I'm about 1600 on lichess and about 1300 USCF, so that 300 delta mentioned earlier appears to me quite accurate.
To clarify #10: both lichess and Chess.com use the Glicko-2 system, though I believe Chess.com uses a slightly modified version.
They are not 'ELO systems', they are rating systems for games between 2 players. Elo (not 'ELO') is one such rating system that is no longer used in chess by anyone notable other than FIDE.
As for the topic: just remember that all ratings for multi-player games are necessarily relative. It depends as much on your skill as it does on the ability of the website's community.
They are not 'ELO systems', they are rating systems for games between 2 players. Elo (not 'ELO') is one such rating system that is no longer used in chess by anyone notable other than FIDE.
As for the topic: just remember that all ratings for multi-player games are necessarily relative. It depends as much on your skill as it does on the ability of the website's community.
Lichess Blitz rating on here is more closer to FIDE and USCF ratings.
Im like 1800 on here and like 1800 on uscf so no difference really
I agree with Hellball in #12: any correlation between lichess ratings and ratings of other sites or organizations is purely coincidental and is subject to change at any time, since the strength of players on this site may vary.
As I understand it, the default rating of new players is 1500 so if lots of strong players join lichess, as their ratings increase the average rating of other players would decrease (and vice versa).
As I understand it, the default rating of new players is 1500 so if lots of strong players join lichess, as their ratings increase the average rating of other players would decrease (and vice versa).
Yeah, I second (third, fourth?) the claim that any rating correlations between lichess ratings and other rating pools, be they national or international OTB organizations, or other online organizations, is mostly coincidental.
The accuracy of human ratings is pretty suspect anyway, since human skill varies so much from game to game, and except for people who regularly play bullet marathons, no one plays enough games to bring the error bars down very low.
Consider that for entities that play at the same level every game (engines), you still need 3000 or so games to get the 95% error bars to +/- 10.
With humans, who play at different levels based on mood, fatigue, distractions, recent study/training, etc., we just don't get enough games in at the same level of play to say much meaningful about our ratings except with tacking +/- 70 or so onto it.
Short story, for a variety of reasons, the ratings don't matter so much, much less how they relate to other equally uncertain measures of playing ability from other rating pools. Just play chess and have fun :)
My two cents, at least.
The accuracy of human ratings is pretty suspect anyway, since human skill varies so much from game to game, and except for people who regularly play bullet marathons, no one plays enough games to bring the error bars down very low.
Consider that for entities that play at the same level every game (engines), you still need 3000 or so games to get the 95% error bars to +/- 10.
With humans, who play at different levels based on mood, fatigue, distractions, recent study/training, etc., we just don't get enough games in at the same level of play to say much meaningful about our ratings except with tacking +/- 70 or so onto it.
Short story, for a variety of reasons, the ratings don't matter so much, much less how they relate to other equally uncertain measures of playing ability from other rating pools. Just play chess and have fun :)
My two cents, at least.
This topic has been archived and can no longer be replied to.