Free online Chess server. Play Chess now in a clean interface. No registration, no ads, no plugin required. Play Chess with the computer, friends or random opponents.
Play
Create a gameTournamentSimultaneous exhibitions
Learn
Chess basicsPuzzlesPracticeCoordinatesStudyCoaches
Watch
Lichess TVCurrent gamesStreamersBroadcasts (beta)Video library
Community
PlayersTeamsForumQuestions & Answers
Tools
Analysis boardOpening explorerBoard editorImport gameAdvanced search
Sign in
Reconnecting
  1. Forum
  2. General Chess Discussion
  3. Ever noticed the "corners" in the cumulative distribution curve?

Well, those corners/peaks/edges are visible in the differential curve as well. I find it remarkable that it can be seen at all full hundreds. I just looked at the blitz distribution btw. because that's the only one I have got and I'm interested in.

I think these are those experts who climb onto or narrow beyond a certain step and then "retire" for a while. Or they play solely unrated, only other variations possibly or open new accounts...

So, if you wanna take over a lot of them in terms of the percentiles you just have to climb over the full hundred number plus a bit more - then you get an extra many of them. :D

I was wondering about the same thing today. For reasons I can't explain, it bothers me that these curves are not smooth.

I think your explanation is correct. However, I think it would be less visible if:
- instead of intervals of 25 elo points, every score is counted separately.
- players with a large uncertainty on their elo score are not taken into account (not sure if the case now or not)

This topic has been archived and can no longer be replied to.