Black magic

One of the defining features of my chess life with the black pieces has been a contradiction at the centre of it. Against 1. e4, with rare exceptions, I've invariably played 1... c5, having apparently come out of the womb knowing how to handle an Open Sicilian structure. But when faced with 1. d4 I chop and change: at one time or another I've relied on the King's Indian, Modern Benoni, Budapest Gambit, Benko Gambit, Dutch, Blumenfeld Gambit, Nimzo-Indian, Slav and Semi-Slav! Quite simply I've never found anything I like (suffice to say I wouldn't dream of playing a lot of those openings now) - if I liked the main line it was too easily avoided, if I could reach my target position it wasn't to my taste, and if I did enjoy it then it was unsound.

After a particularly depressing loss in the final weekend of the 4NCL last season, in which I lost on the black side of the absolutely symmetrical variation of the Exchange Slav, I was idly browsing the bookstall when I noticed that Quality Chess had brought out two brick-like books by Gawain Jones on the King's Indian. I used to play that opening a lot, but as I saw it I'd long since grown up - I held to the conventional modern view that it's one massive cheapo and, whilst a good practical weapon, not entirely correct. Still, the idea gnawed at the back of my mind, for it would certainly bring some more hooliganism into my play, and eventually I decided to take the KID up again for the Major Open.

As it happened no-one played 1. d4 against me in Torquay, so I lay quietly waiting to spring the trap; finally, in today's game, I got the chance. This was a 4NCL game against 2082-rated Dimitrios Zakarian, part of KJCA's match against Oxford. We got crushed 1½-6½ but I picked up a very satisfying consolation win, hours after the rest of the team had finished. The opening worked like a charm, however I needed more than a little luck to finally reel in the full point:

[Event "4NCL Div. 2"] [Site "?"] [Date "2023.02.11"] [Round "5.6"] [White "Zakarian, Dimitrios L"] [Black "Young, Daniel J"] [Result "0-1"] [ECO "E60"] [WhiteElo "2082"] [BlackElo "1985"] [Annotator "Young,Daniel J"] [PlyCount "154"] [EventDate "2023.??.??"] [WhiteTeam "Oxford 1"] [BlackTeam "KJCA Kings"] {Time control: 40 moves in 1 hour 40 minutes + 50 minutes to completion + 30 seconds/move.} 1. d4 {(1:34)} Nf6 {(1:40) Surprise!! As part of my Major Open preparation I'd decided to re-learn the King's Indian Defence, having been worn down by too many dull anti-Semi-Slav lines. This was the 1. d4 opening I played when I first started chess, and intermittently from then until 2015, but I never had much feel for it. But I do at least enjoy the positions, and now that I am a little wiser my results during this revival have been excellent. But all this was unknown to Zakarian - I didn't get to use it in Torquay and it had only appeared in club and county games until now, so he was caught in an ambush.} 2. Nf3 g6 3. c4 Bg7 4. g3 O-O 5. Bg2 d6 6. O-O {In his magisterial repertoire series, Gawain Jones argues that 6. Nc3 is a more precise move order, preventing what follows. Zakarian had played the text move before, and for once my preparation made it to the board. If that's not evidence some dark power was with me on this day, watch what happens later!} c5 7. d5 {7. dxc5 is a serious move, so 6... Nbd7 or 6... Nc6 are viable for must-win games. This is the most ambitious, but} b5 $1 {is a Benko Gambit with a twist. There is a little backstory here: over the new year I was asked by Graeme Buckley to help him prepare for an IM norm APA so that his openings could keep up with the young guns, based on the Tarrasch and the Benko Gambit. My first thought was that making this work was beyond the power of any mortal, but I put a lot of work into the Benko on his behalf, and learnt a lot about an opening I gave up long ago, as well as discovering that it is not as bad as its current reputation suggests. So I was confident going into this structure, and Black has a trick up his sleeve: after} 8. cxb5 a6 9. bxa6 {, rather than making it a pure transposition by taking with the bishop Black can take advantage of his move order and play} Nxa6 $1 {This is a known idea in the regular Benko as well (it's why Black prefers 1. d4 Nf6 2. c4 c5 3. d5 b5 4. cxb5 a6 5. bxa6 g6 instead of 5... Bxa6) - once White has committed to a fianchetto formation the knight has more potential than the bishop on a6 because it can reach b4, with annoying pressure against both a2 and d5.} 10. Nc3 {(1:33)} Bf5 {(1:43)} 11. Nd2 {Aimed against ...Ne4, which would be a helpful exchange for Black, activating both of his bishops.} Nb4 12. e4 (12. Nc4 {Jones does not mention this move, however the long variation} Bc2 13. Qd2 Bb3 14. Na3 Bxd5 15. Nxd5 Nfxd5 16. Nc4 Nc7 $1 (16... e6 $2 {is played more often but only results in static weaknesses after} 17. a3 Qb8 18. Rb1 Nc6 19. Bxd5 exd5 20. Qxd5 $16) 17. Bxa8 Qxa8 $44 {offers Black adequate compensation; his pressure on the queenside files endures and he has won White's d-pawn in return for the exchange. Is the rook on a1 really better than the bishop on g7? }) 12... Bg4 13. f3 $6 {Again walking helpfully into the only move mentioned by Jones, gaining time but weakening the dark squares. The bishop has done its job now so Black does not mind retreating it.} Bc8 $15 14. a3 Nd3 $6 { Unknowingly following the author of my book, but as he points out there is a stronger option - the knight doesn't need to jump in yet.} (14... Nd7 $142 { I did consider this but wasn't sure how to continue after} 15. Nb3 {, defending the a1 rook and clearing a defence of d3. I overlooked} Qb6 $1 16. Kh1 Nd3 17. Qxd3 Qxb3 $15 {with a dream Benko position in which White's queenside pawns are jammed and the pieces easily flow in: ...Ba6, ...c4/... Nc5-d3, ...Rfb8.}) 15. Nc4 {(1:17)} Nxc1 {(1:31) Nevertheless, Black has good compensation for his pawn investment. The g7 bishop on the long diagonal and the heavy pieces on the a- and b-files combine to generate strong pressure on White's queenside pawns. There is also the great paradox of the Benko - even though Black has sacrificed a pawn, endings tend to favour him because the d-pawn is a long-term weakness, and in a simplified position it can be undermined by an ...f5 break. White is also going to miss his dark-squared bishop with all of his central pawns on light squares. In sum, Black can be happy.} 16. Rxc1 $6 {As the knight has no retreat, White could have delayed this move, and indeed should have done - 16. Re1 to avoid the coming pin is safer.} Ba6 17. Ne2 $2 {And now the knight's retreat opens the long diagonal for Black, after which b2 comes under heavy pressure. 17. b3 was preferable although 17... Nd7 still favours Black, with ...Ne5 or ...Nb6 coming up.} Nd7 18. Qd2 Qb8 $1 {An unstereotyped square for the queen, but a good one. Thanks to the threat of ...Bxc4, White is unable to protect his b-pawn and has to return it, but liquidating the queenside does not fully solve his problems because the dark square weaknesses remain, as does Black's piece activity.} 19. b4 Bxc4 20. Rxc4 {(43)} Rxa3 {(1:20)} 21. bxc5 Nxc5 $19 22. Nd4 $2 {A tactical error which should lose at least the exchange.} Qa7 {Improving the queen with tempo, as 23... Nb3 is threatened.} 23. Kh1 (23. Nb5 $143 $6 {loses even more violently than I imagined after} Qb6 $1 (23... Nxe4+ $143 $6 {was my idea, but I missed that White can try} 24. Qf2 {since my queen is also hanging. Stockfish gives the best continuation here as} Qxf2+ 25. Rxf2 Ra1+ 26. Rf1 Nd2 27. Rxa1 Nxc4 $19 {, which is an improvement over the game because the d-pawn is isolated.}) 24. Nxa3 Nxe4+ 25. Qf2 Nxf2 26. Rxf2 Qb3 $19 {and, as always, loose pieces drop off.}) 23... Rd3 $2 {Up until this point I had played a really good game, but now I started to falter with a clear win in sight. The text move wins a pawn because White's queen has so few squares, but great technical difficulties lie ahead. I suspected there was more and searched for it in vain.} (23... Qa6 $142 {The simple attack on the c4 rook seems fairly innocuous at first, but it causes a crisis in White's position because he has to balance its defence with the continuing weaknesses on d4 and d3.} 24. Rfc1 ( 24. Rc2 Rd3 $19 {wins the knight}) (24. Rb4 Nd3 $19 {also forces the rook away from its defence}) (24. Nc2 {counter-attacks against Black's rook, but Black easily evades this with tempo and once again wins material.} Rd3 25. Qb4 Rb3) 24... Bxd4 25. Qxd4 Nb3 26. Qg1 Nxc1 27. Rxc1 Ra2 $19 {The white bishop is no match for a rook with its long diagonal so utterly blocked.}) 24. Qf2 { Zakarian said after that he overlooked my response, but he didn't have anything better.} (24. Qb2 $144 {By contrast, I hadn't found the win against this retreat, but I was sure there had to be one with the queen sat on the long diagonal and sure enough} Rxd4 $1 25. Rxd4 Nd3 26. Qc3 Qxd4 $19 {wins a piece.}) 24... Nxe4 $1 25. fxe4 {(32)} Bxd4 {(55)} 26. Qc2 Ra3 27. Rc7 Qa6 28. Qc4 {Rightly offering a queen trade - White cannot get his pawn back.} (28. Rxe7 $143 $4 {loses to} Ra1 {and White is embarrassed by the defence of his rook. If} 29. Rd1 Rc8 30. Qd2 Be3 $19 {wins material.}) 28... Qxc4 {The computer heartily endorses this as the best move and awards an evaluation of -3.5, but are things really so simple for me? As pieces come off, the opposite coloured bishops cease to be an unqualified blessing for me.} 29. Rxc4 Bf6 30. Rc2 {(13)} Rb8 {(32)} 31. Bf3 Rba8 32. Kg2 Ra2 $6 {A strange decision. Obviously I knew that I wanted to avoid a pure OCB ending, but there's no reason to want to exchange even one pair of rooks.} (32... R8a4 $142 33. Re1 h5 34. Rce2 Bd4 {The engine sticks to its guns, but a human gets nervous when it gives every move a similar evaluation and is just shuffling pieces...}) 33. Rxa2 Rxa2+ 34. Rf2 Ra3 {Of course I wasn't going to make it that easy. But I could not see a way to make progress so opted just to mark time until move 40 and see if White might lash out in some way.} 35. Bg4 {(9)} h5 {(28)} 36. Bf3 Kg7 37. h4 {Transferring the base of the pawn chain from h2 to g3 felt like an achievement, but g3-g4 is coming.} Be5 38. Kh3 Ra1 39. g4 Rg1 40. Bg2 {(3)} hxg4+ {(20)} 41. Kxg4 Kh6 $2 {Add another to the long list of lemons made immediately after the time control! I took some time here and reasoned that it was a good idea to stop White trading more pawns with 42. h5, but missed that this allows him to force the exchange of rooks instead.} 42. Kh3 Bf6 43. Rf1 { Zakarian played this with a flourish, and having realised what I'd done I involuntarily tutted aloud at myself. So he knew that I knew that the position is now dead drawn. But, as John Nunn put it, care is still needed: it only takes a few moves for a position to go from 'drawn' to 'dead'. And I think that Zakarian stopped concentrating at this point, assuming the draw was now in the bag.} Rxf1 44. Bxf1 $11 Be5 45. Kg4 {(51)} Bc3 {(1:04)} 46. Bb5 Kg7 47. h5 gxh5+ 48. Kxh5 Kf6 49. Kg4 Ke5 50. Kf3 {(53)} f5 {(1:05)} 51. Bd3 $2 { Despite the odd geometry that now arises, I should say up front that it doesn't change anything objectively: the resulting position is still a tablebase draw. But I think the punctuation is justified, for why take the slightest chance?} (51. exf5 Kxd5 52. Bd3 Ke5 $11 {would be a trivial draw, even though Black now has a passed pawn: the blockade on d3 is impregnable and there's no way to win the f-pawn.}) 51... fxe4+ 52. Bxe4 Bd2 {Zugzwang!} 53. Bh7 Kxd5 {I did not know the evaluation of this position (as mentioned above, it's still a draw) but I felt that I would probably win if only because of the psychological shock. So it proves.} 54. Ke2 Bg5 55. Kd3 {(53)} Ke5 {(1:04)} 56. Bg8 $6 (56. Be4 {is White's most reliable draw, immediately setting up the frontal attack mentioned below. It's not intuitive to invite} d5 {, when the pawns reach the fourth rank in lockstep, but they won't get any further.} 57. Bg2 Kd6 58. Bf3 e5 59. Bg2 $11 {If the d-pawn pushes then White gets his light-square blockade, and if Black advances the e-pawn instead then of course White simply takes it.}) 56... e6 {I have never made a deep study of this ending, but I was aware of the most important principle: do not let the pawns get blockaded on the opposite colour to the enemy bishop. This small bit of knowledge helped me a great deal.} 57. Bh7 d5 58. Bg8 Bh4 59. Ke3 Bg3 60. Kf3 { (52)} Bh2 {(1:02)} 61. Ke3 Bg1+ 62. Kd3 Kd6 {After some exploratory circles (just to make myself comfortable!) it is time to start advancing the pawns.} 63. Bh7 $2 {As we are now in tablebase territory, I can confidently say that this move ought to lose.} (63. Bf7 $142 {White can still draw with best play. The goal is to set up a frontal attack on the e-pawn before the terrible twins get as far as the third rank.} e5 64. Be8 e4+ {White now has to find a string of only moves, but} 65. Kd2 $8 d4 66. Ba4 $8 Ke5 67. Bc2 $8 Kf4 68. Ke2 $8 $11 {keeps his head above water. Black's king needs to stay in touch with e4 so cannot drive the bishop from its post.}) 63... e5 64. Bg6 Ke6 $2 {For the most part, given that neither side knew much of this ending, it was played to a decent standard - this was the only move I made which changed a winning position to a drawn one.} (64... Bc5 $142 {The bishop needs to be able to drive the white king away from d2 at the crucial moment. It's now not possible for White to set up his frontal attack.} 65. Bf5 (65. Be8 {If we try to follow the same recipe as 63. Bf7 above, then} e4+ 66. Kd2 d4 67. Ba4 Bb4+ {is critical - White's bishop cannot reach c2 in time.} 68. Ke2 d3+ 69. Ke3 Ke5 { The king is soon driven out of e3, and the pawns reach the third rank.} 70. Bb3 Bc5+ 71. Kd2 Kd4 72. Ke1 e3 $19) 65... Bb6 66. Bh7 Ke7 67. Bg6 Kf6 68. Bh7 Bc5 69. Ke2 e4 $1 {The only move, again preventing the bishop reaching the queenside.} 70. Kd2 Ke5 71. Bg8 d4 72. Bc4 Be7 73. Ke2 Bg5 74. Ba6 Kd6 75. Bc4 Kc5 76. Ba6 {Once it's forced to attack from the rear the bishop is no longer so effective.} Bf4 77. Kf2 d3 78. Ke1 Kd4 79. Bb7 e3 {If the pawns get this far then Black is always winning regardless of king or bishop positions (unless of course something is immediately en prise). Here it ends with a nice zugzwang following} 80. Bf3 Bg3+ 81. Kd1 Ke5 82. Bg4 Kf4 83. Bh5 Bf2 $19) 65. Bh7 $2 {(42) I never let go after this. 65. Be8 is similar to 63. Bf7 above and is drawn with best play.} Kf6 {(1:01)} 66. Bg8 e4+ 67. Kd2 Ke5 68. Bf7 d4 69. Bc4 Be3+ 70. Kc2 {(38)} Bg5 {(1:00)} 71. Bb5 Kf4 72. Ba6 Ke3 73. Bb5 d3+ 74. Kd1 Kd4 75. Bd7 {(38)} e3 {(58)} 76. Bg4 Ke4 77. Ke1 Bh4+ {I thought for a few minutes on this move, and to my astonishment Zakarian resigned instantly when I played it. At the board I still was not absolutely certain I was winning, and his father (an FM, who ironically I once got a draw with at a King's Place rapid by holding out in an OCB ending!) was adamant he had resigned in a drawn position. That is not the case and in fact the end is not far off, but I should have been made to prove it!} (77... Bh4+ {Black wins quickly after either king move.} 78. Kf1 (78. Kd1 $144 Kf4 79. Bh5 Bg3 { is a similar zugzwang to the one that concluded the 64... Bc5 variation above - } 80. Kc1 e2 $19) 78... Kd4 {Black's king can't be kept out of d2, and then the e-pawn walks through.} 79. Bh5 Kc3 80. Bg4 Kd2 $19) 0-1

One of my team-mates asked me the next day just how I'd managed to win the opposite bishops ending - my first response was that a magician never reveals his secrets, but on reflection I said I'd won by letting him taste the draw. As I mentioned in the notes, Zakarian defended stubbornly in a very difficult position, but once he forced the exchange of the second pair of rooks he switched off. In that sense the end of this game is similar to the previous one against Nevols. People play worse when you keep them under pressure throughout the game.

As for the opening, rest assured there are more King's Indian games set to appear on this blog soon! When I was younger I didn't grasp this opening at all, but I am much wiser now: I handle the initiative more surely, my understanding of dynamics is stronger and I am a lot less materialistic. That's not to big up my ability in those areas; it's simply that I've got a little older! Ironically I can start playing the King's Indian again precisely because I've now grown up. With the mild exception of 14... Nd3?!, my play was computer-perfect right up until 23... Rd3?, when my lack of tactical vision started to take over. So if there is a lesson from this game for opening choices, perhaps it is that it's worth having a new look at your old favourites - after all, there was a reason you were once drawn to them.

I drew comfortably with a 2200 the following day, and won my next two games in the Surrey League, followed by a memorable night in the Croydon League rapidplay division. I very rarely pay any attention to rapid games any more, but it was a dramatic enough evening that I am going to make an exception.