2

I occasionally play lichess anonymously to just hang out without having to worry about winning and tarnishing my record, however I am bad at guessing players' elo.

Improve this question
2
  • 1
    In my experience from a while back, the player strength is higher the shorter the time control. However, it varies a lot (obviously, as there's no rating to base the matchmaking on).
    – koedem
    Commented Jul 3, 2023 at 8:54
  • 1
    Perhaps I'm picking nits, but LiChess doesn't use Elo. They use Glicko-2: lichess.org/page/rating-systems
    – GreenMatt
    Commented Jul 5, 2023 at 22:30

2 Answers 2

Reset to default
2

My evaluation is that the average opponentthat you will find does NOT have the average elo.

  1. Both in rated and non rated the distribution of games does not match the distribution of players. I am not sure if the statistics are published but in short you will find that higher level players more often play many games (I guess practice does make perfecton average). And this is even stronger biased in bullet games. One reason is that on the high level playing bullet reduces the number of secret bots they may face. Another reason is that many weaker players are simply unable to reliably complete games reliably before running out of time, even if they are winning on the board. So many of the low ratings will be based on just a few games per month.

  2. The lack of match making drives up the curve. Given point 1 we would already expect above average opponents. However in rated that would not be much of a problem as you still get reasonable matches towards your own level. However in unrated lower players will find themselves losing a lot, resulting in fewer of them playing.

Last but not least a proof point:

  1. I play both 1 min and 5 min games anonymously. I find myself winning much more often in 5 minutes games than in 1 minute games and my own difference in plating strength between the two is not enough to explain the difference. I played many games over the years consistently observing this.

In short: the average randomly matched bullet game you will play against a person of above average strength, and even more so in anonymous.

Improve this answer
1

Classical: 1550 Rapid: 1475 Blitz: 1475 Bullet: 1475 UltraBullet: 1325

Improve this answer
4
  • 1
    How did you calculate these?
    – qwr
    Commented Jul 4, 2023 at 15:36
  • 1
    I didn’t. On the website you can look at the bell curve of players. I gave the values at the 50% marks for each category. These are the average ratings for the players. Unless there is a significant difference between anon players and registered players, which I doubt, these numbers should be very close.
    – Matthew
    Commented Jul 5, 2023 at 6:10
  • Why would the anon population distribution be the same as the registered population? I see no reason why this is necessarily the case. E.g. anon could be more new players who don't want to sign up yet. Or it might be strong players who want to hide their new opening idea. Or it might have more cheaters who got their account banned. Now maybe these cancel or maybe there's other factors and the distributions end up similar, but I see no reason to generally assume this to be the case.
    – koedem
    Commented Oct 22, 2023 at 11:51
  • Furthermore, the other answer matches my experience, when I play the odd logged out game, the faster the time control the stronger the opponents and quite significantly so.
    – koedem
    Commented Oct 22, 2023 at 11:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.