23 February 2013

Annotated Game #84: Piece exchanges and draw offers will lose you the game

This was from the first round of the next tournament following Annotated Game #79.  While the previous tournament had a bad vibe to it from the beginning, this one was considerably better overall.  The fact that I was paired in the first round against a strong Expert player was a big benefit for me, as I didn't feel any pressure to win due to the 500-point ratings gap between us.  My opponent on the other hand did not appear to be as happy and looked even less happy after I had built up a fine-looking position out of the opening, an English - Grunfeld defense setup.

Black makes the first significant error on move 12, unnecessarily moving his rook away from the defense of the f7 square.  I was able to take advantage of this and by playing obvious moves had obtained a clearly superior position as of move 15.  At this point my lack of positional judgment starts to show, however, as I choose the wrong square for the retreat of my knight.  This is followed by allowing Black to exchange his so-so knight for my excellent light-square bishop on move 18, which marks the real strategic turning point of the game.  One of the things Class players often lack is a sense of the importance of piece exchanges and this game is an excellent illustration of the consequences.  Black immediately obtains the initiative and the bishop-pair, allowing his pieces to spring to life and target what are now some obvious White weaknesses.

Despite White's forced retreat, Black misses some chances to leverage his positional advantage for tactical gains (including 23...Bb5!) and White re-achieves equality, making a draw offer that is rejected.  Nowadays I've given up the practice of early draw offers in favor of emphasizing mental toughness, but even then I have to admit it was rather rude, given the ratings gap, not to mention being overly optimistic.  Indeed, a few moves later White plays the complacent 27. Nd3? and this time Black does not miss his chance to inflict material losses on White, who eventually loses after trading down into a bishop (for Black) endgame.

Although this was a loss, the game in fact left me in a relatively positive mood for the rest of the tournament.  I had made a 2100+ player sweat through the early middlegame and did not simply collapse after his first counterblow.  This positive frame of mind helped in my later games.  This was also the first tournament I played in after beginning to train with Qigong breathing exercises (part of my Taijiquan martial arts practice), which also appeared to have a positive effect on my mental outlook.  More on that later.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Expert"] [Result "0-1"] [ECO "A16"] [Annotator "ChessAdmin/Fritz/Houdini"] [PlyCount "128"] {A16: English Opening: 1...Nf6 with ...d5} 1. c4 Nf6 2. Nc3 d5 3. cxd5 Nxd5 4. Nf3 g6 {entering a Grunfeld-type structure.} 5. e4 Nxc3 6. bxc3 Bg7 7. Ba3 O-O 8. Bc4 Nd7 9. d4 {White is more or less obliged to play d4 either now or shortly afterwards, as the pawn does little for him on d2 and the d3 advance would leave the c3 pawn undefended.} c5 10. O-O cxd4 (10... Qa5 {used to be the favored move. Here's an example of a White win in the line, in a game between two high-level American GMs.} 11. Qb3 a6 12. Ng5 e6 13. Nxf7 Rxf7 14. Bxe6 Ne5 15. Bxf7+ Nxf7 16. Bxc5 Qb5 17. f4 Qc6 18. f5 gxf5 19. exf5 Bd7 20. Rae1 Qb5 21. Rf3 Re8 22. Rxe8+ Bxe8 23. Qe6 h5 24. h4 Bd7 25. Qe7 Qb1+ 26. Rf1 Qb5 27. f6 Bh6 28. Qe4 Kh8 29. Re1 Qb2 30. Qf3 Qd2 31. Re2 Qc1+ 32. Kh2 Bf4+ 33. g3 Bh6 34. d5 Bb5 35. Re7 Qd2+ 36. Kh3 Kg8 37. c4 Ba4 {1-0 (37) Gurevich,D (2490)-Fedorowicz,J (2560) Durango 1992}) (10... Qc7 {has scored much better for Black and is the recent favorite.}) 11. cxd4 Nb6 12. Bb3 (12. Be2 {is an alternative which would leave the b3 square open for the queen.}) 12... Re8 $14 {this is an error, as it weakens f7 unnecessarily.} (12... Bg4 $11 {is logical, developing with the pin on the Nf3 and is Houdini's first choice.}) (12... a5 $5 {was Fritz's preference, with the idea of pushing ...a4 and dislodging the bishop.}) 13. Rc1 {White moves his rook off of the long diagonal (where it was subject to threat by the Bg7) and onto a valuable open file.} a5 14. Ne5 e6 ( 14... Bxe5 {is Houdini's recommendation, bailing out into a slightly inferior endgame.} 15. dxe5 Qxd1 16. Rfxd1 a4 17. Bd5 e6 18. Bc4 Nxc4 19. Rxc4 $14) 15. Qf3 {any time White can get in this move in a queen pawn opening, usually he's doing well. Here White's pieces are all well-placed, in contrast with Black's uncoordinated forces.} f6 16. Ng4 {White starts going wrong with this retreat to the wrong square.} (16. Nc4 {would exchange off Black's useful knight and leave him backwards in development.} Nxc4 (16... a4 {would offer White the chance to sacrifice a pawn for some strong pressure.} 17. Nxb6 Qxb6 18. Bc4 Qxd4 19. Rfd1 Qb6 20. h4) 17. Bxc4 Bd7 $14) 16... a4 $11 17. Bc4 Bd7 (17... Nxc4 {Black should take the chance now to eliminate White's excellent bishop and obtain the bishop-pair for himself.} 18. Rxc4) 18. Qd3 (18. Bd3 $14 { was a missed opportunity. The bishop is well-placed on d3 and the subsequent minor piece exchange after the text move favors Black.}) 18... Nxc4 $15 { suddenly Black is able to take over the initiative, activating his pieces and targeting White's weaknesses.} 19. Qxc4 Bc6 20. f3 Rc8 21. Rfd1 h5 22. Nf2 Bh6 23. Rc2 $2 {this is one of those cases where the decision on which square to move a piece is very important, without obviously appearing to be.} (23. Rc3) 23... Qd7 $6 (23... Bb5 $1 {would win the exchange now. If the rook were on c3 and Black attempted this, the queen would simply move to b4.} 24. Qxb5 Rxc2 25. Qxa4 Rxa2 $17) 24. Qb4 {I recall offering a draw around here, or perhaps a little earlier, which was naturally rejected by my higher-rated opponent. The focus on achieving a draw was no doubt detrimental to my play.} Bb5 25. Qb2 { overly passive.} (25. Rxc8 {would be the active choice.} Rxc8 26. d5) 25... Bc4 {this seems unnecessarily complicated. I believe he may not have considered my next move.} (25... Rxc2 26. Qxc2 Rc8 27. Qb2 Be3 28. Kh1 Ba6 {and Black has all the chances.}) 26. Bc5 b5 27. Nd3 $2 {this completely ignores my opponent's coming tactical threats.} (27. Qa3 $5 $11) 27... e5 {the Bc5 is now undermined and White will have to lose the d4 pawn.} (27... Be3+ {is a superior way of executing the threat.} 28. Kh1 e5 $17 {and as a result Black can win the d4 pawn with his e-pawn, establishing a dominating position.}) 28. d5 $2 {Not the best move objectively, but it at least requires Black to find the following tactic.} (28. Nf2 $15 {is the best defense here, but White will have a miserable time for the rest of the game anyway.}) 28... Rxc5 $1 $19 29. Nxc5 Be3+ {Theme: Double Attack} 30. Kh1 Bxc5 {Black has two bishops for the rook, giving him a material and in this case positional plus.} 31. Qc1 Kg7 32. h4 f5 33. Rxc4 {I make a desperate attempt to hold the position. It should have been played on the previous move, before the weakening h4.} bxc4 34. Qxc4 Bd6 {at this point the win for Black becomes, as the saying goes, a matter of technique. Black opts for safety first with the blockade of the d-pawn.} 35. Rc1 fxe4 36. Qxe4 Qf5 $6 {this unncessarily lets White have some activity, by leaving the a-pawn and the Bd6 hanging.} (36... Rf8 $5) (36... Rb8) 37. Qxa4 ( 37. Rc6 Qxe4 38. fxe4 Be7 $19) 37... Qf4 38. Qxf4 exf4 39. Kg1 {although the material difference here is slight, White's two disconnected pawns have no hope of survival.} Re5 40. Rd1 Kf7 41. Kf1 Ke7 42. Re1 $2 {I prefer to try my chances in a bishop ending, although this is objectively worse.} (42. Rd4 Rf5 $19) 42... Kf6 (42... Rxe1+ {seems even better} 43. Kxe1 Bb4+ 44. Kd1 Kd6) 43. a4 Rxe1+ {now Black goes for it, entering the final decisive phase of the endgame.} 44. Kxe1 Ke5 45. Kd2 Bb4+ 46. Ke2 Kxd5 47. Kd3 Be1 48. a5 Bxa5 49. Ke2 Bd8 {an example of why it's better to be a piece up in the endgame.} 50. Kd3 Bxh4 {the position is completely resignable now.} 51. Kc3 Be7 52. Kd2 Kd4 53. Ke2 g5 54. Kd2 Bb4+ 55. Ke2 Bc3 56. Kf2 Kd3 57. g3 Be5 58. gxf4 Bxf4 59. Kf1 Ke3 60. Kg2 g4 61. fxg4 hxg4 62. Kg1 Kf3 63. Kf1 g3 64. Kg1 Be3+ (64... Be3+ 65. Kh1 g2+ 66. Kh2 g1=Q+ 67. Kh3 Qg3#) 0-1

20 February 2013

A robust and deep insight on playing style

I finally tracked down the original reference from Playing Styles Deconstructed that underpinned what I termed the impracticality of pursuing a chess "style" for improving players.  The following is from GM Alex Yermolinsky's The Road To Chess Improvement (Gambit 1999).
...One strong GM once told me that during the game we (he meant World Top 100 or so) may happen to know, able to calculate, or in any other way find the best move in approximately 90 percent of positions.  This means that, if an average game lasts 50 moves, there will be 5 times during the game when we won't know what to do!  There comes the most interesting, yet difficult part.  He also said that these moments are very characteristic for a chess-player's style and personality.
I find this to be a robust and deep insight.  In the linked post, the primary lesson I took away was that attempting to conform to some idealized type of playing style ("tactical" or "positional" or whatever), especially at the Class level, would be an impediment to the improving player. What we really need to do instead, along the lines of the above quote, is focus on reducing the number of positions where we don't know what to do!

Another important aspect to the above-quoted idea is the concept that it is normal to have differing opinions over positions, even at the top level and to the point where you do not in fact understand all aspects of the position.  It's been remarkable to me how often the very top-rated players will candidly express this during interviews or in post-game analysis, for example in How Kramnik makes us feel better about chess.  I have seen Carlsen and Anand be similarly self-effacing and the lack of ego in the top ranks these days is very refreshing, as well as motivational for me to see.

A final comment on style: assuming the idea above is broadly true, our personal style will naturally shine through as part of a game's development, giving scope for creativity and individuality.  But  that does not excuse us from increasing our efforts at objective investigation and understanding of truth on the chessboard.

17 February 2013

Annotated Game #83: Rocky Rook Revenge Match

Although the ill-fated Double My Egg Nog tourney was never finished - only a single serving - Rocky Rook was still able to play our second round game.  In the first round game linked, he had won an interesting struggle in the Colle System, featuring two key blunders from myself, so I was looking forward to evening the score with some better play.

In my previous Round Turkey game against Rocky as White (Annotated Game #73), he played a sort of Old Indian type setup as Black.  This time, he started off with 1...g6 and I expected him to go into a King's Indian Defense setup eventually.  Instead, he surprised me with an early 3...c5 and took the game into the Symmetrical English, which is relatively rare at the Class level.  I therefore don't have a lot of experience with it, but Rocky seemed to have even less, so I found that somewhat encouraging heading into the middlegame.

The middlegame opens up after Rocky's 13...b5, which although objectively fine (according to Houdini) I felt played into White's hands strategically.  Essentially Black is forced to drop a pawn as a result of the move, but could have gained full compensation after the variation 16...Rc8, which establishes a strong center and kicks White's pieces around.  In the game continuation, Black remains active, but White is able to consolidate on the queenside and activate the passed a-pawn after Black initiates an exchange of knights.  I was, however, forced to think hard and find "only" moves that would protect material and at the same time give my pieces their necessary activity.

The last phase of the game occurs after Black's 23...Rc2 dangerously unbalances the position.  While the rook is threatening-looking on the second rank, White's attack on Black's back rank comes first and White's minor pieces are able to combine with the queen on an effective attack on Black's king, while White's rook and king hold the defense together.  The most challenging part of the calculation was when I had to find 29. Bd4, breaking a pin on White's queen by force, in order to finish the attack.

Thanks to Rocky for another well-fought game.  The key difference this time was my lack of blundering (always helpful!) and more accurate calculation.  I realized early on in the game that a Symmetrical English would require patience from me in order to eventually make progress, so having that mindset was a useful assist to my play.

Hopefully Rocky and I can do a regular set of matches; it would be good to get a monthly game going, for example.  It makes a difference having someone available to regularly challenge you, since it's not as easy to hide your weaknesses from them as it is with random tournament opponents.

[Event "rated standard match"] [Site "Free Internet Chess Server"] [Date "2013.02.10"] [Round "?"] [White "ChessAdmin"] [Black "RockyRook"] [Result "1-0"] [ECO "A16"] [WhiteElo "1671"] [BlackElo "1701"] [Annotator "ChessAdmin/Houdini"] [PlyCount "61"] [EventDate "2013.??.??"] [TimeControl "3600+5"] {A16: English Opening: 1...Nf6 with ...d5} 1. c4 g6 {in my experience, this usually presages a KID type setup against the English.} 2. Nf3 Bg7 3. g3 c5 { this choice makes it likely that we'll eventually transpose into a line of the Symmetrical English or the Hedgehog.} 4. Bg2 Nf6 5. O-O d6 6. Nc3 O-O 7. d3 { instead d4 would transpose to a King's Indian Defense variation.} Bd7 {I didn't think it was a good idea for Black to choose to place the bishop this early. It weakens b7 and it's not clear what the best square for it is. However, it does position itself to go to c6 and helps keep some control of b5. The database shows White scoring 78 percent from this position.} (7... Nc6 { is the overwhelming choice here and scores only a little over 50 percent for White.}) 8. Rb1 a6 $146 {this lets White get in b4 immediately.} (8... a5 { would restrain b4;}) (8... Bc6 {seems like a logical continuation of the bishop development.}) 9. b4 cxb4 10. Rxb4 Bc6 11. Qc2 {this is where White has to think about his preferred middlegame setup and how to develop the queen and dark-square bishop. The hanging Nc3 usually means that the queen or bishop needs to cover it, or White risks tactics based on a discovered attack from the Bg7.} (11. Bd2 {is a logical alternative, as Bd2 is often played in the main line Symmetrical variations. I was hoping for a more active development of the piece.}) (11. e4 {is also playable, with a lock on the d5 square, and was something I considered. However, the drawback of cutting off the Bg2 made it less appetizing.}) 11... Nbd7 12. h3 {Secures g4, in anticipation of playing Be3.} Rb8 (12... Rc8 {looks like a better spot for the rook, opposite the Qc2.}) 13. Be3 b5 {here I thought that Black was playing White's game strategically, allowing White to concentrate everything on the queenside. White is normally better placed to establish at least a small advantage there, so perhaps more activity in the center was indicated. Tactically, this just seems to lose a pawn, although we'll see later that Black can obtain compensation.} 14. cxb5 axb5 {White now has the more active position and an obvious target in the Black b-pawn.} 15. Nd4 {this is more effective now that the pawn exchange has occurred on b5, as Black no longer has the possibility of capturing on c4.} Bxg2 16. Kxg2 Qa5 (16... Rc8 {instead would allow Black to strike back with a series of threats against White's vulnerable piece positions.} 17. Rxb5 e5 18. Nb3 d5 19. Bc5 Nxc5 20. Rxc5 Rxc5 21. Nxc5 Qd6 22. N3a4 Rc8 $11 {and now play could continue} 23. Qd2 d4 (23... Rxc5 $6 24. Nxc5 Qxc5 25. Rc1 $16 {and Houdini evaluates the R+P as being advantageous against Black's B+N combination.})) 17. Rxb5 $14 {the extra tempo earned by the attack on the queen is very helpful to White.} Qa8+ 18. Kh2 Rxb5 19. Ndxb5 {White's queenside now hangs together well, with the knight supporting each other and the pawn on a2 protected twice. White should still be careful about Black's actions on the queenside, in order to avoid being caught with an overloaded piece or similar tactic.} Rc8 20. Qd2 (20. Qb3 {is Houdini's preference and was the other possibility I considered, since it would have maintained control of d5. However, it would then have allowed a pin of the b5 knight or a free tempo for black with ...Nc5.}) 20... Nd5 {superficially this looks like it generates some pressure for Black on the queenside. However, Black is still playing White's game strategically and the exchange of minor pieces helps make White's passed a-pawn that much more important.} (20... h5 {is an idea from Houdini, looking to stir up some counterplay or at least a distraction on the kingside.}) 21. Nxd5 {forced} Qxd5 22. a4 {clearly best, as it gets the passed pawn moving while also protecting the knight at its outpost square.} Qb3 { I had seen this far after Black's move 20, but had been unsure of how best to continue, so thought for while here.} 23. Qa5 {is the active choice to protect the pawn, as now (so I thought) Black could no longer play ...Rc2.} (23. Qd1 { is much too passive.}) 23... Rc2 {Black plays it anyway. This looks threatening but the rook is needed more on defense than on offense.} (23... Qe6 24. Qa7 h5 25. a5 $16) 24. Qd8+ (24. Qa8+ $5 {I admit I didn't even consider this possibility until after the move was played and it didn't seem like it would make a difference.} Nf8 25. Rc1 Rxc1 26. Bxc1 h5 $18 {White gives up some of his direct kingside attack, but now the a-pawn becomes a monster.}) 24... Nf8 $16 25. Qxe7 Qxa4 {Black's premature pawn-snatching is swiftly punished.} (25... Rxe2 $5 {is better here, although White is still in control.} 26. Rc1 {threatening to go to either c8 or c7 as necessary to support the attack.} Qxa4 27. Nxd6 Qa2 28. Kg1 $16) 26. Nxd6 $18 Qa2 27. Ne8 Rxe2 $2 { this no longer works, as White is a tempo up on the move 25 variation where ... Rxe2 is played immediately. White can now use that tempo to eliminate the Bg7.} (27... Ne6 {would be the way to fight on into the endgame, although it looks bad for Black.} 28. Nf6+ Bxf6 29. Qxf6 Qb2 (29... Rxe2 $2 {a poisoned pawn} 30. Ra1 $18) 30. Qxb2 Rxb2 $18 {and White should be able to convert the two extra pawns into a win.}) 28. Nxg7 Kxg7 29. Bd4+ {the point behind the capture on g7, as the pin on the bishop is now broken by force.} Kg8 30. Qf6 {mate is now inevitable.} Rxf2+ 31. Rxf2 {RockyRook resigns} 1-0

16 February 2013

Book completed - Practical Middlegame Techniques

Today I completed Practical Middlegame Techniques by IM Danny Kopec with FM Rudy Blumenfeld (Cadogan, 1997, First Edition).  There was a second edition released in 2012, with a considerably expanded page count (242 pages versus the original 128) that is referenced in the above link; all comments here pertain to the first edition.

I would liken the effects of working through this book to drinking a large cup of coffee and taking a multivitamin.  It's not going to improve your chess the same way as a full-course meal would, but it should give anyone Class B and below a relatively quick and permanent boost to their understanding and skills.  Although the book is slim, it is content-rich and took me a fairly long time to work through.  (This was in fact my third attempt to do it.)  I'll take the fact I completed it now as a positive sign of my commitment to training, rather than bemoaning my previous laziness.

The contents are divided into three chapters:
  • Chapter 1: Essential Tactical Methods presents a series of common (and not-so-common) mating patterns and then some increasingly complex examples incorporating them.  The chapter moves on to define and illustrate different types of material gain combinations and then finishes with a series of more complex combinations incorporating different themes.
  • Chapter 2: Pawn Structures and How To Use Them, as the authors acknowledge, draws heavily on Dr. Hans Kmoch's Pawn Power in Chess, but then again so do most other books on the topic.  The presentation of pawn structures and how to use different types of "levers" (perhaps best described as pawn advances with a specific purpose, usually challenging another pawn) is concisely done and dispenses with most of Kmoch's weird vocabulary.
  • Chapter 3: The Conditions and Methods for Attacking the King focuses on attacking play and presents some typical maneuvers used, along with identifying the conditions under which they can be used.   
I fit the profile for the best target audience for the book, which would be an intermediate-level player who is self-taught and can benefit most from exposure to the structured concepts in each section.  Someone who has already done years worth of mating and combination study would not find anything new in Chapter 1, although the mating pattern list is probably the best single reference that I've seen on the topic.  I haven't had that kind of rigorous exposure to tactics training, so I significantly benefited from the explanation and examples of mating and other tactical themes.

Similarly, Chapter 2 may not surprise anyone who already has some exposure to the positional concepts of pawn play, but I found the discussion of pawn structures and their static and dynamic qualities to be well worth the time. The examples given, largely game fragments but also including some complete games, were all well-chosen.  For me, I found the illustrations of the "sweeper-sealer" lever (sacrificing a pawn to free its square for a piece), minority attack and techniques for shielding backward pawns to be the most valuable.

Chapter 3 I found to be worthwhile, but more uneven.  The first two parts were the most enlightening, including examinations of things like how to use a space advantages in the attack and particular cases of this advantage such as when White possesses an advanced pawn on e5.  Separate sections focusing on the rook lift maneuver and exchange sacrifice on the long diagonal also were especially useful.  The last part of the chapter ("Radical Approaches") introduces sections on the king hunt, major material sacrifices, and concentration of forces.  Here the examples given struck me as not being as clearly illustrative of the concepts as those in previous chapters.

The wide range of games and game fragments presented in the book is one of its strengths.  They include some classic examples from earlier 20th century play, examples from the authors' games, and selections from more modern and contemporary international tournament practice.  Kopec's annotations are concise and helpful, although there is a lot of "...and wins" type commentary when presenting variations.  It is primarily for this reason that I would say the book is best suited for intermediate players (Class B/C) who will be able to work out some of the tactics not explicitly given in the text.  Of course, this is the case whenever you work through annotated games, so the book is not unique in that respect.

As someone looking for the proverbial shot in the arm for their middlegame technique, I found the time put into the book to be worthwhile and will return to it for reference and to refresh my memory and understanding in the future.

10 February 2013

Annotated Game #82: A good diagnosis

This recent game (part of the Chessmaster ladder series of training games) provided a good diagnosis of my current playing strengths and weaknesses.  The opening phase is strong, with my opponent following a move 8 sideline of the Caro-Kann Panov variation (officially classified as a Queen's Gambit Declined Semi-Tarrasch, which is what it becomes via transposition).  White's strategic error on move 11 with the bishop exchange gives Black relatively easy equality, in contrast with the kingside pressure White normally achieves in this variation.

The early middlegame analysis (moves 12-14) shows some interesting alternative plans for Black.  This for me is often the most valuable part of these training exercises.  Knowing how to play the early middlegame positions, in other words having a good idea of what to do in a position after your opening lines are finished, is a crucial skill and is something that I have often failed to do well.  In this game, my chosen path was not bad, but being aware of the other opportunities in the position will give me an advantage the next time I play a similar middlegame.

The game becomes tactical on move 17 as White drops a pawn with a typical computer handicap move.  However, it was much more interesting than it appeared, as the chosen method of White's piece recapture would have allowed Black to eventually win White's queen with a back-rank pin or check.  This was not obvious, however, and I instead focused on winning the pawn.

By move 24 we have an endgame where Black could have achieved a significantly stronger position by exchanging pawns on f3, inflicting a weakened structure on White and maintaining a strongly supported d-pawn.  Black instead ends up with several weak, isolated pawns that he cannot defend adequately, but is able to capture White pawns in exchange for them.  Black is stopped during the final race on the kingside, where the 2-to-1 pawn advantage is not enough to win in the single minor piece (BvN) endgame.

Diagnosis summary:
  • Solid opening preparation
  • Early middlegame was OK but not optimal, but this is not surprising given my lack of experience with actually playing the position.
  • Although I did not see the full possibilities of 18...Bf4! in terms of trapping the queen several moves later, I at least considered it as a candidate move, which I would not have done previously.
  • Remaining middlegame play was good, including the decision to force an exchange of queens and transition to the endgame.
  • I used my thinking process reasonably well and did not miss any significant threats from my opponent.
  • First endgame strategic decision was incorrect (not exchanging on f3), leading to better chances for my opponent.
  • Later on, my endgame advantage may not have been enough to win against best play, but I passed up several chances to improve my situation.
  • I lack knowledge of correct strategies in BvN endgames.

[Event "Chessmaster: Grandmaster Edition Rated G"] [Site "?"] [Date "2013.02.09"] [Round "?"] [White "Josh - Age 8 (CM Class B)"] [Black "ChessAdmin"] [Result "1/2-1/2"] [ECO "D41"] [Annotator "ChessAdmin/Houdini"] [PlyCount "119"] [EventDate "2012.??.??"] [TimeControl "3600+10"] {D41: Queen's Gambit Declined: Semi-Tarrasch with cxd5} 1. e4 c6 2. d4 d5 3. exd5 cxd5 4. c4 Nf6 5. Nc3 e6 6. Nf3 Be7 7. cxd5 Nxd5 {a position common to the Semi-Tarrasch defense has now been reached.} 8. Bc4 {Bd3 is more usual here.} O-O 9. O-O Nc6 (9... Nxc3 10. bxc3 Qc7 {is an interesting-looking alternative.}) 10. Re1 a6 {end of my personal book. Black has a number of other possibilities here, including ...Bf6.} 11. Bxd5 {Black should be able to better exploit the advantage of the two bishops after this, although the symmetrical pawn structure gives the position a drawish characteristic.} ({ Here's a more classic example of play:} 11. Bd3 Ncb4 12. Bb1 b5 13. a3 Nxc3 14. bxc3 Nd5 15. Qd3 Nf6 16. Bg5 g6 17. Ne5 Bb7 18. Qh3 Bd5 19. Ba2 Rc8 20. Bh6 Re8 21. Bxd5 Qxd5 22. Bg5 Nh5 23. Bd2 Nf6 24. a4 Ne4 25. Qd3 Nxd2 {Uhlmann,W (2555) -Pachman,L (2520) Manila 1976 1/2-1/2 (41)}) 11... exd5 {Black has the pair of bishops.} 12. Ne5 (12. Qb3 {appears the most threatening, but Black has} Bg4 { as an effective countermove.} 13. Qxd5 Bxf3 14. Qxf3 Qxd4 15. Be3 Qh4 16. Nd5 Bd6 17. g3 Qh3 18. Rad1 Be5 19. b3 Rfe8 20. Bf4 Bxf4 21. Nxf4 Rxe1+ 22. Rxe1 Qd7 23. Qd5 Qxd5 24. Nxd5 Kf8 25. Kg2 Re8 26. Rxe8+ Kxe8 27. Kf3 Kd7 28. Ke4 Kd6 {1/2-1/2 (28) Osnos,V-Krogius,N Budapest 1965}) (12. h3 {applies prophylaxis to the ...Bg4 idea, but then Black has time to shore up his position.} Be6 13. Bf4 Bf6 14. Qd2 b5 15. Rad1 Re8 16. Bg5 Nxd4 17. Nxd4 Bxg5 18. Qd3 Qd7 19. b4 Be7 20. a3 Rac8 21. Nce2 Bf6 22. Nb3 Bf5 23. Qg3 Bc2 24. Rc1 Bxb3 25. Rxc8 Qxc8 26. Qxb3 Qc4 {Moliboga,V-Lyrberg,P (2425) Duisburg 1992 0-1 (32)}) 12... Bf6 {first major decision point. Also considered Re8 or exchanging on e5, which are Houdini's top two choices.} (12... Nxe5 13. dxe5 d4 14. Ne4 Bf5 {is the most active option.}) (12... Re8 13. Nxc6 bxc6 14. Bf4 Bf5 {is similar to the game continuation, but with some additional breathing room for Black.}) 13. Nxc6 bxc6 {this pawn structure didn't bother me too much, since I didn't think White could exploit the weakness of the c-pawn and it reinforces the d5 pawn well. The a-pawn is also weakened now, however.} 14. Bf4 {taking advantage of Black's 12th move, since the bishop cannot be easily challenged.} Re8 {played with the idea of exchanging rooks, to take advantage of the fact that the White queen is the next piece in line, so White cannot immediately seize the e-file.} (14... Qb6 {is the more active-looking choice, with threats to the d4 and b2 pawns. White would now be able to penetrate with the bishop on d6, however.} 15. Bd6 Rd8 16. Bc5) (14... a5 {is an interesting idea which I considered at various points, prmarily with the thought of developing with ...Ba6. Houdini likes it because ofthe possibility of developing with ...Ra7-e7.}) (14... Bg5 $6 15. Bxg5 Qxg5 16. Re5) 15. Rxe8+ Qxe8 16. Qd2 Bf5 {I decided getting the bishop out was most important at this point.} 17. Be5 {a typical computer handicap move.} (17. Re1 {is what I had expected.} Qd7) 17... Bxe5 $15 18. Re1 $2 {at the time, I thought that this move pinning the bishop would amount to the same thing as the immediate recapture, with White losing a pawn.} (18. dxe5 Qxe5 19. Re1) 18... f6 { the other option for winning a pawn is ... Bxh2+. I chose this move instead because of the strong center and passed d-pawn that would result.} (18... Bf4 $1 {is the win seen by Houdini. I did actually look at this variation in my calculation of candidate moves, but not far enough.} 19. Rxe8+ Rxe8 $19 { at this point I saw that the Bf4 could not be taken due to the back-rank mate threat, but otherwise thought that the material was roughly equal and would not bring Black an advantage. However, Black can keep challenging the White queen based on the back-rank mate and gain a winning advantage. For example} 20. Qd1 Bc2 21. Qf1 Bd2 {and White cannot avoid the back-rank threat to his queen.}) 19. dxe5 $15 fxe5 20. Qg5 (20. Qe2 Rd8 $15) 20... Qe6 21. Na4 { the obvious plan to take advantage of the c5 hole and also pressure the a-pawn. } e4 22. Nc5 Qg6 {essentially forces an exchange, since Black would threaten .. .Bh3 after a retreat off the g-file.} 23. Qxg6 Bxg6 24. f3 {this marks the beginning of the endgame phase. I should have exchanged on f3, with a superior structure as a result.} Re8 (24... exf3 25. gxf3 Kf7 26. Nd7 $15 Rc8) 25. fxe4 $11 (25. Nxa6 $2 {would be too greedy.} e3 26. Nc5 d4 $19) 25... dxe4 { compared to the line in the variation, Black now has three weak pawns in the center and queenside and White will be able to stop the isolated passed e-pawn. } 26. Nxa6 {I had anticipated this and thought it simply helped Black's cause by breaking up White's pawns on the queenside in return.} Ra8 27. Nb4 c5 28. Nc6 Rxa2 29. Rf1 {threatening a back-rank mate after Ne7+} h6 30. Ne5 Be8 31. b3 Rb2 {Black threatens to win material: Rb2xb3} 32. Nc4 (32. Re1 Rxb3 33. Rxe4 Bb5 $15) 32... Rxb3 $17 {the position at first glance looks great for Black, but he will inevitably lose one of the pawns.} 33. Re1 Bb5 34. Rxe4 Rb4 35. Nd6 Rxe4 36. Nxe4 c4 37. Kf2 Kf8 38. Ke3 Ke7 (38... Bc6 {I find this idea later, but it should have been played sooner.} 39. Kd4 $17) 39. Kd4 (39. g3 {is how White could avoid losing the g2 pawn.} Bc6 40. Nf2 Bd5 $11) 39... Bc6 $17 40. Nd2 Bxg2 41. Nxc4 Kf6 42. Nd6 g5 43. Nc4 g4 {a premature advance of the unsupported g-pawn.} (43... h5 $5) 44. Ne5 $6 (44. Ke3 $5 $11) 44... Kf5 $17 { I had been hoping for this after the g4 push.} 45. Nf7 h5 46. Ne5 $2 {Houdini immediately sees a huge advantage for Black now.} (46. Ke3 Bc6 $17) 46... Kf4 ( 46... h4 47. Nd3 Bf1 48. Ne5 $19 Kf4 49. Ng6+ Kg5 50. Ne5 Kf5 {White's king now cannot move back to help stop the Black pawns without abandoning the knight, while any knight move will allow Black to push ...g3.}) 47. Nd3+ $2 { this would allow Black to make progress with the king.} (47. Ng6+ Kg5 48. Ne7 h4 $17) 47... Kg5 {here I became too conservative.} (47... Kf3 {is clearly better.}) 48. Ke3 (48. Ke5 Bb7 $15) 48... Bf1 {this limits the bishop's mobility too much.} 49. Nc5 Kh4 {what I had intended after the 47th move, although it is obvious that White is now in a much better position to defend.} 50. Kf4 Kh3 {this is now a draw.} (50... Bc4 51. Ne4 $15) 51. Ne4 $11 {one of Black's pawns will now fall in exchange for the h2 pawn. White can then sacrifice the knight for the other pawn if required.} Be2 52. Ng3 Bf3 53. Nxh5 Kxh2 54. Nf6 g3 55. Kxf3 g2 56. Ng4+ Kh1 57. Nf2+ Kh2 58. Ng4+ Kh1 59. Nf2+ { Twofold repetition} Kh2 60. Ng4+ 1/2-1/2

03 February 2013

Studying opening lines: how much is enough?

When creating and then evolving our opening repertoire, whether through use of a simple database or other system, we are faced with an issue that may at first seem simple, but is in fact complex: when to stop?  In other words, how much is "enough" to include, when we are compiling and studying individual opening lines?

Opinions vary greatly on whether opening study should be emphasized at all for players below the professional level, so getting into a philosophical debate over the question is probably not going to be of great use, nor would it resolve anything.  Rather, I'd like to offer some practical observations and guidelines that have helped me in making effective repertoire and study choices.  Regardless of the emphasis you may choose to place on opening study, everybody has to have something prepared for the opening phase of the game.

The primary objective for me out of the opening is to get a playable middlegame position, not to try to achieve a winning advantage.  This is not the most aggressive approach, but it's not a bad one in terms of results, either.  The goals of all my opening lines are therefore to: 1) avoid losing to forced tactics in the opening, and 2) reach a middlegame where I evaluate my chances are no worse than my opponent's.

Goal #1 (not to lose!) is of course the most important limiting factor.  Regardless of your opening choices, there will be at least a few (sometimes many) lines where the opponent can sacrifice material or undertake a provocative maneuver that gives them an early attack.  For example, if there is a tactical sideline with forcing threats that eventually ends up equal after 12 moves into the game, but you only know through move 7, your preparation is both incomplete and risky.  At the Class level, opponents are typically not "booked up" and will play interesting-looking or aggressive moves in the opening rather than "knowing" they ultimately do not work or end up in equality.  This forces us to be prepared not just for lines that theory considers best and popular, but for threatening sidelines.  Examples:
  • Slav Defense, Geller Gambit.  This pawn sacrifice normally gives White an initiative until around move 16-17 and Black can easily misstep if he does not understand the line.  You can see my simul against GM Alex Yermolinsky (Annotated Game #4) for an illustration of how the opening can be prepared (and how the endgame should not be played) in this line.
  • Budapest Gambit (1. d4 Nf6 2. c4 e5).  This whole opening is designed to be aggressive and trappy and is not easily dealt with by White.  In my tournament career, I've more than once been seated at a board next to a game where this gambit had been played and personally observed Black obtaining a won game in less than 10 moves, while White was completely bewildered by what had happened.
Goal #2 (reach an at least roughly equal position) is more flexible but also requires more judgment and evaluation.  With non-tactical sidelines or offbeat openings where there is no immediate threat from your opponent, sometimes just a basic idea of how to proceed can be enough, or perhaps knowing a sequence of 5-6 moves will set you up well for the middlegame after some additional standard developing moves.  In more complex cases, especially in whatever the main line is considered to be, there is more danger of your opponent (especially when you are Black) being able to emerge from the opening with at least a slight advantage.  Greater depth of move knowledge - and knowledge of why the moves are being made - is more critical here.  Examples:
  • 1. b4 (Sokolsky Opening).  The standard advice I've seen is simply to play a Queen's Indian Defense setup against it as Black.  This worked out well in the one game I've ever played against it.
  • Colle System.  I used to stop at move 3 with my preparation, which simply involved playing 3...Bf5 and then exchanging off the light-squared bishop.  This seemed to take most of the potential sting out of the system.  Following my game analysis here, my preparation has been extended out to move 5, for example as in Annotated Game #80, following the recognition from Annotated Games #75 and #78 that Black had to watch the b7-pawn more closely early on.
  • Caro-Kann, Advance Variation.  Most recently shown in Annotated Game #74, there are several early sidelines to know and the main line of the variation is considered the most challenging to Black.
The last big factor I take into account in determining "how much is enough?" is my ability to retain and understand opening lines - which, like everyone's, is limited.  This naturally affects overall repertoire choice, choice of lines and also how far we will study a line until it is considered "enough".  Take the Colle System example above: if I know it will be a more frequent choice of my opponents, then for practical purposes I will devote more time to looking at effective ideas for Black.  However, I am satisfied that I can reach an equal position with my current knowledge level.  So, in the absence of another motivating factor, I am able to safely concentrate on other openings.  I've also deliberately limited the amount of theory to know in the Caro-Kann Advance by choosing the 3...c5 sideline, but it is still a critical variation to know and understand, so requires more attention during opening preparation time.

I look for practical results from opening preparation more than anything else and this approach has worked well, to the point where I can (and should) concentrate on other aspects of my game.  It is also important to highlight that opening study methods can (and should) go far beyond identification and memorization of lines, if you want to truly improve your game.