Sunday, February 22, 2009

The Beginnings of an Improvement Plan

I just got back from vacation which was the perfect opportunity to try out a new improvement plan regiment. Nothing fancy, just a few solid books and my rationale holding them together:
  • Understanding Chess Move by Move, Nunn
  • Learn Chess Tactics, Nunn
  • The Amateur's Mind, Silman
  • Chess Visualization Course, General Tactics, Anderson
  • (finished prior to vacation) Logical Chess Move by Move, Chernev
I generally did a little from each book in chunks of chapters, then moved to the next book. The theme was to alternate between games and tactics. "Move by Move" is full games of course, while "Amateur's Mind" featured parts of games where big plans are (or should be) formed. "Chess Tactics" is divided by theme, based on real games and built up quickly in difficulty. "Chess Visualization" requires the reader to visualize a position after a prescribed sequence of tactical moves. So you didn't have to find the move, but you had to correctly answer questions about the resulting position.

In the past, I'd be blasting through CT-ART 3, or reading some theory book, or some "evaluate this position" book. What I was missing was seeing the whole game, and watching the factors that decide the game be employed by the players, and wax and wane and why.

Of course I was also reviewing some of my past games played at MetroWest CC and Boston University Open. Periodically after going through a few books, I'd play through a game just to see if I could look at the game differently. Indeed, some games I can see through different eyes. Here is a good example (hold your nose):

[Event "MCC Jan 2007"]
[Site "Natick MA"]
[Date "2007.01.02"]
[Round "1"]
[White "Reed, Harvey"]
[Black "Michael, William"]
[Result "0-1"]
[ECO "C68"]
[WhiteElo "1484"]
[BlackElo "1735"]

1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Bxc6 dxc6 5. O-O Bg4 6. Re1 Qf6 7. d3 Bxf3 8. Qxf3 Qxf3 9. gxf3 O-O-O 10. Be3 g6 11. Nc3 Bh6 12. Kf1 Bxe3 13. fxe3 Ne7 14. Ke2 Rd7 15. Rad1 Rhd8 16. Rf1 c5 17. a4 c6 18. a5 c4 19. dxc4 Rxd1 20. Rxd1 Rxd1 21. Kxd1 Kc7 22. c5 f5 23. Ke2 Kd7 24. Kf2 Ke6 25. exf5+ gxf5 26. e4 f4 27. Na4 Ng8 28. Nb6 Nf6 29. Nc4 Ne8 30. b4 Kf6 31. Kg2 Ke6 32. Kh3 Kf6 33. Kg4 Nc7 34. Nd6 Ne6 35. c3 Ng5 

One would expect the simple 36. Nxb7. In fact this was White's (me) goal for many moves, since 22. c5 to be exact. So why on earth would White all of a sudden make a U-turn, which ultimately backfires in a humiliating mess? I think because deep down White didn't trust his own analysis. Probably because he is analytically lazy and therefore doesn't trust himself. The rest of the game continues to disappoint, meeting the predictably bad end in another 12 moves:

36. Ne8+?? Kg6 37. Ng7 {offer draw} Nxe4 38. fxe4 Kxg7 39. Kf3 Kg6 40. Kg4 Kh6 41. Kh4 f3 42. Kg3 Kh5 43. Kxf3 Kh4 44. Kg2 Kg4 45. h3+ Kf4 46. Kf2 Kxe4 47. Ke2 h6 48. Kd2 Kf3 0-1

Several things struck me as I was going through the book regiment:
  • Reviewing whole games is something that I've needed to do for years, but I've always avoided. Starting with the simple Chernev book was the best thing I have done since I started adult onset of chess. I have Howard Goldowsky to thank. He claims to have gotten the "unlearn before you learn" religion from Heisman, and passed it on to me. 
  • Tactics are always important, and there is life beyond CT-ART, although I must admit CT-ART 3.0 is the reason I don't drop pieces left and right anymore.
  • Analysis is hard work. It took me three days to work through game 7 (Kasparov-Karpov) in Nunn's "Move by Move". I feel like I have a lot more patience with analysis, and will start doing deep analysis on my games.
  • Mental toughness can frequently make up for intermitent weaknesses in analysis during a game.
  • The whole imbalance theory of Silman's was starting to make sense by the time I hit the chapter on Initiative. That was a scary feeling.
And then it hit me. Valery Frenklahk, from whom I took half a dozen lessons from in the late 90's, was right: Chess is a game of squares! Just play chess!

Why was I remembering chess lessons from the past decade? I think because he had a style that made a big impression on me, plus he said incomprehensible things to me, and I was determined to somehow figure it out, even if I forgot about it first.

Things are still confusing for me, which is ok. I still need to do analysis of my MetroWest CC game from last Tuesday, but I avoided second guessing, did a little more on the job analysis, and kept my mental toughness, not getting despondent when my opponent made an unanticipated (and aggravating) move. You don't even have to hold your nose...

[Event "MCC Feb 2009"]
[Site "Natick MA"]
[Date "2009.02.17"]
[Round "3"]
[White "Reed, Harvey"]
[Black "Callahan, Daniel"]
[Result "1-0"]
[ECO "E61"]
[WhiteElo "1373"]
[BlackElo "1232"]

1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e3 O-O 5. Nf3 d6 6. Bd3 b6 7. O-O Bb7 8. Qe2 c5 9. d5 e6 10. e4 h6 

With Black's last move he weakened his pawn structure. Previously White had plans focused on a strong center first, then Kingside pressure later. With 10... h6, White can think about breaking up Black's Kingside:

11. dxe6 fxe6 12. e5 Nh5 13. Bxg6 Nf4 14. Bxf4 Rxf4 15. exd6 Bxf3 16. Qxe6+ Kh8 17. d7 Bc6 18. Rad1 Bxd7 19. Qd5 Nc6 20. Qxd7 Qxd7 21. Rxd7 Ne5 22. Nd5 Rg4 23. Bf5 Rg5 24. f4 Rxf5 25. fxe5 Rxf1+ 26. Kxf1 Bxe5 

With White's next move, he nullifies Black's dark Bishop. Next is a plan to exchange the Kingside pawns, leaving all the pawns on side (favorible to White's Knight), which is frustrating to Black's Bishop. As they say, it's now just a "matter of technique". Of course my technique is still simple and crude, so it takes me another 42 moves...

27. b3 Rf8+ 28. Ke2 Bxh2 29. Rxa7 Re8+ 30. Kf3 Bb8 31. Rb7 Rf8+ 32. Kg4 Rg8+ 33. Kf3 Be5 34. g4 Bg7 35. Nxb6 Rf8+ 36. Kg2 Bd4 37. Nd5 Rf2+ 38. Kh3 Rxa2 39. Kg3 Ra8 40. Kf3 Rf8+ 41. Ke4 Re8+ 42. Kf5 Rf8+ 43. Kg6 Rg8+ 44. Kxh6 Rxg4 45. Rb8+ Rg8 46. Rb7 Bg7+ 47. Kg6 Bd4+ 48. Kh5 Rg1 49. Nf4 Rh1+ 50. Kg6 Rg1+ 51. Kh6 Bg7+ 52. Kh5 Rh1+ 53. Kg6 Rg1+ 54. Kf5 Rf1 55. Ke4 Re1+ 56. Kd5 Re5+ 57. Kc6 Rg5 58. Ne6 Rg6 59. Rxg7 Rxe6+ 60. Kd5 Rb6 61. Kxc5 Rxb3 62. Rg4 Rf3 63. Kd4 Rf7 64. c5 Kh7 65. Kc4 Rd7 66. c6 Rc7 67. Kc5 Kh6 68. Kb6 Rc8 69. c7 Kh7 1-0

...and The Quest continues...


  1. Harvey,

    In the first game, getting used to seeing the pawn majority advantage on the queenside is visualization exercise. The pawn structure is a “V” versus (inverted) “V”. Even after the Nxb7 move, the first instinct to get over is the advance of the b-pawn will NOT create a runaway pawn for black. But we’ve been taught that in order to achieve a 3 x 3 breakthrough, the pawns are better off in a line. But here, given the BAD pawn formation for a “typical” break through, the EXCEPTION must come into play and recognized. That exception is the extra c-pawn covering the run-away. Then you get to have a passed pawn. Just to recap: Basic pattern to recognize is standard 3 x 3 break through. First Alarm says you don’t have the ideal formation. Ask: Is this an exception. ( Yes! I have extra pawn and a knight to mess things up).

    In the second game, you showed nice form. It was great you recognized that the time was right to attack the king side once Black played h6. So, some of what you are learning is starting to seep in that thick cranium! Some of the liquidation was unavoidable. In situations when you have the attack, initiative is essential. Remember, you give up initiative if you recapture. Also you have to look at the dynamic piece advantage you have in the section of the board you are attacking. Both your bishops were fired up on key diagonals and a knight and queen ready to drop in like paratroopers. Black only had a rook, a knight and bishop. There’s bound to be a Lasker bishop sac in there somewhere.

    Don’t get out of balance with the Silmanite religion of imbalances. It’s a good basis to evaluate positions in every sense but a lot of the material is beyond the sub-1600 level. You are better off going over whole games on your own without annotations. Try to evaluate them your self. This is the best way to simulate OTB. Only after you exhaust yourself of the whole game, should you back track and look at the annotations in the book. All the games in the books you mentioned can be downloaded as a PGN minus the notes. I am a firm believer in working on your own annotations as the means to getting better. When you do your own games, check them after with an engine or a stronger player.

    Enough of my 2 cents.

  2. Not to muddy the waters....but GM Jonathan Rowson has done great work on chess is so hard to master from a psychological perspective. His latest book, "Chess for Zebras" (Gambit, 2006) has great insights as to why we fail. I don't know why you didn't move Nxb7 vs Bill Michael, but it may have been due to you remembering the maxim that Knights are best placed in the center of the board. Rowson is a proponent of the school of "Unlearning" that you reference. We (most of us adult patzers) have too much knowledge about chess that keeps us from playing better.

    Like Silman, (particularly in his Reassess Your Chess Workbook) Rowson advocates analysing positions from annotated game collections as the best way to improve your play, rather than improving your knowledge. The value is not so much what you learn, but the actual work at the board that you are doing. Like Silman, he suggests going through a game and annotate it, then compare your annotations vs the GM analysis. Spend a couple hours on it, as if it were a tournament game. Of course your annotations will miss a lot, but when comparing your notes to the GM's notes, don't be discouraged, instead feel good about analyzing for 2 hours! Also, comparison (of your notes to the REAL annotations) will, eventually, demonstrate what kind of positions (strategies, tactics) that you do well in, and those that you are lacking.

    Rowson tells of a similar type of training session he had with GM Artur Yusupov in 2003. Rowson was already an established GM with an ELO in the 2500's, but aspired to improve so he wouldn't feel so imtimidated by Super GMs. Yusupov spent a week giving him tough exercises, and Rowson admits to not doing very well -- Yusupov told him that although he knew a lot about chess, that Rowson was lazy when it came to analysing... but Rowson was complimented by his trainer for putting forth such a hard effort for the week.

    So, the moral of the story is that you and Rowson suffer from the same flaw... you're too lazy in your analysis.

    One suggestion if you do try to annotate a game... pick a short one, and don't choose a game full of tactics from Tal or Bronstein. (Even worse, a game where Korchnoi tries to confuse the hell out of his opponent.) My preference are short games (<30 moves) by Smyslov, Capablanca or Karpov.

    By the way, Rowson mentions Mike de La Maza and his book, Rapid Chess Improvement. He has high praise for it, but, feels that while Mike was drilling tactics, he was, more importantly, learning how to think and play.

    Good luck in improving your play. Just remember to focus on improving your play, and not your knowledge. Silman's concept of "imbalances" is only intended to help you play better by helping you arrive at a reasonable number of candidate moves. But imbalances, without the context of an actual game, are worthless. You may have a position, someday, where, in a closed position, the winning move requires you to sacrifice a perfectly posted Knight for a "bad" Bishop, but you won't see it because it goes against what you've learned.