone is cheating when they are in a room by themselves with no camera. If you don't want to believe the accuracy figures then you are also accepting that cheating will go on without repercussion. if you want everyone on camera, then you'll need arbiters to monitor every 5 or 10 boards; and even still won't catch some cheats during the game given what we learned from that incident over summer.
    I would assume chess.com, the other platforms and anyone applying these advanced detection methods has their **** together and have gotten pretty good at catching them by this stage. The rest of us should probably stay out of it but cheating accusations can cause serious harm and the only issue I have is that borderline cases are possible and in such cases maybe chess.com etc. are more likely to side with a known player (note: I'm not saying this is a borderline case). Personally I would have zero issue agreeing with those terms before playing in any online open as I'm unlikely to play well enough to even become a borderline case :D Agreeing to play in front of a camera likewise I'm all for.
    To the other points calling for OTB bans for online cheating - I agree that this may be a necessary deterrent because anyone banned online can just create a new account and have at it again. However, two points on this - I believe the ones saying it are also the ones who would say that online chess and otb chess are totally different animals from almost every perspective. So it would be like handing down a rugby ban for a foul in a football match because the IRFU have a stake in Lansdowne road. And this would also require a huge amount of ICU/FIDE manhours to monitor, research and hand down these bans. I'd go as far as to say that this would become the biggest project of any chess organisation.

    Edit:
    I did mean to mention there somewhere along the line that I'm sure I'm not the only arbiter to keep a mental note of suspicious play from event to event. I also keep a mental note of those who have accounts shut online. Strangely, there has been no crossover as yet between the two lists. But we're talking about small samples.

    I agree that an otb ban is going a bit too far for what seems like an isolated incident but probably would have gone for an analogy involving the legal system instead as cheating in a chess tournament is closer to crime, although I suppose that depends what football team you watch :pac: It's good to hear that arbiters are keeping a keen eye on what goes on though and it should lead to an increased level of monitoring those players when otb play resumes.




  • You're still (a) saying the player is 1700, when they're not, and (b) ignoring the fact that lightning has struck three times in a row here.

    Nobody is saying "that a 1700 player catching an IM or a talented junior napping must be cheating."

    However, it is the case that a 1700 (which, again, this player isn't) playing three very very very very good games (as judged by a computer) to beat 2100, 2500 and 2400 in successive rounds is almost certainly cheating.




  • I'm aware his account wasn't closed until later but is it possible the organizers asked him to not play any further games? Or maybe it was a sudden pang of guilt after all.

    I don't think the organiser will mind me saying, they had nothing to do with it.
    The chess.com accuracy score is certainly evidence, I'm aware of that. It is engine based after all, but it's not a simple exhibit of how often the top engine move was played and I just wanted to clarify that in case some users are confused. I don't know enough about either the chess.com system or the Reagan method to say how similar they are but Danny Rensch does say in the link I posted above that the accuracy score has nothing to do with their actual detection method.

    I would assume chess.com, the other platforms and anyone applying these advanced detection methods has their **** together and have gotten pretty good at catching them by this stage. The rest of us should probably stay out of it but cheating accusations can cause serious harm and the only issue I have is that borderline cases are possible and in such cases maybe chess.com etc. are more likely to side with a known player (note: I'm not saying this is a borderline case). Personally I would have zero issue agreeing with those terms before playing in any online open as I'm unlikely to play well enough to even become a borderline case :D Agreeing to play in front of a camera likewise I'm all for.

    I agree that an otb ban is going a bit too far for what seems like an isolated incident but probably would have gone for an analogy involving the legal system instead as cheating in a chess tournament is closer to crime, although I suppose that depends what football team you watch :pac: It's good to hear that arbiters are keeping a keen eye on what goes on though and it should lead to an increased level of monitoring those players when otb play resumes.

    The feeling of the top arbiter I know is that if they've been banned - they aren't borderline - they're in that .8 and above category I referenced in my previous note. I'd imagine that platforms can't afford to mess around with borderline calls given the litigation that could follow.




  • cdeb wrote: »
    I've two problems with this. First, there's no indication this player is a 1700-1800. Their ICU is sub 1200; their online was sub 1600 at the start of the tournament (and online seems to be quite inflated generally). Where are you getting the extra 200 points from?

    Maybe the extra 200 points is me just being generous to the hypothetical player I'm referring to. Catch up.
    cdeb wrote: »
    In any event, actually a 1700-1800 putting in these results would be outlandish. Again, the game against the WIM who made a hames of things I could allow. But 1800 v 2500 is, in football terms, Brazil v the UAE. Slightly more likely that the UAE might take something off Brazil once than San Marino. But three times in a row? Still no.

    Firsty, the football analogies are great and all but I just don't think they apply here. Sometimes in chess a lead is just unassailable, no matter how many Ronaldos you have and Mubaraks the other guy has. If we can go back to your previous analogy, taking the lead in chess is often literally just like the other team having players sent off one by one by one.
    Secondly, he didn't beat 3 Brazils in a row. He beat a 2100 rated 10 year old, an IM who will probably admit to playing the worst game of her life, and a 2500 GM. In blitz.




  • None of that really matters though. It's still a hugely suspect performance for a 1200. The general odds of an 1800 beating a 2500, rounded to two decimal places, are nil. Literally nil. I can tell you I've gotten plenty of great positions against FMs in blitz, and believe me there are plenty of ways of messing them up. And I'm not 1200 rated. That game against the WIM wasn't exactly queen against king stuff.

    So I still take issue with your comment that people are saying "that a 1700 player catching an IM or a talented junior napping must be cheating" - nobody is saying that. And I still don't know why you're adding bonus points to this person's rating to make your point. I don't see any evidence of "gobsmacking elitism" or "players throwing their toys out of the pram", like you've suggested.


  • Advertisement


  • To correct something I said earlier, I can see that the account I think we're all referencing has been banned. Wasn't the case the day of or day after the event.




  • I am one of the people who would say that online chess, OTB chess, rapid and blitz are all different animals but they are all forms of chess so a chess cheat is a chess cheat in whatever arena he is cheating and since it is impossible to stop players playing online the only sanction can be an over the board one so I reject the football analogy. I don't see what is so harsh or OTT about a ban, if a person doesn't want to get banned then they shouldn't cheat.




  • sodacat11 wrote: »
    I am one of the people who would say that online chess, OTB chess, rapid and blitz are all different animals but they are all forms of chess so a chess cheat is a chess cheat in whatever arena he is cheating and since it is impossible to stop players playing online the only sanction can be an over the board one so I reject the football analogy. I don't see what is so harsh or OTT about a ban, if a person doesn't want to get banned then they shouldn't cheat.

    it really was a terrible analogy.. like, I can even imagine certain types of cheating in football that would get you banned from rugby..

    Still, the manpower issue would be a troubling road to go down without every player buying into rules and regs before entering online events with the understanding that there could be set OTB bans etc. I certainly wouldn't want 10 meetings a year to discuss bans, appeals etc




  • Yeah, I think the point is that you could do it in this case if - let's assume - it's a really obvious case of cheating. But how do you make that process consistent for everyone?

    Do you monitor every online account? Do you ask lichess/chess.com/gameknot/etc to send you a list of all banned players in the past month and review them? (GDPR will love that) And how do you show that player "HahaImCheating" really is ICU member 1234?

    Or does a ban only apply if it's flagged in the Guardian (or happens in something Bunratty online/4NCL online or one of the many European junior tournaments on at the moment)?

    It becomes a bit of a minefield. I don't see how it's practical.




  • cdeb wrote: »
    None of that really matters though. It's still a hugely suspect performance for a 1200. The general odds of an 1800 beating a 2500, rounded to two decimal places, are nil. Literally nil. I can tell you I've gotten plenty of great positions against FMs in blitz, and believe me there are plenty of ways of messing them up. And I'm not 1200 rated. That game against the WIM wasn't exactly queen against king stuff.
    Nil, or 1 in 1000. Two types of people I guess ;)
    I hope you don't think I'm in any doubt to the fact that the guy cheated. My point is that those 3 games in isolation is not enough evidence to conclude it. Nor should his elo rating in isolation be enough. The 3 games and his rating, in the context of the hundreds of other games available on the platform and his relatively high level of activity is what is really damning and I trust that's what chess.com factored into their decision.
    So I still take issue with your comment that people are saying "that a 1700 player catching an IM or a talented junior napping must be cheating" - nobody is saying that. And I still don't know why you're adding bonus points to this person's rating to make your point. I don't see any evidence of "gobsmacking elitism" or "players throwing their toys out of the pram", like you've suggested.

    The 1700 figure was mentioned in the Guardian article and Inquitus' original post on the topic. An extra 100 points were added somewhere along the way. Perhaps you'll disagree but I don't feel like arguing over 200-300 rating points at that level is worth anyone's time. Certainly not as significant as the few hundred between England's most famous junior and IM Bulmaga.
    He was labelled as a cheater before any investigation took place, that's the elitism. Imagine an unknown, but actually quite known and playing in his local tournament, club player having the audacity to beat some underperforming titled players and a 10 year old in some online blitz games. Preposterous.
    The toys out of the pram comment I will apologize for as there's no evidence that was the case.


  • Advertisement


  • I don't think you've argued that the three games isn't enough though (plus two other games earlier in the tournament, one against a 1900 I think). I've seen a player correctly identified as cheating on the basis of one game (correctly because they admitted it). The issue is it's not one game that's being evaluated - it's 25 or 70 moves or whatever it is. 3 games could be 150-200 moves. That's a lot of data points.

    Again, a 1 in 1000 shot I can buy. It's when you do it three times in a row, then the odds start stacking against you.

    The Guardian said he was "under 1700". His online rating was under 1600, and I think that's inflated. Let's be generous and say 1400. I will argue an 1800 with those results would be suspicious, but when you start reducing the rating by 400 points, it does add more to the case unfortunately.

    1400s - let's be generous - simply do not win games like that consistently. You can argue that the WIM defeated herself, but a 1400 would find many, many ways to lose that position. It was a very nice game and arguably not like typical 1400 play. (A 1400 would be inclined to cash in and take the rook on b8 for example). The GM game then - a 1400 does not beat a GM over 70 moves or whatever it was. They will make a blunder in the meantime. It's what being 1400 means.

    There was an investigation - and the site closed the player's account very soon afterwards. Possibly someone reported the games and the site took a day to evaluate them and decide. I don't know of course, but it would certainly fit the known facts.

    There's no elitism at all in suggesting that a 1400 (if that) beating 2100, 2500 and 2400 in a row and having their account closed by the site very soon afterwards may well have been cheating.




  • cdeb wrote: »
    I don't think you've argued that the three games isn't enough though (plus two other games earlier in the tournament, one against a 1900 I think). I've seen a player correctly identified as cheating on the basis of one game (correctly because they admitted it). The issue is it's not one game that's being evaluated - it's 25 or 70 moves or whatever it is. 3 games could be 150-200 moves. That's a lot of data points.

    Again, a 1 in 1000 shot I can buy. It's when you do it three times in a row, then the odds start stacking against you.

    The Guardian said he was "under 1700". His online rating was under 1600, and I think that's inflated. Let's be generous and say 1400. I will argue an 1800 with those results would be suspicious, but when you start reducing the rating by 400 points, it does add more to the case unfortunately.

    1400s - let's be generous - simply do not win games like that consistently. You can argue that the WIM defeated herself, but a 1400 would find many, many ways to lose that position. It was a very nice game and arguably not like typical 1400 play. (A 1400 would be inclined to cash in and take the rook on b8 for example). The GM game then - a 1400 does not beat a GM over 70 moves or whatever it was. They will make a blunder in the meantime. It's what being 1400 means.

    There was an investigation - and the site closed the player's account very soon afterwards. Possibly someone reported the games and the site took a day to evaluate them and decide. I don't know of course, but it would certainly fit the known facts.

    There's no elitism at all in suggesting that a 1400 (if that) beating 2100, 2500 and 2400 in a row and having their account closed by the site very soon afterwards may well have been cheating.

    We have commenters on this very thread, active in Irish chess, who weren't familiar with the player. I'm not convinced his accusers knew about his true strength. So this 1400 figure I'm not agreeing with in any discussion, about hypothetical players or not :rolleyes:

    The odds certainly do stack up but that's raw odds uncolored by any uncharacteristic play from the opponents. And one of those games against a 2100 with an enormous K factor. I don't think anyone is about to suggest an 1800, playing White, beating a 2100 is cause for huge concern.
    Maybe somebody can correct me if I'm wrong but I don't think the Bulmaga game is even significant from a cheating detection analysis as they tend to omit opening play and heavily winning positions. And I don't agree with your point about taking the rook in that game, Nxb5-c7+ winning a queen is also attractive to a 1400 player no? Actually the most unnatural move in that game for me, Qd4 instead of O-O is even an inaccuracy according to chess.com's weak engine, but perhaps not to something stronger, I haven't checked.
    Even the Pert game, the most suspicious, isn't without it's inaccuracies. You can't say somebody is cheating on the basis of a handful of moves. It's absurd. I absolutely detest cheating, in chess, in sport, video games, whatever. But there has to be a better way. If there's a cloud of suspicion hanging over anyone who plays a good game then the chess world won't be a pleasant place to be.




  • We have commenters on this very thread, active in Irish chess, who weren't familiar with the player. I'm not convinced his accusers knew about his true strength. So this 1400 figure I'm not agreeing with in any discussion, about hypothetical players or not :rolleyes:
    That's why we have ratings lists, so we can evaluate players' strength. Their Irish rating is sub 1200. Their online rating was sub 1600 before the tournament, and that's inflated (even the WIM's and the GM's ratings were 100 points higher than in real life). You mightn't like that this player's real strength is almost certainly 1400s tops, but that doesn't make it wrong.
    The odds certainly do stack up but that's raw odds uncolored by any uncharacteristic play from the opponents. And one of those games against a 2100 with an enormous K factor. I don't think anyone is about to suggest an 1800, playing White, beating a 2100 is cause for huge concern.
    Uncharacteristic play is absolutely taken into account in ratings. If I didn't blunder so often, I'd be a GM. But I am prone to lots of errors, so I'm 1800. You could even argue that errors are the only things ratings reflect.

    An 1800 beating a 2100 isn't relevant here as it didn't happen. The 2100's K factor isn't relevant either.
    Nxb5-c7+ winning a queen is also attractive to a 1400 player no?
    And yet they didn't win the queen either.

    I haven't seen a detailed computer analysis of the game (apart from chess.com saying 17 of 18 moves were computer moves). But neither have you, I suspect, so I'm not sure why you disagree with the site closing the account.
    You can't say somebody is cheating on the basis of a handful of moves. It's absurd.
    It's not a handful of moves. Between the three 2100+ wins, it's 132 moves. Add in the first two games and it's 200 moves.

    OK, some of those were opening moves and wouldn't be considered, but you absolutely can say someone is cheating on the basis of, say, 100 moves. I've seen it happen on the basis of fewer moves.

    It's admirable that you want to defend this player so much, but I do think you're being blinded rather too much to the possibility that the account was correctly closed.




  • cdeb wrote: »
    It's admirable that you want to defend this player so much, but I do think you're being blinded rather too much to the possibility that the account was correctly closed.

    I think you're confused




  • Hi All,
    I’m guessing that most people reading this will know who I am, and before I go on with my post here, I’d just like to point out that I am speaking privately here, I’m not speaking on behalf on the Bunratty Organising Committee. I’d like clarify a few things here, if I may….
    • Bunratty announced before the event that prizes would not be paid out until up to 2 weeks after the event, this was to allow time for Chess.com to do their “deep dive” data analysis of the event
    • While the event was going on, no member of the Bunratty Committee was involved in the running of the event in any way, Chess.com ran the event for us
    • After Chess.com finished their analysis of the event, two players had their accounts closed due to breaches of the Chess.com Fair Play Policy
    • One of these was Hari Prasad who withdrew from the event after his round 5 victory over Nick Pert GM, he was not ejected from the event
    • The other player was Michal Surowka who did complete the event, ending up on 6/9
    • As far as I know, both players are members of the Limerick Chess Club
    I’d also like to say that it’s a bit disappointing to read through 3 pages of posts and find the only subject being discussed is possible cheating in the event. All the committee worked very hard to put this event together and we’ve had some lovely emails thanking us for this. We even had one of the prize winners refuse the prize, thanking us for a very enjoyable event and asking us to have a few drinks on him as a thank you. Personally, I think Diana Mirza’s Twitch channel coverage of her participation in the event was very good and captured the spirit of the event very well.




  • Yes it is a pity that the cheating distracted attention from what was a well organised and fun event but I guess the people that deserve the blame for that are the cheats. One would hope that Limerick Chess Club will take strong disciplinary action against these players for bringing their club into disrepute and tarnishing all its members by association. I enjoyed the Bunratty event but found the time control far too short, something like 5 +2 would be much more to my liking


Advertisement