Expert: Louis Lima - 12/8/2009

Question
Hello,
 As my question subject suggests I have problem calculating complicate chess positions. Although I can see a single line in considerable length, when the variations become complicate and the tree of possible continuation become wide(with several options in each move) I cannot remember different position in which branching occured and so my calculation speed and accuracy suffers from that. Is there any good method for calculation or any training to overcome this problem?
 My second question related to first is about situation when there is so many possible move for my opponent in each move(due to the fact that my moves are not forcing). I want to know how to calculate and analyse in such situations too.
Thank you very much in advance

Answer
Dear Sina,

This is one of the best questions I've been asked on AllExperts. It's an excellent question. Unfortunately the concept of calculation in chess is complex and multi-layered. The is simply not a simple and short way to answer this question.  When players ask themselves how to improve their calculational abilities, they usually think that it is just a matter of ply-count (e.g "How can I see ten moves ahead instead of three", or "How can I calculate several variations and sub-variations.") There is much more to this subject though.

In my opinion, it is important to be aware that calculation in chess is a pure and raw skill - meaning that there is no magic bullet or specific approach to calculate better. Instead, it is downright a matter of practice. The more you sit down to analyze positions, the more practice and experience you will acquire. It's a skill that needs to be developed with hard work and practice, practice, practice - always taking an opportunity to analyze and push yourself.

Having said the above, there are several things you can do to improve on your calculational abilities. I am listing them below under random topics. Some of these are tips for implementation, while others are to incorporate in your chess training:

WHEN NOT TO CALCULATE

Part of the battle is to know "when" to calculate. Daniel King in his Chessbase Trainer "Power Play 10" discusses this in some depth. Part of the battle is to know when to calculate specific variations:

1 - When there is a lot of piece interactions (e.g. positions with lots of open diagonals/files, loose pieces - just lots of stuff going on in the position)

2 - When your King is under attack - In these situation is important to look at all captures, all checks, all possible sacrifices)

3 - When exchanges are offered - Before simplifying it's important to assess the endgame that will arise.

4 - When you are finishing off your opponent - when you are winning... that's when your opponent is at his most alert and his best, looking to swindle the game.

In all instances above is important to calculate specific variations. This I think somehow answers your second question. One of the misconceptions between stronger and weaker players is that stronger player calculates more. However, studies have proven that the stronger player doesn't calculate as much. He just simply has a better overall understanding of chess and can calculate what is relevant. Weaker players tend to calculate more useless possibilities, while stronger players ignore certain moves and variations they know are not good to begin with. Thus, the more chess knowledge and experience your acquire, the higher the relevancy of what you calculate will be. We are not like computers who have to process every single possibility in the position.

OTHER METHODS

1. Process of elimination - Basically weeding out bad moves. For instance, if you have three possible alternatives, and two are bad - then you must play the third one, even if you are not sure what the true outcome of the third one will be. There are some highly instructive positions from Daniel King's Power Play 10 that go over this. This I've found really helpful for my game.  

2. Stepping Stones - This I picked from Jacob Aagard's book on Excelling at Chess Calculation. Basically it's to increase your ply count - You calculate a position, and when you reach the end of the variation you spend sometime fixing that position in our mind. Then you start to calculate from that variation. I've tried it, but haven't develop it to say that it has worked for me. I would recommend this book, just for the training positions themselves.

TRAINING METHODS

1. As I said, calculation is a pure raw skill, so there is no magic bullet. Dvoretsky said that improve our calculation one should try to solve exercises, "the main difficulty of which lies not in the search for pretty, camouflaged ideas, but in the need to examine a large number of different branches, which sometimes requires analyzing many moves ahead" - In other words, the answer to your first questions is to "just do it" and do it often. He has several training position in his book School of Chess Excellence 2 (Tactical Play), and School of Chess Excellence 1 (Endgame Analysis). Incidentally, both Dvoretsky and Kotov agreed that one should look for training positions that are comnplex enough and the author has a lot of analyzes on such positions - that allows you to compare the quality of your analysis.  

2. The analysis of forcing variations is great practice. I would recommend the 200 positions in Jacob Aagard's Practical Defense for this, as well as endgames with fewer pieces on the board.

3. King and Pawn endgames are a good way to improve on calculation of several variations. These are complex enough, but also simple enough for the brain because one usually has few pieces on the board with same-move qualities. You can start with tricky King and pawn endgames -maybe one or two pawns, and increase the number of pawns. There are several endgame books with good exercises.

Going back to the beginning, calculation in chess is multi-layered. The more you learn about things like mating patterns, tactical patterns, positional thinking methods, creativity, unforced thinking, visualization, overall chess knowledge - the better your calculation will become.  

Finally, I would suggest you ask this question to several folks to get different answers. I've heard blindfold chess is good practice. Today's software like Fritz or Chessmaster allows you to practice. There is a wonderful blindfold analysis function on Fritz 12. I reviewed the product on www.chesscafe.com  - you can look for it, I think it's in the archives (I write a monthly chess column for them).


Hope this helps!~
Louis