Tuesday, July 24, 2012

Talent = the ability to reason

Due to a discussion with mr. Z about yesterdays post I realized that I had to look again.
So I took this 2435 rated problem and tried my hand in order to see where exactly the failure happens.

White to move.
Last move black was Nxe7
You can find the solution here.

The reason I failed this problem is because I wasn't able to formulate a decent reasoning. Yesterday I argued that this is because my mind is overwhelmed. Today I found that I deceive myself here. The real reason is less flattering. Somehow my mind is lazy. I start with trial and error in automatic mode, but when nothing works, I'm not going to formulate what the problem is but in stead I repeat my trial and error. It seems that I'm not able to leave my comfortzone and formulate what the problem is. At the very moment I do, I'm able to solve the problem. But I cannot set myself to it.

Doing much problems doesn't sound lazy, but in fact it is. Since I did them on autopilot and didn't cross this barrier.


  1. Imagination in Chess [Paata Gaprindashvili, 2004] might help? Read chapter 2 about "Reciprocal thinking": "If you fail to make an idea work we need to stop and ascertain the cause of failure and then attempt to correct our design."

  2. I tried to solve the problem.
    I finally gave up.
    Not beeing able to solve this problem does not seem strange to me, though.
    It is indeed really difficult - it is a check mate in 9, and the 2nd move is a quite move.

    If I miss a puzzle like this one, that I really wonder:

    Whenever I do it wrong (and I did several times, today was no exception) only then I clap my forehead and think "oh no! Not that simple one again!!"

    I hope I can eliminate these blind spots by repeating failed puzzles. (at least to some degree).

    But failing the 24xx rated puzzle you showed here doesnt seem to be totally "stupid".
    I dont mind failing those (as long as I am not a master).
    And you know what? keep that very 24xx rated puzzle on your watching list. I could imagine it will climb even higher in the future. (only 6 solvers so far, so the 24xx rating isnt so stable).

  3. @Aox,

    Indeed, that advice helps, although it still ain't easy

  4. @Munich,

    It doesn't bother me that I fail this specific problem, but that I'm not able to build a decent logical reasoning.

  5. I did not try the exercises of the book but there are many and the methos is further developed..

  6. Hm, failing logical reasoning.
    Here I have a more simple puzzle, which is "only" at 1912 CT Blitz rating.

    I just failed it. The reason is, that I could not derive the solution by logical reasoning:

    And I know why I could not: I did not know the mating pattern, white is threatening.
    This is already a hint for solving the puzzle. You need to know that white is threatening check mate. It must be part of the logical reasoning. If you miss this detail, you wont find the right move. If you know that white is threating check mate, then the puzzle becomes pretty easy. You know then what to do.

    What am I trying to say? Logical reasoning has a lot to do with knowing what patterns are in the position.
    I know, that sounds terrible close to "back to learning patterns", with the answer "I did already solve 100.000 puzzles, I know most patterns".
    True. We know most of them. The often even disturb when trying to find the solution. Which is the right pattern? several patterns seem to scream at the same time and some of them are misleading.
    Only logical reasoning can find the right move order and sort out wrong/ irrelevant patterns.
    Nevertheless: you need to know them in order to decide which pattern is relevant and which is not.

  7. @Munich,
    You will never learn enough patterns within 100 years to cover all those possibilities.

    Logical reasoning is the safety net that comes into action when pattern recognition fails.

    Logical reasoning is assisted by trial&error at lightning speed. You must learn to play the moves in your head in a Troyis kind of way. If it becomes totally effortless to look what happens after the check, you will find the problem of mate without pattern recognition and without logical reasoning.