Wednesday, September 12, 2007

Paralysis by analysis




















This is about the 10th day that I look at the same position. I have trouble to come any further. With the list of 3 items (invasion, overloading counterattack) it is possible to find out the goal towards you have to work.

But the problem is how to reach that goal and especially which move order is best? This is where actual calculation comes in.

When I try to calculate the moves the problem is that I try to accomplish 3 things at the same time. I calculate my own attack with my opponent's possible responses, I calculate the counterattack of my opponent with my possible responses and I try to do the bookkeeping to see if I'm actually ahead or behind in material. This is too much for my poor short term memory and I simply paralyze.

Your comments on my previous post are helpful. Montse pointed out that it is all about threats, and he is right. But how can I simplify these 3 tasks (my attack, his counterattack, bookkeeping) I try to accomplish? Finding an answer would mean a quantum leap ahead.

4 comments:

  1. Hi Tempo,
    you are dividing the analysis in three parts here: 1) analyzing your attack, 2) the counterattack possibility, and 3) bookkeeping. In my imagination, bookkeeping works differently from the first 2: you visualize each position and evaluate it. Then you chose the move leading to the position(s) with the best evaluation from your point of view. I believe that this is one major difference between patzers and GMs: GMs can evaluate positions almost instantly.

    ReplyDelete
  2. Sciurus,
    yes, that's right. I imagine there are methods one can use for that. But I have to find out what shortcuts I can use.

    When there is a pawnrace in the endgame, I'm inclined to move my pawn, then move my opponents pawn, then mine again etc.. All before my minds eye. That is very difficult since I tend to forget where I was. As Likesforest pointed out you can use the different method of just counting. That is a method that is far less taxing for the mind. But you can use this method only if you know the preconditions. It makes a difference who starts, for instance. And if there are impediments along the road etc..

    I imagine that such simplefying methods can be found in this complex middlegame position too. I'm trying to define them.

    ReplyDelete
  3. What are the traditional methods used to attain the necessary skill to solve these types of problems? I've always hoped that it would be enough to solve a fair amount of tactics in a relatively long time, maybe in combination with calculation exercises a la Kotov. I can't say that I have enough experience to say if this method would be effective or not, but you (Tempo) seem to have been solving tactics for quite a while now. What are your thoughts on this?

    ReplyDelete
  4. SP,
    I'm not sure if people overcome this impediment usually by exercises. Which is something quite different than to say that it is impossible to overcome it by exercise.

    What I mean to say is that the usual plateauing of people is determined by their natural ability to overcome this impediment. The way people usually train doesn't have an effect on this ability.

    I thought for long that it would suffice to do an enormous amount of simple problems in order to learn to do the more complex problems. I have proven by experiment that this is not true.

    The method of Kotov is nice, but he was already a master when he started with it. I don't think there are ecercises advocated for this specific impediment.

    My rule now is simple. Before I can do this fast I must be able to do it slow. To be honest, I have still difficulty with this problem even after 10 days. Of course I know all the lines by heart by now. But that isn't the point. I have not found a reliable method to solve this problem in such a way that I can be sure that if I see the same problem within a few years that I will still know how to solve it. Or a similar problem tomorrow.

    And if I can't solve it slow, for sure I cannot solve it fast!

    I'm experimenting with different counting methods. So far I haven't found a method that satifies me.

    ReplyDelete