Saturday, January 08, 2011

Skill assisted thinking.

Can I?

After 6 years of blogging the gist of chess improvement seems to boil down to two area's of improvement: knowledge and skill. Both have to be taken care of.

In my case, at this moment, positional knowledge is the area where I can gain the most. I'm busy to make a summary of Silman's HTRYC 4th edition, so I consider positional knowledge to be taken care of.

I did write a lot about skill-assisted thinking in the past, without ever finding an appropriate method to train it though. Today I was thinking about a certain section of De La Maza's articles:

"You can refine this experiment further by creating two computerpersonalities, one that can
see three moves ahead but has no positional knowledge and the other that can see
two moves ahead and has complete positional knowledge. The tactical
personality, which can see three moves ahead, will win the vast majority of the
This is a key lesson: all of the positional knowledge in the world is worth less
than the ability to see one move ahead. In other words, given the choice between
being able to see five moves ahead in every position and having no positional
knowledge and being able to see four moves ahead in every position and having a
GM's positional knowledge, you should choose the former."

This is a very intriguing text. I haven't tried it myself, but I can imagine that it is true. If so, it sheds a certain light on positional knowledge. What are the means by which positional knowledge affects your chess? In fact it is an attempt to see ahead by means of probability. I cannot calculate enough ahead to known where the best position of my bisshop is, but I can put in on a square where it is active. The probability dictates that there it has the greatest chance to attribute succesfully to the game.

The words of DLM suggest that the nett effect of moving by means of probability is less worth than the ability to really see ahead. DLM narrowed that down to tactics, but I don't think that is necessary.

Do we find confirmation for this point of view? What was the amount of moves that a computer could see ahead by brute force when they begun to outplay even the positional grandmasters? What was the amount of moves that a computer could see ahead by brute force when they begun to outplay us? It should be possible to create a graph with rating on the one side and the playing depth on the other side.

If this is true then the decisive feature in chess is the ability to see ahead. Positional knowledge is a helpful means to emulate this.

(Sorry Chesstiger, again I sacrified clarity for speed)


  1. " ...It is worth pointing out that, even for improving depth of search, it is more efficient
    improve one’s knowledge base than one’s ability to look-ahead. Assume that there
    e, on average, 35 legal moves in a position, each generating a new position or a new
    de in the search tree (cf. De Groot, 1978). Without knowledge, it is hard to decide
    mong the possible choices. Assume also that a large amount of knowledge, encoded as
    unks and templates, reduces this choice to 4 plausible moves, on average...."
    Training in Chess:
    This text and others suggest that the calculation ability from a master is not that different to an expert or even a high class player

    Usually you calculate only the forced lines and the moves of both sides wich go ahead with your positional evaluation.

    The evaluation functions of the open source engines are complex (timeconsuming), with a more simple evaluation it would be possible to calculate deeper:
    There is a lot of positional knowledge in this function

  2. " ...It is worth pointing out that, even for improving depth of search, it is more efficient
    improve one’s knowledge base than one’s ability to look-ahead.

    Well, I have no valid reason to not do both. So I do.

    The blindfold chess makes me feel stronger during games, I need less time and energy to see the lines so I assume it works for now. It is fun, even.

    I will have a look at your links.

  3. I read this somewhere, dont know where.

    "EFFORTFUL STUDY is the key to achieving success in chess, classical music, soccer and many other fields. New research has indicated that motivation is a more important factor than innate ability."

  4. To make an estimation about the relevance of positional skill and knowledge, the following test is helpfull. Take a book position that you have studied (from HTRYC 4). Adjust the playingstrength of you chessengine to a fixed ply-depth (start with 3). play against the machine and try to win. As soon as you start to win raise the ply-depth with 1. Stop as soon as you are unable to beat the computer. At that moment you know that the computer has topped your playingstrength considerately.

  5. Looking forward to your report on Tata and your findings.

    It would be interesting to hear if you feel that strategy,planning and restricting your opponents play are larger factors in this years games.

    Please let us know if you think working with the Silman book has significantly changed your thought process and playing style. I wonder if 1 tournament will be be a sufficently large enough sample size ?

    Good Luck!

  6. A video about the Silman thinking system.

  7. @Phaedrus
    That is a very clever idea. I have to think about what it exactly implies. It sounds like a good way to improve your "technique" too. I will experiment with it in due time.

  8. @Tak,
    Thank you, those video's are very clarifying!

  9. Silman thinking system has been stopped in HTRYC4. See page 13