Saturday, August 27, 2011

Stepping stones
























.
.
.
Allthough the chess season hasn't started yet, it is allready itching. The Tata tournament is lurking beyond the horizon. Time to make a study plan for the next half year. First let me distill a final conclusion from my previous posts:

Conclusion.
In essence there is no difference between analysing a current position and a future position. Since a position is a position. The only difference is the base from which you start. With a current position you can interrogate the physical board for the exact position of the pieces. For a future position you have to interrogate the mentalization in your mind.

Question.
This leads to the following question. How do you form a mentalization of a future position?
There are a few options.
  • Stepping stones.
  • Chess memory.
  • Mentalization.
Stepping stones.
Jon Tisdall advocates a technique with stepping stones in combination with blindfold chess in his book "Improve your chess now". This technique sounds convincing enough, allthough I couldn't find proof that it actually works on the web. The reason I like it is that Tisdall's solution seems to fit my description of the problem so well.

Chess memory.
Before the summer break I had started with a nifty little program called "chess memory". It shows a chess position for 10 seconds which you have to reconstruct from memory. At first I could only remember positions of 6 pieces (=what fits in my short term memory). But after two weeks this developed to positions of 10 pieces. What I like about this program is that you see chunkbuilding at work. In order to remember more pieces, you must group them in a chunk en store the chunk as a whole in your memory. Susan Polgar showed in a documentary of the BBC ("My brilliant brain") that she can reproduce an entire position of 32 pieces after seeing it for only 3 seconds. Being able to do so seems to be a good way to solve the question I defined. The chess-memory program I mentioned seems to train this isolated aspect.

There are a few facts that cast some doubt on this solution though:
  • The quality of the chunks I build doesn't feel all that well. I start binding together pieces with "a knight's distance", "a long knight's distance" (1 square diagonal, 2 straght), "a very long knight's distance" etc..
  • There is scientific evidence that intermediate players could learn to memorize entire chess positions this way in about 50 hours of training without any significant difference in their playing ability.
  • chess memory seems to be a side effect of other skills.
Mentalization.
This is about translating a chess position into previously acquired concepts. See the scientific base here. I discussed this with mr. Z and we did some experiments. This idea seems to be the most promising. It means that we get round visualization. This theory covers the most facts and there is scientific backup. The downside is that we have to invent a trainingsmethod based on this theory yet.

What now?
Before the summer break I was busy with method 2, training chess memory. During the vacation I inclined to method 1, stepping stones. Hence the title of this article. But now I'm thinking about it while writing, I see that method 3, mentalization is the most in line with my previous thoughts. So method 3 it will be.

Preparation before training.
In september I will have another geology excursion.
I will start in octobre with serious training for the Tata Steel chess tournament.
This means that I will have about a month to invent a training method based on mentalization.

5 comments:

  1. Chess memory is related to "chunking" but these chunks should have "quality" to improve the play. I think the training of easy tactics might create some ( tactical ) of these chunks. If you do easy tactics ( maybe like Bright Knight ) and improve the speed of recognition, you are forced to recognise pattern ( chunks ) wich have a chess-meaning.
    Such a frame of tactical chunk's might help to implement other chunks too?

    ReplyDelete
  2. Well, read this to understand why I don't feel much enthousiasm for your idea.

    ReplyDelete
  3. Your exercises where tactic exercises. To do Heisman "speed"-tactics is pattern recognition. All tactics you did where never intended to solve the problem in (virtually) zero-time = recognise them from LTM in zero-time.

    Its not about solving: its about remembering

    Bright Knight showed that its possible to memorise thousands of pattern in a relatively short time.

    ReplyDelete
  4. I was busy with it, but my latest rantings lead me to the conclusion that there are two areas:
    Analysing a position and creating a mental stepping stone (= a future position).

    The greatest gain can be acquired by the second area. Speed tactics lie in the first area.

    I always follow my latest conclusions. My conclusions change by the day, BTW.

    That doesn't mean that I don't come back to area one some day. But that day simply isn't now.

    If you think that area one deserves more investigation NOW, feel free to lend a helping hand:)

    ReplyDelete
  5. I wonder if your "mentalization of a future position" is the same as Rolf Wetzels APROP= Ability to PROject Positions? In his Book Chess master at any age he suggest some drills for APROP:
    Look at any game at, say 10 moves, close the book and rapidly construct the position an a diagram/board. As you improve try 12 moves..

    ReplyDelete