Cheating in chess

Discuss anything you like about chess related matters in this forum.
NickFaulks
Posts: 8453
Joined: Sat Jan 02, 2010 1:28 pm

Re: Cheating in chess

Post by NickFaulks » Tue Aug 03, 2021 11:03 am

Nick Burrows wrote:
Tue Aug 03, 2021 10:53 am
I'm playing a guy in some daily games on chess.com. He has a rating of 2150. His blitz rating is 1450 from 1500 games. Is this rating disparity plausible?
In daily games, I believe you are allowed to shunt the pieces around on a practice board. If so, then definitely yes.
If you want a picture of the future, imagine a QR code stamped on a human face — forever.

Roger de Coverly
Posts: 21301
Joined: Tue Apr 15, 2008 2:51 pm

Re: Cheating in chess

Post by Roger de Coverly » Tue Aug 03, 2021 11:11 am

Nick Burrows wrote:
Tue Aug 03, 2021 10:53 am
I'm playing a guy in some daily games on chess.com. He has a rating of 2150. His blitz rating is 1450 from 1500 games. Is this rating disparity plausible?
Analysis and consultation of "books" is permitted in daily chess, even if engine use is banned. Before computer engines it wasn't uncommon for high ranking postal players to be not that good at faster forms of chess. Supposedly chess.com monitor daily games for engine usage. A while back though ejh reported that his account had been flagged presumably for engine use even though he'd only followed some deep lines in recently published books.

Perhaps take a look at some of the blitz games to see how your opponent loses.

Nick Burrows
Posts: 1705
Joined: Sat Aug 14, 2010 12:15 pm

Re: Cheating in chess

Post by Nick Burrows » Tue Aug 03, 2021 12:05 pm

Both ratings appear to be legit. Which means some MUCH weaker players are superior in assessing positions and that memory counts for the vast majority of rating difference.
I suppose it's possible that he's spending an inordinate amount of time on each move.

Stewart Reuben
Posts: 4542
Joined: Tue Apr 03, 2007 11:04 pm
Location: writer

Re: Cheating in chess

Post by Stewart Reuben » Tue Aug 03, 2021 12:15 pm

NICK >He has a rating of 2150. His blitz rating is 1450 from 1500 games. Is this rating disparity plausible?<

i suspect my blitz rating would now be much lower than my standard play rating now. That is why I gave up that form of chess. Is his age known? If he is cheating, using a computer for slow chess, why is his rating not much higher? Once your suspicions have been aroused, it is very difficult to recover.

NickFaulks
Posts: 8453
Joined: Sat Jan 02, 2010 1:28 pm

Re: Cheating in chess

Post by NickFaulks » Tue Aug 03, 2021 12:21 pm

Nick Burrows wrote:
Tue Aug 03, 2021 12:05 pm
some MUCH weaker players
You are defining "weak" to mean "no use at online blitz". Perhaps that is the modern definition.
If you want a picture of the future, imagine a QR code stamped on a human face — forever.

Thomas Rendle
Posts: 469
Joined: Tue Aug 10, 2010 8:31 am

Re: Cheating in chess

Post by Thomas Rendle » Tue Aug 03, 2021 12:37 pm

Nick Burrows wrote:
Tue Aug 03, 2021 10:53 am
I'm playing a guy in some daily games on chess.com. He has a rating of 2150. His blitz rating is 1450 from 1500 games. Is this rating disparity plausible?
I coach some older players and often the methods they use to improve in slower games make little or no difference to their blitz ability (or can even be harmful if they're carefully checking moves instead of going with the flow). I'm inclined to give players in daily chess the benefit of the doubt on this one!

Pete Heaven
Posts: 58
Joined: Sun Dec 08, 2019 9:47 pm

Re: Cheating in chess

Post by Pete Heaven » Fri Aug 13, 2021 8:01 pm

It's been months since my last substantive post on cheating in online chess but, with the controversy in the British still fresh in people's minds and prominent in today's Guardian column, I thought I'd just provide some food for thought. After all, people may wonder how the 'leader' after round 8 of the main event managed to evade detection by the anti-cheating software (and, in his case, Zoom) for so long - there was the qualification event too, of course.

I'd just like to give one example, one which reinforces Leonard's comment that 'suspicious results need to be investigated earlier'. Although my example does not involve the battle for the British Championship itself, which puts it more under the radar, it is far more extreme and does not require the word 'earlier' that Leonard chooses, a word that I believe has been superfluous in 2021. I've kept everything as vague as possible except a few key numbers, since the identity of the person is irrelevant. What is relevant is how obvious it can be that you're cheating and how long you can do it for, without being caught.

By the way, this player is not a junior, improving or otherwise; he is rated in the bottom half, OTB, in his club. His rating has been pretty stable over many years. He is a cheat but is far from the worst in the UK, in my opinion.

Early on in lockdown chess, he played a few 4NCL tournaments/league games without much success, although far from a disaster. He had a small minus score across twenty odd games. He finished above his seeding in one event and below in the remainder. His performance rating was around 100 ELO less than his OTB rating. The key number here is 67.5, his median centipawn score across these games, the number which encapsulates the accuracy of his games.

Fast forward to today. He has played in ten more similar tournaments, plus further league games. These were all subject to the same anti-cheating scrutiny. The opposition has been stronger, around 240 ELO points on average, so a huge step-up in fact. He has, nevertheless, increased his score from under 50% to nearly 85%, finishing out of the top few places in his events only once and winning some of them. His rating performance, as will be obvious, is many hundreds of points higher in these events and many of the games are absolutely superb. In these games, his median centipawn score was 22, with over 60% of his games being 23 or less.

For people not familiar with what a 67 or a 22 centipawn score means, here is some help, using two of my own statistics. Firstly, my online bullet centipawn score is in the low to mid-60s, over more than a thousand games, which is pretty poor but I am getting on a bit. At the other end of the time controls, in OTB classical chess, centipawn scores for each season of my life are routinely between 30 and 34, almost always, since I was 15 or 16. Again well over a thousand games. In my two best seasons, I achieved lower than 30, but only just, courtesy of them being the two seasons with the highest proportion of short draws, a foolproof way of keeping one's centipawn score down. My lifetime median is 31.

I have also looked into some titled players' OTB centipawn scores and none of them match 22 as a median score, not even close.

So, the person in my example has gone, pretty much overnight, from being worse than my blunder-ridden online bullet play to being incomparably better than my best ever OTB classical season....and I've had a lot of seasons...and a lot longer than 45+15 in each game to find better moves....and, better still, OTB I've not had to play anyone cheating which is another major help in keeping my centipawn scores down. Finally, my OTB rating, despite being in freefall for years, is still several hundred points higher than this person.

I have presented things above to make it crystal clear that the person concerned is cheating. Also, I have analysed the games and any tiny chance he isn't melts away rather quickly. In fact, I thought his cheating was blindingly obvious many months ago.

My question is - how can one get through TEN consecutive tournaments like this, including British Championships, 4NCL congresses plus some league games, without triggering some sort of exclusion? Why is the anti-cheating software failing to such a mind-boggling degree?

It may be that the answer to this is that the software isn't failing at all but the spectre of legal action from people claiming a false positive is enough to leave the 4NCL/ECF/WCU...or whoever, with no intention of excluding anyone from a tournament. If this is true, shouldn't people be told? At least they would know what the unwritten rules really are before parting with their entry fees and committing their weekends.
At present, the Fair Play rules merely exist as a deterrent.

In the uevent that anyone chooses to dispute the contents of this post, I would be grateful if they started with my example above. After all, the whole concept of Fair Play would be a joke if just this one example were true.....wouldn't it?

===================================================================================================

One final note - I have a lot of faith in centipawn scores, having looked at thousands of games. For example, I recently looked at a large number of players, across a wide range of grades, and looked at their last 30 recorded OTB games. My aim was to see. despite 30 not being a massive sample, how close their ordered centipawn scores would be to their grading order. It turns out that ordering people by OTB centipawn scores also orders them almost exactly in OTB grade order as well.

However, online centipawn order = OTB grade order? Not so much. Not so much at all.

User avatar
JustinHorton
Posts: 10364
Joined: Mon Aug 04, 2008 10:06 am
Location: Somewhere you're not

Re: Cheating in chess

Post by JustinHorton » Tue Aug 17, 2021 11:56 am

Pete Heaven wrote:
Fri Aug 13, 2021 8:01 pm
My lifetime median is 31.
My mean score over my past thirty lichess games, all played at 7+5, and neglecting one terminated after three moves, is about 30.46, I think. (I think my median score might be 25 although I also suspect you shouldn't be calculating that over that small a sample of games.)
"Do you play chess?"
"Yes, but I prefer a game with a better chance of cheating."

lostontime.blogspot.com

User avatar
JustinHorton
Posts: 10364
Joined: Mon Aug 04, 2008 10:06 am
Location: Somewhere you're not

Re: Cheating in chess

Post by JustinHorton » Wed Aug 25, 2021 10:23 pm

I know nothing of Call Of Duty: Warzone so while this may be relevant, also it may not
"Do you play chess?"
"Yes, but I prefer a game with a better chance of cheating."

lostontime.blogspot.com

Matt Bridgeman
Posts: 1077
Joined: Wed Aug 03, 2016 9:21 pm

Re: Cheating in chess

Post by Matt Bridgeman » Wed Aug 25, 2021 10:31 pm

I did spot on Simon Williams’ You Tube page that his latest video was trying to do an informative speed run by starting out playing 800-1000 rated players on chess.com
...he then promptly loses to two 800’s with 99.5% accuracy ratings!

Matt Bridgeman
Posts: 1077
Joined: Wed Aug 03, 2016 9:21 pm

Re: Cheating in chess

Post by Matt Bridgeman » Wed Aug 25, 2021 10:38 pm

JustinHorton wrote:
Wed Aug 25, 2021 10:23 pm
I know nothing of Call Of Duty: Warzone so while this may be relevant, also it may not
That’s quite a big number. 700,000 accounts banned. So just shy of the total for the last British Online Championship. Lol

User avatar
Joey Stewart
Posts: 1860
Joined: Wed Apr 11, 2007 2:35 pm
Location: All Of Them
Contact:

Re: Cheating in chess

Post by Joey Stewart » Thu Aug 26, 2021 2:42 am

It is quite relevent as it shows just what a large percentage of the video game playerbase indulges in cheating and with chess being a bit trendy right now we have picked up a lot of new players who are coming into the game with the mindset that it is ok to behave this way too.
Lose one queen and it is a disaster, Lose 1000 queens and it is just a statistic.

Joseph Conlon
Posts: 339
Joined: Thu Jun 06, 2019 4:18 pm

Re: Cheating in chess

Post by Joseph Conlon » Sun Sep 26, 2021 3:27 pm

Sorry to bump this, but I remember some discussion of what an 'engine move' is, and I had an example yesterday of something that I thought at the time was definitely an 'engine move' (and on reporting lichess agreed with me). So posted here for interest.

The position below, I am White (and clearly losing), Black to move played 29. ... Re1 - which I think very few human players would choose. The move depends on seeing 30. Kxe1 Qxg3+ 31. Kf1 Qxf3+ 32. Kg1 Be4 and correctly evaluating it as an unavoidable forced win for Black. On the other hand the move is (together with the more natural 29. ...Re6 or 29. ... Re7) one of the top three engine choices.
Screenshot 2021-09-26 at 15.26.23.png
Screenshot 2021-09-26 at 15.26.23.png (125.33 KiB) Viewed 823 times

Roger de Coverly
Posts: 21301
Joined: Tue Apr 15, 2008 2:51 pm

Re: Cheating in chess

Post by Roger de Coverly » Sun Sep 26, 2021 3:43 pm

Joseph Conlon wrote:
Sun Sep 26, 2021 3:27 pm
Black to move played 29. ... Re1 - which I think very human players would choose.
I suppose you mean very few. Looking at the position though, if as Black you were feeling pessimistic about your position, you might consider .. Re2+ with the idea that after Kxe2, you take the Knight with check and may have perpetual check opportunities. Having seen that, the idea of going one move further down the board may come to mind, particularly if .. Be4 ideas had also been considered.

Anyway I thought that lichess and chess.com were sufficiently respectful of probability theory not to base bans on a single move or single game. That's what they claim anyway.

Eric Gardiner
Posts: 115
Joined: Sun Dec 06, 2009 10:42 am
Location: Hull

Re: Cheating in chess

Post by Eric Gardiner » Sun Sep 26, 2021 3:59 pm

How much time did Black have at this stage? If short of time, an innocent explanation is that the intention was ...Re2+ but the mouseslip ...Re1 occurred. It looked to me like Black had compensation for the exchange after ...Re2+; Kxe2, Qxg2 - so it might have been an intended winning rather than drawing attempt. Checking with an engine confirms that Black is winning in this line.

Post Reply