Detecting Cheaters

@RED-WARRIOR777 said in #17:
> can u paste that user here

I am not sure if thats allowed according to the forum etiquette - would that fall under public shaming?

@BorisOspasky said in #18:
> @broken_pawn_hell Excellent job IMHO. It’s taken me two years to climb about 300 points to1400 (but I don’t play thousands of games) so why should genuine players be disadvantaged by damned cheaters? Good job!

Thank you :)! IMO it heavily depends on how much you play and how you analyze your games - I did that consistently now for the last months. At first I was only better than <10% of all players in both rapid & blitz and thought I am a lost case... but now I got addicted to chess :D. I am still worse than more than half of players on this page, but I hope to get past that point in the near future :)

#21: That Would Be Public Shaming; Send It In A DM.

Well done on reporting them and catching them:)))

Good for reporting - report, and let lichess look it at it as they have access to way more back-end stuff than we do playing.

Some people get mad when they lose, and assume everyone who beat them is cheating, but if you are calm about it, then the human brain is an incredibly good neural network and there are lots of small ‘tells’ that one can use that, added together, somehow ‘work’ - at least in that they agree with lichess’s assessment of engine use. More than 90% of the time I have reported someone they then subsequently got banned.

The one strong tip I have is if you are playing a suspected engine user, is always play to mate and don’t resign. Engines do fancy mates a queen up that humans would never consider, who just exchange off and take material. And naive engine users can’t speed up and blitz out a win when (eg) a rook up in an endgame, which any ‘real’ 2400 lichess can.

The best way to catch a cheater is if you lose to them. If you lost, he was cheating. Hit report button for mental satisfaction knowing nothing will happen. Feel better.
Play new game. Very easy.

@broken_pawn_hell You are going to find a lot of apologists defending cheaters but I go with my gut feelings.

In the thick of it, we sense the tone and play change in our opponent, it's like another gear turns on, usually after a blunder or mouse slip or anything to justify their cheating, I'd imagine.

The truth is it's very hard to catch a smart cheater. I look at win percentage the most. Move times are obvious but the biggest thing that gets cheaters is impatience and dupers delight, where they start to get bold and greedy.

Anything much above a win rate of 70% is suspicious and raises eyebrows because that means they are winning a ton of games as black.

Also look at the average rating of their opponents, in their defense because it's possible to manipulate ratings gradually by hunting for weak players.

In short, there are ways to game the system and then there is cheating.

Your arguments seem sound but I'm a sucker for graphs, stats and visual evidence too.

:D.

Ok on a serious note. If Lichess really wants to catch and ban cheaters. They need to adopt the "secret shopper" strategy. Send full strength Stockfish bots to go play in human games. If a human loses to them they don't lose any rating points. If someone beats the bot... They're cheating. Auto-ban. No moderator necessary. They could reduce the playing strength to give the cheaters the ability to beat it with weaker engines, but keep it tough enough to never lose to a real human. So maybe about 3100 ELO or so. A decent engine like Komodo and older versions of Stockfish could beat the bot, but no human would be able to without cheating.

I bet this bot would clean this place up real quick. Make a few of them and turn them loose.

#1 Dang, you did the mods' job yourself!

I totally agree with @Ganjman 's idea. I don't understand the thumbs down on his comment.

Cheating has become so big that it's time to treat it seriously, otherwise online chess is doomed.

I personally am very close to giving up because the number of cheaters is so great.

the current measures are obviously too ineffective.