Wednesday, January 02, 2019

BADPDO. The arise of the patterns.

The new training regimen yields new insights. New insights are accompanied by new geometrical patterns. What am I talking about?

When you know that you have a double attack at hand, you look for LPDO, Loose Pieces on the verge to Drop Off. A new insight is that you should look for BADPDO too. B.A.D. Pieces Drop Off.
These are (at least) two common patterns that accompany that insight:

Loading a battery from behind

Loading a battery from upfront

You will find these type of patterns  time and again.

Another insight concerns the tree of scenarios that emerges from a double attack. Attacking two targets at the same time, makes from those targets two potential desperado's. It is necessary to check if those desperado's can refute your attack.

Just to give you an idea.


12 comments:

  1. For almost every tactical motiv there are a set of typical refutations, pin, unpin, double attack, undouble attack asf

    ReplyDelete
    Replies
    1. Exactly. And almost none of them are automated!

      For the native English speakers: motiv should read as motif.

      Delete
    2. I did some statistics gathering on Chess tempo's problem set this evening. I was somewhat surprised at what I found.

      (I switched to Courier New font for the following table to avoid the zigzag associated with a proportional font. Hopefully the table will align and make sense. If it doesn't, copy it out to a Word document and use Courier New font.)

      Tactical Problem Average Average Percentage
      Theme Count Rating Moves of Total
      Exposed King 25548 1488 2.7 12.6%
      Fork/Double Attack 14333 1558 2.8 7.1%
      Defensive Move 13484 1625 2.2 6.6%
      Sacrifice 11066 1573 3.1 5.4%
      Coercion 10818 1556 3.2 5.3%
      Hanging Piece 10315 1427 2.0 5.1%
      Pin 9619 1639 2.7 4.7%
      Mate Threat 9523 1757 2.9 4.7%
      Distraction 9216 1630 2.9 4.5%
      Advanced Pawn 7546 1629 2.9 3.7%
      Zwischenzug 7102 1766 3.1 3.5%
      Discovered Attack 7029 1616 2.8 3.5%
      Clearance 6407 1639 3.2 3.2%
      Attraction 6080 1629 3.3 3.0%
      Counting 5509 1517 2.6 2.7%
      Unsound Sacrifice 4544 1618 2.3 2.2%
      Capturing Defender 4094 1659 3.0 2.0%

      (I did every category of tactical puzzle, but only listed the top 20 categories above.)

      That was NOT the breakdown I was expecting for the top 20 largest categories of puzzles. Pins and forks are considered to be the most common tactical themes. I also suspect there is some skewing associated with people resigning prior to the "visibility" of certain standard mates during the games. A lot of the stock mate patterns were below 1% of the total (with several at approximately 0% [rounded to 1 decimal place]), in other words, virtually "noise."

      There are 3 tactical themes per puzzle, on average. The average number of moves ranges from 2.0 to 3.9.

      Delete
    3. Oops! Not only did the table skew, but I left off 3 rows. Blame on looking at the row number in Excel, forgetting that I had offset the rows downward. Sorry about that. Here are the remaining 3 rows (skewed, of course):

      Overloading 3815 1662 2.8 1.9%
      Quiet Move 3360 1847 3.3 1.7%
      Skewer 3034 1600 3.0 1.5%

      Delete
    4. @Robert, very useful! I will focus one the first 9 themes. Those make up 56% op the problems.

      Delete
    5. Great stats Robert! :). Thanks a lot! Now I can see what motifs are really pretty repetitive :).

      Delete
    6. https://chesstempo.com/chess-tactics/103618
      1. Qd7 would be a strong double attack un the Bf7 and Rc8
      but there are methods to un-double attack
      Rf8 would remove one target and defend the other but there is Qe8 to , a "double defense" ;)

      Delete
    7. An insight I just gained from Aox's problem:

      In the encircling motif, there are two components: (1) an immobile target and (2) superior force on the target square.(LPDO is a special case of this motif, in which there are initially NO defender(s) of the loose piece.)

      The general approach for handling the encircling motif is to eventually establish a favorable attacking/defending ratio on the target square. This can be accomplished by either (1) adding attackers or (2) removing defenders, either through capture(s) of the defender(s) or by diversion of a defender or by overloading one or more of the defender(s) with Functions.

      The insight concerns this situation, where there are an adequate number of defenders available (in this case, both the Black Queen and Rook). By exchanging on the target square, the Black King now becomes the pivot for a fork. The Black Rook remains LPDO, but the dynamics have changed. There is no longer anything to protect by the Rook move, and it's White turn to move, so it is eliminated. Now a check on d7 forks the King and the Rook, and there is no longer a defense by the Black Queen: it cannot block the check AND simultaneously protect the Rook.

      Another way of looking at the situation is to (mis)apply Nimzovich's 'Zerstorung der Blockadeur'! (destruction of the blockade). In this case, the Black Queen can "block" the planned attack against f7. By removing the Bishop, a more amenable target (the Black King) must replace the Bishop on the target square. In this case, the blockade refers to the desired tactical theme (the fork), not to a static target such as a blockaded Pawn.

      Delete
    8. Another insight from this puzzle (103618): observing the intimate connection between strategy and tactics.

      The stem game for the puzzle (also found on chessgames.com with no annotations) is:

      Amin, Bassem (2650) vs Hesham, Abdelrahman (2375)
      Date: 2013-07-11
      Event: Universiade Men, Kazan RUS
      Round: 4.19
      Result: 1-0
      Opening: Sicilian Defense, Modern Variations (B50)
      Problems: 103618

      1. e4 c5 2. Nf3 d6 3. g3 g6 4. Bg2 Nc6 5. O-O Bg7 6. c3 e5 7. d3 Nf6 8. a3 O-O 9. b4 Nd7 10. Nbd2 b5 11. a4 cxb4 12. axb5 bxc3 13. bxc6 cxd2 14. Bxd2 Nc5 15. Ba5 Qe8 16. c7 Be6 17. d4 exd4 18. e5 dxe5 19. Nd2 e4 20. Bb4 Qb5 21. Qb1 Qb6 22. Bxc5 Qxc5 23. Rc1 Qe7 24. Qxe4 Rae8 25. Rxa7 Qb4 26. Qd3 Bc8 27. Ne4 f5 28. Ng5 Bh6 29. h4 Bxg5 30. hxg5 Re1+ 31. Rxe1 Qxe1+ 32. Bf1 f4 33. Qc4+ Be6 34. c8=Q Rxc8 35. Qxd4 Bf7 36. Rxf7 Kxf7 37. Qd7+ Qe7 38. Qxc8 Qxg5 39. Bc4+ Ke7 40. Qc7+ Ke8 41. Qxf4

      Problem 103618 occurs after Black's 35. ... Bf7 (the "0th" move of the problem).

      As I was contemplating that ramifications of this problem, and playing over the game score, I was struck by the connection to Nimzovich's My System, Chapter 3, The Seventh and Eighth Ranks. One of the "elements" of STRATEGY is an invasion of the 7th or 8th rank with a Rook. Lots of good things usually ensue from accomplishing this goal. After 9. b4, the stats are highly in White's favor (White win: 50%, Draw: 16.7%, Black win: 33.3%) but there are only 18 games in the Chess Tempo database. White gets an advanced passed Pawn on the 7th rank fairly early in the game at the cost of a Pawn, then promotes it, setting up the problem position with what Nimzovich called "the 7th rank absolute". It is not surprising that Black succumbs to a tactical assault very shortly thereafter. Black had some (insufficient) compensation for the lost piece, which White promptly snuffed out.

      So, the strategy of playing to eventually control the 7th rank with a Rook paid tactical dividends.

      "In general I consider that in chess EVERYTHING RESTS ON TACTICS. If one thinks of strategy as a block of marble, then tactics are the chisel with which a master operates, in creating works of chess art." - Tigran Petrosian

      Delete
  2. If I remember well, Richard of Chess Tempo has somewhere described how he harvested his problem database. Does anybody know where?

    ReplyDelete
    Replies
    1. Found it!

      Link: Chess Tempo FAQ

      Topic: How are tactical problems generated?

      The Toga II chess engine is used to evaluate positions from real games and a purpose built application examines the computer analysis and directs the engine in order to discover tactical positions and their resulting moves. Finding the positions and moves takes a great deal of computing time, therefore the tactics finder is designed to run on multiple machines with a central controller aggregating the results. Some attempt is made to determine ambiguous positions and these are discarded. The tactics finder application is not perfect and "bad" or incorrect problems can slip through, when found these problems are used as test examples to help improve the quality of the problem generator.

      Delete
    2. If you do an advanced problem search, you can search for the minimum rating of the players.

      For instance:
      total problems: 67,044
      rating above 2200: 47,370

      Delete