How do YOU analyse master games from books?

I am getting frustrated with trying to get to grips with games from my books, i have some fantastic games collections, but i dont think im getting the most from them. How do you read/analyse your games collections? should i be playing through a game very slowly (1-2 hrs) ? trying to pretend im playing from the winning side? or should i play from the losers side? should i read through the game and play over it very quickly, then go back and look at critical positions? basically i want some good advice from some strong players, how to get the most from my games collections...please! :-)

Play out the game, and see what you would have done in those positions.

same problem here,

I find it really boring instead,,

keep trying,,

you will learn

Start by scanning through the book, playing through the games quickly to get a feel for it then...

A few writers suggest you should start with games that use openings with which you are familiar, playing through the first lot of moves quite quickly until you get into the middlegame. Then you analyse the game as if you were playing it yourself - it doesn't matter which side you choose. Analyse the strengths and weaknesses of the position for both colours and decide what move you would make (obviously this works best if you can cover up the moves in the book!) with a couple of other strong candidate moves (making notes on why you chose the move you did), uncover the move in the book. If it is the same as yours then move on, if not, re-evaluate the position until you understand the GM's choice then move on.

This is a very long-winded and thorough approach, taking a lot of patience, but it is the way to get the most from a good games collection.

Just playing through top quality games quickly gets something from a collection - some games are a work of art, even when played through quickly - you watch the dance of the pieces in awe.

Hi gumpty, I would like to be able to do that too, read the games and be able to understand it. And with jonnyjupiter's tip- ". Analyse the strengths and weaknesses of the position for both colours and decide what move you would make (obviously this works best if you can cover up the moves in the book!) with a couple of other strong candidate moves (making notes on why you chose the move you did), uncover the move in the book. If it is the same as yours then move on, if not, re-evaluate the position until you understand the GM's choice then move on." I liked this ideas.

Very interesting thread! It is easy to fall into the "learning-by-nodding" trap but I do think that it is usefull to skim trough a lot of master games using "your own" openings.

You are 2000+ equivalent FIDE so I won't presume to export my practice to you, but I can tell you that Ken Smith in the back of his old thick catalogs had a section on improving one's play, and that he recommended spending 10-15 minutes on a game, proposing that quantity made up for quality. And do a lot of them.

He was 2365 FIDE "back in the day."

I do that and engage in long analysis as well, exposing one side's moves one at a time after selecting what I think is an appropriate one. Like an ungraded quiz.

The slow way exercises your calculation, the fast way exposes you to the way structures and different kinds of positions are handled. The fast method may be better suited to players who don't know as much as you?

Someone--I forget who it was--said to pick a player from a tournament book and work through (slowly I presume) all of his  first. Then pick another player. Maybe you ought to pick a beloved player and work through heaps of his games and figure out his approach?

Was that helpful at all or was it all stuff you had already pondered?

Someone--I forget who it was--said to pick a player from a tournament book and work through (slowly I presume) all of his  first. Then pick another player. Maybe you ought to pick a beloved player and work through heaps of his games and figure out his approach?

I think Jesper Hall said that from one of his book ?

 I don't have any Jesper Hall so the idea  was either picked up from him or he got it somewhere down the line.

edit: Actually I think this idea is from Ken Smith as well.

I was told long ago to have two board to play with--one for the game and the other to go over the annotations. Also I have an index card that I use to block out the next move and I try and play the game as one player. As mentioned above I try and look at games with openings that I play or am interested in. This is what i do for my serious study.

For other books I play through them quickly trying to pick up patterns or tactical ideas. These games take me 10-15 minutes to go through and I do this when really trying to understand an opening or other theme.

I think a lot of valuable advice has already been collected in this thread. Both playing through many games quickly for pattern recognition and working through a game very thoroughly have their merits. Which method you choose mainly depends on what you want to practice: If you want to to learn to play a certain opening or a certain kind of position the skimming method is appropriate, while the slower one will help you with calculation skills and decision making, for example.

One little remark about my own style in replaying master games: I try not to move the pieces on the board when looking at the sidelines. Instead I try to understand the variations by playing them through in my head. Of course, if I haven't grasped why a line works, I will still have to play it out on the board.

All this advice is geniuos, what I always try to do is pick a game of one of my favorite players or a specific opening, and in every move think, not what I would have played, but what was the idea behind the move the grandmaster made and find the logic behind it so that when I play a similar position I will not look so much for a similar move, but attempt to cary out a similar plan with corresponding ideas that I know have worked in that kind of position or pawn structure.

Here is what I do:  I pick an unannotated game that's about 23-26 moves long.  I prepare to have 4-6 hour free time.  I make notes on all the moves that I know for a fact that had a purpose.  For those moves that are beyond my chess intellect I question mark and return to at a later time in the session.  That's how I progress, but there may be an exception in procedure.  By the way the game can't be  a drawn game. 

Many of these methods will be useful.  The study alone is beneficial.  Going through a game after some time, and experience, so see it in a new light is often helpful.

Also one other thing gumpty, 'you have all answers that you need to know is in your old games'.

I have seen many different methods recommended.

A popular one is the 'guess-the-move' method, where you play through the opening moves, then cover it up (easier thing to do is bring up the game in Chessbase and use the training tab) and come up with a move, check it against what was played, play the reply or possibly repeat the process for the other side, then repeat this process for the rest of the game.

Some recommend playing through them fairly quickly to try and expose yourself to as many top games as possible, possibly reading the notes or not.

Bronstein recommends a 3 step process:

1. Play through the game quickly, not stopping for more than a couple of seconds at any spot. If there is any point where you want to linger then make note of it. Take a 10 minute break or so, go over things you remember about the game in your mind.

2. Go through the game a 2nd time somewhat slower, marking down any spots you missed first time round.

3. 3rd time, go straight to the points that you marked and give your imaginative and creative energy free reign. Write your findings in a notebook. (This could probably be combined with the Dvoretsky/Yusupov idea of positional sketches).

As pointed out above there is the approach that Silman gives for those wanting to improve their calculation skills, which is a modified guess the move, where you spend 30 minutes or however long you need coming up with candidate moves and variations for those moves, select your move, write all your analysis down in a notebook and then check your move against the move played.

I believe Aagaard suggests writing down something that you learned from the game after playing through it.

Yusupov has recommended playing through the game until you reach a diagram in the book. At that point begin to analyse, and then compare what you come up with to what is in the book. Then continue on until the next diagram.

I think the most important thing is to do whichever way you find most enjoyable, perhaps mixing up different methods depending on your mood.

Lately I have been bringing up the game in chessbase and doing the guess-the-move routine on the first pass in the training tab, then I will play through the game a 2nd time reading the notes in the book as I go.

its been 10 years since i studied anything but i saw a recomendation to J.Silman's books, I totally agree to buy his books...i read my "reassess your chess" so much it fell apart!  i dont think i can add anthing else to the wealth of knowledge already proffered on this thread... great advice chess.com members, thats why i love this site