Saturday, September 15, 2012

Category "Incomprehensible"

I have invented the category "incomprehensible"
It contains problems with moves that I will not find in an OTB game since they don't make sense to me. Take for instance this problem:

White to move.
You can find the solution here.

The first move is Rcd7+
I am not able to formulate a narrative that describes in a positive way why this move is so good. What does it accomplish?
I can only formulate it in a negative way: every other move is bad. Such moves simply blur my mind. There is no way I can find such moves myself as long as I don't see what it does.


  1. If every other move is bad than its an "defence" move?

  2. Yeah, kind of voluntary zugzwang :)
    But what does the move do?

  3. Removing defender: the king is forced to the king-wing and the rook gets captured because the king can go to c7 and discouvered attac...
    Many ways to look at a tactic...
    I would say Kc7 unpins, attacks ... thats a massive activity gain, before nothing was possible= no activity

    Do you feel any improvement through your training by now?
    I would think ~200 problems is not enough to get a "break through"

  4. By the way i have often the same problem: i dont get it. Then i play with the engine around. The engine takes my part and i try to defend. Just the change of seat is helpful and playing around sometimes to.

    But even then... my brain is to small...

  5. @Aox,
    It is not about quantity. High rated problems differ so much that they can't be compared by geometry. It is about finding the shortcomings in your mind and overcoming them. That is deliberate practice.

    Memorizing the 178 problems is only the first step. Now I must process the acquired knowledge. The more I organize it, build structures and lay connections, the more cues and chunks my mind will install. Hence the chance that I will remember it otb will improve.

    It is about basic skills that you need in every position. If you are not able to recognize an overworked pawn, you are clueless when you reach a position where it plays a role. Processing knowledge is to find out: how do I recognize an overworked pawn. Only then it becomes position independant. And will there be a transfer of knowledge to other positions.

    The problem with incomprehensible moves is that they don't trigger an emotion. With a normal move there is an emotional reward like the feeling "that's winning". Without such emotion, I can't learn it. It is not transferable. In 5.6% of the problems I encounter such moves. So I ask myself "what is happening here and how can I overcome it?"

    Another point is that I often underestimate a kingside invasion. Here the question is justified "how do I recognize the danger?" By investigating those positions and find out what I overlook I might be able to overcome this.

  6. Just had a look at my statistic: about 11% of the problems i get served i do "bad" and i would say 50% of them are wrong/stuped/unclear/strange.. thats ~ your 5.6%.
    But i doubt that this is a "new" information, this might be "normal" anyway, just astonishing that my low rated set has ~~similaritys to your high rated set.

    I have the bad feeling that you (i) can create new structures and connections but only at the cost of other (aproved old good) structures and connections. Not used structures are only available in our youth. It seems to be necessary to reprogram neurons which had been used for other things >>but<< chess or we simply circle around the same quantity of neurons.

    We will see...

  7. Are you suggesting that learning to recognize an overworked pawn can cost you the ability to recognize a knightfork? Or that you have to give up your ability to parallel parking in order to extend your chess skills?

  8. I do not believe that the defence given is the critical one. The Black K can retreat to c8. White has Rc7+, followed by Rxh7, but Black can play Rxd5+. The resulting pawn ending is won for White though.

  9. But actually, that is even a pattern!

    Two points:
    a) a fork
    b) the black pawns on the queens wing dont let the black rook chicken out.
    the white rook is keeping the black king on the back rank, so he cant make way for the trapped black rook.
    Black is helpless against the threat of a rook manoever Re7xh7-h8+ with a skewer and winning the black rook.

    The thing you can learn here is, that the white rook is not only nailing the black king on the back rank, but that this can lead to skewers on the back rank.
    I have seen this idea several times, so I call it a pattern. But because of the variation with the fork (which is actually played out as main variation) this puzzle is very difficult and hence it has a high rating. If I come across a similar puzzle with the same back rank skewer idea I will post it here.

  10. @BK,
    true, I found that line too. I found in the end a few things that blacks first move accomplishes. It annihilates the bishop, so there come backrank threats in the position. But the problem still remains a bit vague to me.

  11. @Munich,
    not quite. The skewer doesn't work here due to Rd8+ and black comes a full rook ahead.

  12. Most incomprehensible moves become clear after extensive study. But not always.

  13. Tempo: It is difficult to understand your single Rd8+ move. I cant see a critical variation where one or the other black rook gives check on d8.

    I made a comment in the comment section of this puzzle, and made the skewer pattern the main variation, but also covered 2 side lines.

    I guess now it becomes clear. This isnt really unsharp. It is difficult - yes. But the difficulty is to see all variations which have all on its own pretty clear patterns. You miss one pattern - you fail the puzzle.
    There is this fork (which is played out by the CT player)
    Then there is the discovery with Nf6+.
    And most important (I think) is the skewer pattern thread with Rxh7 and Rh8+ next.

    This puzzle is the prototype of the argument that difficult puzzles are made of several (easy) patterns.
    This argument is not always true (there are high rated puzzles with just one pattern), but here in your example it is very true.

  14. @Munich,
    can you give me the line with the skewer?

  15. I have looked at the comment you made at CT. I see what you mean. Indeed there is a skewer which is remains only a threat in most lines. I had ofc. seen that. In fact I had seen all patterns. Yet I had a feeling of vagueness by the first move. You give up the exchange with a reason you can only find after processing a lot of variations. It is hard to describe what the move actually does. So I'm forced to look at a counter intuitive move which doesn't seem to have much justification beforehand.