I think of analysis as the ability to mentally move through variations, recognize tactical ideas in those variations, and draw accurate conclusions regarding the fitness of the resulting positions.

I have heard it said that the ability to perform proper analysis is the greatest factor in chess strength up to the USCF 2000 level. How can one improve this skill?


Based on Tony Ennis' and Goerge Jempty's answers I have devised the following training plan.

  1. Using CVT Online I am performing board vision drills on a daily basis.

  2. I am solving K+P ending problems using progressively complex problems (i.e. increasing the number of pawns) from a variety of sources in my head.

  3. I am using Chess Success: Planning After the Opening and a set of blank boards to go through the games "blindfold" - meaning in my mind. Once I get to a diagram I am drawing the position in my head using the blank board and then comparing it to the one in the book. I am also doing this with any variations that don't have diagrams. But that requires a manual check.

  4. When I am solving my typical tactic and other puzzles I am making an effort to move through my variations with my eyes closed and also performing my check of the final position for checks, captures, and threats without looking at the board as well.

  5. I am also doing quick drills using the Fritz software from chess base to quickly recognize checks, captures, and threats. This training material is under the "Training" tab of the CHessBase user interface of their engine software.

In the next 90 days I will edit with any thoughts or considerations that I have regarding the results of this training plan. Thank you to all who answered.

Related thread: How to get out of "Hope" chess into "Real" chess


There are no shortcuts.

  1. Work tactical problems

  2. Read highly regarded books to learn why you should make moves. I wouldn't necessarily read, "GM X's Best 50 games", but instead would seek out books about tactics or positional training. "My System" is an example, though probably dated.

  3. Get an instructor. He can tell you what you're doing wrong and how you can fix it. Figuring this out yourself is time consuming at best.

  4. Play a lot, working to apply what you're learned. Don't play for rating points, or even to win, as such. Play to learn.

  5. Analyze your games with a machine to see what you - and your opponent missed. Work hard to figure out why the machine thinks your moves were substandard, and how it's suggested moves prevent the problems caused by your moves.

  • 1
    Tony, I especially appreciate the comment on not playing to improve your rating or even win. Good advice! – Robert Kaucher Jul 24 '12 at 15:01
  • +1, I agree with not playing to improve your rating as well. As Nike says, "Just Do It" – xaisoft Jul 25 '12 at 14:52
  • 2
    +1 for 2, 3 and 4. Be careful the machine can easily give you optimal moves when you're able to read a 8 move long var, but uneasy to understand, and not applicable elsewhere. – Nikana Reklawyks Nov 18 '12 at 7:14

Try your hand at blindfold chess. It stands to reason, if you can visualize things without sight of the board, it will improve your visualization with sight of the board. Speaking from my own experience, when I did this, my rating jumped from the 1700s into Class A and this included mowing through 2000+ rated players in at least one tournament.

There are a number of different ways to attempt blindfold chess:

  • Some versions of ChessMaster software have an option for this.
  • You can use Mephisto dedicated chess machine, and simply press the indicated squares on the board, without placing pieces there.
  • I read somewhere, perhaps right here on chess.stackexchange, that fics has an option for this
  • 1
    A pointer to the question George has in mind at the end of his answer: chess.stackexchange.com/q/823/167 – ETD Jul 24 '12 at 15:18
  • Here are few blindfold exercises: chessvideos.tv/blindfold-chess-quiz.php .Maybe this can be helpful. – Akavall Jul 24 '12 at 16:23
  • Chessmaster is great, I use it all the time. – xaisoft Jul 25 '12 at 14:52
  • @George Jempty, I would like to select both your answer and Tonny's but since his is a little more complete I am going to select his. But you certainly got my upvote for this. – Robert Kaucher Jul 26 '12 at 20:10

The way to improve analytical skills "over the board" is to improve those skills generally.

First, analyze your own games after the fact. Second, read analyses of master games. At each point, try to guess the next move before looking at it. You will make a lot of "mistakes" early on, but improve over time.

The more practice you have with "offline" games, the more you will benefit in over the board situations.


As an experienced player I would say that the basics to improve in that way is first of all to have a solid base on resolving chess problems of mid-game and have your openings learned, with lot of games over them, to have the enough experience to develope the opening by the right way.

By the other hand, speed-games makes the player think faster and it consolidates your knowledge of mid game, leading you to think faster from time to time while playing more faster games.

Of course if you don't have a solid knowledge of the middle game, playing like this just will make you lose all the games and don't even notice the mistakes you're doing. So you shouldn't make the second step without the first.

This will give you the hability to think a bit faster, so you can think about different variants of the game in the same time you would have probably thought about half of them.

Apart from this, it's always important to analyse the normal time games with a computer, so you can see the mistakes and the best branches you should have taken.

  • You: "speed-games makes the player think faster and it consolidates your knowledge of mid game, leading you to think faster from time to time while playing more faster games." GM Vladimir Kramnik: "Playing rapid chess, one can lose the habit of concentrating for several hours in serious chess. That is why, if a player has big aims, he should limit his rapidplay in favour of serious chess." (source) – SecretAgentMan Jan 28 at 3:09

I am rated 2200 USCF and 2100 FIDE. Look, this is a question that is often asked by lower rated players. Yes, stronger players have better analytical skills and you need to develop them. But it is wrong to think you can train chess vision by simply doing blindfold exercises or working from a tactics book. The key is to first develop a base of chess understanding (no, this is not off the topic) and then proceed to develop a battery of chess knowledge through reading many many (many) games. Vision comes from learning chess positions. You need to understand that even the world's top players can't visualize random positions well. They visualize and analyse well because they have seen the positions before. A knowledge of positions and the best moves played in those positions helps to develop visualization skills and helps to trim the tree of analysis to a manageable level. Trying to imitate a computer is a sure way of getting bogged down. Also, tactics books are great, but they train only a certain type of knowledge. More than 50% of chess moves are quiet moves and most tactical errors extend from inferior positions obtained by playing inferior quiet moves. Learn all types of moves by reading games played by the world's best players. And don't forget the endgame. This is where the basics of chess come into sharpest focus.


Practice your visualization skills. Opportunities lie in any chess book with a game score in it. Start with a position (the starting position is fine, but harder), and play 4-5 moves for each side. Don't check the position just yet. First, try to find all of the captures, checks and threats (CCT's) available to the player who's now on move, and write them down.

Then proceed to that position on a board (wood or electronic - doesn't matter), and verify a) that you had the pieces on the right squares, and b) your findings of the CCT's.

When you are good at this, you'll be going longer and longer before having to verify the board position (10 moves a side, and eventually a whole game). A master has to be able to visualize well; many make a point of being able to replay an entire game in their head, including memorizing the moves, not reading them.

Getting good at tactics requires being familiar with tactical patterns, which means you have to see them, and then learn them. Tactical problems are a good way to learn them, but make sure you don't think longer than 5 minutes about the candidate moves; if it takes longer than that, you're still in learning mode for this tactic, and it's more productive to watch the demonstration than to bang your head against the wall for 20 minutes trying to invent an unfamiliar solution from scratch. Then, run through it a couple of times to familiarize yourself with all the ideas it encompasses. Finally, if you can, put the puzzle through an engine, and try out some of your alternative ideas. That way, you'll see why they don't work (if you haven't already), and that's just as valuable.

If you do the work to learn the solution, then the next time you see a similar problem, you'll solve it within 5 minutes.

I agree with the poster who suggested playing "Solitaire Chess" with an annotated game, trying to predict the moves for one color every step of the way. But that's going to be difficult if you're a beginner or low intermediate player (say 1400 or below); it only becomes manageable once you have a decent toolkit of ideas for what to do in various kinds of positions. So, until then, just play through the games, trying to understand each move's purpose. A good "Move-by-Move" book will help at this stage, and there are many to choose from.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.