Sunday, November 11, 2007

Count your attacks

Precondition: No tactical niches like traps or promotion.

Positional moves.
A move that doesn't attack a target and is based purely on positional considerations I wil call a positional move.

When one target is attacked with a move the move is called a mono-move. The defender can always do something about it. But that will cost him a move. A move that defenses against only one attack I will call a mono-move too.

When two targets are attacked simultaneously with one move, there is a great chance that there is no move that meet both attacks. Since mono-moves are much more common than duplo-moves.
A move that attacks two targets simultaneously is called a duplo-move. The only way to gain wood in a forced way is by a duplo-move. A move that defenses against two attacks I will call a duplo-move too. In stead of counting your beans you should count attacks.

Difference between tactical problemsets and games.
Tactical students, like we all are, have a distorted picture about reality. First, a tactical problemset is always about duplo-moves. Otherwise it wouldn't be a tactical problemset. Secondly, you jump in the middle of a position. To evaluate that position, beancounting often will be necessary.
In real games it is different. Positional moves and mono-moves are the bane of the game. With positional moves and mono-moves there is no necessity to count beans. When a piece is attacked you have to defend it immediately. Otherwise you will lose wood. At the utmost the defender can postpone the defense by means of a counter attack. Bottomline is that there will always come a moment that the defender will have to defend his piece. That's why you will always know, after a series of positional moves and mono-moves, that every piece is well defended. Otherwise you would have lost it already. Any beancounting will tell you that. So there is no need to count your beans.

Only when the first duplo-move is played, attacking two pieces simultaneously, one of the pieces can become underdefended. If the defender has only a mono-move to his disposal, he is in trouble.

During the the game, you can replace the whole beancounting stuff by "duplo-move-awareness". Before the first duplo-move is played, there can't simply be no unbalancies in the form under-defended pieces.

If there is an unbalancy in the form of under-defended pieces before the first duplo-move, there are two posibilities:

You have blundered. You have forgotten to answer a mono-attack with a mono-move. You have answered an attack with a positional move. If you blunder, you deserve to lose.

That leaves only one other possiblity: if your opponent has blundered by answering a mono-attack with a positional move and you didn't notice it. That is the only situation where beancounting can give another answer than one would expect. Then again: you deserve to lose. That's not the kind of chess that is useful to talk about.

If we forget about blunders, you can simply know that all pieces are well defended. The pieces will be protected exactly sufficient, or they will be overprotected. The price you have to pay for this knowledge is that you have to be aware of all duplo-moves before they are played.

Know your duplo-moves.
In order to see the duplo-moves coming, you must know them well. There are 3 kind of duplo-moves:
  • Double attacks
  • Discovered attacks
  • Pins/skewers
I will deal with these duplo-moves in a separate post.

Dear reader,
I hope you can look through the dullness and semi-intellectual highbrowedness of these posts. What I'm describing is almost too trivial for words. But that is what I'm after! Triviality. Since triviality means ignoring the irrelevant. And ignoring the irrelevant is relieving your short term memory. These posts are about finding out what is irrelevant. Which worries in my game can I let lose without being punished for it. Which phantoms can I prove to be unreal. So I don't have to worry about them anymore in the future.


  1. lovely. thank you. im able to follow along at a distance, but not deeply, but with sincere interest in your work.

    i also thank you for your kind personal notes, for not quitting on me.

    [this is your post, so the 'me part' most briefly: these times are rocking me, but, despite that, my faith is very strong, what will be for the best will happen, and there is a higher order, a higher power. there is an inteligent universe, i feel.]

    warmest regards to you, great hopes for your chess, wellwishes to your family, and success in what you define as important to you, and regards to all your excellent kind readers.

    love dk

  2. another thing which is trivial but maybe could be mentioned here, is that every duplo-move is preceded by a mono-move, and it's actually not the duplo-move you need to adress, but the mono-move leading to it. the defending duplomove against an attacking duplo I would treat as a rare exception, so the 'last chance' mono-move really is the one we should be focusing on.

    I think this probably the biggest difference between a tactical problem and a real game. in a tactical problem the last mono-move defending against the unstoppable duplo is already missed, so you won't see that move in any of the thousands of problems you solve. naturally we all know it's there, but in practice we easily miss it and blunder. I know I miss loads of those in blitz, even though I see the first tactical move in a problem fairly well. I think there is a difference (which playing a lot of blitz can cure).

  3. WW,
    that's not trivial at all.
    Before the duplo-attack you can have a few preparatory moves which are forced mono-attacks, and before that the attacker plays his last positional move or mono-move which the defender has to defend with a mono-move.

    The point is, you must see the duplo-moves coming long before. But how? What are the characteristics?