Lichess to FIDE rating converter

Hey everyone.
For those of you who have an established, ACCURATE, fide rating, I would be interested to know how accurately this thing judges you. Its a fide rating guesser, based on your lichess ratings. My current suspicion is that it guesses too low by somewhere between 100-200 points, but I could be wrong about that. Anyone who wants to, test it, and then tell me how accurate it is for your rating. Also, I hope this doesn't count as advertising since its not a lichess thing, and its not my work. If someone thinks it does, let me know.
DISCLAIMER: I DID NOT design this thing. That was someone else's work, but its available for anyone's use.
ethanlebowitz.github.io/RatingConverter/index.html

i think it gives indeed a too low number. Agree with 100-200 points

@ChessMathNerd it depends on the person that the computer is judging. For @EwoudUtrecht it might be too low, but for somebody who only plays blitz mindlessly and does not study, it is likely that the rating converter will be too high. I know a lichess user who has a 1300 Fide and 2100 lichess blitz and 1950 lichess rapid.

"I know a lichess user who has a 1300 Fide and 2100 lichess blitz and 1950 lichess rapid."
that's most likely because they didnt play OTB for a long time, due to COVID.

@EwoudUtrecht they had similar stats in 2019.

@ChessMathNerd said in #1:
> Hey everyone.
> For those of you who have an established, ACCURATE, fide rating, I would be interested to know how accurately this thing judges you. Its a fide rating guesser, based on your lichess ratings. My current suspicion is that it guesses too low by somewhere between 100-200 points, but I could be wrong about that. Anyone who wants to, test it, and then tell me how accurate it is for your rating. Also, I hope this doesn't count as advertising since its not a lichess thing, and its not my work. If someone thinks it does, let me know.
> DISCLAIMER: I DID NOT design this thing. That was someone else's work, but its available for anyone's use.
> ethanlebowitz.github.io/RatingConverter/index.html
its almost accurate see this id @ minidora he says 1950 and converter 1976 but with 1 it can't be defined we need more samples .

The methods used look ok, I suppose that if the fide ratings listed by lichess users on their profiles are at least vaguely meaningful, this estimator should work quite well. One possible source of systematic error might be in the fact that not all fide ratings are as up to date as lichess ratings, and, because most players tend to improve over time, this might result in lower estimates.

The way to create a justification would be to find out your predicted rating, then go get a rating, and report back on its accuracy.

Until a lot of people do that it has no statistical justification.

@jeffsonadam Yeah that is a serious potential issue. It said drnykterstein was around 2600.

@ChessMathNerd said in #9:
> @jeffsonadam Yeah that is a serious potential issue. It said drnykterstein was around 2600.
Oof!