Monday, January 07, 2019

Learning the tactical ABC

Time and again, it is proven that I am not very good at the recognition of basic tactical elements. The choice of my problem sets reveals that clearly. I do 10 low rated problems well, and then I encounter a low rated problem that takes me ages.

That is a clear sign that I don't master the basics as well as I like to think. Take for instance the following position:

White to move






8/5k2/5ppp/4r3/4nRP1/4N2P/5PK1/8 w - - 0 1
[solution]

The first time it took me 1:41 to solve. Average solving time is 0:54. Rating is 1508.
The combination consists of three basic elements:
  • a tempo move (Nc4)
  • a target exchange (Rxe4)
  • a duplo attack (Nd6+)


19 comments:

  1. well its the average time, average means you are 10 times quicker and 10 times slower. Some of the low rated problems are in real more complicated, better player see more problems and calculate longer. And with more pattern you know as more you have to check.
    One way do see this puzzle is: Ne4 is weak and there is a night fork at d6 reachable in 2 moves, both are HE's of Chuzhakin

    ReplyDelete
  2. The funny thing is that the average solving time of 1500 to 1900 rated problems is roughly the same (30 sec).

    ReplyDelete
  3. easy a higher rated player is quicker ( at the same puzzle ) and there are more pieces at the higher rated puzzle on the board. in this case its compensating exactly

    ReplyDelete
  4. But you are slower at the higher rated piuzzle, everything is ok ;)

    ReplyDelete
  5. PART I:

    The following is kind of a "stream of consciousness" that came to me as I looked at the first problem. It seems sequentially "logical" because that's implied by writing it down as a narrative. A lot of the thoughts became conscious simultaneously, so it's not a literal step-by-step "thought process." Perhaps there is some nugget in there that might help someone else to "look" at a puzzle differently, and therefore improve the tactical vision. Or, maybe not; it seems that so much of what we do when looking at tactics is idiosyncratic and unique to each of us based on our prior experience. Please do NOT get the impression that I'm saying this problem is "easy" nor that anyone else should "go and do likewise."

    First problem:

    (1) White to move.

    (2) Material is balanced.

    (3) There is no potential mating attack and no (immediate) Pawn promotion for either side. Therefore, the puzzle is about gain of material.

    There is only one "target": the BNe4 is B.A.D. (attacking/defending ratio = 1/1). The White Knight is a potential additional attacker on e4, but it will take 2 moves to get it into a direct attacking position. Since the target is a Knight, that target Knight can simply capture the additional Knight attacker whenever it presents a direct attack on e4.

    There is an idea that has already been discussed here. When using the "encircling motif", "tit for tat" capturing on a square removes an attacker AND a defender. There is a pin on f6, but only the White Knight can attack that Rook. What's the most forcing move (CCT)? Try 1. Nc4. Switching sides, Black cannot abandon the BNe4 by counterattacking the White Knight with 1. ... Rc5, because the White Rook will protect WNc4 after 2. Rxe4; the Black Rook must remain on the e-file. At this point, the potential Knight fork on d6 becomes "visible" as a first order THREAT (there is not yet a direct attack on e4) of a Knight fork against the f7 and e4 squares, combined with a forcing threat to the Black Rook. The Black Rook must move to avoid the direct attack from the Knight. Let's pick an arbitrary "safe" square: 1. ... Re6. Now that the White Knight is now in place for a fork, removing the "defender" of e4 is simple: capture on e4 with 2. Rxe4, forcing Black to recapture with 2. ... Rxe4, thereby losing control of e4. 3. Nd6+ forks the Black King and Rook, winning a piece.

    ReplyDelete
    Replies
    1. "it seems that so much of what we do when looking at tactics is idiosyncratic and unique to each of us based on our prior experience. "

      That is exactly what I was trying to say. The trouble is to find out what is exactly your personal list of failure. Only then you can try to give yourself adequate feedback.

      Delete
  6. PART II:

    Second problem:

    Before clicking on the link, the "solution" was already primed by Tempo's "hint": same principle.

    (1) White to move.

    (2) Material is balanced.

    (3) There is no potential mating attack and no (immediate) Pawn promotion for either side. Therefore, the puzzle is about gain of material.

    At first glance, the immediate "target" is the B.A.D. Black Knight on d5. The Black Rook on a8 is LPDO. This time there is the possibility of "ganging up" on the target square: the White Queen and the White Knight on b1 can be thrown at d5. However, Black has adequate defensive resources (e7-e6 and Bc8-b7) to maintain the Knight on d5. So, that's not it (on preliminary assessment). What's the most forcing move (CCT)? Checks aren't available, and simply capturing with 1. Bxd5 Qxd5 does not gain material. Try 1. Nc6 threatening the Black Queen, with a potential check on e7. Black must maintain a defender of the BNd5 (and the WBf4 prevents the Queen from moving to d6), so 1. ... Qd7. But now we have the THREAT of a Knight fork on e7 if we can cajole the Black Queen to d5. We can FORCE the Black Queen to recapture on d5 to maintain material balance (or else simply lose a piece) AND simultaneously "protect" the saucy Knight on c6 with 2. Bxd5. Since Black does not want to lose his Queen to a potential fork, the BNd5 is lost. The only "question" is whether White can extricate the White Knight (or gain some other advantage from the two minor pieces in the middle of Black's game) or lose the White Bishop on d5. 2. ... e6 allows White to win a Rook with 3. Ne7+ Kh1 4. Bxa8. 2. ... Bb7 looks to be the most likely continuation. White can defend the WBd5 AND "threaten" f7 with 3. Qf3. He can always "bail out" by playing the White Knight to e7 with check and capturing on b7.

    I have no idea if this helps or not. However, I am convinced that "I [also] am not very good at the RECOGNITION of basic tactical elements." Or at the very least, not very good at stitching them together into a coherent tactical plan of action in any and all situations.

    ReplyDelete
    Replies
    1. CCT is calculation, to see the fork in 2 moves is patternrecognition or boardvision. But to be aware of the most forcing moves is always good ;)

      Delete
  7. I believe that the reduction of complexity is essential to get on the track of these personal failures:

    low rating
    max 2 moves
    themed

    In the specific cases above, my problem seems to be to see the duplo attack when one of the targets is not yet in place.

    ReplyDelete
    Replies
    1. Without the reduction of complexity I would never have noticed the similarities between the two positions.

      Delete
    2. reduction of complexity are the 5x5 minitactics of ct-art and endgame-tactics
      if you are looking for easy tactics at ct.. you can reduce the number of pieces too in the search for puzzles and you may look for the average time. Recognition is fast.
      So clear examples for pattern = : < moves, < time, < pieces

      Delete
  8. It says that the tactic is 1332 rated, and that most people solve it in 2:38. I solved it in 2:37 (so, average). It's hard to guess what there numbers mean. My live rating would be just over 1860 USCF, and I may not have found this tactic, had I not known that it was one. 1332 problem? Very strange. I would say that the 1332 might find the tactic, because that might be there primary skill when it comes to chess, simple tactics. It would also depend on whether or not the 1332 player is playing slowly enough to spot it. If this is a 1332 problem, then clearly one would benefit more by studying endgames, for example, if virtually anybody is solving this problem. Funny that 62% solves it. If the problem is really 1332, then you'd think that 80% would be solving it.

    ReplyDelete
  9. If someone is rated 1332, then he is not beating 80% of the players which are rated 1332. This problem is served people with an average rating slightly above 1332 so the score ist 62%

    ReplyDelete