Ranking players past and present

While walking my dog, I have been thinking about how to rank the great players, both past and present. I use the game review feature quite a lot in my games and can see that this is also often used in live streaming events for top tournaments - including the Champions Chess Tour. I wonder whether it is possible to do the same for the great players of the past, thereby coming up with a more accurate way of ranking the players. It would certainly be a better way of ranking players than the ELO rating system, for instance. We have access to most, if not all, the “classical time control” games of the great players, dating back over the past 150-200 years. I think it would be a fascinating exercise to use the game review system to rank players according to accuracy, brilliant moves, mistakes, etc. Of course, there are caveats, such as the fact that  time controls were often different in the past - especially pre-WW2 - as well as other factors, such as advantages that players have had over their predecessors, like advances in access to chess information over time, books, databases and computer technology.

Anyway, I think it would be a very interesting and, perhaps, valuable exercise.

Cheers

Mike (mjk1962)

This has been done by calculating rating backwards using historical results:
http://www.chessmetrics.com/cm/CM2/PeakList.asp

Chessmetrics is interesting but, again, is based on ratings. My main point is that ratings are only one aspect of ranking players It is a quantitative method rather than qualitative. Performance based on how well a player does in a tournament or a match does not necessary provide much guide to how well a player performed in the games themselves. The game review used by Chess.com may provide a much better guide about how good a player is/was since it is based on the quality of the games played - using such criteria as accuracy, brilliancy moves, good moves, mistakes, blunders etc. I think this would be a fascinating exercise to do on the games of the greats of past and present.

@4

"Chessmetrics is interesting but, again, is based on ratings." ++ Yes, it is based on results.

"Performance based on how well a player does in a tournament or a match does not necessary provide much guide to how well a player performed in the games themselves." ++ That is subjective. A player may play one good move after the other, and finally overlook a back rank mate and lose. That is a loss and the oversight undoes all good moves before.

"The game review used by Chess.com may provide a much better guide" ++ No, that is not better at all. If I play a weak player, I may well reach 100% accuracy, all good moves, no mistakes. Against a stronger player, I may well drop below 50% accuracy.

"accuracy" ++ Accuracy is a measure of how much you play like an engine

"brilliancy moves" ++ That is an aesthetic measure

"good moves" ++ All moves that are neither mistake nor blunder are good moves.

"mistakes" ++ That is key: he who loses made the last mistake.

"games of the greats of past" ++ That is regularly done now. Example: https://en.chessbase.com/post/endgame-riddle-fischer-spassky-game-3-solution

"and present" ++ That is regularly done now. Example:
https://en.chessbase.com/post/tata-steel-chess-2024-r13

You make many good points. The game review by Chess.com is, of course, not without its faults since it is based on a chess engine. I often play games where it ranks a move as a mistake but it appears to be the subjective opinion of the engine, not least because its alternative move recommendations don't always seem to be any better or, objectively, better than the move played.

Nevertheless, overall, I still feel it might provide a better understanding of past players' standards. ELO type ratings don't take into account the quality of the play, while using a "game review" approach may provide a better understanding of the games played. For instance, by averaging out how well they performed in the games based on the number brilliancy, good moves, misses, blunders and so forth, you will be able to better gauge a players' qualitative standard of play.