A lot of top coaches, especially from ex-USSR, recommend studying the classics, rightly so. But few give a clear-cut method to do it yourself. This thread offers some thoughts and guidelines about the subject.

1) A big difference must be made between four categories : a) The Romantic Era, up to Steinitz
b) The Classical Era, up to WWII
c) The Modern Era, up to Kasparov
d) The Computer Era, post 1990s
I think a) can be dispensed with, so did Dvoretsky.

2) To find good games for b), a quick search from Chessbase or free alternatives such as http://chessgames.com , or even the lichess Masters database, is enough. Don't focus on move orders but on the logic and psychology of moves played. Lasker, Capablanca, Flohr, Rubinstein.

3) For the same period, if you want to train your dynamic sense, Keres, early Botvinnik, Alekhine, Euwe are good models.
Post WWII, Korchnoi, Tal and Spassky are also excellent dynamists.
Smyslov, Petrosian and late Botvinnik are excellent for the technical aspect. Fischer too.

4) Now, before we go into the Computer Era, how do we select the most instructive games ? Start with the longest ones (60+ moves) in structures that you want to play. 10-15 games per structure is already excellent. Then go to 30-50 moves games, then under 30. See why below.

5) This way, you can understand the long-term general goals of the structure before going into crazy complications will help you decide when to hit the brakes and when to go turbo.
Respect for statics will help make dynamics work, or tactics always work in good positions.

6) The Computer Era has the issue of too much data, as more high-level games are played and the level of play can only be understood through what players learned from earlier periods. It may be possible to get to a 2200+ understanding through those periods' study alone.

7) Again, focusing on a few players, ones that share something with us (style, nationality, personality...) will help create a modern hero, preferably alive. However, mimicry is rarely successful and you should continue looking into various influences.

8) I only cited a few players for the divide Builders/Destroyers. Obviously styles are more complex than this. A nice feature of looking at a great player's games is that you'll encounter other great players that you may then choose to follow in turn. Examples following.

9) Karpov's loss to Geller, Capablanca's to Reti, Steinitz's to Blackburne will show you the limitations of a great player's approach and help you find balance again. If you only like uncomplicated slow middlegames, your opponent may not let you have that, or crazy attacks.

10) Finally, Focus On
b) Classical Era : Trades and pawn play
c) Modern Era : Transformation of Advantages, the element of Time (dynamic factors) against Space (static factors)
d) Computer Era : The level of precision needed compared to earlier at the top.
END OF THREAD