Wednesday, January 30, 2019

Weird things can happen

From time to time I'm presented with a problem that takes me an absolute weird time to solve. Take for instance the following diagram:

White to move
4r2Q/p2r1k2/1p6/3b2Pp/2q1P3/2B2PK1/8/8 w - - 1 1
[solution]

It took me a shocking 6 (!) minutes to solve this problem. I kept repeating the wrong moves over and over again in my mind.

Now I have made a new pattern of it. The queen as a hammer, and the two squares behind the king as an anvil. Belief me, I have used this pattern very often since I invented it. I just look for the squares that are suited as an anvil to crush the king against. It is much easier when you know what you are looking for.

4 comments:

  1. This is a clear checkmate puzzle: Black is 2 rooks up ( counting material first )

    The knowledge of the standard checkmate patterns helps
    its a swallow's tail mate

    https://www.youtube.com/watch?v=Xy9SYWolhH4

    ReplyDelete
  2. The framework that is needed to retrieve the right patterns decays over the years. Since you know the patterns are still there, you have the false impression that the framework for retrieval is in good shape.

    I have solved thousands of swallows tail mates over the years, and the pattern is very well known.

    My method reveals the decayed frameworks and restores them.

    ReplyDelete
  3. Paraphrasing Shakespeare: "To 'see' or not to 'see', that is the question."

    There are several named checkmates which reduce to the same "idea":

    Épaulettes Mate
    Guéridon Mate
    Swallowtail Mate
    Dovetail Mate
    Dovetail Bishop's Mate
    Escalator Mate (Closest to the pattern in the problem)
    Triangle Mate
    Vukovic Mate
    Box Mate

    There are other stock checkmates which also have elements of these mates, for example, the King and Queen checkmate against the lone King.

    Paraphrasing The Bard again: "A rose [checkmate] by any other name would smell as sweet."

    While I was studying the "Box" (the enclosure that must exist or be constructed and enforced around the enemy King in order to checkmate it), I noticed something common to all these patterns involving a Queen checkmate (named above). A Queen can control all of the squares around a King except for two. Those two squares will always lie on the two uncontrolled (by the Queen) squares which are a Knight's distance from the focal point, which is a square near the opponent's king position that we want to focus on with our pieces. Once you "see" that relationship as a pattern, you can "see" the pattern when it occurs (or can be reached in the reasonable future), regardless of the piece position/configuration. That (to me) is the essence of pattern recognition: "seeing" the underlying pattern that is common to these specifically named checkmates, without requiring a recognition of the specific checkmate per se.

    As Aox noted: (1) White to move, and (2) no possibility of winning material or Pawn promotion, therefore it must be about checkmate.

    There are two potential checkmates with the King in the "Box": (1) BKe7 and (2) BKe6. The focal point in both cases is f6. The only piece which can effect the checkmate is the Queen, so the required sequence of (two) moves is to force the Black King to either e6 or e7. 1. Qg7+ Ke6 (forced) 2. Qf6#.

    The key is to "see" the "Box" and how to create and close it off. The Bishop cannot be used for this, but only for protecting the Queen as it tries to "up close and personal" with the enemy King.

    The same considerations allow for variations on a theme by using other pieces to control those two squares that cannot be controlled by the Queen, and which are a Knight's move from the eventual focal point.

    By analyzing in this way, the number of distinct patterns which must be remembered (i.e., triggered by pattern recognition) are reduced to a minimum.

    The names are unimportant for recognition purposes. It also is unimportant to categorize the various mates by incidental details (such as differentiating based on whether the two squares are a Knight's distance away from the focal point and on a rank or file, or on a diagonal), as long as the underlying pattern can be recognized.


    ReplyDelete