You know the problem. You have been studying chess for several years now. Yet somehow you don't seem to improve. You have bought Michael de la Maza's book and read it. You have read "Improve Your Chess Now" and "The Inner Game of Chess". You have read "Think Like A Grandmaster". These books inspired you. But they didn't make you stronger. Why not? Why do you still blunder in strong positions. Why did you miss that fork when you were playing a young and talented junior who knows absolutely no chess theory (ouch that hurts)????
Today, cognitive science has helped us to understand this. The reason strong players analyse better than weak players is because of a principle known as "chunking" - just another name for pattern recognition. To improve, you need to learn more patterns or chunks.
I will explain what chunks are in a minute. To become a master, it is estimated that you need to have about 50,000 chunks in your memory. However, herein lies the problem as well. To gain 400 rating points, you probably need to learn 10-times as many chunks or patterns as you know now. A player rated 1800 FIDE may easily recognize 5,000 chunks... 1400 FIDE would be 500 chunks and so forth.... So as you improve, the rate of progress decreases because you need to know much more. Does that sound familiar?
Strong players can visualize the board better than weak players because they can divide the board up into big chunks. A chunk is a pattern. A grouping of pieces that is seen as a single big chunk. Let me give you the very basic idea here. Suppose I asked you to remember the following string of numbers: 121643681. Could you do it easily? OK, now try chunking the information. Remember the numbers in groups. 121 64 36 81. Is that easier? OK, now suppose I tell you that each group is the square of some simple number (121 is the square of 11, 64 the square of 8, 36 of 6 and 81 of 9). Now it is much easier again, because you recognize the patterns. They are pre-existing patterns that you have already learnt.
A castled king position for example (Kg1, Rf1, Pf2, Pg2, Ph2, Nf3) might be a chunk. It is easy to remember isn't it? Despite the fact that there are 6 pieces on 6 out of the 64 squares on the chessboard, you could set the position up in a flash if asked to do so. Someone who has never seen a castled king configuration (with a trusty knight on f3) would have trouble remembering it, but for you, it is easy, right? Here is the key to visualization. All you need to do is be familiar with the chunk.
You store chunks in long-term memory but you process them through short-term (working) memory (involving brain structures such as the prefrontal cortex and the hippocampus). It has been estimated that you can store up to about 7 chunks in working memory at any one time, but the most recent studies suggest that it is closer to 3 chunks, not 7. For that reason, to visualize the entire board and hold it in working memory, you need to see the entire position in only 3 chunks at most. Therefore, chunks need to be quite large. More realistically, you probably work with 3 chunks at any one time and process information rapidly between working memory and long term memory in order to visualize the entire board.
You need to see and be familiar with groupings of pieces. Knowing typical pawn structures helps in particular. For example, as a junior, I spent a lot of time studying the isolated queen pawn positions that arise out of the QGD and Caro-Kann openings. There are typical placements of both pawns and pieces. You get to know where they belong and can visualize their positions (and where they can move to with effect) fairly easily.
This then is the key to becoming a strong chess player. Learn chess positions, and learn lots of them. And of course, learn what moves work in these positions and what moves are poor. The learning of patterns or chunks is dependent on two basic principles: repetition and timing.
Studies show that frequent repetition of a training stimulus helps in learning the stimulus pattern. For this reason, I recommend the study of related structures, which helps to reinforce the concept better. Study games with similar pawn structures. Note that the key here is to learn patterns. Solving problems helps to learn a pattern, but my recommendation is to learn the pattern first and then reinforce this with problem solving. In other words, if you want to learn a tactical pattern, don't spend too much time solving it in the first instance. It is better to spend a minute or two on it, look up the solution and then come back to it later (in 5 minutes) and see if you remember it.
Recent published electrophysiology experiments from Mu-ming Poo's group shows that maximum reinforcement is achieved when the stimulus is repeated within 3-5 minute intervals. My recommendation for study then is to examine a large number of master games, all with similar or related structures, reading through them 2-3 times at 5 minute intervals. Read through the game quickly and then read through slowly studying the notes. The third time, see if you can recall the moves of the game and remember the main strategic and tactical themes in your mind's eye.
Of course it is also possible to train working memory, but only in a limited way. Recent studies suggest that working memory can be improved by exercises, but it is fairly pattern specific. That means that the training task must be similar to the memory task. One way to do this is to try blindfold chess. Set up a chess board in front of you and read through the first 10-15 moves of the game. Now stop moving the pieces and keep reading the game. Can you visualise the next 10-20 moves? With practice you should be able to do this. Bear in mind that this exercise is like flexing your muscles. It gets you in shape but ultimate progress in chess study depends upon your ability to learn new patterns.
There are basically 6 types of patterns you need to learn: 1) Opening patterns, 2) early middlegame patterns (where planning is tightly controlled by the specific pawn structure), 3) positional patterns (learning to place pieces and pawns on their optimal squares), 4) tactical patterns, 5) Strategic endgame patterns and 6) technical endgame patterns. So which should you study first? A chess trainer can help you there, by showing you where you are weakest. If you don't have a trainer, then a close analysis of your own games may help you to decide what you are weakest at.
A good one I read is about using "chunks".
You know the problem. You have been studying chess for several years now. Yet somehow you don't seem to improve. You have bought Michael de la Maza's book and read it. You have read "Improve Your Chess Now" and "The Inner Game of Chess". You have read "Think Like A Grandmaster". These books inspired you. But they didn't make you stronger. Why not? Why do you still blunder in strong positions. Why did you miss that fork when you were playing a young and talented junior who knows absolutely no chess theory (ouch that hurts)????
Today, cognitive science has helped us to understand this. The reason strong players analyse better than weak players is because of a principle known as "chunking" - just another name for pattern recognition. To improve, you need to learn more patterns or chunks.
I will explain what chunks are in a minute. To become a master, it is estimated that you need to have about 50,000 chunks in your memory. However, herein lies the problem as well. To gain 400 rating points, you probably need to learn 10-times as many chunks or patterns as you know now. A player rated 1800 FIDE may easily recognize 5,000 chunks... 1400 FIDE would be 500 chunks and so forth.... So as you improve, the rate of progress decreases because you need to know much more. Does that sound familiar?
Strong players can visualize the board better than weak players because they can divide the board up into big chunks. A chunk is a pattern. A grouping of pieces that is seen as a single big chunk. Let me give you the very basic idea here. Suppose I asked you to remember the following string of numbers: 121643681. Could you do it easily? OK, now try chunking the information. Remember the numbers in groups. 121 64 36 81. Is that easier? OK, now suppose I tell you that each group is the square of some simple number (121 is the square of 11, 64 the square of 8, 36 of 6 and 81 of 9). Now it is much easier again, because you recognize the patterns. They are pre-existing patterns that you have already learnt.
A castled king position for example (Kg1, Rf1, Pf2, Pg2, Ph2, Nf3) might be a chunk. It is easy to remember isn't it? Despite the fact that there are 6 pieces on 6 out of the 64 squares on the chessboard, you could set the position up in a flash if asked to do so. Someone who has never seen a castled king configuration (with a trusty knight on f3) would have trouble remembering it, but for you, it is easy, right? Here is the key to visualization. All you need to do is be familiar with the chunk.
You store chunks in long-term memory but you process them through short-term (working) memory (involving brain structures such as the prefrontal cortex and the hippocampus). It has been estimated that you can store up to about 7 chunks in working memory at any one time, but the most recent studies suggest that it is closer to 3 chunks, not 7. For that reason, to visualize the entire board and hold it in working memory, you need to see the entire position in only 3 chunks at most. Therefore, chunks need to be quite large. More realistically, you probably work with 3 chunks at any one time and process information rapidly between working memory and long term memory in order to visualize the entire board.
You need to see and be familiar with groupings of pieces. Knowing typical pawn structures helps in particular. For example, as a junior, I spent a lot of time studying the isolated queen pawn positions that arise out of the QGD and Caro-Kann openings. There are typical placements of both pawns and pieces. You get to know where they belong and can visualize their positions (and where they can move to with effect) fairly easily.
This then is the key to becoming a strong chess player. Learn chess positions, and learn lots of them. And of course, learn what moves work in these positions and what moves are poor. The learning of patterns or chunks is dependent on two basic principles: repetition and timing.
Studies show that frequent repetition of a training stimulus helps in learning the stimulus pattern. For this reason, I recommend the study of related structures, which helps to reinforce the concept better. Study games with similar pawn structures. Note that the key here is to learn patterns. Solving problems helps to learn a pattern, but my recommendation is to learn the pattern first and then reinforce this with problem solving. In other words, if you want to learn a tactical pattern, don't spend too much time solving it in the first instance. It is better to spend a minute or two on it, look up the solution and then come back to it later (in 5 minutes) and see if you remember it.
Recent published electrophysiology experiments from Mu-ming Poo's group shows that maximum reinforcement is achieved when the stimulus is repeated within 3-5 minute intervals. My recommendation for study then is to examine a large number of master games, all with similar or related structures, reading through them 2-3 times at 5 minute intervals. Read through the game quickly and then read through slowly studying the notes. The third time, see if you can recall the moves of the game and remember the main strategic and tactical themes in your mind's eye.
Of course it is also possible to train working memory, but only in a limited way. Recent studies suggest that working memory can be improved by exercises, but it is fairly pattern specific. That means that the training task must be similar to the memory task. One way to do this is to try blindfold chess. Set up a chess board in front of you and read through the first 10-15 moves of the game. Now stop moving the pieces and keep reading the game. Can you visualise the next 10-20 moves? With practice you should be able to do this. Bear in mind that this exercise is like flexing your muscles. It gets you in shape but ultimate progress in chess study depends upon your ability to learn new patterns.
There are basically 6 types of patterns you need to learn: 1) Opening patterns, 2) early middlegame patterns (where planning is tightly controlled by the specific pawn structure), 3) positional patterns (learning to place pieces and pawns on their optimal squares), 4) tactical patterns, 5) Strategic endgame patterns and 6) technical endgame patterns. So which should you study first? A chess trainer can help you there, by showing you where you are weakest. If you don't have a trainer, then a close analysis of your own games may help you to decide what you are weakest at.