cheating detection

Pretty hard for a few people to review a million FREE games.
Start a team and tournaments with other honest players at your rating level. If enough members think a player is dishonest boot them from team/tournaments.

@for_cryingout_loud said in #5:

Only a fool
You also could've used better language.

@Munich said in #1:

occasionally I got cheated, I reported the cheater, and many of them got banned.

But for some weeks now, I report more and more suspicious games, and none of those game lead to a ban.
Yes, I agree, it is sometimes not for sure if someone has cheated, and we all can have excellent games.

I dont expect every report of mine will lead to a ban. But so many players have recently scored against me high accuracies with very low average centi-pawn loss - and none of those reported games resulted in a ban.

Isnt cheat detection meanwhile half-automated? A few months ago, it often only took seconds for a cheater to get banned after I reported him.

I wonder - anyone else experience that more and more games have too excellent play? The less cheaters get detected, the less they get banned, the more auditiously they start cheating. I am not singeling out anyone, just wonder if the administration or automatic cheat detection is "under water"?

ah, and casual games - is cheating allowed here?

I keep seeing people annoyed at cheaters like you and this is what I dont ever get : If your opponents are cheating, then why are they not already like 3000+ yet with a massive winrate? If they are losing games on purpose and staying at the same rating, why do you think they cheat for? and whats the difference between them and like a 2100 level chess.c*m bot?

you are right, if cheaters continue to cheat, their rating should continue to climb.
However, even though this is not a metric on its own, I noticed that cheaters often have fresh accounts. I started to abort games when I notice that they are like "2000?" rated with only 30 games played. Though, if that happens too often, I risk myself a penalty time out for aborting too often games.

Many cheaters have not made it even into the 2000s, previously. What I experience at the (around) 2000 rapid rating level are maybe players that had previously been caught more early, and thus they never reach the 2000+ ratings. But as of late, they seem to have done so.

I also think that lichess check routinely any player above 2500 rating (as not too many have that), maybe by first calculating their performance over their last 20 games., and if the performance is then a 3000+ performance, maybe then an automated quick check is done. How exactly cheaters are caught - I dont know. Lichess keeps that secret, so cheaters do not know how to circumfence cheat-detection.

But ye, cheaters hardly reach 3000+, at some point they get caught before that.

@Munich said in #14:

I started to abort games when I notice that they are like "2000?" rated with only 30 games played.
What exactly is the logic behind this supposed to be?

Though, if that happens too often, I risk myself a penalty time out for aborting too often games.
Which is only fair. Aborting games because of some absurd "signs of cheating" criteria is a bad habit that should be punished.

@mkubecek said in #15:

What exactly is the logic behind this supposed to be?

read the comment before: cheaters would get to rating 3000+ eventually, but certainly not if they have only played a few games.
Many cheaters get banned when the reach, Idk, maybe 2300 or 2500 (I dont know where, but they usually never get to 3000). At some point lichess seems to automatically detect those fresh accounts with 2500 rating and performance of over 3000+ over their past 20 games or so.

But when such players play me, they have not reached a very high rating such as 2500. So, even though most are innocent, I abort games against accounts that want to play me but have a questionmark behind their rating (I gave "2000?" as an example). The question mark indicating their rating isnt stable, and the few games indicating they are a fresh account. If you get to a rating of "2000?" within only a few games, their performance is considerably higher.

Not playing such players saves time, as I dont need to then bother to report them.
really, fresh accounts have a bigger share of cheaters than well settled accounts. Those of you who regularly report cheaters can certainly confirm that more of half of those reports are made against players who have played only a few games (=fresh accounts).

@Munich said in #16:

Many cheaters get banned when the reach, Idk, maybe 2300 or 2500 (I dont know where, but they usually never get to 3000). At some point lichess seems to automatically detect those fresh accounts with 2500 rating and performance of over 3000+ over their past 20 games or so.
Pure speculation.

The question mark indicating their rating isnt stable, and the few games indicating they are a fresh account.
It took me two months to get stable rating (classical, that is; my rapid rating is still provisional after a year and a half). And 30 games is usually something like two months for me.

If you get to a rating of "2000?" within only a few games, their performance is considerably higher.
Not really. I won my first two games against opponents rated ~1500 and ~1570 and after just those two game I had (provisional) rating of ~1780. If I happened to win two more, I would likely be quite close to 2000. I'm pretty sure if you started from scratch now, you would reach 2000 before making your rating stable. How would you like being automatically treated as a cheater by other players and having games aborted for absurd reasons?

really, fresh accounts have a bigger share of cheaters than well settled accounts
Pure speculation again. Or do you have some hard data confirming that? And even if it is true, do you really believe the fraction is big enough to justify your behaviour?

Those of you who regularly report cheaters can certainly confirm that more of half of those reports are made against players who have played only a few games (=fresh accounts).
I thought you were talking about users who are cheating, not users reported for cheating. It's not nearly the same. The sad truth is that some people tend to report others for all kind of weird reasons.

You may not remember but your rating was also provisional once. Actually, some of your ratings are provisional right now - and two of them are "2020?" and "2156?". Would you say your opponents should treat you as a potential cheater and abort games against you just because of that?

@mkubecek said in #17:

Pure speculation.
find me a banned account who got that far? read discoooooord's comment, he has a good point indeed.
@discoooooord said in #13:
why are they not already like 3000+ yet with a massive winrate?

@mkubecek said in #17:

And 30 games is usually something like two months for me.
unfortunately I cant quick-check how long someone needed. the longer (like in your case) the more likely he isnt a cheater. most players are not cheaters. However, if a cheater gets banned, he respawns with a fresh account. Most cheaters have fresh accounts as they usually get banned at some point. So the proportion of cheaters is highest among the fresh account population. This is bad for the fresh accounts that are honest fair players. And freshers do have probably a higher amount of aborted games because, like me, people are not so fond of playing fresh accounts.
Still, sorry, I fare better if I dont play fresh spawned accounts, and most are only a few days old if they have a "?" behind their rating (but again, I cannnot quick-check that, and time ticks down when I need to decide to abort or not to abort a game).

@mkubecek said in #17:

(provisional) rating of ~1780. If I happened to win two more, I would likely be quite close to 2000.

Which isnt so likely, is it? Possible yes. Likely? increasingly not so much. It might be hard to believe for you, but even if you manage that, if you put more similar accounts into one statistic sample you will discover that the likelyhood of cheaters increases in this sample.

But if it is not a cheater, it is likely an underrated player, which is also not great to play against. I prefer opponents with a stable rating, it just gives me peace of mind.

@Munich said in #18:

most players are nnot cheaters
And yet you treat them as such. :-(

However, if a cheater gets banned, he respawns with a fresh account.
Not nearly always.

Most cheaters have fresh accounts as they usually get banned at some point. So the proportion of cheaters is highest among the fresh account population.
Here you are silently assuming that cheaters - or at least big majority of them - cheat from the very start. According to people who have real data, that's not exactly true. And many cheaters who are willing to talk about it say that they only started cheating to "even the odds", i.e. out of frustration that too many of their opponents are cheating. So convincing oneself that cheaters are everywhere around seems to be common first step on one's path to the Dark Side.

I need to decide to abort or not to abort a game).
That's an easy decision: don't abort games unless you have a serious reason (e.g. unexpected real life stuff that cannot/shouldn't wait). And no, a question mark is not a serious reason. Such behaviour is awful and you should stop.