Tuesday, June 19, 2007


The position below is a great exercise of Kotov's book "Think like a grandmaster.

Diagram 1

Black to move, white to win.

White threatens Qh6 which leads to mate or big material gain.
Black has 5 reasonable defenses:

1. Kh8 (14) to defend h7 with the rook
2. f5 (19) to let the queen help in the defense
3. Bxd5 (45) to capture in important attacking piece
4. Rae8 (15) to defend f6 with a rook
5. Rfe8 (29) ditto

The figures between parentheses are the amount of moves of which the main branch exists (usually split in different branches of a higher order).
The idea is to analyse all logical lines until white wins.
I suggest you try it for the first option of black: 24. . . . Kh8 which has only ca 14 moves.
You can write down all the importants lines you find.
[24. . . . Kh8 25. Bc5! Qe6 26. Be7!! Bxd5
(26. . . . Rg8 27. Bxf6+ Rg7 28.Qg5 Rag8 29. Ne7! Qe7 30. Bxg7+ wins the Queen while every other 29th move wins easy)
27. exd5 attacks the Queen and threatens mate at h7]

To analyze all important lines you have to investigate about 125 moves, which will probably take you a few hours. And even then you will probably find that you missed quite a few beautiful combinations. This is really a great exercise! Which brings me to the following inevitable conclusion: there is no difference between calculation and blindfold chess.


  1. PS One of your least favorite authors, Buckley, asserts "Doing calculations is simply playing blindfold with the board in front of you, so you should learn blindfold to improve at analysis".

    Tisdall makes the same point. I think you will really like Tisdall chapter 2.

  2. As I commented on BDK's blog not too long ago, I play much better blindfold chess if I'm allowed to look at an empty board while I play. Of course, this is cheating in blindfold chess. But this is not cheating in regular chess. So I think there is still some difference between really playing blindfold and calculation in a regular game.

    One awful symptom in calculation in a regular game is that the seeing the pieces can cause a mis-calculation! Most chess players I know have at least once mis-calculated because they thought a piece was still on it's original square at the end of the line even though it moved early in the line.

    How about this for a related conclusion: Since you have branching points in your analysis tree, calculation is like playing a blindfold simultaneous exhibition!

  3. Ok, I will change my definition in
    "calculation equals to blindfold chess with cheating" if that makes you guys happy. The essence is that there is no difference.

    Loomis, your remark about a simultaneous exhibition is very important. The focus of conscious attention works in a serial manner by nature. Since it makes use of the short term memory it can handle only one thing at the time. That's why a single forced line can be followed for multiple moves with relative little effort. But the tree structure is more appropreate for a parallel approach. That is the problem we have to solve, how do you handle that?

  4. "But the tree structure is more appropreate for a parallel approach. That is the problem we have to solve, how do you handle that?"

    I don't think we can handle it, to any great extent. Forcing lines? Sure, we can go out 12 ply or more, but calculating the results of non-forcing moves to any great depth is impossible given the combinatorial explosion that is chess.

    I believe there are benefits from practicing thinking techniques and analysis, but for tournaments there are even greater constraints. 125 positions? Three hours? Not happening when my time control is 90/40 SD 30. I'm going to lose on time! Perhaps it would be better to set a limit of 15 minutes and check to see if you come up with all the reasonable 3 ply sequences.

    For myself, I am trying to learn to calculate, something more like 5 first moves to a depth 3 ply (15 total positions). Why? I'm finding that I (and my opponents) lose most games as a result of rather simple miscalculation.

    I'll be interested to read more of your research into this topic. Thanks.

  5. Errr, "90/40 SD 30" should be "40/90 SD 30" (I'm not playing blitz!! ;)

  6. Correct me if I'm wrong, but in the Melody Amber "blindfold" tournaments don't the players look at an empty board?

    I think "blindfold" no longer has a literal connotation.

  7. As a matter of fact the "blindfold" games use a computer:

    "The blindfold games will be played with computers. An empty chessboard will appear on the player’s screen. The moves are made with a mouse or with the keyboard. Each move will appear on both screens. After confirmation of receipt, the opponent’s clock will start and the move will disappear. However, the arbiter and the spectators can follow the entire game on monitors that are not visible to the grandmasters."