If someone wants to play on a chess site, where the resulting ratings closely follow those from otb US Fed. ratings then it is chess.com. Lichess offers a lot for free to enjoy without question. But regardless of the reasons, accurate ratings according to those obtained from otb is not one of them. My aim is not to provoke but rather to understand why, and I suppose no one can honestly explain it
> My aim is not to provoke but rather to understand why, and I suppose no one can honestly explain it
you are doing very badly then. again, it is easy to explain: following otb ratings is not the goal of lichess rating. the goal is to predict winning chances. lichess ratings are *very* accurate, and in my experience way more accurate than fide ratings, at least at the lower levels.
you are doing very badly then. again, it is easy to explain: following otb ratings is not the goal of lichess rating. the goal is to predict winning chances. lichess ratings are *very* accurate, and in my experience way more accurate than fide ratings, at least at the lower levels.
@Lachesis said in #21:
> My aim is not to provoke but rather to understand why, and I suppose no one can honestly explain it
Dude, I already answered the question: Prof. Glickman recommends 1500, so 1500 is recommended:
> If the player is unrated, set the rating to 1500 and the RD to 350. Set the player’s volatility to 0.06 (this value depends on the particular application).
http://www.glicko.net/glicko/glicko2.pdf
lichess.org/forum/lichess-feedback/rating-should-not-start-at-1500#6 is my previous honest, accurate, and complete answer to your question, along with a link to reference materials authored by the world's leading expert in chess ratings. Half this forum is getting sidetracked with the apples, oranges, etc. discussion which is a red herring; 1500 is used because 1500 is correct **by definition**.
> My aim is not to provoke but rather to understand why, and I suppose no one can honestly explain it
Dude, I already answered the question: Prof. Glickman recommends 1500, so 1500 is recommended:
> If the player is unrated, set the rating to 1500 and the RD to 350. Set the player’s volatility to 0.06 (this value depends on the particular application).
http://www.glicko.net/glicko/glicko2.pdf
lichess.org/forum/lichess-feedback/rating-should-not-start-at-1500#6 is my previous honest, accurate, and complete answer to your question, along with a link to reference materials authored by the world's leading expert in chess ratings. Half this forum is getting sidetracked with the apples, oranges, etc. discussion which is a red herring; 1500 is used because 1500 is correct **by definition**.
@Lachesis Nope. Cheese.com is as off as any online website to federations. I have students that are 1000 points different cheese.com vs federation. Stop planting false ideas.
DrHack if you look at the cross tables for G/15/10 US Federation tournaments played at chess.com you can see the close correlation between the ratings otb and online.
@Lachesis That's great. A friend of mine built a website to compare ratings across websites and federations. Here's the data he collected: www.chessratingcomparison.com/ I don't see any closer correlation between anything.
Thanks DrHack for the link.
This topic has been archived and can no longer be replied to.