Cheating detection absent in lichess

#1 and #20

2 sides of the same coin
both prove the point of why sample size matters and data accuracy

If i had a super computer that could go through all 96mil games from August
i would i am interested to see how good lichess detection is but we must also bare in mind the accuracy of whatever software we use

@Morobiondo said in #15:

You seem to not have used the report system much - True, I never used that because I think it is not right to report on people
just on a suspicius basis

then you're all the more not entitled to your claims (I mean, you aren't anyway, because your attitude is simply not true). Reporting by suspicion and gut feeling is everyday business and mods get along well with it.

As far as I'm aware you don't get a notification in lichess whereas you do on chesscom.

Perhaps you've noticed more there.

@Dean179 if you report then you get a Someone you reported was banned. If you lose to a cheater, you get a refund under certain conditions.

@Deadban said in #7:

He's not pointing out that chesscom is free, because it's not, but that he's on the free plan.

You mean the Ad Bombardment Plan.

Refund policies on Lichess and chess.com are (probably) different. I do not know where chess.com's refund policy is published, but Lichess' is described in the FAQ and I believe that chess.com does give out rating refunds much more often then Lichess. As written above, if you report players you will get a notification if such a player gets banned. If you do not report players, it might still happen that some of your opponents get banned for fair play violations, but unless you check the account states of your previous opponents or you qualify for a rating refund you might not learn about those bans.

@Morobiondo said in #1:

Over 6K games played and no chetating detected. In chess.com (free) they detect and find this kind of abuses...

You might wanna rethink this: https://www.chess.com/de/member/vladimirkramnik bio.

Lichess seems to have a pretty good cheat detection. They just don't shout it in everyones' face when they ban someone.

@for_cryingout_loud said in #19:

last month ther was 96,118,124 games
6000/96,118,124 * 100 = 0.00624231908% of the games
to call that Representative of the user experience would be foolish
as it is a fraction of a percent of the games of 1 month

Plus as i alluded to before 6k games with no cheaters does not tell us anything
all 6k games could have cheaters that were not found
or some or none

and without knowing that we cant tell if 6k being none is good because you played none
or bad because none were caught.

That's not how statistics works.

A representative sample can absolutely stand in for the entire population. Twenty thousand randomly-chosen games would, 99% of the time, have a rate of cheating within 1% of the actual value for all games that month. Within 2% of the value, 95% of the time? Just a couple thousand games. Within 5% of the value, 95% of the time? Four hundred games.

Their "6k games with no cheaters" claim is not a representative sample, but that's not an indictment of statistics—just of them. They seem to be unaware that only the last 40 games within 3 days of being caught are eligible for refunds. I'm sure some of those games were against cheaters who were eventually caught, but not within 40 games / 3 days of that game.

@Eiszett said in #29:

That's not how statistics works.

A representative sample can absolutely stand in for the entire population. Twenty thousand randomly-chosen games would, 99% of the time, have a rate of cheating within 1% of the actual value for all games that month. Within 2% of the value, 95% of the time? Just a couple thousand games. Within 5% of the value, 95% of the time? Four hundred games.

pls site source and funny enough you have been proven wrong in this thread
#21 i meantion the fact of how a person with 6k games found none but a person with 3k gets them every 200 proving that it is to small of a sample size but if your sample size for your data is a fraction of a percent and you want to tell me it means something it does not. Never mind the problem of all the data being from 1 person or the fact we still need to analyses the data

Their "6k games with no cheaters" claim is not a representative sample, but that's not an indictment of statistics—just of them. They seem to be unaware that only the last 40 games within 3 days of being caught are eligible for refunds. I'm sure some of those games were against cheaters who were eventually caught, but not within 40 games / 3 days of that game.

true but it could be 0 or all or somewhere in between without anaylising the data we are just wasting time