Saturday, May 14, 2011

Method of exclusion redux.



















.
.
.
Little recognition.
So far there has been little recognition of the supertrick idea. I have this idea for about 10 years now. I can be wrong of course. I'm not bound to any opinion. But as long as I have nothing better I stick with it. When I look back at those pesky 8-13 yo kids I used to encounter during summertournaments, I simply cannot imagine that their deliberate practice is decisive. Be it in quantity or quality. I always interviewed them afterwards to find out how they were working. I cannot discover any flaw or leak in my reasoning either.

Guinea pig.
The area of investigation has narrowed down vastly. I'm looking for the skills that maintain a stable representation of a future chessposition from where I can start analyzing. A stable platform or "stepping stone" (Tisdall) from which I can manipulate pieces without interference or a fading memory. Little is known about these skills and information is contradicting. I have to rely on the method of exclusion again and will have to act as a guinea pig myself.

Visualisation.
The method of visualization is already eliminated by both my own experience as by the arguments of scientific papers I have read lately. There is no need to visualise a physical board and pieces. In stead you have to encode the position by concepts, relationships and such. "Mentalization". You know where the pieces are and what they are doing without actually seeing them.

Blindfoldchess.
Blindfold chess is excluded too as method to adress the needed skills, as I already mentioned in a previous post.

Chess memory.
Admittedly chess memory hasn't the best papers to be the skill we are looking for. It seems to by a side effect of the skill we are looking for and not the skill itself. Yet I can't dismiss it solely based on theoretical arguments. So I started a program to improve my chess memory in the way prof. Adriaan de Groot suggested it. I use positions with few pieces abstracted from real chessgames and look at it for 10 seconds. Then I try to set up the position by memory. The average amount of pieces I can remember is about 6. Which is the amount I can store in my STM, I guess. Thus there is no or little use of chunking obviously. Maybe that has to do with the fact that with only a few pieces there are few recognizable chunks possible since the pieces appear somewhat scattered over the board. I don't know, but I'm going to find out. When my memory improves, I will use more pieces.

Ideas for other skills that might be relevant for future experiments in this context, anyone?

17 comments:

  1. Something in between tactics, visualisation and memory

    http://www.fileden.com/files/2011/4/11/3113768/visualwizedemo.swf

    Such exercises could be produced looking up the games behind some tactics. But i think Standard -puzzels at CT are somehow the same.

    ----------------------------

    The pattern you try to memorise have to be connected to a meaning.
    At least you should use "real" endgame-positions maybe mate in x's, or promotion in x, Or Pawnbreakthroug in x....
    Why try to memorise "nonsense".
    -----------------------------

    I think the best would be, to learn (anotated) GM-games with your openings. You benefit in many ways at the same time. Study game, read anotation, analyse moves and give a reason to every move, learn game by heart.

    ReplyDelete
  2. mentalize a game by 2 weak player move by move: find the blunder, there is more than one.

    ReplyDelete
  3. As far as i remebee prof. Adriaan de Groot suggested not to do this type of training, it was a type of testing the chess-memory. But i had an idea a moment ago. He said that bad player ( like we, sigh ;-) forget especially the pawns. Might be a good training to try to remebre "only" the pawnstructure at the beginning. The pawnstructure is important for the strategys of the game..

    ReplyDelete
  4. @Aox,
    I did the visualwize demo. It doesn't seem to add anything beyond CT indeed.

    The pattern you try to memorise have to be connected to a meaning.

    The odd thing is that it doesn't have the same sensation as memorizing. Doing it for a week now the result seems to be building chunks indeed. I can remember 1 piece more at everage, solely due to thinking in groups.

    I think the best would be, to learn (anotated) GM-games with your openings. You benefit in many ways at the same time. Study game, read anotation, analyse moves and give a reason to every move, learn game by heart.

    That is a balanced approach of deliberate practice indeed. But I'm looking for extremes at the moment in order to isolate the skill. Otherwise you don't know to which element you should attribute the progress. By using methods everyone uses you only can expect average results.

    ReplyDelete
  5. @Anon,
    I don't know if I can make the difference between mentalization and memorization during the exercise. Those two aren't the same. Memorization feels as the lazy men's way.

    Maybe "chess memory was the wrong title".

    ReplyDelete
  6. As far as i remebee prof. Adriaan de Groot suggested not to do this type of training,

    You are right, I took some liberty there.

    Might be a good training to try to remebre "only" the pawnstructure at the beginning.

    I read somewhere that chessnovices could learn to remember a position "De Groot-style" in about 50 hours. So I'm not going to worry about which pieces to learn first, I will do them all.

    That article said that there was little effect on their chess performance. Hence I'm doing this to get more "inside information" of the dynamics.

    I discriminated 3 area's:
    A analysing present position
    B evaluation and movement from present to a future position.
    C analysing a future position.

    I think pawnstructure exercises could be a good supplement to A.

    ReplyDelete
  7. ". Otherwise you don't know to which element you should attribute the progress."

    How do you want to know, if any progress you will have, was from chess-memory or from you tatics-few-exercises?

    You should do a check on your present abilitys first

    ReplyDelete
  8. That's simple. The tactical-vision exercises adress A not C.

    C is without any doubt the weakest link in my calculationprocess by far. So I'm first going to fix this before I get back to these A exercises again.

    In stead of "method C" I will talk about creating a stable stepping stone. That's more clear, I hope.

    ReplyDelete
  9. There is a guy who says Tactical-few exercise would adress C too

    "Yesterday I won my first game solely due to the new acquired skills to see knightforks and pins fast and easy. I played against an opponent who is tactically quite strong. It worked exactly as Dan Heisman said it would work: There was a series of captures with a knightfork involved. At the end I had the possibility to pin a knight and win a piece due to the pin. Nothing complicated, everything pretty straightforward. But I had seen the pin at the end of the series of captures which my opponent had not."

    A tactical analysis in a future position.

    "
    C is without any doubt the weakest link in my calculationprocess by far. So I'm first going to fix this before I get back to these A exercises again."

    So you think your tactical exercise dont improve your OTB? ?? ???

    If your main weakness is realy C, then your Standard-rating at CT should be "low" compared to your Blitz-rating ( in comparison to other players ). I think your Enggamerating should be worse too ( long calculation ). Are you shure C is your weakness. Khmelnistkys Chess Exam is wrong at calculation. People who did Chess Exam #1 and #3 (Fischer) in short time difference showed a big gap at "calculation".

    ReplyDelete
  10. I didn't want to make matters too complicated in the comments.

    You can divide C in two:
    C1 Create a stepping stone
    C2 Analyse a future position

    C2 = A
    I elaborated on that somewhat in my previous post.

    I can't create a stable stepping stone since it simply fades away in seconds. That is how I know I'm bad in it.

    So you think your tactical exercise dont improve your OTB? ?? ???

    Don't panic, it is only a temporary delay. Tactical exercises DO work as I have proven. But it makes more sense to fix my weakest link before I go back to these exercises.

    ReplyDelete
  11. I agree if it could be isolated and you can beat it with a hammer then you have to try something. On the other hand, reduction is often better than isolation. We will see I guess.

    ReplyDelete
  12. Reduction: Do the same as in the real thing, with but fewer / shallower elements etc (= Scale down.)
    Isolation: isolate an element and practice it separately.

    ReplyDelete
  13. I am not sure if it is lkie aox syas, myabe tihs txet mhgit hlep thinknig auobt it. Hevweor, I am not sure how to get it into cteoxnt with ptrtean rciietongon and chess. The pretatn is: the odrer of leetrts in a word is not ionatprmt for the txet to be rlabedae, the olny reqrmiueent wluod be taht the fisrt and last lteetr of ecah wrod are kpet in the rhigt place. The rest can be a ceopmtle mses and slitl be rabdelae. The rhrcaesrees bveiele this is becasue we do not read every leettr iidldliuavny but rahetr the wrod as a wohle.

    ReplyDelete
  14. You may check your weaknesses now here:
    http://www.chessik.com/index.htm

    ReplyDelete
  15. http://www.youtube.com/watch?v=WREgHsTr5yE&feature=related

    ReplyDelete
  16. About STM and so on:

    http://www.cogmed.com/wp-content/uploads/2010/02/Cogmed-White-paper.pdf

    http://www.ritterbrainbuilding.de/trainingsdiamant.htm

    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2383929/

    http://www.mental-aktiv.de/mental-aktiv/Mentaltraining/Trainingsubungen.php

    http://cmslive3.unibe.ch/unibe/philhuman/psy/apn/content/e5622/e7317/e7414/files7415/BT_n-back_e_eng.swf

    http://cmslive3.unibe.ch/unibe/philhuman/psy/apn/content/application/braintwister/index_eng.html

    ReplyDelete