Tuesday, June 21, 2016

Sniffing glue

I have gathered about 70 puzzles which I failed. The next step is to milk every drop of knowledge out of these positions. For now I skip the mates, since they need a separate approach. I tagged the puzzles according to my own system. There are eight main tactical themes by which one can gain wood:
  • Double attack
  • Discovered attack
  • Pin
  • Skewer/Röntgen attack
  • Simultaneous attack
  • Trap
  • Promotion
  • Incorrect sacrifice/hanging piece
Whenever wood is gained, at least one of these themes must be present. In a combination, these themes can be connected directly. For instance, a discovered attack has two targets, and one of the targets can be the attacking square of a double attack.

But the main themes can be glued together by a series of different themes too:
  • Coercion
  • Redeployment
  • Distraction
  • Blockade
  • Interference
  • Overloading
  • Removal of the guard
  • Exchange target piece
  • Capture
All these glueing themes have in common that they preserve the initiative, which means they are a member of CCT. I study the glue.

Sniffing glue


  1. "Directing attacker to attacking field (don't know the English term)" - do you mean DEFLECTION? (to take away the defender from its duty). Or is it different term (situation)? Can you provide a simple example (FEN + variation) of this?

  2. The first move for white in http://chesstempo.com/chess-tactics/109389

  3. The first move is a "redeployment" (using the check on the Black King to gain the tempo needed) of the Knight to a square which "forks" the BBc6 and BQd7, but the "fork" is not the intended goal. After the "redeployment" of the White Knight to e5, the Black Queen is overloaded: it has to defend both Bishops. However, the two Bishops are located a Knight's distance apart, so the Queen is not capable of defending both simultaneously. The initial "clue" is the B.A.D. (B.A.D. = "Barely Adequately Defended"; Heisman's terminology, I think) BBe7, combined with the "obvious" fork if a White Knight can reach e5 without allowing Black to do anything useful in the meantime; the initial check serves that purpose. The White Knight then "attacks" BBc6, while the WBc5 attacks BBe7.

    I considered the term "dislodging maneuver" but rejected it because the purpose of the maneuver is NOT to dislodge the Black Queen from a crucial defense. Obviously, the Black Queen still "protects" both Black Bishops from e8 just as well as it does from d7. In either case, two Black pieces are attacked, with only one defender capable of doing the job. Black is forced to allow the redeployment maneuver because of CCT: (1) check, followed by (2) Equal or Greater Threat (capture of the Black Queen). The double attack occurs at the end of the forcing maneuver.

    Since there is no "sacrifice," (a requirement for a combination according to Botvinnik's definition), there is just a redeployment maneuver which ends in an overloaded Black Queen. (I'm not enamored with Botvinnik's definition, FWIW.)

    The maneuver of the White Knight is somewhat reminiscent of Capablanca's win of a crucial Pawn in the game Capablanca vs. Yates, New York 1924, beginning with 40. Nc3. White wins the Black a5 Pawn at the end of the maneuver. There is considerable controversy over whether or not it is a combination or a forced maneuver. Lasker and Romanovsky consider it to be a combination; obviously, Botvinnik and others did not.

    Link: Capablanca Vs. Yates - "Wheeler Dealer"

  4. Oh, one more thing: it is crucial to capture the BBe7 FIRST; otherwise, Black can capture WBc5 while the WNc6 is still hanging. 1.Nf7+ Kg7 2.Ne5 Qe8 3.Bxe7 (NOT 3. Nxc6??) Qxe7 4. Nxc6 Qb7 5. Qa4 with a safe extra piece.

  5. I have just tested some ideas at CT. I noticed that whenever I recall the position and understand it fully - I can solve it within 3-5 seconds (I mean mate in 2-3). However when the position is not at my stored patterns database - I have to decide what has to be done to achieve the goal.

    What I am trying to say is: every time I see the position familiar (exact) to these from my chess database - I do not have to think at all - just to recognize the specific position and recall the solution. But when I see the unfamiliar position - I have to work it out and the degree of mistakes (errors) grows a way more! That's why I could achieve a very high score at the tests which contains the positions very similiar to these I have analysed and stored in my chess database.

    1. And that puts you easily on the wrong foot. You are talking, if I interpret your words well, about memories here. Memories are not patterns. You need an infinite amount of memories to cover all situations. I can describe almost all combinations at CT with only the 18 patterns as described in the post. What we have to learn is to recognize those 18 patterns fast. I try to accomplish that by dissecting every combination into its elements in every puzzle. Your memories should serve as cues for fast retrieval of the 18 patterns. That are the kind of memories you need.

  6. I do not know if you understood my intention right. I wanted to share the info that whenever you recall/recognize the position in exactly the same as you learnt it (by heart)... you can find the solution in a nanosecond. Of course I do not mean to memorize all the universe of positions to recall these! I am just curious how much memory is resposible for finding the best solution in a position we do not know (or at least have not memorized yet).

    What I noticed is I can recognize most (simple) tactics when playing fast blitz (or bullet) game online. I do not know what is the reason for that, but I can feel a tiny progress (or at least better and faster recognition of tactics when being low on time).

    1. Your observation and my somewhat dogmatic answer might lead to an interesting experiment. I studied 70 positions in depth, as described by this post. I couldn't help but learn the solutions by heart too. I noticed the same as you describe, I can solve the positions within seconds. The weird thing is that I remember the positions and the move sequences, which I haven't studied, but not what I have actually studied: the 18 tactical elements. I'm going to take this rote learning one step further, I am going to memorize the tactical elements of which the combinations exist too. after all, this is what I said to you:

      Your memories should serve as cues for fast retrieval of the 18 patterns. That are the kind of memories you need. Let's see if that works.

    2. IF you compare the problem referenced above (Link: Redeployment Maneuver) with the forced maneuver used by Capablanca against Yates, New York 1924 (referenced above), you will find that memorization of one position will provide virtually no clues about how to proceed in the other. Dr. Lasker uses the Capablanca-Yates position to illustrate a combination, with the following comment: "White makes a combination by which he gains an important Pawn."

      As Dr. Lasker put it (and which I have quoted before):

      "The motifs of a combination, in themselves simple, are often interwoven with each other. What is it that unites the multiplicity of motifs? We call it the "idea." Motifs, as, for instance, a simultaneous attack against several pieces or the encircling of the hostile King, are tricks of the trade, technicalities. The idea which links the motifs is artistic, it creates something that had never before been there. Motifs can be taught, ideas must be discovered by original effort. Ideas come from nowhere, they are sudden inspirations; the place of motifs is definite: the memory."

      It seems obvious that if you have memorized a specific position, you should be able to recall the "solution" instantaneously. However, being able to recall a specific position will NOT enable you to quickly solve a different position, especially one in which the "idea" is similar but the configuration of pieces, the motifs available and the precise sequence of moves are all totally different (i.e., as in the two examples cited).

      As I was reading Temposchlucker's description and the comments, trying to find the appropriate English term ("redeployment maneuver"), without intentionally searching my memory I recalled the Capablanca-Yates maneuver, and the similarity of the underlying ideas. IMHO, THAT TYPE OF RECALL is the desiderata, not recall of the accidental specifics of either position. It is the idea of redeploying a piece to a more useful position utilizing available forcing moves, not that the end result wins a piece or a Pawn. That idea is "plastic" and can be applied as needed. As Temposchlucker rightly observed, memories should serve as CUES for the appropriate retrieval of the 18 patterns. In order for that to work appropriately, the memories cannot be exact in any relevant sense (i.e., the exact position of the pieces, the exact order of moves, exactly the same motifs, etc.). That is where the effort resides, in making analogies at the appropriate conceptual level and thereby sparking the unique "idea(s)" that is applicable to that specific position.

      Preaching to the choir. . .

    3. Let me chime in...

      In the metaphor of recognition of patterns in a cloud, we don't have a clear picture of the mammals in our database. If you have a clear image of the archetypical rabbit in your database, you can recognize it everywhere. Be it in a cloud, an ink spot, a tree bark, an oil smear.

      Evidently we don't have a clear image of the 18 most important archetypical tactical themes. We have difficulty to see the difference between a rabbit and a rodent, let alone the difference between a rabbit and a hare.

      That implies, we don't learn anything (at least not enough by far) from the way we treat puzzles. We get mesmerized by the geometrics of the position, but we don't see in an archetypical way. We see pieces and squares, but not the functions of the pieces and the squares. We must learn to see the ideas hidden in the position. We must build our archetypical tactical themes database. That shouldn't be too difficult, with only 18 important themes...

  7. Could you present the archetypical images of these "only 18 important themes..." I am wondering what are they - if they are just the abstract concepts related to the group of positions (like pins) or do you mean some type of patterns or anything else?