@gradienthinking, we already had a chat offline so I won't repeat again here :) But for all the other users, feel free to analyze the above game, but I was not able to find anything suspicious in it. Also, as @gradienthinking would also agree now, you need to consider several factors while detecting the cheat. It is definitely not an easy task, even the Lichess system is not 100% perfect. So it is best to put your trust in Lichess mods and the system, it is not perfect, but it is the best you can do.
@sgtlaugh said in #71:
> @gradienthinking, we already had a chat offline so I won't repeat again here :) But for all the other users, feel free to analyze the above game, but I was not able to find anything suspicious in it. Also, as @gradienthinking would also agree now, you need to consider several factors while detecting the cheat. It is definitely not an easy task, even the Lichess system is not 100% perfect. So it is best to put your trust in Lichess mods and the system, it is not perfect, but it is the best you can do.
Agreed :).
@Cedur216 said in #65:
> in fact the bottom paragraph shows you're also siding with the trolls because you're just pretentiously making things up. Your "testing" is surely a humbug
I am not pretentiously making things up. I am a trained scientist with statistical knowledge. The point about testing was to illustrate that people in denial about cheating are just pretending they know the facts. I was trying to illustrate that NOBODY can be certain (including me). That is why I clearly tried to differentiate what I admit is opinion from fact. Stop trolling me!
> any Lichess mod can confirm that the vast majority of players is honest and the vast majority of cheat reports sent is inaccurate
Yes, the majority of players are probably not cheating. That doesn't have anything to do with what I said. I said in my opinion that cheating is more widespread than many people think, and almost certainly more undetected than people believe.
I base that opinion on the fact that it is obviously easy to cheat (i.e. almost impossible to detect) by only cheating for the occasional move or during the opening (presumably book moves aren't going to be flagged as cheating normally).
>
> And I guess there's safe grounds to assume that cheat detection is pretty close to perfect. @sgtlaugh by the way, when you also confirm that Lichess is not infested, why do you thumb up posts that support claims of the opposite?
It isn't just my opinion. A lot of top players (GMs) are fed up with how common cheating is online. See Hikaru, Naroditsky, Simon Williams, Caruana et al.
And, scientifically, it definitely isn't safe to assume that cheat detection is pretty close to perfect. If you don't understand why that is, then you don't understand science.
The problem isn't just cheating on this site. But also that mods easily lie about anything just to let everybody think that the people with authority have it under wraps.
@C-Bear said in #74:
> The problem isn't just cheating on this site. But also that mods easily lie about anything just to let everybody think that the people with authority have it under wraps.
They also delete posts, or the posts mysteriously vanish, with no explanation as to the reason, justification as to the motivation, or reference to the material to which they object.
@Pimander said in #58:
> I didn't say I know, I said in my opinion. NOBODY knows how many cheat undetected because that is what undetected means.
>
>
> If an "expert" says they know how many undetected cheaters there are, then they are not an expert, they are an idiot. Undetected means you don't know who they are.
Sure as a scientist you must understand the difference between knowing who the undetected cheaters are, and estimating how many of them are there. While the former is clearly impossible (by definition, as you so shrewdly observed), the later is very well possible. Mathematical models used by chess.com may or may not be sufficiently accurate, but if their estimates are off, it is not because such estimation is impossible in principle.
I totally get that it is possible to estimate using statistical models.
The problem with almost all such models is that the starting point of any model has to be based one or more assumptions. If the assumption is wrong (i.e. or hasn't been proven scientifically) then the model can be mathematically correct but the estimate completely wrong.
This topic has been archived and can no longer be replied to.