Chessmetrics

Sometimes I use Chessmetrics, not to get any retro-rating, but to look for events, tournaments, matches some player might have participated in that I may have missed. 
A few minutes ago I was looking to find possible listings on a handful of American women players from the 1940s-1970s.  Not a single women I was interested in was listed, although most had elo ratings over 2000.  Then, just out of curiosity because I noticed the name, I clicked on Max2 Lange.  This would be the son of the famous Max Lange who was also a chess player and a Go player.  He is listed there with a rating of 2463 - which I thought seemed rather high.  Investigating further, not a single event, tournament, match or even a game is given - so the assigned rating was based on . . . absolutely nothing.   So much for Chessmetric historical ratings.

I was just peeking into the site. It is all a silly idea, if you ask me, to compare old and modern chessplayers like that. It is all based on "by-eye" measurement no matter how scientific or matematical they make it sound. Silly to compare Aaron Nimzowisch to Boris Gulko.

Chessmetrics is a good site for results but it does have some glitches. The only result listed at Chessmetrics for Max is Hilversum 1903, and the other tournaments he played are missing, so I guess the rating is based on Hilversum only, where he performed 2508 against rated opponents (with the Chessmetrics system). The player has been discussed at Chessgames.com, quoting "sneaky pete":

"From an article by Peter Guetler in Kaissiber # 13, January/March 2000: Dr.Max Lange I (1832-1899) and (no relation) Dr.Max Lange II (1883-1923). The second Max Lange was born in Stettin, Germany (now Szczecin, Poland) in 1883. Mathematician, friend of Em. and Edw. Lasker. Played in some minor tournaments (or lower sections of major tournaments) between Hannover 1902 and Mannheim 1914. Moved to Japan in 1920 (considered Go more interesting than chess). Died during the Kanto earthquake of 1923. Author of <Das Schachspiel und seine strategischen Prinzipien>, Leipzig, 1910"

http://www.chessgames.com/perl/chessgame?gid=1279356

Max the Second was a rather strong player that might deserve to be ranked #54 at his peak. In Coburg 1904 he finished 5th of 17, 0.5 ahead of a young Nimzowitch. He also played for example in Berlin 1907, and was 2nd of 15 behind Treybal (Edward Lasker was 9th). Gino di Felice lists a dozen tournament results for him between 1902 and 1914.

The fact that so many events aren’t included in the Chessmetrics database makes it rather unreliable, but I guess including all minor tournaments would be tough work.

Little Lange's results at Hilversum is based on 2 wins against chessmetric-rated opponents although there were 14 players (Lange scored 6.5/14).  We don't even know if the games were actually played or who those players were. We also know nothing about the players he lost to nor the other players he won and drew against.  So, from the info given it would seem that Max2 Lange's assigned rating is based on very nebulous and shaky, at best, data.  Although the only tournament lists was in 1903, Lange has no rating in 1903, but shows up in the 1904 list with no competitions to account for it.  It's impossble to determine anything about Little Lange from chessmetrics.  I don't feel he was as strong as chessmetric indicates, but my opinion is no more supportable than chessmetrics.   Chessmetrics might have used other data - or might not have - who knows?   The point, however, isn't Lange's assigned rating or how strong he was or wasn't. the the fact that chessmetric included him in its database, based on almost no (given) data, yet excluded women like Gresser, Vines, Karff, Lane and Belcher - all of whom, in the least,  approached 2100, and were probably underrated since with the possible  exception of Vines they mostly played in women's tournaments and about whom we have many games and results.