Cheater gameplay? - part 1

@elektron2kim said in #18:
> I just experience wrong numbers. Good players at 800-1000 and why don't they move up then? Find it more fair in tournaments. Might stop with quick pairing or expect the stats to be shot down forever. I don't mind losing 200 rating and move up later, however, it's looking bad in my opinion.

I don't know who you are playing, so I can't say exactly - but I do know I was 900 for a long time despite being closer to 1100 in strength because I barely played any games. After I decided to play in a lot of tournaments, I gained 200 rating in one day. I can't prove anything, but that might be what's happening maybe if your opponent has only played 6 games. I remember I was extremely low rated because I lost six games in a row when I first signed up to Lichess.

They could also be sandbagging or doing something else, in which case don't listen to me.

@cherrytea78 said in #21:
> I don't know who you are playing, so I can't say exactly - but I do know I was 900 for a long time despite being closer to 1100 in strength because I barely played any games. After I decided to play in a lot of tournaments, I gained 200 rating in one day. I can't prove anything, but that might be what's happening maybe if your opponent has only played 6 games. I remember I was extremely low rated because I lost six games in a row when I first signed up to Lichess.
>
> They could also be sandbagging or doing something else, in which case don't listen to me.

I appreciate your input.

Such an effort! :-)

About analysis with STOCKFISH:

- when tactical moves are coming, STOCKFISH can help very much...

- when positional moves have to be found, STOCKFISH has difficulties to understand what to play...
In this case, I analyse my games by forcing STOCKFISH to look at some particular moves... & not always to look at the moves that it wants to force me to look at... After a while, it recognizes that some moves - it didn't see at first - are better than moves it wanted to "play".

Once, I got an issue with some guy during a game (5'+3"): they quit the game - after having lost a piece without any compensation - during exactly 29 seconds... & come back... then beat me really fast, without any time spent for every move.
Soooo nice... :-)

Very cool blogpost. I'm not sure how the current cheating policy works but you could definitely apply the same approach to all time formats and help feed a ML model with these features

interesting blog

@oortcloud_o absolutely amazing stats

can you debunk some myths about cheaters like:

there is more cheaters at 1500 since more people start there.
would really love to know how true that is

and a hard one
has the amount of cheaters increased over time with lichess
and/or has lichess cheat detection gotten better

for this i recommend pgnspy github.com/MGleason1/PGN-Spy
as third party eval for cheat to see how accurate lichess is
but it might be hard to compare and produce an accurate statement with it
but with your skills as shown in your blog you can do it

Oh wow I love these graphs! Very nice post! :D

Es una pena perder tu cuenta de lichess , siendo baneado y puesta la marca roja ,deberían ser mas equitativos con eso. Hay jugadores con muchos puntos elos los cuales pueden hacer trampa y nunca los banean solo por ser patrones de la pagina con unos juegos tan perfectos que solo la maquina los juega así, desmeritan porque tienes una cuenta sin ser titulada y piensan que por eso no eres bueno y tienes que hacer trampa en tus partidas.

Es lógico que las partidas bala o blitz son imposible de que las ganes perfectas mas cuando juegas con un jugador fuerte hay si debe haber errores e imprecisiones , esos que ganan esas partidas si son tramposos en poco tiempo no puedes calcular y hacer las mejores jugadas y obtener la máxima precisión.

si van a banear que los baneen a todos no solo por preferencias de jugadores , per di mi cuenta y eso que jugaba partidas clásicas y rápidas donde puedes jugar mas preciso porque hay mas tiempo , pero debo decir que lichess es la mejor plataforma su jugabilidad y grafica son geniales ademas de ser gratis . Excelente blog me ha gustado leerlo.

Good example of use of graphics to give a sense of the information that can be inferred from lichess data, when asking well posed questions and choosing clever end-points, clear ones.

Presentation of data analysis wise, also good display of the nature of the data spread. And of the methodology steps and various angles.. a good basis to start discussions based on more than one person's experience. Factors to be debated, other variables to be included? well how could we discuss, if the first proposal did not give handles for the thinking.. so . good blog. I should read it for the results too. later.