21 October 2020

Video completed: "Why You Should Always Muddy the Waters When Losing" by Tatev Abrahamyan

"Why You Should Always Muddy the Waters When Losing" is the eighth video in the Chess.com series by Tatev Abrahamyan. This is an excellent practical idea, both for while you are at the chessboard and to keep in mind during post-game analysis of your own games. You do not in fact get any points for losing more correctly, as the engine analysis scores of your games might imply. So to give yourself a chance to turn the game around - which means giving your opponent more chances to blunder - you may need to take more risks, in order to make your opponent's path to victory less clear. Specifically, this involves trying to generate the possibility of some counterplay, in order to complicate the situation. (This idea is also closely allied to the art of swindling.)

The first example is from a classic game, Amos Burn - Frank Marshall. Marshall, in trouble, gave up a rook on a8 (which wasn't doing anything anyway) in order to open up the center and further expose White's king. Black is lost in the long term, in terms of counting material, but moves in the short term from defender to attacker and effectively takes the White queen out of the game. Abrahamyan points out that for psychological reasons, these kinds of inflection points in games can sometimes trigger immediate blunders, since your opponent will have trouble adapting to the new situation and mentally switching gears.

The second example is GM Nana Dzagnidze - GM Marie Sebag, from the 2019 Cairns Cup. Here Black's position is "miserable", with one bishop completely out of the game in the corner and the other not doing much, with White about to penetrate on the queenside threatening to win a pawn. Instead of opting for static defense and continued positional torture, Black sacrifices a knight for two center pawns, immediately giving her excellent control of central space and activating her two bishops. Again, White goes from attacker to defender and has to completely rethink things. In the actual game, White eventually won, but as Abrahamyan puts it, Black was at least alive for much longer and was playing for three results rather than just two (loss or draw, with no counterchances).

The last example is IM Davaademberel Nomin-Erdene - GM Irina Krush (the video graphic intro reverses the order, although the audio makes it clear Krush has the Black pieces), from the 2018 Olympiad. Black is up a pawn, but similar to the previous example she has a bishop locked away at a8 and White is bearing down on the queenside, this time with two advanced connected passed pawns. Black (according to post-game discussions) felt she had messed up the game and just making natural moves would lose. Black pitched a pawn to try and gain activity, which gave White the opportunity to capture it with the wrong piece. Because of this, Black was able to sacrifice a second pawn, opening up her bishop on the long diagonal and getting her queen into a kingside attack, after which White blundered under pressure.

16 October 2020

Training quote of the day #33: Axel Smith

The best way to practise endgames is to decline draw offers, play long games and analyse the endgames afterwards...There are between a dozen and a hundred endgames that are, depending on your level of ambition, useful to know by heart...As I see it, it's enough to carefully study those endgames once and later review them just once a year or something like that.

From Pump Up Your Rating by IM Axel Smith 

14 October 2020

Video completed: "Why You Should Always Pause during Your Calculation" by Tatev Abrahamyan

"Why You Should Always Pause to Calculate in Chess" is the seventh video in the Chess.com series by Tatev Abrahamyan. (An alternate title, a little more reflective of the actual content, is given in the series list, "Why You Should Always Pause during Your Calculation"). The main point she introduces is that you should not make assumptions about your opponent's responses while calculating a line - for example, by assuming that "automatic" recaptures will happen. To minimize blunders and maximize your own opportunities, instead take pauses to mentally visualize these positions and check for more options.

The first example game is GM Alejandro Ramirez - IM John Bartholomew. In the actual game, Black (who lost) played an "automatic" recapture of a minor piece, thereby missing the opportunity to play an in-between move that would have protected his en prise b-pawn, while still keeping White's bishop trapped and defenseless.

The second game is GM Hikaru Nakamura - GM Ding Liren, from the Ivory Coast Grand Chess Tour 2019. Ding quickly played a forcing line involving piece exchanges in the center and an attack on White's queen, but missed a back-rank tactic where Nakamura could leave his queen en prise and recover a pawn, ending up with a winning position.

The third example is a 2019 game featuring GM Evgeny Shtembuliak and Abrahamyan herself. White has what looks like an overwhelming attack in exchange for having sacrificed two pawns. Black looked at a couple of tactical defensive tries, calculated that they did not work, then lost quickly. Post-game analysis showed that she could have combined the tactical ideas with a brilliant intermediate move - admittedly, a difficult-to-find queen sacrifice - that would have resulted in a drawish rook endgame.

This video at 7 minutes is shorter than most in the series, but I think it is the right length to highlight the concept and help make it stick as part of a player's thinking process. Previously I've highlighted the problems involved in always playing "automatic" or "obvious" moves without examining other possibilities, especially the role of powerful in-between moves that are not necessarily obvious.

12 October 2020

Annotated Game #254: Experimenting is never a complete failure

This last-round tournament game was my first experiment with 1. d4, with the intention of playing the Stonewall Attack. Not entirely unexpectedly, I never actually got to that particular opening, but into a sideline more like a Colle System. The positive aspect of the experiment was getting a nice opening advantage by around move 9. The negative aspect was not understanding what to do after that in this type of position and losing a miniature. 

I recall reading an account of GM Viktor Korchnoi being asked by someone, who was experiencing some trepidation, when would be the right time to play a new opening for the first time in a tournament. That person was afraid they would not be sufficiently prepared and would lose. Korchnoi's reply was along the lines of "You play it, you lose, so what? Just go out there and play it!" He himself had a wide repertoire and played a number of different openings, sometimes successfully, sometimes not. His point was that there is never a perfect time to start playing something new, so you might as well go do it and start learning from experience. Remembering that, I didn't feel so bad.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Class A"] [Result "0-1"] [ECO "D00"] [Annotator "ChessAdmin/Komodo 13.2"] [PlyCount "36"] [EventType "simul"] [EventRounds "5"] {[%mdl 8192] D00:1 d4 d5: Unusual lines} 1. d4 d5 2. e3 {the recommended move order for the Stonewall Attack} c5 3. c3 (3. dxc5 $5 {should be strongly considered here, as Black is effectively playing a Queen's Gambit a tempo down. White actually scores poorly overall in the database, but the higher-rated players have a big plus score with it.}) 3... Nc6 4. Bd3 {looking for a Stonewall continuation. This has the advantage of controlling f5, so Black cannot develop the bishop to the b1-h7 diagonal.} c4 {this wins a tempo on the bishop and gains space. However, structurally it benefits White to close the center by better enabling kingside play, and leaves the bishop on its preferred diagonal.} 5. Bc2 b5 {yet another pawn move in the opening. White should be able to take advantage of this.} (5... e5 $5 {is the engine recommendation.} 6. b3 cxb3 7. axb3 $11) 6. Nf3 {we now have a Colle System structure, where the main idea is to play an e3-e4 break. It could have been done immediately as well.} g6 {this does not seem like the best place for the bishop.} 7. O-O {a bit too cautious.} (7. e4 {played now would take advantage of Black's lack of development, specifically White's better control of the e4 square.} Bg7 8. O-O $14) 7... Bg7 (7... Nf6 $5) 8. e4 {well-played now, as Black has passed up her chance to control e4.} dxe4 $2 {if one goes by the principle of just counting tempi, then this move makes sense, as it forces the bishop to move for a third time. However it greatly improves the position of the piece at the same time and is yet another pawn move by Black.} (8... Nf6 $14) 9. Bxe4 $16 Bb7 {here is where I had a good think, but did not follow up most effectively. The fact that the Bb7 is undefended can be exploited by immediate action in the center, as the Nc6 is effectively pinned.} 10. Bf4 { while this is a normal developing move and not bad in itself, it misses the opportunity for more forcing play.} (10. Ne5 {Black in fact cannot just exchange the piece to solve her problems. I did not fully investigate the possibility, dismissing it too early. The key line is somewhat complex, however.} Nxe5 11. Bxb7 Rb8 12. dxe5 Qxd1 13. Bc6+ $1 (13. Rxd1 {is also good}) 13... Qd7 14. Bxd7+ Kxd7 $16) 10... Nf6 {White may still have a slight pull here, but Black has repaired her position.} 11. Bxc6+ {a little too quick to exchange.} (11. Nbd2 $5 {would at least develop another piece while waiting to see what Black does next.}) 11... Bxc6 $11 {Black now has the pair of bishops and the position is very balanced.} 12. Ne5 Bb7 {again a long think here and an incorrect plan.} 13. b3 $6 {this is awkward and not forcing enough. The pawn advance leaves the c-pawn significantly weakened.} (13. a4 {is the way to target the pawn chain, at its base, without creating more weaknesses.}) 13... Qd5 {I saw this only after the fact. It was more of a psychological blow than a real attempt at advantage, but the surprise combination of threatening mate on g2 and asserting control over the queenside squares was nonetheless disheartening, especially considering the advantage I had in the opening.} 14. f3 $11 {the correct response, blocking the long diagonal. Opening the a7-g1 diagonal is not an issue here, since Black has no way to take advantage of it. However, it leaves the e3 square, which should have been recognized earlier.} cxb3 (14... O-O $11) 15. Qxb3 $6 {this results in an unfavorable hanging pawns structure for White. Black controls the c4 square and the White pieces are not well placed to support the c/d pawn pair.} (15. axb3 {should be investigated more closely, comments Komodo via the Fritz interface.} Nh5 16. Be3 $11) 15... O-O $15 {Black completes her development, leaving her with no problems and good play in the center and queenside. Meanwhile, I start falling behind in my own development, with disastrous consequences.} 16. Nd3 $6 {this unnecessarily retreats the knight and loses time. Ironically, part of my consideration for this move was to protect the Bf4, but in fact it jeopardizes the piece.} (16. Re1 $11 {is one alternative, activating the rook on the half-open file.}) (16. Na3 {looks awkward but at least would get the knight out. Putting it on d2 would actually be more awkward in reality, as it would have no good other squares to go to.}) 16... Qf5 $17 {another queen move that I missed, attacking the Nd3 and Bf4. This forces me to protect the Nd3.} 17. Qd1 $2 (17. Rd1 $17) ( 17. Nc5 $2 Bd5 $19) 17... Nd5 $19 {my opponent ably finds the killer move in response. I cannot adequately protect both the Bf4 and the e3 forking square.} 18. Bd2 $4 {covers the bishop and the e3 square, but forgets about the now-hanging knight.} (18. Be5 Ne3 $19) 18... Qxd3 0-1

09 October 2020

Using tactics training to strengthen your thought process

Chess tactics training and practice is necessary at all levels; even master-level players will often do problem sets as a warm-up before tournaments. When a player is in the process of gaining strength during their chess studies, the main purpose of learning tactics is to progressively unlock different possibilities on the chessboard. This includes:

  • Recognizing common and uncommon mating patterns
  • Fundamental tactical ideas, such as double attacks/forks, pins, bank rank vulnerabilities, and common sacrificial themes
  • Employing more advanced concepts, such as deflections, clearance sacrifices, the power of intermediate moves in a sequence, and forcing your opponent into zugzwang
Tactics training therefore feeds the pattern recognition process, which is a largely unconscious cognitive function, along with the mostly conscious and effortful process of calculation. (Temposchlucker's posts on "System 1" and "System 2" provide a number of illustrations.) Because of the complexity of chess, in most games we cannot simply remember what to do in a particular position. We can nonetheless recognize similar elements leading to tactical opportunities: poor king safety, hanging pieces, controlled files and diagonals, etc. The more we understand how these elements interact, the more detailed our "automatic" pattern recognition becomes and the easier it is to find specific tactics over the board.

The tactics training process is a necessary but not sufficient condition for gaining additional chess strength, however, past a certain point. Fundamentally, this is because the majority of positions we encounter in a game cannot be "solved" in the sense of there being a forced line leading to mate or material advantage. In many cases there is also no consensus even among professionals on the "best" positional move, as handily illustrated by the "Beat the Masters" series at The Abysmal Depths of Chess blog. This means we have to employ a more general thinking process when actually playing, as opposed to when doing dedicated tactics drills.

From longtime observation of the chess improvement community, it seems that bridging the gap between tactical problem solving and general chess thinking is one of the main hurdles in sustained chess improvement. This is especially the case in adult improvers, who tend to learn by applying systemic thought processes, rather than by intuitive mimicking of whatever works in a particular situation. This parallels how adults learn languages and other complex skills versus how children do it.

The problem is that chess contains both tactical and non-tactical positions, which have different "solutions" and therefore require different mental approaches. Or do they? Naturally it would be best to have a general thought process which takes into account all types of situations. The simplified thought process I generated and its companion, what to think about on your opponent's time, have held up pretty well for me - when I have the energy and focus to appropriately apply the principles. (Energy management in chess is another subject that is worth addressing in more depth.) One of the early positive results of having a more explicit thought process, I found, was to measurably boost my tactics puzzle ratings. This was a positive sign that I was on the right track.

Lately I have more consciously applied general thought process principles to tactics puzzles, most of which are drawn from real games, finding that it has helped both the tactics results and to reinforce the positive evolution of the thought process itself. Specifically, one of the main weaknesses I have identified in my play is an inconsistent recognition of my opponent's resources and threats. This is a common phenomenon among chessplayers, who naturally focus on their own possibilities, and is currently one of the top things holding me back in the middlegame. (Materialism is now less of an issue.)

Beginning each tactics puzzle with the question "What are my opponent's threats?" and identifying all of them has been a simple but helpful change in this regard. "What does my opponent want to do with their last move?" is a related question, also helpful in identifying their strategic plans. In the context of tactics training, you are not likely to be looking for strategically good or prophylactic moves - unfortunately, since that is a major element of gaining chess strength - but identifying threats to your position is always necessary. (Calculating your mate in three is not helpful if your opponent threatens a mate in two, in other words.)

Building the habit of looking at the opponent's moves first as part of tactics training may seem like a simple enough concept, but I expect it will have more profound effects on my general thought process as well. As with all good habits, with repetition comes improvement and greater automaticity, requiring less energy to consciously manage the thought process.

01 October 2020

Annotated Game #253: Time for a win

In this third-round tournament game, I finally get a win out of playing a move-order trick in the Classical Caro-Kann. This is not an opening trap, since the resulting positions are equal, but it is designed to take White out of known territory and get an equal game for Black without much effort. This does not necessarily make it easy to play, however, and some knowledge and experience is helpful; see Annotated Game #244 for a previous game in this line against Expert-level opposition.

In contrast with the previous game, I do well in developing and avoiding unhelpful piece exchanges. While there were a number of places where I could have improved play - move 14 is the most instructive example, for positional reasons - I did not blunder and was able to pursue logical middlegame plans that suited the position. Overall, the game is a classic example of how building up pressure on an opponent significantly increases the chances of them eventually cracking, which is what happens here.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "Class B"] [Black "ChessAdmin"] [Result "0-1"] [ECO "B18"] [Annotator "ChessAdmin/Komodo 13.2"] [PlyCount "68"] [EventType "simul"] [EventRounds "5"] {[%mdl 8192] B18: Classical Caro-Kann: 4...Bf5 sidelines} 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Bf5 5. Ng3 Bg6 6. Nf3 {this allows the following move-order trick, as Black foregoes the standard ...Nd7, which is what most White players expect.} Nf6 7. h4 Nh5 {blocking the h-pawn from further advancing and threatening to exchange on g3, which would mess up White's pawn structure.} 8. Nxh5 Bxh5 9. c3 {the pawn move reinforces d4 and opens the d1-a4 diagonal for the queen.} (9. Be2 {is what my Expert-level opponent played previously, going for normal development.}) 9... e6 {opening the way for developing the bishop. Since this and ...Nd7 are both going to be played to develop Black's pieces, move-order does not seem critical.} (9... Nd7 {instead was played in this top-level game.} 10. Be2 e6 11. Bf4 Be7 12. Ng5 Bxe2 13. Qxe2 h6 14. Nf3 Qa5 15. a3 Qf5 16. g3 Nf6 17. O-O-O Qe4 18. Rde1 Qxe2 19. Rxe2 h5 20. Bg5 Ng4 21. Bxe7 {1/2-1/2 (21) Grischuk,A (2775)-Anand,V (2767) Stavanger 2019}) (9... Qc7 {is another option here.}) 10. Be2 Be7 {immediately pressuring the h4 pawn.} 11. g3 {deciding to not worry about protecting the h-pawn, but of course at the cost of weakening the kingside light squars.} O-O 12. Bf4 Nd7 13. Qc2 Bg6 { I did not think the bishop had a great future on the h5-d1 diagonal, so retreated it with tempo, also in part to see where White would commit his queen.} 14. Qd2 {this is a slight waste of time, as the queen could have gone to d2 on the previous move.} (14. Qb3 $5) 14... Nf6 $6 {not a major error, but the knight is actually less helpful on f6 than on d7, where it guards e5 and would support a ...c5 break.} (14... Bf5 {I considered, since it would restrain a g3-g4 advance and not make the bishop a target of an h4-h5 advance.} ) (14... h6 $5 {would be a good restraining move, covering g5, while also giving the bishop a square on h7.}) 15. Ne5 $14 {my opponent correctly occupies the central outpost.} c5 {a thematic pawn break.} 16. Nxg6 hxg6 { White now has the pair of bishops.} 17. Be5 (17. dxc5 $5 Bxc5 18. Qxd8 Raxd8 $14 {and White would consolidate a small positional plus, namely the two bishops and a 3-2 queenside pawn majority.}) 17... Qd5 {in the Classical Caro-Kann, the queen sometimes is able to (or needs to) go to d5, in order to keep Black active and make threats. The queen's strength is enhanced by White's weak light squares, which likely made my opponent unwilling to castle.} (17... cxd4 {Komodo favors exchanging in the center first, to take away White's ability to exchange on c5.} 18. Bxd4 Qd5 19. O-O $11) 18. Rg1 (18. O-O $5 $14 {looks like a viable alternative.}) 18... Rfd8 $15 {this strongly activates the rook on the d-file. White now has to be careful of tactics.} 19. Qe3 {removing the queen as a d-file target and overprotecting the Be5.} cxd4 ( 19... Nd7 $5 {is another possibility, keeping the central tension.}) 20. Bxd4 Qa5 {protecting the a-pawn and also restraining White from queenside castling. It is now difficult for White to come up with a useful plan.} 21. Bf3 {White threatens to win material: Bf3xb7} Rd7 $11 {this shows a lack of creativity, as the b-pawn is tactically protected.} (21... Nd7 $5 {is an interesting move, since it now threatens ...e5, controls c5, and clears f6 for the bishop. Snatching the b-pawn is not good for White, for example} 22. Bxb7 Rab8 23. b4 ( 23. Bf3 e5 24. Bxa7 Rxb2 $19 {and Black has multiple threats that cannot be met, for example ...Nc5 followed Rdd2, or simply capturing the a-pawn if the Ra1 moves. The main problem for White is that the bishop on a7 has no squares left and cannot be further protected.}) 23... Bxb4 $1 $19) 22. a3 {freeing the Ra1 to move, while further weakening the light squares.} Rad8 {creating tactical possibilities in the center. The pin on the c3 pawn is now a factor.} (22... Nd5 23. Qe4 Rad8 $15) 23. Rd1 $2 {again, my opponent avoids castling into a weakened light square complex. However, keeping the king in the center is worse.} (23. O-O-O Bxa3 {is something my opponent may have wanted to avoid. The engine shows it as equal, but Black is going to have all the fun, as White has to find difficult moves. For example, the straightforward} 24. bxa3 Qxa3+ 25. Kc2 Qa2+ 26. Kc1 {runs into} Rxd4 27. Rxd4 Rxd4 28. Qxd4 Qa1+ 29. Kd2 Qxg1 30. Qxa7 b5 $17 {and White's less secure king gives Black an edge.}) 23... Qa4 {perhaps not the strongest move, but it increases the pressure on White, specifically on the d1 square, and poses my opponent additional problems.} ( 23... Nd5 $5) 24. Rc1 (24. Kf1 Bc5 $11) 24... Qb5 {a simple threat to the b-pawn, but giving up the potential tactical control of d1.} (24... Qb3 { is a better version of the idea. For whatever reason, I was concerned about the queen being trapped.}) 25. Rb1 $2 {this loses and is a blunder. However, it is also an illustration of how keeping strong pressure on an opponent, even if not making the "best" moves, makes their defensive task difficult and often leads to them cracking.} (25. Qe2 Qf5 $15) 25... Bxa3 {there are more complicated ways to take advantage of White's last move, but I chose the simplest path.} (25... e5 $5 {would create a number of threats, but is not that easy to see, or at least was not for me at that point in the game.} 26. Qxe5 (26. Bxa7 Rd3 $1) 26... Qd3 $1 {is the key}) 26. Bxa7 $2 (26. Be2 Qf5 27. Qd3 $19 (27. Bd3 Qa5 $19)) 26... Bxb2 $19 {here the simplest way is the best, as ...Bxc3+ is threatened and White's king lacks squares to run to, forcing the response.} 27. Rxb2 Qxb2 28. Bd4 Qb1+ 29. Bd1 Ng4 {the idea of bringing the knight into the attack is correct, although Komodo prefers to send it to e4.} 30. Qg5 (30. Qe2 {I saw failed to} e5 $1 {and the bishop can no longer block the rooks on the d-file.} 31. Qxg4 exd4 32. c4 d3 {is what I looked at, which is completely winning.}) 30... e5 (30... Rxd4 {is perhaps less complicated, but I saw the text move would win, with the same idea as in the above variation.} 31. cxd4 Rxd4 32. Qxg4 Rxg4 $19) 31. Qxg4 exd4 32. c4 Qb4+ { picking up the c-pawn is unnecessary, but I saw the simple winning strategy of just running the b-pawn up the board, which is unstoppable.} 33. Kf1 Qxc4+ 34. Kg2 b5 (34... b5 35. h5 gxh5 36. Qxh5 Qd5+ 37. Qxd5 Rxd5 $19) 0-1