My Blitz rating deviation is 45, and I wanted to find out how that compares to other players who are roughly at my rating. I know about lichess.org/stat/rating/distribution/blitz which is a terrific visualization of the distribution of ratings, but is there something similar for the distribution of rating deviation?
I don't think the information would be interesting.
High deviation is typical. It will apply to the vast number of accounts whose owner has dabbled with Lichess in the past, often years ago, but no longer plays. So the distribution curve across all accounts, or across all accounts with a rating close to yours (which is what you say you want), will just climb immediately and sharply as the deviation increases.
To give any sort of relevance of the data to what you are looking for, you would need to limit the pool of players to those who are currently active. But this is an artificial limitation. What is "active"? Played rated during the last month? Logged in during the last month? Or six months, or year? However you define active also impinges on the data you are compiling, because being active is directly related to having a low rating deviation.
High deviation is typical. It will apply to the vast number of accounts whose owner has dabbled with Lichess in the past, often years ago, but no longer plays. So the distribution curve across all accounts, or across all accounts with a rating close to yours (which is what you say you want), will just climb immediately and sharply as the deviation increases.
To give any sort of relevance of the data to what you are looking for, you would need to limit the pool of players to those who are currently active. But this is an artificial limitation. What is "active"? Played rated during the last month? Logged in during the last month? Or six months, or year? However you define active also impinges on the data you are compiling, because being active is directly related to having a low rating deviation.
the other reason why it probably would not be very interesting is that lichess has a floor for the rating deviation, which is 45 - source: github.com/lichess-org/lila/blob/89e1093521a53ee68bfe2b3740a2dceb1bb2ff5c/modules/rating/src/main/Glicko.scala#L67
a lot of players reach that floor.
a lot of players reach that floor.
I guess it would show that it would show an average higher RD on the "x00" anomalies, which would be a nice confirmation for the theory of people resting or leaving just after reaching those magic numbers.
@nadjarostowa said in #4:
> I guess it would show that it would show an average higher RD on the "x00" anomalies, which would be a nice confirmation for the theory of people resting or leaving just after reaching those magic numbers.
that's a good point, that would actually be interesting.
> I guess it would show that it would show an average higher RD on the "x00" anomalies, which would be a nice confirmation for the theory of people resting or leaving just after reaching those magic numbers.
that's a good point, that would actually be interesting.
This topic has been archived and can no longer be replied to.