Try also Edo Historical Chess Ratings which, while similar to chessmetrics, uses different methods. It focuses on pre-1900 chess and the information or data is based more on the times rather than in trying to apply 20th+ century standards on older chess. ?In my opinion, the statistical results, which often differ substantially from chessmetrics, are more accurate and meaningful. However, for post 1900 players, chessmetrics is the gold standard.
Chessmetrics
Sort:
Because I have no life I have spent a few hours on Jeff Sonas' Chessmetrics site and its fascinating, even if you have a hard time accepting the pre-ELO ratings. What I don't get is how the base ratings are arrived at. For instance there are examples of players in the 19th century who play a single tournament or match and immediately go to 30th in the world or something. Then there is the strange situation that players can go up and down in the ratings even when they don't play, in fact some players highest rankings don't appear on their player pages because they happened at a time when they weren't playing.
The most glaring thing I have noticed is Emmanuel Laskers rating dropped 113 points between 1903 and 1904, apparently based on the fact he 'only' scored 2.5/6 in a match against Chigorin. In the site notes it says that they don't make huge ratings adjustments based on a small sample size of games so what happened in Lasker's case?
All the same its a fascinating resource for Chess History buffs.