You are better than % of players.

Hello, how is this stat (in title) calculated?
Because my friend dannier_uwu has (at the time of writing this) rapid rank 239 243 and weekly rating distribution says there's 472 177 rapid players this week. So I would assume he would be better than [100*(472177-239243)/472177] % = 49,3 % of Rapid players. But his profile says he's only better than 33,5 % of rapid players.
He's new player (at least on lichess), but his rating deviation is 58.49 (calculated on 37 games). Not sure if that matters.
Thanks for the answer.

Not sure if that can explain the comparably huge difference, but the “better then % players” only takes those players into account that were active in the category in the last week. Maybe the absolute rank looks at a different set of players.

Though actually on the rating distribution curve it also says "xxxxxx players this week". Otherwise the number would skyrocket indefinitely.