Pitfalls of Computer Analysis

After having worked through and published over forty annotated games, in the process employing what I'd consider to be a simple and straightforward method of computer-assisted analysis, I've acquired an idea of some of the major pitfalls, as well as the benefits, of using chess engines.  In sharing the following observations, it's worth noting that they are offered from the point of view of an improving Class player, rather than that of a professional chessplayer or computer programmer, whose needs and approaches to using computers may well be different.

• Ironically in my view, one of the most pernicious pitfalls for the improving player is psychological: the sense of failure that can result after seeing computer analysis repeatedly demonstrate wins missed or ways that a loss could have been avoided.  How often do you see players comment about how stupid they feel or how bad their chess is following a game?  It's important to have the mental toughness to stop kicking yourself afterwards and truly learn from your mistakes, rather than waste mental energy on self-denigration.  Let the chess engine serve its function as a constructive tool to be used for improvement purposes.
• While outright blunders and forced wins are easily highlighted by engine analysis, the "best move" in a position is often not so easy to determine.  This may even be impossible, given broadly conflicting plans and ideas: see what super-GM Kramnik has to say about this.  In practical terms, engine evaluations that differ by 0.1 pawns are essentially equivalent and I would not sweat too much over a difference of up to 0.2 pawns.
• Only looking at a single visible line of analysis (the engine's top choice) can therefore be misleading when trying to understand a position; unfortunately, that is usually the default setting for a lot of software programs.  For analysis purposes, a player needs to have multiple move choices (lines) visible in order to better comprehend the potential of a position; having 3 lines visible is my preference.
• Trusting the initially displayed engine move evaluations can be counterproductive to analysis and misleading.  Engines need time to sort through possibilities in non-forcing situations and at least 1-2 minutes should be allowed when looking at a move using "infinite analysis" before proceeding.  Stepping through subsequent moves of key variations is also very important and may change the evaluation of the line, after the engine is able to calculate further ahead in it.
• Computers are of less practical help in objectively evaluating best opening play, due to the sheer number of possibilities and lack of forcing tactics leading to an advantage.  The best use of an engine in the opening phase is instead to search for and evaluate alternative moves that don't appear in, or are neglected by, opening theory; this is how the professionals use them to find theoretical novelties.
• Playing a move suggested by an engine is dangerous when you don't understand why it should be played.  This is one of the main reasons that blindly following engine recommendations in the opening can be counterproductive, even if they are objectively "best" from the engine's point of view.  Ending up in a middlegame position that you don't know how to play is often a quick route to defeat.
• In situations without forcing tactics, different engines may have significantly different assessments of the position.  It's therefore important to understand as best you can the particular biases and limitations of the engines you use, and/or use multiple ones for analysis.  Common biases include over-valuing material and mis-evaluating endgames in the absence of tablebases.  For example, having initially run through my past tournament games with Fritz and now using Houdini to assist in analyzing them, I've run across several instances of key positions where Fritz preferred a different move, while Houdini has supported my original evaluation and move choice.

In the end, the point of seriously analyzing your games (or anyone else's) is to achieve a greater understanding of them and through that, of the game of chess in general.  It's therefore wise to remember that the computer is there to support and asisst you, not to function as some sort of ultimate arbiter of what is correct and should always be played.

Annotated Game #43: New tournament; Caro-Kann Classical

After Annotated Game #42, about two years passed before I played in another tournament, due to my work and life circumstances.  However, this next tournament turned out rather better than the last one and I at least held my own overall.

This first-round game features the Caro-Kann Classical, with a relatively unchallenging sideline (7. Be2) chosen by White.  Like most variations chosen by Black in the Caro-Kann, the Classical Variation is solid rather than unbalancing, so deviations by White from "book" play allow Black to more easily reach equality, rather than offering the chance at an advantage.  Black is assessed by the engines as equal on move 12 and by move 14 I would say has successfully taken over the initiative, along with having an advantage in piece coordination.  Some subtle inaccuracies in piece placement (18...Nb6) and then choosing to dominate the wrong file means that Black is unable to turn his initiative into anything concrete, although he is certainly no worse.  The Bishop vs. Knight ending that occurs after a series of exchanges illustrates a typical Caro-Kann piece imbalance, where Black's knight and pawn placement are sufficient to contain White's bishop.

Games that lack a lot of fireworks can still be useful (perhaps sometimes more useful) to draw lessons from. In this case, the analysis shows where Black could have better placed his pieces in the early middlegame, specifically the queen's knight and the doubled rooks, something which will better inform my future play.  This is also a relatively rare example of a use of the ...e5 break in the Classical variation, where ...c5 is more usual, and is a good illustration of how it can be set up and employed.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "Class B"] [Black "ChessAdmin"] [Result "1/2-1/2"] [ECO "B18"] [Annotator "ChessAdmin/Fritz/Houdini"] [PlyCount "85"] {B18: Classical Caro-Kann: 4...Bf5 sidelines} 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Bf5 5. Ng3 Bg6 6. Nf3 Nf6 7. Be2 {a passive continuation} e6 8. O-O Bd6 { normally Black plays Be7 in this variation, but here is able to seize the h2-b8 diagonal and preempt Bf4.} 9. Re1 O-O 10. Nh4 {now out of the database. This allows White to eliminate Black's light-square bishop without in turn trading his off (as is usually done on d3). However, this is also a loss of time for White and Black continues developing.} Nbd7 11. Nxg6 hxg6 12. Bf3 { Although White still has this bishop, it's doing very little, given Black's dominating pawn structure on the light squares.} Qc7 $11 13. Nf1 Rfe8 14. g3 { evidently the point of Nf1.} e5 {this and the ...c5 break are the two main freeing moves for Black in this variation. Here the piece configuration means that ...e5 will be more effective.} 15. Be3 exd4 16. Bxd4 Be5 {Black is fine with trading bishops, since White's dark-square bishop is more active overall and particularly threatening on d4.} 17. Ne3 Rad8 {effectively bringing another major piece into play and creating latent threats along the d-file.} 18. Bg2 Nb6 (18... Nc5 {is a more active placement for the knight and possible due to the discovered pin on the Bd4.}) 19. Bxe5 {breaks the pin by force, due to the simultaneous attack on the Qc7.} Qxe5 20. Qc1 Qc5 21. c4 Re7 (21... Nbd7 {with the idea of repositioning it on e5 is better, as the knight while on b6 is essentially out of play.}) 22. Qc3 Rde8 {dominating the d-file instead would seem to be of more use.} (22... Red7) 23. b4 Qh5 24. a4 {this invites Black to do the effective repositioning move ...Nbd7, which he however passes up.} Ng4 25. Nxg4 Qxg4 {now Black instead has exchanged off his more effective minor piece.} 26. Rxe7 Rxe7 27. Bf1 Nd7 {now it's clear Black needs to do something with the knight.} 28. Re1 Rxe1 29. Qxe1 Nf6 30. Qe3 b6 31. Qe7 Qd7 32. Qxd7 Nxd7 {we now enter a drawn minor piece ending. White's bishop has no prospect of being able to invade Black's camp to attack his pawns.} 33. f4 Nf6 (33... c5 {would be simpler, leaving the knight on d7 while there is still tension between pawns on the queenside and not giving White any potential openings there.}) 34. Bg2 c5 35. bxc5 bxc5 {Black now has to watch the queenside carefully, although it's still a draw.} 36. Kf2 Kf8 37. Ke3 { inviting the knight check on g4.} Ke7 {winning the h2 pawn didn't seem worth the risk of allowing White's king to penetrate my position, although according to Houdini White couldn't get too aggressive.} (37... Ng4+ 38. Kf3 {Houdini 1. 5a w32:} (38. Ke4 f5+ 39. Kd5 Ne3+ 40. Kxc5 Nxg2 41. Kd6 Ne3 42. c5 Ke8 43. Kc7 Nc4 44. c6 a5 45. Kb8 Nd6 46. Kc7 Ke7 $19) 38... Nxh2+ 39. Ke4 Ng4 40. Bf3 Nf6+ 41. Ke5 Ke7 42. a5 Kd7 43. Bd1 Kc6 44. Bf3+ Kc7 $11) 38. h3 Kd6 39. g4 Nd7 40. Bd5 f6 41. Bf7 Nf8 42. h4 a5 43. h5 {and White will not be able to make any progress against Black's pawns on the dark squares.} 1/2-1/2

Annotated Game #42: The Punishing Slav

The following game occurred in the last round of this particular tournament and was a much-needed win.  Even with the victory, I ended up at -1 (win-loss) for the tournament along with three draws, all of which were with significantly lower-rated players.  Not a terrible overall result, but one that served to underline the declining trend in my chess performance, primarily due to the lack of systematic practice between tournaments and only playing seriously around once a year.  This was not enough exposure to chess to either improve my skills or keep my mind in good practical shape for OTB play.

Nevertheless, this game is a good example of what I was capable of when warmed up.  White with 5. d5 enters a falsely seductive line against the Slav, one which you won't find in any opening books, for good reason.  Black correctly reacts aggressively and soon regains the gambit pawn, along with forcing the win of the exchange (although alternative play on move 9 is objectively better).  Despite Black's material advantage, White has good attacking possibilities and some compensating factors, leading to a tactical tussle.  White leaves his remaining rook out of the fight - a classic amateur mistake - and then Black is able to force enough simplifications to segue into a clearly won endgame.

The Slav has a deserved reputation as a solid defense, but it can also be a punishing one when White does not respect it, as this game shows.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "Class B"] [Black "ChessAdmin"] [Result "0-1"] [ECO "D20"] [Annotator "ChessAdmin/Fritz/Houdini"] [PlyCount "82"] [EventDate "2003.??.??"] [TimeControl "240+2"] {D10: Slav Defence: cxd5 (without early Nf3) and 3 Nc3} 1. d4 d5 2. c4 c6 3. Nc3 dxc4 {an independent move that tests White's setup with Nc3.} 4. e4 e5 5. d5 {an aggressive-looking move which however can land White in trouble quickly. } Nf6 6. Bxc4 {this is the basic idea by White, to quickly regain the gambit pawn with what appears to be a strong point at d5.} b5 {this counterattacking move is considered best by Houdini and has a high scoring percentage in the database.} (6... Bb4 {is the other popular choice here.}) 7. Bb3 b4 {the necessary follow-up to the previous move, pushing away the defender of e4.} 8. Nb1 {a poor third choice for a knight retreat. Two example of games that went differently:} (8. Nce2 Nxe4 9. Nf3 Bc5 10. Be3 Bxe3 11. fxe3 Nc5 12. Rc1 Nxb3 13. Qxb3 cxd5 14. Qxb4 Nd7 15. Nc3 a5 16. Qa3 Qe7 17. Qxe7+ Kxe7 18. Nxd5+ Kf8 19. Rc7 f6 20. Nd2 Rb8 21. Ne4 Kf7 22. Nc5 Rxb2 {Szollosi, L-Garcia Palermo,C/ Berlin West 1984/MCD/1/2-1/2 (51)}) (8. Na4 Nxe4 9. Qe2 Nf6 10. dxc6 Bd6 11. Nf3 O-O 12. O-O Ba6 13. Bc4 Bxc4 14. Qxc4 Qc7 15. Bg5 Qxc6 16. Qb3 Nbd7 17. Rac1 Qa6 18. Bd2 Qb5 19. Rc4 Rac8 20. Rfc1 Rxc4 21. Rxc4 Rb8 22. Ng5 Qd5 { Sabolik,F-Kijac,M/SVK 2000/EXT 2001/1/2-1/2 (55)}) 8... Nxe4 9. Nf3 {now out of the database.} Bc5 {this aims for the following line, in which Black wins the exchange. Houdini sees nothing better for White than entering it.} (9... Ba6 $1 {immediately would instead prevent White from castling and the bishop's dominance on the f1-a6 diagonal would prove difficult to challenge.}) 10. O-O Ba6 {forcing the win of the exchange, as White cannot interpose on the diagonal and Re1 fails to ...Bxf2+} 11. Nbd2 Nxd2 {necessary, otherwise the Ne4 is hanging.} (11... Bxf1 $2 {leads to nothing, notes Fritz.} 12. Nxe4 Bxg2 13. Kxg2 $18 {and White has a big attack coming while Black is underdeveloped.} (13. Nxc5 $6 Bxf3 14. Qxf3 Qf6)) 12. Bxd2 Bxf1 {now that Nxe4 is no longer possible for White, Black can take the rook.} 13. Qxf1 O-O {White has compensation for the exchange, notably in his advantage in development and in possessing the two bishops. He also has the potential for an attack on Black's king. Houdini rates the position as equal.} (13... cxd5 {is inadvisable.} 14. Qb5+ Nd7 15. Nxe5 {and Black will lose the d-pawn and have his king exposed in the center.}) 14. Qc4 (14. Rc1 {would activate the rook and immediately increase the pressure on Black. The text-move looks dangerous, but White in effect chooses to conduct the attack while not having his rook in play.}) 14... Qb6 (14... Qd6 {would be more solid.}) 15. Nxe5 Bxf2+ 16. Kh1 Qd4 {this defensive move is considered best by Houdini.} (16... a5 {was Fritz's choice, which Houdini considers significantly better for White.} 17. dxc6 Ra7 18. Rf1) 17. Nxf7 {this doesn't turn out to be good for White.} (17. Bxb4 c5 18. Bc3 Qxc4 19. Bxc4 {is considered equal by Houdini. White is still down the exchange, but the bishop pair and strong positions of his minor pieces compensate.}) 17... Qxc4 {leads to a winning advantage for Black.} (17... Rxf7 18. dxc6 Qxc4 19. Bxc4 Nxc6 20. Rf1 Ne5 21. Bxf7+ Kxf7 22. Rxf2+ Kg8 {draws.}) 18. Bxc4 Kxf7 $1 $19 ({Inferior is} 18... Rxf7 19. dxc6 Bh4 20. c7 $15 { says Fritz.}) 19. dxc6+ Ke8 20. Bxb4 (20. Rc1 a5 21. Bb5 Ke7 22. Bg5+ Ke6 { was White's best chance, keeping Black under greater pressure.}) 20... Rf4 ( 20... Nxc6 {is clearly inferior, notes Fritz.} 21. Bxf8 Ne5 22. Rf1 Nxc4 23. Rxf2 $16) 21. b3 $6 (21. Rc1) 21... Rxc4 {Black makes the decision to simplify into a won endgame with a large (piece for pawn) material advantage.} (21... Nxc6 $5 {is considered better by the engines.} 22. Bd2 Rd4 23. Bc3 $19 Rd6) 22. bxc4 Nxc6 {now White's possible queening and mating threats have disappeared.} 23. Ba3 Rd8 24. Bb2 Bd4 25. Re1+ {sidestepping further trade of material.} Kf7 26. Rf1+ Kg6 27. Ba3 Re8 {Black's control of the f2 square and the e-file prevents White from being able to stop Black penetrating on his second rank.} 28. Bd6 Re2 29. a3 Ra2 30. h4 Ra1 {Black opts for further forced simplification.} 31. Rxa1 Bxa1 32. g4 Be5 33. Bxe5 {this hastens White's demise, but there was little he could do at this point anyway.} Nxe5 34. c5 Kf6 35. Kg2 Nxg4 36. Kf3 Ne5+ 37. Ke4 Ke6 {now Black's king is centralized and in tandem with the knight can neutralize counterplay from White, then go after White's weaknesses.} 38. Kd4 h6 {done to win in the simplest way possible, without allowing any hope for White.} (38... Nf3+ {makes it even easier for Black, comments Fritz.}) 39. Ke4 g5 40. h5 (40. hxg5 hxg5 41. Ke3 Nc4+ 42. Kf3 Nxa3 $19) 40... a5 41. a4 Nc6 {reaching a zugzwang, where White is forced to let Black penetrate.} (41... Nc6 42. Kf3 Kd5 $19 {and if White tries to queen a pawn, then} 43. Kg4 Kxc5 44. Kf5 Kd5 45. Kg6 g4 46. Kxh6 g3 47. Kg6 g2 48. h6 g1=Q+) 0-1

Tournaments: The Joy of Battle

This was conceived in response to some recent posts in the chess blogosphere about players retiring from the OTB tournament field of battle.  I'm very much a supporter of people doing what they themselves are most interested in doing; so for those who don't wish to participate in tournaments, may they thoroughly enjoy what they do in the chess world.

This post, however, is for the tournament warriors - past, present and future.

* * * * * * * * *

Walking into the venue for the first time.  (If it's a hotel, checking in with a sense of purpose.)  Perhaps other players are there in the entryway, sporting various types of chess gear.  A sense of anticipation is palpable.

(Unpacking in the hotel room.  Laptop, travel set and a book or two are carefully placed on the table.  The Endgame Clothing wear is separated from the rest of the clothes and hung with care.)

Supplies must be gathered for the upcoming campaign.  Juice drinks for careful consumption during the games, to ward off decision fatigue.  Snacks for eating during a very long game, or if there is not enough time between rounds to go for a full meal.  Warriors fight on their stomachs, as well as their brains.

The tournament gear is packed and at hand.  Set, board and clock are in their carrying case, ready to be drawn like swords from their scabbards.

(Shall it be the stairs or the elevator, to reach the tournament hall?  Taking the stairs, the body is warmed up and the blood flows for the coming struggle.  In the elevator, the mind is calm while watching others nervously or excitedly enter on the trip down.)

Carefully pushing through the throngs around the pairing sheets and wall charts.  Letting the most impatient clear the way, then moving softly, like a cat, to see who your next opponent is.  Recording the opponent's name, color choice, and board number on the scoresheet, which is otherwise blank - although not for long.

Heading for the tournament hall, feet treading the ground with a sense of purpose.  An impression of organized chaos, as players find their places and the round comes closer to starting.  Your board - the most special of them all, it seems - is waiting.  Feeling the thrill of getting in the ring with a higher-rated player, or a sense of confidence tinged with healthy respect for a lower-rated opponent.  

The joy of seeing the battle open as the first move is played.  Strategic choices are made and the duelists start to feel out each others' defenses.  Will it be a furious, sharp encounter?  Or a subtle exchange of thrusts and parries, slowly building to its conclusion?  The game is there to be created, its rhythm constantly ebbing and flowing.

The outwardly restrained jubilation of a well-deserved win.  The relief of a not-so-well-deserved win.  The acknowledgement of a draw.  The resignation of a loss.  All possible, all accepted beforehand by the warrior, who nonetheless strives for victory.  Respect for an opponent's efforts shown, as the field of battle is departed to mark the results.  Seeing how the competition stands and weighing your chances.  Resolving to play well and fight hard the next round.

Packing at the end of the campaign.  Reflecting on the results and lessons.  Eager to get back in the fight, after a well-deserved rest.  Thinking of preparations for the next campaign... 

Annotated Game #41: Stymied in an English-KID

This next tournament game is illustrative of how certain apparently subtle decisions can turn a potentially winning positional advantage into a draw.  My opponent transposes into a King's Indian Defense (KID) setup against the English, but from moves 8-11 succeeds neither in disrupting White's plans for queenside expansion, nor in establishing any kingside counterplay of his own, giving White an advantage.

White's move 12 indicates that he is looking to open up the a-file with pawn play, rather than fix Black's weaknesses and exploit them with his knight.  With some help from Black, White executes an advantageous piece trade (knight for bishop) on e6, but then errs by also trading his dark-square bishop for Black's. After seizing the a-file with the help of his monster Bg2, White is unable to exploit it after Black redeploys his forces to prevent further penetrations.

I've previously run across this problem in the English, where I'm able to achieve a large queenside space advantage, but then lack the wherewithal to make any further progress.  The improvements and other potential avenues of play found in analysis (particularly on moves 12 and 19) should help me overcome this problem in the future.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Class B"] [Result "1/2-1/2"] [ECO "A26"] [Annotator "ChessAdmin/Fritz/Houdini"] [PlyCount "49"] [EventDate "2003.??.??"] [TimeControl "240+2"] {A26: English Opening vs King's Indian with ...Nc6 and d3} 1. c4 Nf6 2. Nc3 g6 3. g3 Bg7 4. Bg2 O-O 5. Nf3 d6 {Black elects to go to a KID setup. Another option would be to transpose to a Symmetrical English with ...c5} 6. O-O e5 7. d3 Nc6 8. Rb1 Nh5 (8... a5 {restraining White's b4 is the most popular and effective move here, according to the database.}) 9. b4 Ne7 (9... f5 {is the logical follow-up to Black's previous move and is by far the most played here.} ) 10. Qb3 {now ...f5 would expose Black's king on the a2-g8 diagonal.} c6 { preparing to fight for d5.} 11. a4 {this prepares the b5 push and allows development of Ba3 without getting in the way of the a-pawn.} (11. b5 { immediately is the other main option, which would result in a different queenside pawn structure.}) 11... a6 {this does not prevent White from further advancing his queenside pawns and also leaves a hole on b6. Black at this point needed to get more pieces in play rather than make additional pawn moves. } 12. b5 (12. a5 $5 $14 {would allow White to take better advantage of Black's pawn structure deficiencies and the hole on b6, also freeing a4 for the knight. }) 12... axb5 13. axb5 Be6 (13... h6 {would prevent a future Ng5, allowing Black to develop Be6 without it being challenged.}) 14. Ba3 (14. Ng5 {is preferred by the engines in order to force the bishop off of its valuable diagonal.}) 14... c5 (14... h6) 15. Ng5 {White finally takes advantage of the undefended g5 square, along with a discovered attack on b7 by the Bg2.} Rb8 ( 15... Bc8 {is the best defense, preserving the bishop for future use.}) 16. Bc1 {played in case of ...Bh6 by Black. However, taking immediately on e6 was better.} b6 {again passing up the chance to retreat the bishop.} 17. Nxe6 fxe6 $16 {now the Bg2 is an unchallenged monster and White has excellent prospects on the a-file.} 18. Bg5 Bf6 (18... h6 {immediately challenging the bishop was better, as after} 19. Bd2 {White appears to have gained little, although Black's kingside is slightly weakened as a result.}) 19. Bxf6 {this gives up part of White's positional advantage.} (19. Bd2 {retreating the bishop was best, as White retains much more scope for his dark-square bishop than does Black. Also, Black's Bf6 blocks the activity of the Rf8 and takes away a retreat square from the Nh5.} Bg7 $16) 19... Rxf6 (19... Nxf6 {gets the knight back into the game and doesn't leave the rook on an awkward square. Evidently with the text move, Black had thought he might try doubling heavy pieces on the f-file at some point.}) 20. Ra1 {time to seize the open file, with Black unable to challenge on a8 thanks to the Bg2.} Ng7 (20... Rf7 21. Ra7 $16) 21. Ra6 {while this isn't bad, compare with} (21. Ra7 Nc8 22. Ra2 Ne7 {preventing the Bc6 penetration} 23. Rfa1) 21... Ne8 22. Rfa1 Nc7 23. Ra7 Nc8 24. R7a2 Ne7 25. Ra7 {unable to find a way to make progress, White takes the draw.} (25. Ne4 $5 {is what the engines prefer.} Rf7 26. Qb2 $16 Nf5 27. e3 {is one possible continuation and White has a strong positional plus, but it's still not clear how he can break through Black's defenses.}) 1/2-1/2

Annotated Game #40: Be careful what you wish for

This game in the 5. Qf3 sideline of the Caro-Kann Classical is most notable for how Black repeatedly did not take advantage of a number of opportunities presented by White in the late opening and middlegame phases.  However, at this point (round 6 of the tournament) I was hoping to stabilize my performance and was clearly looking for a draw.  Be careful what you wish for, since you just might get it.

Before my most recent chess training period (starting with the establishment of this blog), I had relatively little notion of the importance of active piece play.  Starting on move 9, by which point Black had more than equalized, I pass up several chances to improve my piece activity and create multiple threats.  I like to think that now I would recognize at least some of the better candidate moves, given a better understanding of the importance of piece play and how to widen the move selection process.

In the end, the draw wasn't a bad result, but it illustrates another psychological trap - for Class players especially - which is to settle for a result that is less than the position merits, either through mis-evaluating the position or from lack of a winning desire.  In the long run, for the improving player I believe that realizing potential wins is just as important as avoiding losses.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "Class B"] [Black "ChessAdmin"] [Result "1/2-1/2"] [ECO "B18"] [Annotator "ChessAdmin/Fritz/Houdini"] [PlyCount "53"] [EventDate "2003.??.??"] [TimeControl "240+2"] {B18: Classical Caro-Kann: 4...Bf5 sidelines} 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Bf5 5. Qf3 e6 {this is an immediate and effective antidote to White's queen sortie.} 6. c3 Nd7 7. Bd3 Bxe4 {Black exchanges, believing that White's light-square bishop is not going to pose much of a threat, given the pawn structure.} (7... Bg6 {is the other option for Black.}) 8. Bxe4 Ngf6 9. Bg5 Be7 (9... Qb6 {is preferred by the engines. Normally I don't like early queen sorties, but this one makes sense, given the Nf6-Nd7 setup and the chance to inflict some structural damage on White. One possible continuation:} 10. b3 Qa5 11. Bd2 Nxe4 12. Qxe4 Nf6) 10. Bd3 h6 (10... O-O {is a necessary move anyway and also less committal.}) 11. Bh4 O-O 12. Bg3 Qb6 13. Qe2 Rad8 14. Nf3 c5 { this is the classic pawn break in the Caro-Kann Classical. Black gets more scope for his pieces and challenges White in the center.} 15. h3 {not sure what White was protecting against, as Ng4 does nothing but hit air.} (15. O-O $5 $11) 15... cxd4 $15 16. Nxd4 Nc5 17. Bb1 {this retreat doesn't make much sense, among other things locking in White's rook on a1.} (17. Rd1) 17... Rd7 ( 17... Na4 {is immediately found by the engines and creates multiple threats, while tying White to the defense of b2.} 18. Nb3 Qa6 19. Qc2 (19. Qxa6 bxa6 20. O-O Nxb2 $17) 19... Bd6 20. Bxd6 Rxd6 $17) 18. O-O Rfd8 19. f3 {while this controls e4, it also opens up the key g1-a7 diagonal and leaves g3 unprotected. } (19. Be5 a5 $11) 19... Nd5 (19... Nh5 {instead immediately exploits the hole on g3.}) 20. Kh1 Bf6 21. Bf2 Nf4 {the knight hops into an even better square} 22. Qc2 (22. Qc4 Bxd4 23. cxd4 Qa6 24. Qxa6 Nxa6 $15) 22... Ncd3 (22... Nfd3 { is the correct knight choice.} 23. Nb3 Nxf2+ 24. Qxf2 Nxb3 25. Qxb6 axb6 26. axb3 Rd2 27. Be4 (27. Ra2 $2 Rd1 28. Kg1 Rxf1+ 29. Kxf1 Rd1+) 27... Rxb2) 23. Nb3 (23. Be3 $5 $17 {was possible in this line, as the Nf4 will now be left hanging by any move by the Nd3.}) 23... Nxf2+ 24. Qxf2 Qxf2 25. Rxf2 Rd1+ 26. Kh2 Kf8 (26... Nh5 $1 {threatening Be5} 27. g3 g5 {controlling f4} 28. Bc2 Rxa1 29. Nxa1 Be5 $17) 27. Bc2 {and I take a draw in a slightly advantageous position.} 1/2-1/2

Tournament Preparation: Mental Toughness

This companion post to Tournament Preparation: Chess Skills discusses some techniques that players can use to mentally prepare themselves for success at a tournament.  Most importantly in practical terms, mental toughness and improved focus will enable a player to better leverage their existing chess skills, which in turn will lead to a higher performance level.  Furthermore, good mental preparation will help form a bulwark against negative thinking and results, while also helping one recover more quickly from a bad outcome.

Below is a set of conscious decisions that can be made, or attitudes that can be adopted, regarding your chess game.  Using these techniques has assisted me in combating negative and unproductive modes of thinking.  In fact, I credit them as being a key part of my improved chess performance since this blog was started, which has included beating the strongest opponent (2100+) faced in my chess career.  I believe that the techniques listed here will all contribute to a player becoming mentally tougher at the board, as they address (directly or indirectly) a number of the unhelpful fears, anxieties and fantasies we may have about our tournament performance.

• Treat each game as an individual chance to excel at an activity you enjoy, while accepting the fact that you may lose.  If you lose, take away lessons from it both in terms of chess skill and mental preparation.  Do not waste energy on excuses, whether or not they are justified.  If you win, enjoy your victory but be sure to identify what your opponent did to allow it, since you did not win the game alone.  In this way, each game will become another stepping stone on your path to mastery.
• Consider every opponent as worthy, but able to be defeated.  Do not worry about your opponent's rating during the game, as this becomes a source of fear and loathing.
• Go into a game with the objective of playing well and looking to win, rather than to play perfectly; no one plays perfectly, so do not bother to attempt it.  Having an understandable winning plan or spotting a crucial tactic is what is needed for victory, not what your computer engine says afterwards is the best move.  (This is also a key underlying premise of Chess for Tigers.)
• Resolve to play what the position demands.  If your opponent's position is vulnerable to an attack, then go on the offensive; do not play passively.  If your opponent's attack will end up coming first, or your attack has failed and and a counterattack is imminent, find the mental coolness necessary to defend well.  It requires mental courage for both attack and defense.
• Do not deliberately aim for a draw from the start of a game, regardless of your opponent's rating or your tournament standing.  If you play your best and press any advantages you are able to obtain, you are more likely to achieve what you need and may in fact win.
• Resolve not to offer a draw to your opponent unless the position on the board is in fact completely drawn.  This will contribute to a winning mindset and to not being afraid to play out any position.
• Do not think about the probable future result of the game while it is in progress; think about what you (and your opponent) can do with the position on the board.  There is a big difference between what should be a winning position and what is actually a won position; do not confuse the two.
• Do not think about your current or future tournament standing during the game.  The position in front of you is not affected by your current score or by how many future points you may fantasize about winning.
• Consciously accept your overall limitations while playing in the tournament, so that you are not distracted by your perceived failings as a chessplayer; address them later as part of your long-term training program.  During a tournament, you will not be able to remember by rote all of your opening lines; this is normal and will not make you ignorant and helpless.  You are also unlikely to make major new breakthroughs in middlegame or endgame knowledge while the tournament is in progress; instead, recognize that what you know gives you enough skills to play well and win against comparable opposition.

These practices can of course be difficult to adhere to, especially the ones where you resolve to not think or do something. The point here is to have a conscious goal of ignoring certain things which will only serve to distract you from the central activity of playing chess to the best of your ability.  When these distractions do occur, with your additional mental toughness you can identify them and then mentally set them aside, thereby returning to a more productive thought process that is focused on the position in front of you.  Again, perfection in thought is not the goal, just as perfection in play is not attainable.  However, an end result of more productive behavior and thinking at the chessboard (or computer, for online tournaments) is quite attainable.

There are also broader practices involving mental strengthening and stimulation which can be applied to chess.  I find especially helpful ones which help calm the mind and allow you to perceive situations more objectively, which improves your analysis and judgment at the board.  One should follow the dictates of the position, whether it tells you to attack or defend.  It is a great deal easier to hear what it is telling you when your mind is not generating extra noise.  For more on this, along with some of my own observations on cross-training and the Kung Fu of chess, it's worth highlighting that GM Nigel Davies has made some parallel observations on his Chess Improver blog.

Tournament Preparation: Chess Skills

After long and sometimes hard experience, I've come to the conclusion that the most effective pre-tournament preparation consists of sharpening and focusing what you are (or should be) doing for your longer-term training efforts.  This contrasts with the more common pre-tournament routine in which over the space of a week or two (at best) or a couple days (more often) players mostly spend time on openings and doing tactical drills, then put everything aside until the next tournament.

This type of staccato and rushed approach resulted in little success for me as an adult player.  During the scholastic phase of my career, I played in tournaments quite often, so without really trying I had constant exposure to new chess concepts and practical lessons, even though my (self-taught) training was not systematic.  There's a lot to be said for simply playing a lot of longer time control games, which looking back on it now was probably my best chess improvement practice.

Now with only sporadic participation in tournaments and a longer-term goal of improving my overall game, I've had to come up with something different for a training regimen, which has also meant a revised approach to pre-tournament prep work.  The primary principle I use both for long-term training and in preparing for tournaments will be a familiar one to many: train the way you fight, then fight the way you train.  This means that any training method used should accurately reflect, at least in part, the tournament game experience.  Conversely, it also means that when in a tournament game, a player should rely on their training when making decisions, rather than impulsively "winging it" when faced with an unclear situation.  One common example of this phenomenon is choosing to abandon your opening preparation when faced with a particular opponent.  This typically occurs when there is a large ratings gap and a player feels that their openings aren't good enough (if the opponent is higher-rated) or that the opponent (if lower-rated) can be easily beaten in an unfamiliar line.

Although I've codified things in a checklist format below and have indicated my own particular preferences, I want to clearly distinguish between the "what" and the "how" of tournament preparation.  This means that it is more important for a player to train the various skill sets before a tournament, rather than how exactly they go about it.  So within each category there will naturally be a number of options available regarding the materials and tools to be used for training.  Methods can also vary greatly from player to player; the below is simply what has proven to work best for me.

Pre-Tournament Checklist

1. Tactical exercises.  Perform at least one set of exercises for 10-15 minutes a day, or multiple sets if desired, with breaks in between them.  Important: the exercises' objectives should be randomized.  This means avoiding sets of "mate in 3" problems or the like.  Your thought process will benefit most from having to figure out the best move from an original position, with no hints, as this is the reality of a tournament game situation.  Limiting the time spent on individual sessions helps keep the mind active and engaged throughout the process (a state of "mindfulness"), aiding in the longer-term retention of tactical concepts and patterns.  Finally, focus on relatively simple (up to three-move) combinations and motifs, as this is what you will see the overwhelming majority of the time during actual play.  I primarily use the Chess Tactics Server, since it meets all of the above criteria.


2. Training games.  Play at least one slow (60 5 or 45 45 or higher time control) game per week against opposition of generally comparable strength.  Tournament rules should apply (game clock, touch move, no takebacks allowed).  Colors should be varied, either randomized each time or generally alternating over several games (no more than two Whites in a row, for example).  Early resignation is not allowed, except for clearly losing circumstances (loss of a major piece, or a minor piece with no compensation whatsoever).  I use Chessmaster: Grandmaster Edition for my computer opponents, playing in rated game mode and always playing the recommended next opponent after a game, a practice which adjusts their playing strength according to your performance.  The Chessmaster opponents' Elo ratings are not always good reflections of their practical performance, but I've found that my own rating generated by the program is remarkably accurate.


3. Opening preparation.  Critically review all of your opening repertoire lines, concentrating on: a) the variations most likely to be played by your upcoming opposition; b) the most theoretically critical ones; and c) the most dangerous ones if you don't know/remember the theory.  All of the above are reviewed using chess books and articles (including videos), as well as database searches, so that ideas can be studied and not just lists of moves.  Taking the Caro-Kann variations as an example, I focus on the Main Line and the Advance Variation, with some attention paid to the Panov-Botvinnik Attack and the Exchange Variation (relatively common at the Class level), and finally the Fantasy Variation (as a potentially dangerous line).  All the others get at least one run-through in my repertoire database so that I can be reminded of key ideas and concepts of play.


4. "Big think" activities.  While I'd consider the above three items to be fundamental, there's also an important role to be played by other activities that will energize your chess mind for the tournament experience.  Going over master-level annotated games (preferably ones played by the author) is one example, as you are thereby directly exposed to winning concepts and effective thought processes via the game commentary.  Other examples would include tackling middlegame and endgame books in a serious frame of mind, with the goal being to work through one particular book before the tournament.  The lessons will then be fresh in your mind and you are more likely to see opportunities to immediately apply them in your games.

Ideally an improving player is already doing all of the above types of things as part of a long-term training program.  In that case, preparing for a tournament becomes more a question of choosing to increase the focus on certain priority items over a shorter time period.

On a related note, I believe that a balanced approach to training chess skills will be the most productive for achieving good tournament results, rather than concentrating solely on one aspect such as tactical training or opening preparation, even if someone only has a short time period to put in the effort.  This is because successful tournament play demands a broad range of skills and practical preparation, rather than just knowledge of chess theory or performing well in rote drills.  As part of a balanced approach to pre-tournament preparation, it's also worth mentioning that players can choose to undertake certain mental preparations, which can be just as critical to one's tournament results as pure chess skill.  This is the subject of a companion post on mental toughness.

Finally, as with training in general, I believe people can and should have different approaches for tournament preparation, based on their learning styles and life circumstances.  One of the things I appreciate about the chess improvement community is the opportunity that it offers to better understand and then compare/contrast others' best practices, so I can incorporate them (or modified versions of them) into my own program.  So new ideas - at least ones that are new to me - and informed commentary about tournament preparation are always welcome.

EDIT: See also Tournament prep is less about the chess, more about you 

Annotated Game #39: Rook Endings Are Always Drawn

The next tournament game, following on the heels of my most instructive loss, is typical of my play during that period of my chess career.  The opening is an English Four Knights with a slightly advantageous, or at least comfortable, position for White once the early middlegame is reached.  I don't have a particularly good idea of where to place my pieces, for example the bishop on move 12, the rook on move 14 and the queen on move 19.  These sorts of small positional errors accumulate and when White follows a skewed plan to push e4 using the f3 pawn as a support, tactics as a result appear for Black, who emerges up a pawn.

By the end of the exchanging sequence on move 26, we have a double rook endgame where Black has a central passed pawn and another one easily created on the queenside.  However, White is not lost, as his king is closer to the action and his rooks are active.  These sorts of endgames are notoriously difficult for the side with a material advantage to win; the move required for Black to win on move 35 would have taken a great deal of accurate calculation and nerves to play at the board.  White is able to exchange a pair of rooks and eliminate the central and queenside pawns in exchange for Black picking up two kingside pawns, which however means that Black's pawns will not be able to breach White's defenses.  By move 41 the draw has taken shape, with White's rook in an ideal position to prevent Black's king from getting into the fight.  By move 51 the draw is concrete, although it takes another 30 moves for my stubborn opponent to concede the fact (amusingly with both of us missing a threefold position repetition as well).

The main lessons I drew from the analysis of this game were:

• Knowing an opening variation well is not enough, as some idea of the early middlegame requirements for the position is necessary to maintain momentum and accuracy of play.
• Many times it is sufficient simply to follow a plan of improving the position of each of your pieces in order to increase their activity.  This is especially a good rule of thumb for when you know the opening well but lack experience with the resulting middlegame.
• The defender in a rook endgame should never despair when he has active pieces that can get behind the enemy pawns and harass the king.  After all, all rook endgames are drawn, as the saying goes...

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Class B"] [Result "1/2-1/2"] [ECO "A28"] [Annotator "ChessAdmin/Fritz/Houdini"] [PlyCount "161"] [EventDate "2003.??.??"] [TimeControl "240+2"] {A28: English Opening: Four Knights Variation} 1. c4 e5 2. Nc3 Nf6 3. Nf3 Nc6 4. e3 d5 5. cxd5 Nxd5 6. Bb5 {this is the key move in the variation, which changes the game from being a reversed Sicilian, and an important reason why I play the variation with 4. e3.} Nxc3 7. bxc3 Bd7 8. O-O Bd6 9. d4 exd4 10. cxd4 (10. exd4 {is rated roughly equal to this by Houdini and it's probably a matter of taste which is played.}) 10... O-O 11. Bd2 {now out of the database, with this simple developing move. Here a variety of moves have been played, with e4 being the favorite and h3 played by two GMs.} (11. h3 Qe7 12. Bd3 Nb4 13. Bb1 f5 14. Bb2 Nd5 15. Ne5 Bxe5 16. dxe5 Be6 17. a4 Rfd8 18. Qf3 g6 19. e4 fxe4 20. Bxe4 c6 21. Qg3 Qf7 22. f4 Bf5 23. Rae1 Re8 24. e6 Qc7 {1-0 Franco Ocampos,Z-Mompo Ballester,V/Mislata 1994/EXT 2000 (24)}) 11... Ne7 (11... Bf5 { the engines agree this is best for Black, seizing a key diagonal and preventing Rb1. It looks anti-positional because of the pawn structure weakness after} 12. Bxc6 bxc6 {but Black's bishop pair and piece activity more than compensate.}) 12. Bxd7 (12. Bd3 {is more patient and prepares Rb1 while seizing the b1-h7 diagonal.}) 12... Qxd7 13. Qb3 b6 14. Rfd1 {this puts the rook on what is currently a less-than-useful file.} (14. Rac1 {places the other rook on its most promising file and postpones deciding where the Rf1 should go.}) 14... Qf5 15. Bb4 Bxb4 16. Qxb4 Rfe8 17. Rac1 h6 {this doesn't appear to do much that's currently useful for Black.} (17... Qe6 $5) 18. Qa3 { the queen needs to be moved to a more useful square, this puts pressure on the a-file and helps cover d3.} Nd5 19. Qb3 {this negates one of the points of having played Qa3 in the first place, the queen exerting pressure on the a-file. Now Black is able to move his rook to a more useful square.} (19. Ne5 $5) 19... Rad8 20. Ne5 a5 21. f3 {White sees the value of pushing e4 but stubbornly refuses to move the rook, instead weakening his pawn structure and the e3 square.} (21. Re1 Qe6 $14 22. f4 c5 23. dxc5) 21... Qe6 (21... a4 { the engines have no problem spotting the weakness on e3 and this deflection tactic.} 22. Qa3 (22. Qxa4 Nxe3) 22... Nxe3 23. Qxe3 $11 f6 24. f4 fxe5 25. dxe5 c5 $11) 22. a3 {cleverly thinking this prepares the way for e4 by taking away the b4 square from the Nd5. This has tactical problems, however.} (22. e4 {is in fact best, as now there are no longer tactics involving Nxe3 and Black doesn't have time to undermine the d4 support point for the Ne5.}) 22... f6 ( 22... c5) (22... a4) 23. e4 {this just loses a pawn, reflecting a counting error on White's part (I used to be bad at counting tactics and it's still an area for improvement.)} (23. Nc4 {simple is best.} a4 24. Qb1 $11 {as} Nxe3 { fails to} 25. Re1) 23... fxe5 $17 24. exd5 Qxd5 25. Qxd5+ Rxd5 26. Rxc7 exd4 27. Kf2 b5 (27... Rc5 {would increase Black's advantage.} 28. Rb7 Rc2+ 29. Kg3 Ree2) 28. Rdc1 {doubling rooks here unfortunately does nothing for White.} (28. Rd3 {immediately blockades the d-pawn and makes Black's life harder.}) 28... Kh7 (28... Re3 {is the nastier response that Black missed.}) 29. Ra7 b4 (29... d3 {passed pawns must be pushed!} 30. Re1 Rc8 31. Rxa5 Rc2+ 32. Ke3 {forced, in order to prevent having to give up the rook for the d-pawn.} Rxg2 $19) 30. Rcc7 {so after a couple of inaccuracies from Black, White gets an (admittedly easily parried) threat going.} Rg8 31. axb4 axb4 32. Rab7 d3 33. Ke1 {this looks best, in order to stop the d-pawn, but eventually would lose with best play on Black's part.} (33. Rd7 {is what the engines show after some calculating, although with Black evaluated as still a pawn to the good.} Rxd7 34. Rxd7 $17) 33... d2+ 34. Kd1 b3 35. Re7 {otherwise Re8 and then Re1 wins for Black.} Rd3 {this move only draws.} (35... Rc8 {wins, although it would be relatively difficult for a human to submit themselves to the next sequence.} 36. Rxg7+ Kh8 37. Rh7+ Kg8 38. Rhg7+ Kf8 39. Rbf7+ Ke8 40. Re7+ Kd8 41. Rg8+ Kxe7 42. Rxc8 Rb5 43. Kxd2 b2 $19) 36. Re2 Rc8 37. Rxd2 Rxd2+ 38. Kxd2 Rc2+ 39. Ke3 Rxg2 40. Rxb3 Rxh2 41. Rb7 {despite Black's passed h-pawn, this should be a draw, since White's rook is ideally placed to harass Black's king and pressure the pawns from the side.} h5 42. f4 Kg6 43. Kf3 Ra2 44. Rb6+ Kh7 45. Kg3 (45. f5 {seems to resolve things quicker for White.}) 45... Ra1 46. Kg2 g6 47. Rb7+ Kh6 48. Rb6 Ra4 (48... Kg7 49. Rb7+ Kf6 50. Kg3 $15) 49. Kg3 (49. f5 Kg5 50. fxg6 Rg4+ 51. Kh3 Kh6 $11) 49... Rd4 (49... Ra3+ 50. Kg2 $15 {would keep alive Black's chances, at least temporarily.}) 50. Ra6 (50. f5 {White keeps missing this idea.} h4+ 51. Kf3 Kg5 52. Rxg6+ Kh5 $11) 50... h4+ (50... Rd3+ 51. Kg2 $15) 51. Kg4 {White now has a completely drawn position.} h3 52. Ra3 h2 53. Ra1 Rd5 54. Rh1 Rd2 55. Kh4 Re2 56. Kg4 Rd2 57. Kh4 Rc2 58. Kg4 Ra2 59. Kh4 Kg7 60. Kg3 Ra3+ 61. Kg4 Ra2 62. Kg3 Kf6 63. Rxh2 Ra4 64. Kg4 {there's now no point at all for Black to play on.} Ra1 65. Rb2 Rg1+ 66. Kf3 Rf1+ 67. Kg4 Rg1+ 68. Kf3 Kf5 69. Rb5+ Kf6 70. Rb2 {Threefold repetition, missed by both players.} Rd1 71. Rb6+ Kg7 72. Rb5 Kh6 73. Ra5 Rd7 74. Kg4 Rd1 75. Ra6 Rf1 76. Rb6 Rg1+ 77. Kh4 Re1 78. Kg4 Kg7 79. Kg5 Rg1+ 80. Kh4 Rh1+ 81. Kg4 1/2-1/2

Book completed - Starting Out: The Caro-Kann

Today I finished going through Starting Out: The Caro-Kann by Joe Gallagher (Everyman Chess, 2002).   This is actually the third time through the book for me, in addition to having used it as a reference when looking at particular variations.  However, this was the most comprehensive run-through that I've done of it, so I believe it's worth reporting on, especially since I haven't discussed it before in-depth.

Per my usual practice, I went through the book with a chess set in front of me, playing through all of my chosen lines and illustrative games.  This time, I also at least read through most of the sections on alternate lines, even if they're not included in my existing repertoire.  I found the whole process most useful in: 1) re-exposing myself to the middlegame plans in each key line, and 2) re-evaluating certain variations I play and opening myself to alternatives that I had previously dismissed. 

The book is not meant to provide a comprehensive treatment of the opening in all lines, but it is still the best single-volume introductory work on the Caro-Kann that I've seen, with excellent coverage of some key variations.  The author is not himself a player of the defense, but rather is a 1. e4 player and has prepared a number of different variations over the course of his career against it.  While at first this lack of direct experience as Black seemed a little strange, I now in fact greatly appreciate it; the author's evaluations come across as more objective than those in many opening manuals and the point of view is more balanced between White and Black.

A more general review of the book can be found at the above link; here are my own observations:

• Formatting: I find the single-column format used by the Starting Out series to be very readable and it is designed for larger amounts of text, rather than variations.  I found the diagrams  in this book to be generally more illustrative of key positions and less awkwardly placed than in other books in the series that I've seen. 
• Typos: I only found two move typos, which is on the high end as far as editing quality for chess books go.
• Classical Variation: this was both the best (in my opinion) and personally most useful section of the book.  While some of the earlier sidelines (such as 5. Nc5) aren't fully examined, the overall treatment of the variation is quite meaty and the various different approaches for both sides are explained well.  This is a variation in which theory is usually well-developed until at least moves 10-12, so I appreciated the author's efforts to highlight typical middlegame plans and ideas, rather than simply giving an evaluation at the end.
• Other Main Line variations: the author provides similarly deep treatment (two chapters) of the 4...Nd7 variation, which Karpov popularized, but only one chapter on the 4...Nf6 lines.  This arrangement is warranted, given the popularity and perceived effectiveness of the variations in question.  For example, if you're interested in the Tartakower Variation (5. Nxf6 exf6) there isn't a lot of material.  However, it's probably not in the author or reader's interest to dedicate much time to inferior and unpopular variations.
• Advance Variation: this is now the most popular at the GM level and the book provides a good treatment of both the main response (3...Bf5) and the alternative pawn sacrifice line (3...c5).  The latter section isn't comprehensive enough to provide a basis for a repertoire, especially given the evolution of the line since the book was published, but it is a good place to get started.   Given the relative lack of attention to the 3...c5 lines, it's also useful to see an author (coming from the White perspective, no less) treat them seriously.
• Panov-Botvinnik Attack: another strong section, where all three of Black's main alternatives (5...e6, 5...g6 and 5...Nc6) are presented with some depth.  For Class-level players, I would say that enough material is given here to provide a solid foundation for a repertoire.  Some of the 5...e6 lines will transpose into heavily-studied opening positions with a lot of branches, which means they are outside the scope of this book and will require more individual research by the player.
• Fantasy Variation: this variation (3. f3) is a personal favorite of the author and the section on it was particularly welcomed by me.  Theoretical works on the Caro-Kann generally ignore or quickly dismiss the variation, despite the fact that it can be quite dangerous for the unprepared player.  As a side note, it can also be relevant for defending against the Blackmar-Diemer Gambit (1. d4 d5 2. e4) since I decline it by playing the Caro-Kann (2...c6) and BDG players will likely continue trying to play in the spirit of the gambit with 3. f3.
• Miscellaneous Lines: the Exchange Variation, 2. c4, and the King's Indian Attack are covered here.  While each only gets a small section, there is enough material to flesh out a Black repertoire against each line.  The KIA against the Caro-Kann simply isn't very good; as long as Black knows the best moves early on against it, there are no challenges.  The Exchange Variation requires more preparation as Black, given that more theory has been devoted to it (mostly in the 1970s).  2. c4 is also something Black players should be prepared for, although it's much less theoretical.

Annotated Game #38: A most instructive loss

This fourth-round tournament game is a strong contender for my most instructive loss.  My opponent, perhaps around twelve years old, played the Panov-Botvinnik Attack against my Caro-Kann, which transposed into a Queen's Gambit Declined-type position; the computer in fact classifies it as a Queen's Gambit variant.  The opening goes eleven moves before leaving the database, something of a rarity at the Class level.

The middlegame features a tense duel between White's pressure and Black's countering moves.  Black makes some inferior moves in the early middlegame (moves 15-16) but White then becomes overeager and plays a premature rook lift on move 17.  The pendulum shortly afterward swings back in favor of Black, although after some back-and-forth the position simplifies into what should be a drawn rook and minor piece endgame.  Shortly after this occurs, I play carelessly and White immediately takes advantage of this, creating multiple threats against my pawns that I cannot parry.  Although I hold out until we reach a bishop endgame, White gains a decisive advantage and I resign.

During the post-mortem analysis in the skittles room, my opponent's teacher/trainer sat in and provided some useful pointers.  My opponent was originally rather cocky about his position in the opening, but his trainer then corrected his evaluation and pointed out how Black was doing just fine until the endgame error.  Because of the key nature of this opening system and the typical tactical and strategic themes that were shown during this game, I feel like gained a great deal from playing and then analyzing it.  Interesting how a loss can become a gain, in that respect.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "Class B"] [Black "ChessAdmin"] [Result "1-0"] [ECO "D26"] [Annotator "ChessAdmin/Fritz/Houdini"] [PlyCount "83"] [EventDate "2003.??.??"] [TimeControl "240+2"] {D26: Queen's Gambit Accepted: 4 e3 e6 5 Bxc5 c5 sidelines} 1. e4 c6 2. d4 d5 3. exd5 cxd5 4. c4 {the Panov-Botvinnik attack} Nf6 5. Nc3 e6 {the other two moves played here are Nc6 and g6; the text move is probably the most conservative one.} 6. Nf3 Be7 {the other choice here is Bb4, which gives the position more of a Nimzo-Indian flavor, as opposed to the text move, which is more akin to the Queen's Gambit Declined.} 7. Bg5 {cxd5 is the main line.} dxc4 {here castling is by far the most popular option. I played the text move based on a recommendation from the book "The Caro-Kann in Black and White"} 8. Bxc4 O-O 9. O-O Nc6 10. Re1 Re8 {here b6 and a6 are the most played moves, looking to develop the light-square bishop. One sample game:} (10... a6 11. Rc1 b5 12. Bd3 Bb7 13. Bb1 Rc8 14. h4 g6 15. Ne4 Nxe4 16. Bxe4 Bxg5 17. hxg5 Qd6 18. d5 exd5 19. Bxd5 Nd8 20. Rxc8 Bxc8 21. Re3 Ne6 22. Rd3 Qc7 23. g3 Bb7 {1/2-1/2 Adams,M-Dreev,A/Las Vegas 1999/CBM 72 (23)}) 11. a3 a6 {first move out of the database.} 12. Qd3 b5 13. Bb3 Bb7 14. Bc2 g6 {while somewhat ugly, this is a common theme in this variation, with g6 necessary to stop a battery on the b1-h7 diagonal. In return for weakening the squares around his king, Black blunts the force of White's pieces.} 15. Rad1 Na5 (15... Rc8 {would have been better here, getting the rook to its most useful file immediately and deferring placement of the knight.}) 16. Ne5 $14 {White makes the obvious knight sortie to a central outpost.} Rc8 (16... Nd7 {would immediately challenge the outpost and prove exchanges, reducing White's central and kingside pressure.}) 17. Re3 {a little premature.} (17. Qh3 $5 $14 {is recommended by the engines.}) 17... Nh5 18. Rh3 (18. Bxe7 $5 {preserves more of White's attacking chances.} Rxe7 19. d5 exd5 20. Qd4 $11) 18... Bxg5 $17 19. Rxh5 {Demolition of pawn structure} f5 (19... gxh5 {naturally doesn't work:} 20. Qxh7+ Kf8 21. Qxf7#) (19... Rc7 {is found by Houdini, guarding against the mate and therefore allowing the pawn push f6 in this variation.} 20. Rh3 f6 21. Nf3 Nc4 22. Nxg5 Nxb2 23. Qg3 fxg5 $17) 20. Rh3 $15 Bf4 21. Qe2 Qg5 22. g3 Bxe5 23. dxe5 Red8 {wrong rook, according to Houdini. (Isn't it always the wrong rook?)} 24. f4 (24. Rh4 Qe7 $15) 24... Rxd1+ 25. Bxd1 Qd8 26. g4 Qd4+ (26... fxg4 27. Rd3 (27. Qxg4 Qd4+ 28. Kf1 Rf8 29. Qxe6+ Kh8) 27... Qb6+ 28. Qf2 Qc6 { is much preferred by Houdini.}) 27. Qf2 Qxf2+ {this simplifies into what should be a drawn ending.} 28. Kxf2 $11 Rc7 29. gxf5 gxf5 30. Rd3 Nc6 $2 { an example of lazy play where I didn't think about my opponent's possible threats (failure to falsify).} (30... Nc4 {is obvious and best.} 31. b3 Nxa3 32. Rd8+ Kf7 $11 {according to Houdini.}) 31. Bb3 $16 Kf7 (31... Bc8 {doesn't work either after} 32. Rd6) 32. Rh3 (32. Rd6 Re7 33. Nd5 exd5 34. Bxd5+ Ke8 35. Bxc6+ Bxc6 36. Rxc6) 32... Kg6 {deciding to give up the e-pawn.} (32... Re7 33. Rxh7+ Kf8 34. Rxe7 Kxe7 {would have been somewhat better, as White's h-pawn has a long way to go.}) 33. Bxe6 Nd4 {this will shortly land the knight in trouble.} (33... Ne7) 34. Rg3+ Kh6 35. Ba2 Rd7 (35... a5 $142 $18) 36. Rd3 { taking advantage of the knight's lack of flight squares to pin it against the Rd7.} Bc6 37. Ne2 {White decides to simplify down.} Nxe2 38. Rxd7 Bxd7 39. Kxe2 Be8 40. Be6 {it's now all over, as Black's pieces cannot defend against all of White's threats.} Bg6 41. b4 Kg7 42. Bc8 1-0