Surprise!
I'm easy to surprise. And then again, I'm surprised that I am so easy to surprise by something that actually isn't surprising at all. It feels as is I'm rather naive, chesswise.
It reminds me of an old post about focal gamma bursts. Where brain scans of amateurs showed that the they see every position as new, while in grandmasters the Long Term Memory and the region of complex motor skills showed much activity.
 |
| White to move |
3r2k1/ppq2pp1/3Np1b1/1N1nP2p/3Q4/Pn3B1P/5PP1/4R1K1 w - - 0 1
[solution]
Treating it as a tit-for-tat problem, the first move was not difficult to find.
1.Qxd5
But I was surprised by the answer of black
1. ... Qa5
Yet this is completely logical. I continued against Stockfish.
2. Qd1 Bc2
And again I was completely surprised.
3.Qe2 Bd3
And again I was completely surprised.
What does this show?
Apparently my trial and error habit is very strong. I only focus on my own moves and am completely blind for what my opponent can do. Despite that my tree of scenarios prescribes to have a look at my opponent's position every now and then.
The good news is:
- It totally explains why I suck at chess
- It is not rocket science
- The fix is a matter of discipline and not of learning something new
- I finally am quite aware of the scope of the problem and the importance
- It totally explains the "trick" of child prodigies
Tempo sez:
ReplyDelete"All the tactical elements are not too difficult to SEE."
We don't SEE because we don't LOOK. We don't LOOK because we are so focused on our own schemes that we fail to remember that the opponent ALWAYS gets a vote - every other ply.
How hard is it to project ONE PLY into the future?!? It only involves mentally shifting ONE PIECE from one square to another. Surely that is well within our mental capabilities as adult chess improvers to SEE ONE changed piece position at a time!
Perhaps we should approach SEEing into the future (calculating ahead) by focusing on mentally moving just one piece, combined with focused attention on the actual board position as it is in order to ground what we SEE.
IT [the actual position] IS WHAT IT IS.
We can take advantage of that fact. We have no need to visualize more than one piece at a time as we mentally change the given position. We have no need to try to mentally toss ALL the pieces around randomly in our heads by trial and error, trying to hold all those changes in mind simultaneously. The hard thing to do is to totally switch perspectives from "our" side of the board to the "opponent's" side of the board AFTER EACH PLY. We MUST SEE the same kind of things [PoPLoAFun, squares, lines of attack, tactical themes/devices, local areas of tension, interconnections, etc.] from the opponent's viewpoint every other ply.
That switching of perspective seems [to ME] to be the "trick" of the child prodigies; they do it without thinking, as if it is the most natural thing in the world to do. They also are totally unaware that they are doing it, which is why (after they attain IM or GM status) they can't tell us adults what the "trick" really is!
As you stated:
"The fix is a matter of discipline and not of learning something new."
I must read this again, since I seem to be focused on my own schemes. You might be on to something.
DeleteGM Alexander Kotov, in his landmark book Think Like A Grandmaster suggested the following Blumenfeld Rule:
DeleteWhen you have finished analyzing all the variations and gone along all the branches of the tree of analysis, you must FIRST OF ALL write the move down on your score sheet using long-form algebraic notation, BEFORE you play it.
That advice is now against the FIDE rules regarding recording of moves:
Article 8.1.2 It is forbidden to record the moves in advance, unless the player is claiming a draw according to Article 9.2, or 9.3 or adjourning a game according to Guidelines I.1.1
What was the INTENT behind this suggestion?
Then when you LOOK at the board again, with your move written down but not yet played, you will be looking at it not with the eyes of one gazing into the distant future but with the eyes of someone who is present in the tournament room, of someone who recognizes the PRESENT REALITY and the worries of the present moment.
This is your first step on the road that brings you back to the present. Even now you must not hurry to make your move. SPEND ANOTHER MINUTE LOOKING AT THE POSITION [AS IT IS]—you won’t have cause to regret it—and LOOK at the position through the eyes of a patzer. Imagine you are not a grandmaster or master [NOT hard to do, since we ARE club players!], but a mere beginner. Am I threatened with mate in one? In two? Is my queen en prise, or my rook? Am I blundering a pawn away? Such an elementary check will almost certainly save you from an immediate oversight on the first move and is a sound supplement to the deep analysis you have just finished.
By following this Blumenfeld Rule you will combine depth of thought with practical accuracy and blunder-free play.
Going back to the REPL suggestion, I add this modification to the process:
1) RECOGNIZE one or more salient cues in the current position. (The “current position” may be the actual position or a mental projection into the future; that matters not.)
— Ask yourself FIRST: Was the immediately preceding move SAFE?
2) EXAMINE/EVALUATE the ramifications (lines of attack, interconnections, additional “resulting” salient cues, etc.) associated with the specific local area of tension defined by the salient cue.
3) PROJECT the ideas that you SEE outward from the local area of tension along a principal variation.
4) LOOP: LATHER, RINSE, REPEAT until there are NO more salient cues AND the investigation reaches quiescence in ALL variations.
I guess it is not so much the trial and error in itself that is the problem. Since I assume that that is how system 1 works. The problem is when it has no direction. When nobody is home to correct it.
ReplyDelete"Perhaps we should approach SEEing into the future (calculating ahead) by focusing on mentally moving just one piece, combined with focused attention on the actual board position as it is in order to ground what we SEE."
ReplyDeleteThere is a lot of merit in that remark. We must learn to see what he piece does. I see a move, I'm enthousiast about it because it solves the problem I see. But then I totally forget to look at the consequences. Even more so, I am surprised that there even are consequences. Which is actually rather strange, if you think about it.
To me surprises are rarely unexpected.. Inspector Clouseau
ReplyDeletePART I:
ReplyDeleteFEN: 3r2k1/ppq2pp1/3Np1b1/1N1nP2p/3Q4/Pn3B1P/5PP1/4R1K1 w - - 0 30
The game from which this puzzle was taken is given below. I set both GM Stockfish and GM Komodo to analyzing this position yesterday and let them run over night. Given the strength of these two SUPER GMs, I expected similar mainline continuations after the first White move—but that’s not what happened. I’ll defer that aspect until later.
Look (again) at the position as given (without the context of the actual game).
Since both Queens are under attack, the right to move first becomes critical. The White Queen can do virtually anything, grabbing whatever Black piece that can be captured, confident in the fact that the Black Queen is still under attack.
What violent attacking move springs immediately to mind? 30. Qxd5
According to my suggested REPL loop approach, we now should switch perspective and LOOK at the “new” position from Black’s perspective.
NOT SO FAST! Remember that GM Aagaard admonition to “SLOW DOWN!”? Instead of immediately switching perspective 1-ply in the future, we should FIRST consider ALL the “wonderful” [sarcasm] things that 30. Qxd5 actually does BEFORE we “explore” the variation(s) resulting from it, and NOT merely rely on the fact that we have grabbed a piece [hopefully] for nothing.
Capturing with the Queen and leaving it susceptible to immediate recapture is a very large commitment, predicated on the Black Queen NOT being able to escape attack with additional punch. Are there any other vulnerabilities in the White position? YES—the WRe1 is LPDO. Can the Black Queen attack that vulnerable LPDO? YES—30… Qa5 eliminates the White attack on the Black Queen AND counterattacks the LPDO with threat of check! Any move that accomplishes multiple things at once must be good! If we LOOK at the position AS IT IS, we should be able to SEE these considerations—thereby avoiding “surprise” before we even get to the first switch in perspective.
NM Dan Heisman wrote an excellent thought-provoking book The Improving Chess Thinker. Chapter 2: Thought Process Basics contains a thorough discussion of problems and solutions associated with thought processes. He gave an excellent summary of five ideas that should be contained in any simplified thought process (along with 12 potential problems and the results of those problems that could occur while trying to use that simplified process):
“Thus we can summarize good, simple chess: First, SEE if there is a tactic for EITHER SIDE; if so, address it; if not, maximize the activity of your pieces and minimize your opponent’s. You can play pretty well if you just follow that advice! A similar statement is Take your time [SLOW DOWN!] to do the best you can at keeping your pieces as safe and active as possible—and your opponent’s pieces the opposite.”
Problem #4 is:
“4. You don’t pay ENOUGH attention to your opponent’s moves. You concentrate mostly on what you are doing. Result: Your opponent often SURPRISES you and you can’t meet — or sometimes even SEE — his threats.”
Once we SEE that Black can counterattack a LPDO while also leaving our Queen hanging (two important things to consider before deciding on this plan), we should NOT be surprised by the move(s) following 30. Qxd5.
The essence is to take into consideration ALL of the surface-level clues (both positive AND negative for BOTH players) BEFORE beginning the “calculation process” (whatever that might be).
PART II:
ReplyDeleteGM Stockfish “thinks” the Principal Variation should be:
D42 +4.46 30.Qxd5 Qa5 31.Qd1 Nd2 32.Re3 Nc4 33.Qe1 Qxe1+ 34.Rxe1 Nb2 35.Be2 Rd7 36.Rc1 Bd3 37.Kf1 a6 38.Nd4 g5 39.Nb3 Kh7 40.Rc8 b6 41.g3 Kg6 42.h4 g4 43.Rc6 Ra7 44.Nc1 Bf5 45.Rc8 Rd7 46.Ke1 f6 47.Nxf5 Kxf5 48.Rc2 Na4 49.exf6 b5
GM Komodo “thinks” the Principal Variation should be:
D33 +4.13 1.Qxd5 exd5 2.Nxc7 Nc5 3.Nxd5 Nd3 4.Ne7+ Kh7 5.Re3 Nxe5 6.Rxe5 Rxd6 7.Bxb7 Rb6 8.Bc8 Rb1+ 9.Kh2 Rf1 10.f3 Rc1 11.Bb7 Kh6 12.Nxg6 Kxg6 13.Be4+ Kf6 14.Rxh5 g6 15.Rb5 Rc3 16.a4 Rc7 17.g4 Kg7 18.Ra5 Rd7 19.Kg3 Rc7
[Note: the reason for the difference in move numbers is because GM Stockfish is analyzing from the actual game score, and GM Komodo is analyzing from the problem position as if it is the first move.]
I think it somewhat humorous that White has a significant (winning!?) advantage after 30. Qxd5 [+4.00 at least] and yet still manages to LOSE the game in 11 more moves—snatching defeat from the jaws of victory.
It ain’t over ‘til the fat lady sings!
[No offense intended to the two lady players involved.]
I don’t see myself calculating 19 moves into the future—EVER—regardless of the process I use.
Kosintseva, Tatiana (2496) vs Gunina, Valentina (2509)
Date: 2013-12-15
Event: SportAccord Blitz Women 2013, Beijing CHN
Round: 11.2
Result: 0-1
Opening: Caro-Kann Defense, Advance Variation, Short Variation (B12)
Problems: 105530
1. e4 c6 2. d4 d5 3. e5 Bf5 4. Nf3 e6 5. Be2 Ne7 6. O-O c5 7. dxc5 Nec6 8. Be3 Nd7 9. c4 dxc4 10. Na3 Bxc5 11. Bxc5 Nxc5 12. Nxc4 O-O 13. Qc1 Qc7 14. Qe3 Nd7 15. h3 Bg6 16. Rac1 Bh5 17. Rfe1 Rfd8 18. b3 Nf8 19. Nd6 Qe7 20. Nc4 Nb4 21. a3 Nd5 22. Qe4 Bg6 23. Qg4 Nd7 24. Nd4 Rac8 25. Bf3 Nc5 26. Nd6 Rc7 27. N4b5 h5 28. Qd4 Nxb3 29. Rxc7 Qxc7 30. Qxd5 Qa5 31. Qd1 a6 32. Na7 Bc2 33. Qxc2 Qxe1+ 34. Kh2 Qxe5+ 35. g3 Nd4 36. Qe4 Nxf3+ 37. Qxf3 Qxd6 38. Qxb7 Qd7 39. Qb6 Qd4 40. Qxd4 Rxd4 41. Nc6 Ra4 White resigned
PART III:
ReplyDeleteHere are a couple of examples (by World Champions, no less) regarding “surprises” that occurred when a World Champion failed to SEE “obvious” surface-level clues. It didn’t end well for either of them.
1. Petrosian-Bronstein, Amsterdam 1956
FEN: 1rb2r1k/1p1n2q1/p2Q2p1/P2Npn1p/2P1N2P/6P1/1R3P2/1R6 w - - 0 36
Black's last move was 35... Nf5. The greatest defensive player in the history of chess carefully looked around the board at the surface-level clues at ply 0—and played 36. Ng5. What did Black play in response?
2. Christiansen-Karpov, Wijk aan Zee, 1993
FEN: r2qkb1r/p2p1ppp/1pb1p3/7n/2P1P3/P1N1B3/1PQ2PPP/R3KB1R b KQkq - 3 11
The World Champion ignored a surface-level clue (the LPDO BNh5) and failed to consider THAT when he made his next move. White had just played 11. Be3 and Black replied 11… Bd6. What did White play for his 12th move, ending the game immediately?
GM John Nunn in his book Secrets of Practical Chess: A Chess Olympic Gold Medal winner explains how you can dramatically improve your chess results, Chapter 1: At the Board, section Warning Signals had this to say about surprises:
“One of the most common reasons for blundering is failure to take warning signs into account. Tactics very rarely strike out of a blue sky. There is normally some underlying weakness, visible beforehand, which the tactic seeks to exploit. When you SEE such a potential vulnerability, it pays to look very carefully for a possible tactic by the opponent exploiting this weakness.”
“(V)irtually any tactical weakness can give the opponent a chance for a combination [presuming he can SEE it!]. Note that by ‘tactical weakness’ I do not mean isolated pawns, bad bishops and other such long-term weaknesses; these are ‘strategic weaknesses’. A tactical weakness is a short-term [maybe the very next move!] vulnerability such as an undefended piece, a potential pin or fork, or a trapped piece. Tactical weaknesses may exist only for a single move, so if your opponent creates one, you should look for a way to exploit it straight away.”
“Surprises” happen to all of us; just a lot more frequently with some of us!
It indicates a deeper lying problem. I don't usually see what the pieces are doing. Well I see what they are doing to some degree, but not all important consequences. I must reason to find the consequences. For that I must be triggered, otherwise I simply don't look for them. It is a very low level problem.
DeleteIt is like reading a text from Greek where you must look up every letter in a table. Some letters you recognize, others you don't. If you don't know a few letters of a word, you guess the word.
After you know the word, you try to figure out the meaning. What you tend to forget is to think about the consequences, since the process itself consumes too much resources.
The reason that my rating is not too shabby after all, is that I often find a way to repair the damage.
The good news is: there is ample room for improvement.
PART I:
ReplyDeleteI’ve mentioned the following game in previous comments, but did not include either the critical position nor the game score, which is below. The purpose of this ‘exercise’ is to SEE ALL the things going on through the various moves. Surprisingly (perhaps), it is NOT that complicated to figure out. We don’t have to be grandmasters or even masters to SEE everything that is important.
FEN: q5k1/3bbpp1/p3pn2/1p4N1/1P2N2Q/6P1/P3PP1P/2B3K1 b - - 0 24
In this position, White just recaptured a Rook on c1 with 24. Bxc1. I’m starting with this position merely because on the surface, it appears there is not much tactically going on—but there is.
Statically, Black has the two Bishops for a pawn, so material is fairly even. King safety is approximately even. The two Bishops are not well positioned and somewhat blocked by the kingside pawn structure. White’s pieces are aggressively placed in the Black kingside vicinity.
Dynamically, Black has the right to move first. There are two strategic considerations: (1) Improve the worst placed piece (the BQa8), and (2) try to utilize ‘threats’ to gain/maintain the initiative and set up a tactical shot.
Nothing says “centralization” like putting the tallest piece into the center of the board. Perhaps 24… Qd5 grabs the attention first. BEFORE MOVING OR CALCULATING AHEAD, let’s look at potential tactical ramifications that result from that move.
Mentally visualize the Black Queen sitting on the d5-square. That is the only change we have to visualize in the board position.
WHAT CAN WE SEE TACTICALLY — FOR BOTH PLAYERS?
There is an immediate fork threat on d1, potentially grabbing a ‘free’ Bishop. The White Knight on e4 just became B.A.D. There is the possibility of adding the BBd7 to the ‘attack’ on e4 but no way to ‘attack’ the two defenders directly (too bad that Black doesn’t have an h-pawn?). White is forced to respond to the threat.
White can eliminate the threat to the WNe4 by simply capturing on f6 with check. Or, he can protect the d1-square while making a counterthreat against the Black Queen. The first option does not appeal to Alekhine because it heads toward an endgame with pawns on both sides of the board in which the two Bishops may develop a lot of activity. Besides, doing two things at once increases the complications (tension) in the position and Alekhine is in his element doing that. So, 25. WNc3. We now have two pieces moved MENTALLY on the board position. Look again at the resulting position from Black’s perspective.
WHAT CAN WE SEE TACTICALLY — FOR BOTH PLAYERS?
White now threatens to capture the Black Queen AND prevents the fork on d1 AND removes the Knight from the B.A.D. e4-square. However, it also makes the WNc3 LPDO. Black can ‘attack’ it AND add ‘pressure’ to the e2-square, making it B.A.D. So, 25… Qe5.
PART II:
ReplyDeleteWHAT CAN WE SEE TACTICALLY — FOR BOTH PLAYERS?
Here’s a surface-level clue: If your Knight is sitting on the same color square as either the opponent’s King or Queen AND there is a possibility of forcing the other piece to a square of the same color, there is a potential Knight fork lurking in the position. White’s Knight and Black’s Queen are sitting on dark squares. The “missing” h7-pawn is another clue, because there is a LoA h4- h8.
Figure it out (which Alekhine and Euwe did not). The tactical sequence is not that difficult to SEE and work through MENTALLY.
Note that getting to this point only required shifting two pieces MENTALLY. The combination involves SEEing three more pieces moving MENTALLY.
Sometimes, chess is not that complicated. We MUST be vigilant after every single PLY for tactical opportunities!
Alekhine, Alexander (2700) vs Euwe, Max (2650)
Date: 1937-11-11
Event: World Championship 17th, NLD
Round: 16
Result: ½-½
Opening: Catalan Opening, Open Defense (E02)
Problems: 118221, 118016225, 120746465
1. d4 Nf6 2. c4 e6 3. g3 d5 4. Bg2 dxc4 5. Qa4+ Nbd7 6. Nf3 c5 7. Nc3 a6 8. O-O Be7 9. dxc5 Bxc5 10. Qxc4 b5 11. Qh4 Bb7 12. Bg5 O-O 13. Rad1 Qc7 14. Rc1 Qb6 15. b4 Be7 16. Rfd1 Bc6 17. Be3 Qb7 18. Rxd7 Bxd7 19. Ng5 Qb8 20. Bxa8 Qxa8 21. Nxh7 Rc8 22. Ng5 Rc4 23. Nce4 Rxc1+ 24. Bxc1 Qd5 25. Nc3 Qe5 26. Bb2 Bc6 27. a3 Bd6 28. e3 Qf5 29. e4 Qg6 30. f3 Nd7 31. Nh3 f6 32. Nf4 Qf7 33. Qg4 Ne5 34. Qxe6 Nxf3+ 35. Kf2 Qxe6 36. Nxe6 Nd2 37. Nd4 Nxe4+ 38. Nxe4 Bxe4 39. Ke3 Bb7 40. Nf5 Bc7 41. Bd4 Kf7 42. Bc5 Kg6 43. Bd6 Bd8 44. Nd4 Bd5 45. h3 Bb6 46. Bc5 Bc7 47. Ne2 Bc4 48. Nf4+ Kf7 49. h4 g5 50. hxg5 fxg5 51. Nh3 Kf6 52. Nf2 Kf5 53. g4+ Ke6 54. Nh3 Kf6 55. Bd4+ Kg6 56. Ke4 Bf1 57. Nf2 Bg2+ 58. Kd3 Bd5 59. Nd1 Bf3 60. Nf2 Bf4 61. Be3 Bg3 62. Kd4 Bb8 63. Kd3 Be5 64. Bd4 Bf4 65. Be3 DRAW
Knight forks are a breed apart. And should have a training regimen of their own. I have a problem set with knights moves. Yet I'm still vulnerable in this area.
DeleteRepeating NM Dan Heisman’s “recipe” for good, simple chess:
ReplyDelete1. SEE if there is a tactic for EITHER SIDE.
2. If so, address it with your net move(s).
3. If not, maximize the activity of your pieces (keeping them safe and active as possible) and minimize your opponent’s pieces activity (make them vulnerable and inactive as possible).
TAKE YOUR TIME [SLOW DOWN!].
While searching again in Michael de la Maza’s little book “rapid chess improvement - a study plan for adult players” regarding his recommended thought process, I had an insight regarding what he did NOT EXPLICITLY tell us.
De la Maza’s eight-step procedure for OTB play is:
1. Make a physical movement. Shuffle your legs, shift in your chair, move your arms up and down, or wiggle your toes.
2. LOOK at the board with Chess Vision. [the eagle’s eye view]
3. Understand what the opponent is threatening. [the eagle’s eye view]
4. Write down the opponent’s move.
5. If the OPPONENT has a serious threat, then respond.
6. If not, calculate a tactical sequence [for US].
7. If no tactical sequence EXISTS, implement a plan based on the following:
1) Improve the mobility of the pieces. [Start with the worst-placed piece].
2) Prevent the opponent from castling.
3) Trade off pawns.
4) Keep the Queen on the board.
8. Make sure I have pressed the clock.
As GM Nunn noted (in a review of de la Maza’s book in Secrets of Practical Chess), some of these suggested steps violate FIDE rules for tournament play, and may seem somewhat bizarre (step 1).
If you examine these steps in comparison with Heisman’s simple process, you’ll find that the two processes are essentially the same—where it counts!
What I previously missed (or are possibly imagining now) is the purpose behind some of these steps. I think those “bizarre” steps may be “triggers” to reset the attention after each move, to avoid going into tunnel mode (which usually results in unpleasant surprises). De la Maza actually states this purpose a little later in the book. (As usual, I didn’t previously connect these particular dots while reading it.)
But you must FORCE YOURSELF to work through the entire eight-step process ON EVERY MOVE; otherwise you will develop poor habits and MISS TACTICAL OPPORTUNITIES when they arise.
SURPRISE!!
I know I’ve usually just learned [KNOW THAT] about these various thinking processes to the “familiarity” level, but never applied one of them with diligence ON EVERY MOVE [KNOW HOW].
I think I need to change my approach to incorporate some kind of “triggers” that will bring me out of and then back into focused attention MOVE BY MOVE.
I'm not there yet. I totally focus on the process for tactics.
DeleteI have ideas for the rest of the game, of course. But I must first have the tactics straight. In an OTB game I focus on making an exchange plan. I try to attack the King, but if it doesn't work out, I base my play on making the right exchanges.
As long as I don't look for the consequences of moves but only at what they are evidently doing, the focus area is my habits.
Correction:
ReplyDelete"2. If so, address it with your net move(s)."
should read:
2. If so, address it with your next move(s).
While mulling over the REPL idea, I realized there is a connection to earlier advice from GM Davies:
ReplyDeleteThe REALITY is that YOU’VE GOT TO MOVE THE PIECES AROUND THE BOARD AND PLAY WITH THE POSITION. Who does that? Amateurs DON’T; GMs DO.”
I deliberately did NOT include “moves” per se in that REPL formulation. Why not?
Because the point is NOT trying to SEE LEGAL concrete moves and variations projected into a Tree of Analysis, but, instead, to SEE at a more abstract conceptual level (based on relationships between pieces and squares, interconnections, functions, etc.).
As previously noted, the idea is to mentally reposition ["shuffle"] pieces [“move the pieces around the board and PLAY with the position”] as needed to clarify the potentialities in the position currently imagined in the mind’s eye. IMHO, there will usually only be a few pieces at most that need to be “imagined” in new locations. That’s the reason for emphasizing using the actual position as the starting point, and (in the imagination) shuffling only those pieces required to SEE the consequences of the idea(s). By NOT trying to keep strict account of the piece movements (IE, the Tree of Analysis), there is reduced load on working memory.
Since all of this “SEEing” is done in the mind’s eye, the only actual move that must be remembered is the first move at the start of the conceptual sequence.
MOST IMPORTANTLY: After making the finally selected move and getting the opponent’s response, REPEAT THE SAME REPL PROCESS FROM THE “NEW” POSITION ON THE BOARD!! Do NOT rely on your imagination, and just lackadaisically play whatever you “thought” should be an entire sequence of moves. You may have overlooked something that only becomes clear when you actually see the position on the board.
Remember GM Larsen’s axiom: Long variation, wrong variation.
You are only limited by your imagination — but most people cannot “imagine” how limited their imagination really is!