Rating inflation?

I don't remember when we started having the rating distribution graph here, but some months passed. My perception is that we're having a rating inflation here on Lichess.

I mean: at now, I'm 1912 and "better than 87.8%" of lichess players who played "this month". But, if I'm not wrong, I would have been ~90% with the same rating some months ago.

Developers, can you show the rating distribution "by month", so we can see if it's happening a rating inflation?

The average score for classical players 50% is ~1612 this month. Was it lower in the previous months?

Players, what's your perception?

P.s. I don't care too much about ratings, but I'm curious about rating inflation: it happened and it's happening OTB, I think it's happening online, and here, too.

Regarding my own rating I had the same thought lately, as it seems for my rating to be about 50 - 100 points higher than it used to be a year or so ago. And I dont really think that my chess improved so much lately.

I wouldn't be surprised if there was inflation, but I would like to point out that an increase in the average rating doesn't necessarily indicate inflation.

Fewer low rated players playing classical time controls or strong players newly joining the site or returning from inactivity can be reasons why the average might increase without the rating actually being devaluated.

If the Gaussian bell is shifted to the right without a change in skewness the ratings are "inflated". The problem with this description is that "inflation" refers to the same unit.

Take a look at Kyle; he is playing chess occasionally and poor Kyle does not improve. Lets say this is a fact. If Kyle's rating somehow exceeds the boundaries of the confidence interval for the null hypothesis: "Rating/s are not inflated". Then you still can't make a statement about other players. With the above example you get an impression about the statistical problem you are facing.

personally I'd assume;
more active players use multiple accounts.
more active players improve their rating.
more higher rated players join the community.

You'd need monthly data for identical players over a longer period of time. But, you also need to show that the skill does not change. Therefore you'd need data for the same players from another website or OTB.

Greetings