Transfer by analogy
The main problem is the transfer of knowledge gained from one position to another. Let'see if we can get that to work.
 |
| Diagram 1. White to move |
[solution]
[FEN "2Rr2k1/1p1Prp1p/pq3Q2/4n3/P7/6PP/6B1/5R1K b - - 0 0"]
The move 1.Rxf6 retaking the queen immediately doesn't accomplish anything. That raises the question, can the retake be postponed, and is there something that can be achieved by that?
 |
| Diagram 2. White to move |
The move 1.Rxd8+ fetches the rook and has enough additional punch (the check) to maintain the postponement of the retake of the queen. It is a bit deceptive though, since black has the double whammy 1. ... Kg7. Which serves two goals:
- skedaddling the king out of harms way
- protecting the queen
 |
| Diagram 3. White to move |
2.Rg8+ maintains the postponement and clears d8 for promotion. At the cost of giving the rook back. Since 3.d8Q+ promotes with check, the postponement of Rxf6 is still maintained.
What can be learned from all this? Which knowledge might be transferable to similar positions?
I reckon the following:
- When a direct take back doesn't get you anywhere, it's worth looking for postponement
- An intermediate move must contain enough punch to maintain the postponement. In this case, only a check is powerful enough
- In stead of trial and error, it is worthwhile to look at what you can possible gain with the postponement. The conquering of Rd8 doesn't work, but the promotion does
- Postponement moves can be treated move by move. Each move is a stepping stone. Is the pressure still maintained? Then you can look at the next move.
This is a perfect illustration of why a "pieces on squares" (POS) approach to learning "patterns" combined with a formal "step-by-step" logical thinking process do not significantly improve our chess SKILL. The thinking SKILL which must be gained is independent of the specific locations of pieces in specific positions; in short, it is only through the process of abstraction and analogy that we gain SKILL. Memorization of this position may (or, more likely, may NOT) embed something useful for future problems or games. However, the process of thinking through the ramifications and implications WILL embed something very useful into long-term memory, which can be recognized instantly by analogy using System 1. This SKILL is compact, easily triggered, and is quite useful in many varied circumstances. The process you demonstrated to arrive at the final four points is the process we must learn to apply consistently in every position. Just trying to memorize the four summary points in isolation will NOT improve SKILL. Context is required, and only analogical thinking can provide context and thus put the appropriate "trigger" for System 1 into long-term memory.
ReplyDeleteA sense of satisfaction and release of tension makes it much more difficult to force ourselves to FOCUS and dig deeper into a position when we do not practice "seeing" beneath the SURFACE until we find the ESSENCE of the position. The surface ALWAYS provides sufficient "clues" to find the essence in every position. We fail only because we are just too "lazy" to dig out the essence; we console ourselves that it is just "too hard" for our present level of skill. It is NOT too hard! It merely takes TIME and a stubborn unwillingness to give up just because it's hard.
Look at this problem from the usual chess amateur perspective: White has a pin on BRd8 AND can capture the Rook with check. (I presume that Black just captured the White Queen on f6; otherwise, the position makes no sense.) If we followed a logical step-by-step process (ala Heisman), we KNOW" that we should always look FIRST at the last move played by the opponent. OH NO! We've just lost our Queen! Can we immediately checkmate? NO. Okay, then we had better re-establish material equality and recapture his Queen (if possible). So, first thought - 1. Rxf6. WHEW! I thought I had lost my Queen for a second. WE USUALLY S-T-O-P LOOKING for a "better move" at this point, make the move, hit the clock and lean back, satisfied that we have "seen" what is required by the position. (I can pat myself on the back for my great move.) NEXT PROBLEM!
And nothing useful gets added to our repertoire of chess SKILLS.
The key is to learn something that can be used in FUTURE positions from every problem and every game that we play. (That's why it is so important to analyze our own games. Too bad we don't apply that same logic to analyzing problems.) Acquisition of SKILL can only occur through regular exercise of the process of abstraction and analogy. The harder it is to extract the essence, the greater the likelihood that we learn something very useful from the effort, which WILL get stored in long-term memory, providing the needed "trigger" to System 1.
PoPs, LoAs, and Funs are a way of concentrating focus on the surface features which provide "clues." Forcing moves (CCT, captures) and tactical devices/themes) are also surface features. The essence is ALWAYS buried in the interrelationships, which are NOT necessarily specific contacts between specific pieces. What is going on in one part of the board may seem unconnected to what is happening on another part of the board, but it is the whole rather than the individual parts that is important, that must be "seen" as the essence.
Every move is the most forcing move.. a no brainer
ReplyDeleteThe essence is ALWAYS buried in the interrelationships, which are NOT necessarily specific contacts between specific pieces.
ReplyDeleteI always wondered why my usual logical reasoning was so useless for chess. This new area of analogy and concepts might be a realm where it CAN be applied. It might even spark a new interest in how tactics actually work.
Which brings me to the Pomodoro technique. In an ideal world, you are so interested in what you study, that you don't need techniques to focus your attention. I will give it a try, though, just to find out if my focus comes anywhere near what I think it is.
Much to my surprise, my usual average focus time exceeds the Pomodoro timer setting by far. So I don't need one. Interest is a better motivator than a timer, apparently.
DeleteWhat might help is that my social life on social media is sub standard. I have no friends on Facebook, most forums have nothing to add but repeat the same tile wisdom over and over again.
I never call someone by archaic telephone so nobody calls me. I don't like watching television. We have no children.
I only get lots of emails from all over the world. But those can simply wait.
FWIW:
ReplyDeleteGM Valeri Beim postulates that there are two basic methods for "seeing" what is required - (1) System 1 pattern recognition, and (2) System 2 logical analysis. In some cases, it may require application of BOTH methods. If (1) is not happening, then we are stuck with (2). They are NOT mutually exclusive, but a lot of players tend more toward one method than the other.
Gm Beim's book How to Calculate Chess Tactics - A revealing look at the nuts and bolts of chess thought has a fairly detailed description of the method of logical analysis. Another of his books The Enigma of Chess Intuition - Can You Mobilize Hidden Forces in Your Chess? contains a general method for logically examining a position PRIOR TO beginning calculations.
While training, the relevant question is:
What is it about the position that triggered the player (who may be YOU) to "see" the combination?
That "triggering information" is what must be ferreted out via logical analysis DURING TRAINING. Training focuses on the process of thought (NOT a mechanical step-by-step process). While playing, one takes into consideration the practicality associated with the available time for contemplation.
In my comments, I have made several references to those books and the insights available, so I won't repeat any of that stuff now.
PART I:
ReplyDeleteAnother example of how to look for postponement.
FEN: 1k2r2r/1b2n1p1/pp2p3/4q2p/7B/1N1B1Q2/PP4PP/2R3K1 w - - 0 23
White to move.
Black has a material advantage.
The first thing that grabbed my attention was that Black has just "discovered" an attack on the White Queen. So, either White will have to move his Queen out of the attack, or will have to counter with an EQUAL OR GREATER THREAT Since there is no direct check, that only leaves an attack on the Black Queen. Lo and behold! White can PIN the Black Queen to the Black King with 1. Bg3!
BUT, but, what about the attack on the White Queen?! Here it is crucial to observe the difference between the two attacks. Black has NO defenders of his Queen AND it is ABSOLUTELY pinned against the Black King. White's Queen, on the other hand, is attacked AND defended (by WPg2), but the White King is not involved. Ergo, White has time for a postponement, capturing second with CHECK, which defers Black from escaping with his Bishop. Black loses the Bishop, giving White an advantage - two very active Bishops in an open position versus a Rook (I prefer to think of the inactive BRh8 as that Rook) and Pawn. Given that the two Bishops are usually given an extra half-point count, White has both a positional and material advantage. GM Stockfish values it at 5.71 after thinking about it for nearly an hour.
1. Bg3 Bxf3
2. Bxe5+ Ka7
3. gxf3 wins
PART II:
ReplyDeleteAs can be seen from the game score below, just because you have an advantage at one point does NOT mean you have a "won" game: YOU STILL HAVE TO WIN IT! White lost on time!
Game:
[Event "Rated Rapid game"]
[Site "https://lichess.org/RVg9JtHb"]
[Date "2021.01.02"]
[White "bichopebitou"]
[Black "Humangazor"]
[Result "0-1"]
[UTCDate "2021.01.02"]
[UTCTime "14:10:48"]
[WhiteElo "1972"]
[BlackElo "1867"]
[WhiteRatingDiff "-7"]
[BlackRatingDiff "+25"]
[Variant "Standard"]
[TimeControl "600+0"]
[ECO "D00"]
[Opening "Queen's Pawn Game: Mason Variation"]
[Termination "Time forfeit"]
[Annotator "lichess.org"]
1. d4 d5
2. Bf4 { D00 Queen's Pawn Game: Mason Variation } Nc6?! { (0.00 → 0.63)
Inaccuracy. c5 was best. } (2... c5 3. e3 cxd4 4. exd4 Nc6 5. Nf3 Bf5 6. c3 e6 7. Nbd2)
3. e3 e6
4. Bd3?! { (0.56 → 0.00)
Inaccuracy. c4 was best. } (4. c4 Nf6 5. Nc3 Bd6 6. Bxd6 Qxd6 7. Nf3 O-O 8. a3 a6)
4... Bd6
5. Bg3 Nge7?! { (-0.16 → 0.67)
Inaccuracy. e5 was best. } (5... e5 6. Ne2 Nf6 7. c3 e4 8. Bc2 O-O 9. O-O Bg4 10. Bxd6)
6. c3?! { (0.67 → -0.25)
Inaccuracy. Nf3 was best. } (6. Nf3 Nf5 7. c4 dxc4 8. Bxc4 O-O 9. Nc3 b6 10. a3 Bb7 11. e4 Bxg3 12. exf5 Na5)
6... a6?! (-0.25 → 0.65)
(Inaccuracy. e5 was best. } (6... e5 7. Nd2 e4 8. Bb5 O-O 9. Ne2 Nf5 10. Bxd6 Qxd6 11. O-O)
7. Nd2 h5?! (0.28 → 1.27)
{Inaccuracy. Bxg3 was best. } (7... Bxg3 8. hxg3 e5 9. Be2 O-O 10. dxe5 Nxe5 11. Ngf3 N7c6 12. Nb3 Ng4 13. Rc1 Nf6 14. c4)
8. Bh4?? (1.27 → -0.27)
{Blunder. Ngf3 was best.} (8. Ngf3 Nf5)
8... Qd7? (-0.27 → 0.95)
{Mistake. e5 was best.} (8... e5)
9. f4 Na7?! (0.63 → 1.56)
(Inaccuracy. Nf5 was best.} (9... Nf5)
10. Ngf3 b6? (1.89 → 3.25)
(Mistake. c5 was best. } (10... c5 11. Ne5 Qc7 12. O-O Nac6 13. e4 cxd4 14. cxd4 Nxd4 15. Rc1 Ndc6 16. Kh1 f6 17. exd5)
11. Ne5 Bxe5
12. fxe5 c5
13. O-O Bb7
14. Qf3 O-O-O
15. Qxf7 Nac6
16. e4?! (4.73 → 3.34)
(Inaccuracy. Qxg7 was best. } (16. Qxg7 Rdg8)
16... cxd4?! (3.34 → 4.49)
(Inaccuracy. g5 was best. } (16... g5)
17. cxd4?? (4.49 → 1.37)
(Blunder. exd5 was best. } (17. exd5)
17... Rde8?! (1.37 → 2.01)
(Inaccuracy. dxe4 was best.} (17... dxe4 18. Bxe4 Qxd4+ 19. Kh1 Qxd2 20. Bxe7 Nxe7 21. Rac1+ Kb8 22. Bxb7 Kxb7 23. Qxe7+ Qd7 24. Rf7)
18. exd5 Qxd5
19. Rac1?? (1.59 → -2.46)
(Blunder. Qf3 was best. } (19. Qf3 Kb8)
19... Kb8?? (-2.46 → 1.65)
(Blunder. Qxd4+ was best. } (19... Qxd4+)
20. Nb3?? (1.65 → -7.17) { Blunder. Be4 was best. } (20. Be4 Qxd4+)
20... Nxd4?? (-7.17 → -1.87)
(Blunder. Nxe5 was best. } (20... Nxe5)
21. Rf3? (-1.87 → -3.54)
(Mistake. Rf2 was best.} (21. Rf2 Rhf8)
21... Nxf3+
22. Qxf3 Qxe5?? (-3.39 → 4.65)
(Blunder. Qxf3 was best.} (22... Qxf3)
23. Bg3 Bxf3
24. Bxe5+ Ka8
25. gxf3 Nd5
26. Be4 Rhg8
27. Rc7 Rc8
28. Re7 g5
29. Rxe6 Rgd8
30. Rxb6 Ka7
31. Bd4?? (8.15 → -2.32) (Blunder. Rh6 was best.} (31. Rh6)
31... Nxb6
32. Bxb6+?? (-2.76 → -7.01) (Blunder. Be3 was best.} (32. Be3 Rd1+)
32... Kxb6
33. Kf2 Rf8
34. Ke3 g4
35. fxg4 hxg4
36. Nd4 Rc1
37. Ne2 Re1
38. Kd2 Rh1?? (-6.49 → 0.12)
(Blunder. Ra1 was best.} (38... Ra1 39. Nc3 Ka7 40. Ke3 Rb8 41. Bf5 Re1+ 42. Kf2 Re5 43. Nb5+ Kb6 44. Na3 Rxf5+ 45. Kg3)
39. Bxh1 Rf2
40. Ke3 Rxh2
{Black wins on time.} 0-1
PART I:
ReplyDeleteMore on the idea of "postponement" - or actually, a somewhat different view of the process. In some ways, this reminds me of the view held by Nimzowitsch about the opposition; he declared it to be obsolete in light of his theory of the blockade. Either viewpoint will enable the student to figure out what to play in pure King-and-pawn endgames.
Ref.: Combinations in the Middle Game, Igor G. Bondarevsky
(GM Bondarevsky was the trainer for Boris Spassky from 1961 onward.)
As Temposchlucker noted regarding postponement (leaving out specifics regarding the example position):
An intermediate move must contain enough punch to maintain the postponement.
It is worthwhile to look at what you can possible gain with the postponement.
Postponement moves can be treated move by move. Each move is a stepping stone.
Bondarevsky looks at possible postponement through the concept of "compulsion". That concept is familiar: EQUAL OR STRONGER THREAT, coupled with a hierarchy of (1) check; (2) mate threat; (3) capture of material; and (4) threats (in general that are at a lower level of compulsion than the first three levels listed). The COMPULSION hierarchy is usually denoted by CCT.
The active side all the time compels definite replies
First and foremost every checking move is the highest degree of compulsion. Then the number of possible replies can be sharply restricted by a threat of mate, though it is not as forcing as a check. In fact when one is threatened with mate one can still undertake any advantageous operation as long as it is accompanied by checks. Hence when playing a combination one must NOT reckon that a threat of mate will AUTOMATICALLY force the opponent to defend against it.
Example 1 (Diagram 9)
FEN: 3r1bk1/2r2pp1/p3n1P1/1p6/4P3/qPpB4/P1P2P2/1KN1Q1RR w - - 0 1
Black has an unstoppable mate threat - except it is White to move.
PART II:
ReplyDeleteThe defender's replies are considerably reduced by an attack on various pieces or pawns or by the threat of such an attack. Moreover it is understandable that an attack on the queen forces its defense or the creation of a counter-threat GREATER IN STRENGTH OR EQUAL TO an attack on the queen, such as a check or a threat of mate. Of course a THREAT OF EQUAL STRENGTH would be a counter-attack on the opponent's queen.
Example 2 (Diagram 10)
FEN: 2br2k1/pp3ppp/2p2n2/3r1q2/8/1PP3BP/PQ3PPN/R3R1K1 w - - 0 1
By playing 1.Nf1, White threatens the fork 2.Ne3. [GM Stockfish prefers either 1.c4 or 1.Nf3.] Black must defend against this by moving his queen or his rook on d5 otherwise he would lose the exchange.
But we have not exhausted the various means of compulsion. The CAPTURE of any piece or pawn for the purpose of exchanging also compels the opponent either to make the one move that enables him to re-establish material equality or to choose for the same purpose one of a restricted number of moves.
Example 3 (Diagram 11)
FEN: r2q1rk1/pppbbppp/2np1n2/1B2p3/3PP3/2N2N2/PPP2PPP/R1BQR1K1 w - - 0 1
White now wins by a forced manoeuvre which uses exchanges as a means of compulsion.
Finally there is one more means of compulsion which is linked with the very essence of combination, namely the sacrifice. The fact that a sacrifice forces definite replies is clear without examples, as we have already seen cases of this.
Hence there is huge arsenal of the means of compulsion. The important point is merely to choose the MOST EFFECTIVE MEANS in each concrete position without over-estimating the available degree of compulsion.
PART III:
ReplyDeleteExample 4 (Diagram 12)
FEN: 4r2k/pp2bpp1/2p5/2n5/8/P2r2NP/1PRR1PP1/5N1K w - - 0 1
Black here supposed that White could NOT make a combination with the idea of destroying the defense of the BRd3 by 1.Rxc5 and convinced himself of this by the following line of reasoning: "If rook takes my knight [1.Rxc5] I obviously don't recapture with the bishop [1...Bxc5] as then my rook is unguarded and en prise to his rook. Instead I take his rook on d2 [1...Rxd2] and in order not to lose material he will have to play 2.Nxd2 [ASSUMED COMPULSION] when I shall play 2...Bxc5 and so finish up with the exchange to the good. [WHAT IS THE FATAL FLAW?] All this line of reasoning is INCORRECT as Black did not consider ALL White's possible replies after 1.Rxc5 Rxd2 from the point of view of the GREATER DEGREE OF COMPULSION. Considering the position carefully we see that White can play a move of the HIGHEST degree of compulsion - 2.R-h5+ - getting his rook away from the attack of the bishop. Black must defend against the check and instead of Black winning the exchange White wins a piece. Moves like 2.Rh5+ are called INTERMEDIATE [Zwischenzug] checks; one must always take account of them in calculating combinations.
Of course it is not just an intermediate move that can cause a breakdown in our calculation-any reply which has a greater degree of compulsion is by its nature an intermediate move. Obviously we must pay attention [FOCUS!!] to all such moves. In connection with the fact that the moves of a combination have different degrees of compulsion we have the case of the "desperado" piece.
Example 5 (Diagram 13)
FEN: r4nk1/1pp1rppp/1bn3b1/1q6/p3N2Q/P1R2BNP/1P3PP1/2B2RK1 w - - 0 1
[THIS IS A PERFECT EXAMPLE OF POSTPONEMENT!]
1.Rxc6
White plays a combination based on destroying the defense of the BRe7, as if 1...bxc6 2.Qxe7. However after the capture of the knight Black is not COMPELLED to take the rook and can look for a move of EQUAL OR GREATER DEGREE OF COMPULSION which MAY allow him to guard his rook indirectly.
[The analysis of this combination is very instructional-and left as an exercise.]
We have already said that an attack on a piece or pawn is a means of compulsion. It is very useful to dwell in greater detail on the various sorts of attacks which are produced by different moves. One must note that THE MOST VALUABLE MOVES FROM THE POINT OF VIEW OF COMPULSION (as a rule, of course, but NOT always) ARE MOVES WHICH ARE LINKED WITH ATTACKS ON TWO OBJECTS INSTEAD OF ONE [double attacks]. In such cases the means of defence are more limited. In some positions there are no means of defence at all.