What is the most efficient way to study GM games?

Please share the method that works best for you. 

I would like feedback on ways to improve the way I study GM games. The method I use is this:

1. Find collection of well annotated GM games of early masters (Morphy, Steinitz, etc...) such as "Masters of the Chessboard" by Reti.

2. Play over the games with a real board.

3. Try to guess the move that the Master played and analyze why my move is or isn't the right one.

4. Go through all variations given and any moves that I would play or I'm curious about until I understand them to the best of my ability.

5. Note the moves that I got wrong with a hypothetical reason why and any new ideas I got from the game.

I would say on average that this takes me 2 hours/game. I haven't noticed an ability for me to retain anything noticeable about pawn structures (despite making a few deliberate efforts in this regard) by doing this and I kind of think that my best bet would be to study them separately to get a "foothold" per se on the ideas and by doing so I would be able to notice when those ideas are applied and make for a more productive experience in that regard. All in all, it seems that the main thing that I get from my study method is an increased imagination and a few ideas on how to adapt my thinking in a way to allow more logical chess. I think that I could get more out of these games if I tweaked my study patterns a bit but I'm not sure what parts it is missing and I would appreciate any insight on this matter. Thank you for your time.

all you say is fine.   in addition, I like to stop at key positions and ask myself if I not only understand the GM's strategy, but also if I would feel confident playing his/her moves........as I would like to say, if presented with certain positions, I would play like the GM.      I have never done this with Tal-type sacrifices, but have been successful with other famous players, ie; Botvinnik, Alekhine, Fischer........

many great biographies out there.    they are some of the best books to learn from, ie; famous players sharing their thoughts on their games......

 That is interesting. I never really ask if I would feel confident because most of the time after I analyze the move until I understand it there is usually no doubt that their move is way better than mine. The problem is when they throw in some really crazy imbalances or compensations and say that they are better which is clear to them but no where near clear to me and very hard to figure out why. 

ahhhhh, so in those cases -- where you don't agree with their evaluations they are better, you would not play their way.      sure, some of the time, their annotations are not clear to you.      But I would guess, as you improve, more of their annotations will be clear to you.   however, granted, some annotations may be unclear to all.

That reminds me when I was a young teen trying to figure out how White  could castle in the Ruy Lopez leaving his "e" pawn en prise to the Black knight on f6.    I spent hours on that  ( some might say I was thick.......lol ),  but I remember how hard that was for me to grasp that by playing pawn to d4 and in some cases  R-e1, White was perfectly fine and still trying to maintain an opening edge.

with motivation to improve and stick-to-itiveness to see things thru, you will continue to improve.

It's usually the same thing in the scenario that I gave too. "Now the purpose of the move played 10 moves ago is clear." 

I'm learning that GM's tend to milk a certain aspect of a position with every move that they play and I haven't learned how to be so proficient yet. I sometimes think "You're kidding, there is no way that this works." and quickly put it in Houdini or the like and play through as many different variations that I can think of with its help until I somewhat grasp the concept if I can. 

It reminds me of when I play a computer program and I think a computers move is one of the mistakes (the typical obvious blunder computers make in an attempt to simulate the set playing strength) and suddenly my pieces can't get anywhere they need to be.

I try to study people whose playing style is appealing to me so it is usually true that I would play that way if I was good enough.

I'm not sure of your training history, K-Brown, so this may not apply, but let me put it out there anyway.  I would suggest playing through a large quantity of verbally annotated games before going too deep into the guess-the-move aspect.

If you had to do some (advanced) math homework, you wouldn't just look at the question, guess, look at the answer key and then try to figure out where you went wrong.  That's probably the worst way to learn math.  You would get a lesson from a teacher who points out the general ideas, and then you'd use that as a guide to help.

The same is true for chess.  Take many 'lessons'.  Observe a large number of games, see the annotations, start learning the patterns.  Once you have that, guessing the move will be easier and make more sense.

Also, when studying games, I tend to do two things.  First, I try to give a verbal summary of the game afterwards.  Something like, "White had a lead in development in the opening but started to attack before castling.  Black was able to use a tactic against the King to win material, and then he defended until he managed to simplify into a winning endgame, where his extra exchange won all of White's pawns against White's bad Bishop."  Doing this helps me learn the main lesson from the game (ie, make sure to castle before launching your own attack, and Rooks are very strong in the endgame).

Second, and this may be a little advanced for your level, but I try to determine the exact point the loser lost the game.  Where was the decisive mistake, in other words.  Sometimes it is obvious, where one error allows a tactical refutation.  Sometimes it's harder, where an innocuous pawn push creates permanent weaknesses in the resulting endgame 20 moves later.  I use a computer to help here, where if it says +2 or more, I consider myself right.

tl;dr: Review a large quantity (50+) of verbally annotated games before guessing the move; create a verbal summary of every game; try to determine the exact reason the loser lost the game.

The things you suggested are good tools as well but they have a different purpose than what is best for overall understanding in my opinion. I want exposure to as many pawn structures, positions, etc... as possible.

@SmithyQ 

Very good insight! The summary of the game afterwards is something I haven't been doing and is a great idea. I usually point out the ideas that I found interesting or instructive but giving a description of the whole game has to be better than that. 

To clarify on the guess the move aspect of how I study, in key positions there is usually a good amount of annotation with it so when I see such annotation or anytime I read the word mistake, that is the main times that I employ this tool. What sucks the time down is when I get interested in a position. When I was in Trigonometry some 10 years ago a similar thing would happen. Trigonometric sequences were one of my favorite things. There was a homework problem that I spent probably 1 hour on before I solved it. It was only a matter of time before I would find the pattern, I was very confident in that. The next day when we were going over the answers, when that question popped up I was the only one who raised their hand to answer. The teacher said that she couldn't even solve it so she threw the question out and then she called on me and I gave the answer. I tried to explain how I got the answer but no one could grasp it. So, interesting positions to me are positions that I think it is very likely I will find the "pattern" if I'm not lazy and put in the work. If a position is too complex, I will acknowledge that and not "waste" my time per se. It seems that I come across interesting positions every 3 games or so that I study. Some games I'm done with in 30 minutes and others take 4 hours. That isn't very efficient. I think this will help me a lot:

1. Find collection of well annotated GM games of early masters (Morphy, Steinitz, etc...) such as "Masters of the Chessboard" by Reti.

2. Play over the games with a real board.

3. Go over all annotations

4. Determine the critical positions and the move that lost

5. Summarize the game and lessons learned

For me the most efficient approach to GM games is to follow online live comments and analysis of some of the top tournaments done by one of the top guns. E.g. Peter Swidler

Without that I would miss the majority of all ideas and stratetic concepts laying hidden. The insight gained by following the line of thoughs from a strong grandmaster is sometimes really mind-blowing. 

I download pgn files of up-to-date games from TWIC to go over eventually. I find that modern GM games are usually too advanced for me (or am I mistaken?) so I’ve been sticking with the early masters and it seems that I learn/retain more from them.

SmithyQ for the win!!!   Never heard of his approach before and I like it a lot!

I have a personal problem with an aspect of verbal annotations. I think IM Bartholomew is one of the best in this department and my problem is that he makes it look so simple to play chess and I always feel like I will crush the next person I play after watching any of his videos. All his comments are very logical and “straightforward” and it makes me want to bang my head against the wall when I can’t do the same. 

All joking aside, it does kind of frustrate me but it is still usually very instructive.

K_Brown: I think "Guess the Move" is a great approach to studying GM games. However it is time-intensive and if one is interested in acquiring patterns, there's something to be said for going over more games more quickly.

I recommend Fred Mellender's "Guess the Move" app which allows one to study any game in PGN format with one's guesses scored by chess software. 

https://sites.google.com/site/fredm/  (Go to bottom of page and click on down-arrow at far right.)

Chessgames.com also has a "Guess the Move" feature available for a large number of GM games with a scoring feature.

K_Brown, there's something to be said for your preference for studying older games Guess-the-Move style. Modern GM games are tough to follow. FredM's app gives you points for making moves which are reasonable though may not be up to current GM standards. This may help you from getting too discouraged.

Very interesting, thank you.

Hello @K_Brown! Check out this article on how to analyze chess games: http://chesswinning.com/how-to-analyze-chess-games/

One suggestion is to do it blindfolded. Doesn't matter if you can't see it till the end. Just try. It will improve your visualization.

Sounds like you are thorough.  Don’t forget to play the openings you are looking at in games.  Looking at the GM games will be more valuable if you play a lot.  Learn by doing.  For example, if you play daily chess with 1 or 3 day per move you can actively apply ideas you are learning from the games you review.  You can also try ideas in shorter time control games.  Repetition is key unless you are a savant/genius