The chess struggle in practice

For the fourth round at the Kingston Invitational I was paired with the white pieces against Tim Seymour, rated 2220. This game required a lot of effort in advance, but that was no bad thing: I was on a high after the earlier win, it was a warm summer evening as well as a chance to dive deep into an interesting opening line, and altogether not a chore to work into the night. We can draw satisfaction from the entire creative process of a chess game, and that includes preparing for it. And on a more practical level, sometimes the results have been impressive...

These days you can't count on a tournament player having a weak opening to home in on, so you need to be savvy to prepare well. I usually start by playing over my opponent's recent games. This is purely impressionistic, done at blitz speed without an engine: what are their hopes and fears, their likes and dislikes, and how do they win or lose games? It's amazing how much can be learned from even 20-30 games; in this case I noted that Seymour likes reliable classical systems, can lose the thread in a sharper struggle, and has a very narrow repertoire but knows it well. That made it easy to discard 1. e4 once I turned to the database since it would lead to a Berlin Defence, which plays to his strengths. It was easier to inject some spice using the English Opening, and that spice was what eventually won me the game. Even though I couldn't find an advantage, I could find a route to a position where I was more comfortable, and it showed:

[Event "Kingston Invitational"] [Site "?"] [Date "2024.08.14"] [Round "4.17"] [White "Young, Daniel J"] [Black "Seymour, Timothy P"] [Result "1-0"] [ECO "A13"] [WhiteElo "2017"] [BlackElo "2220"] [Annotator "Young,Daniel J"] [PlyCount "155"] [EventDate "2024.??.??"] [SourceVersionDate "2024.08.18"] {Whatever the results, tough pairings were bound to keep on coming. That said, I was feeling quietly hopeful for this one: whilst I'd been warned beforehand that he knows his openings very well, I'd given them 'the treatment' overnight and did not come to the board unarmed. Having also seen a few tactical accidents in his games, I could definitely see a path to winning.} 1. c4 {(1:30)} e6 {(1:30)} 2. g3 d5 3. Bg2 Nf6 4. Nf3 dxc4 {Seymour is a solid player, and this is reputed to be one of the most solid defences to the English/Reti. Nevertheless it was precisely here that I chose to give battle. I've been fascinated by these ...dxc4 lines ever since I discovered them, and how White can often create 'invisible' compensation for a gambit pawn in positions where Black has no weaknesses and the two armies haven't come into contact. In that spirit, I was happy to follow my own old recommendation and played} 5. O-O {, which I analysed in the notes to Young-Picot, Alexander Cup 2019. In that game I played the usual 5. Qa4+, which recovers the pawn and avoids any risk, but costs two tempi and develops the queen prematurely. The text move intends to capture the pawn more harmoniously later, or goad Black into compromising himself to cling onto it.} a6 {When I annotated the Picot game this line was still very fresh, and I was surprised to conclude that the text is Black's only way to equalise; I didn't think it would be the first move to come to mind if caught by surprise. A lot more games have been played since then however (even if 5. Qa4+ remains far more popular) and what was once secret knowledge is now well established.} 6. Qc2 c5 7. Ne5 Ra7 {Not the first time Seymour has played this move. It's computer-approved and he won previously, but if you're going to be this unbendingly static in your opening choices then you need either exceptional feel for the positions or a very low-risk repertoire. And whilst ...Ra7 has a logical basis, intending to meet a capture on c4 with ...b5 and ...Bb7 to neutralise the dangerous long diagonal, ignoring development for four moves running is inherently risky. So...} 8. Rd1 $5 {The ambush! I got this idea from Ntirlis's English repertoire book, where it's mentioned as a sideline without analysis. It seemed a good surprise weapon, having been played just once a decade ago, but in any case White is justified in making a real gambit of it to exploit his lead in development: the next move will be d2-d3(d4), abandoning efforts to recover the c4 pawn but forcibly opening the centre.} (8. a4 {is more usual.} b6 {Here in both Hill-Seymour, Kingston Invitational (!) 2023 and Amato-Seymour, 4NCL 2023, Black faced 9. Qxc4 and scored 1.5/2. Instead the main line is} 9. Na3 Bb7 {, fulfilling the strategic aim of his seventh move. Ntirlis continues with the sequence} 10. Naxc4 Bxg2 11. Kxg2 Be7 12. b3 O-O 13. Bb2 Nd5 14. Qe4 {, intending d2-d4, and claims an edge for White, but I was in a more ambitious mood.}) 8... Nbd7 {After the game Seymour said that he'd analysed 8. Rd1!? before, since it's the move suggested by Stockfish. When I heard that, I realised the pre-game warning was well justified, though I already had an inkling since he proceeds to reproduce the engine's main line over the next few moves. Even so my surprise soon yielded a huge time advantage as Seymour tried to recall his analysis.} (8... Be7 {is the other move that springs to mind, hurrying to get the king out of the centre. I believe this is also playable, though quiet development allows White to demonstrate his idea in a purer form than the more forcing game continuation.} 9. d3 (9. d4 $144 {is also possible; this is likely to transpose since} cxd4 $2 {compliantly opens all the lines whilst accelerating White's development even further.} 10. Be3 Bc5 11. Qxc4 {Black can't preserve his material advantage in view of} b6 12. b4 $16 {, therefore he will be left without anything to make up for his stationary pieces.}) 9... cxd3 10. Nxd3 (10. Rxd3 {can also be considered, however it's not a real tempo on the queen since} Qc7 11. Bf4 Nh5 $8 {is a nuisance. The tactical point is revealed after} 12. Nxf7 Nxf4 13. Nxh8 Nxd3 14. Qxd3 g6 $44 {, when White has regained his pawn and deprived Black of castling rights, but his knight is trapped and will have to be given up on g6. This position could certainly be investigated further but I wasn't able to find more than a repetition for White.}) 10... O-O 11. Bf4 Nbd7 12. Nc3 $44 {This is a good example of the 'invisible' compensation I mentioned earlier. Black's extra pawn looks healthy enough, but try playing some natural moves against the engine and see what happens! Ideas such as Ne5 followed by Na4-b6, piling up against the pinned knight on d7, promise White a dangerous initiative. A possible continuation offered by Stockfish is} Qe8 13. a4 b6 14. Bc6 Bb7 15. Ne5 {, and already Black is forced to find the unnatural 15... Ba8 to stay afloat - not} Bxc6 $4 16. Nxc6 Rb7 17. Nxe7+ Qxe7 18. Bd6 {winning the exchange.}) 9. d4 cxd3 $146 {A new move, though one familiar to both players.} (9... Nxe5 $143 $2 {was played in the stem game between two strong GMs, however the combination of a cramping pawn on e5 and yet another tempo for White proved too much to bear.} 10. dxe5 Nd7 11. Nd2 b5 12. a4 Qc7 {In Salgado Lopez-Pelletier, French Team Ch. 2013, White instructively built the initiative with} 13. Ne4 $1 {(remember this idea for later!)} Nxe5 {and now the most powerful would have been} 14. axb5 $146 {(White was successful with 14. Bf4 in the game, which is also pretty good)} axb5 15. Rxa7 Qxa7 16. Bf4 $18 Ng6 17. Qd2 {and Black's lack of development is fatal in light of White's control of the open central file, for example} Be7 18. Nd6+ Bxd6 19. Bc6+ $1 Kf8 (19... Bd7 20. Qxd6 $18 {is also decisive as the black bishop is pinned to mate on d8}) 20. Qxd6+ Kg8 21. Qd8+ Nf8 22. Bd6 {winning at least a piece.}) 10. Rxd3 {(1:32)} Nd5 {(1:22) Using tactics to hold his position together: now the knight on e5 is hanging and ...Nb4 is threatened, so White has to simplify and ease the pressure slightly. The next few moves are forced on both sides.} (10... Be7 $143 $6 {White gets a good game if he is allowed to proceed with normal development, for instance} 11. Bf4 {, with ideas of Nc4-d6, and if} Nh5 12. Be3 Qc7 13. Nxd7 Bxd7 14. Rc3 $14 {. White regains his pawn whilst retaining his open lines and more active pieces. Note that the greedy} b6 15. b4 Bf6 $2 {doesn't work because of} 16. bxc5 Bxc3 17. cxb6 {and too many pieces are hanging.}) 11. Bxd5 Nxe5 12. Bc6+ Bd7 13. Bxd7+ Nxd7 {At the end of the forcing variation Black has succeeded in exchanging off two pairs of minor pieces, including the 'Catalan bishop' on g2. Nevertheless, given White's lead in development, and therefore how long it will take Black to contest the d-file, his problems are not yet over.} 14. a4 {According to Stockfish this is a 0.00 position, and my preparation had confirmed the accuracy of that evaluation. Objectively then my search for an advantage was a failure, but consider instead that White has a whole pawn's worth of compensation in a position where nothing is happening and there is no forcing line for Black to rely on. That seemed like an excellent practical weapon, and my instinct was vindicated by the course of the game.} b6 15. Bf4 {(1:32)} Be7 {(1:05)} 16. Nd2 {Up to here was still preparation, and I was blitzing to make sure Seymour knew it: I've found it useful to have a reputation as someone who lures opponents onto ground where mines have been laid around the entire perimeter. Perhaps his next move should be viewed in that psychological context.} g5 $2 {In a position where he is 45 minutes behind on the clock and White has an automatic plan of strengthening his bind with Nc4 and Rad1 if left unmolested, Black gives in to stress and lashes out with an aggressive-looking move. In truth, it merely creates further weaknesses whilst turning the safety of his king into a permanent problem.} (16... O-O $142 {surely ought to be automatic. Black has enough time to get out of the pin on the d-file.} 17. Nc4 Qc8 {The follow-up I'd prepared, following the computer, was} 18. Bd6 {(an operation to make the invasion square on d6 available to White's rooks)} Bxd6 19. Rxd6 Rb7 20. a5 bxa5 (20... b5 $143 $6 21. Nb6 Nxb6 22. axb6 $14 {favours White despite his material deficit, because the backward pawn on a6 is weak, Black's pieces are passive and he struggles to play around the b6 pawn.}) 21. Rxa5 {Black's pawn is devalued now that his queenside has been splintered. However, he can equalise with the tactical sequence} Nb8 22. Nb6 Qc7 23. Qxc5 Qxc5 24. Rxc5 Rxb6 $1 25. Rxb6 Nd7 26. Rcc6 Nxb6 27. Rxb6 $11 {with a drawn ending, which my opponent would certainly have deserved if he had found all this. Alas, there's many a slip twixt the cup and the lip...}) 17. Be5 Bf6 $2 {I expected this move, challenging my bishop and claiming the long diagonal, however after my reply Black faces a choice of poisons: to lose even more time in an under-developed position preparing an escape route for his king, or leave it in the centre to face the fury of White's heavy pieces. 17... O-O really was better, even though 18. Nc4 essentially leaves White a tempo up on the previous variation.} 18. Bd6 Qa8 {At the time this did seem like a clever defence. Black unpins his knight, prevents my knight landing on e4 and defends the a7 rook so that 19. Nc4 can be met by ...b5. Even so, White's initiative should already be decisive.} (18... Be7 $144 {was what I expected, however Seymour still liked his position and didn't want to allow a repetition. Nor did I! In this case White can continue calmly with} 19. Ne4 O-O {(finally, but Black has tarried for too long)} 20. a5 b5 21. Qd2 h6 22. Rd1 $18 {, when Black is bound hand and foot.}) 19. Qc4 $2 {Unnecessary prophylaxis against ...b5, which dilutes White's tempo advantage.} (19. Rd1 $142 {White has no need to fear Black's threats and should instead develop his last piece. I considered this, but after} b5 {I overlooked the less natural} (19... Be5 20. Nc4 {gives Black no respite because of the tactical point} Bxd6 21. Nxd6+ Ke7 22. Nxf7 $18) 20. Rf3 $1 {, indirectly pressuring f7, locking Black's knight in place and allowing mine to invade via e4.} Be7 21. Ne4 f5 {Taking on d6 would hang this pawn, but after} 22. Nxc5 Nxc5 23. Qxc5 $18 {Black's position is riddled with weaknesses and his king remains caught in the centre.}) 19... Qc6 $2 {The sincerest form of flattery! The queen is recentralised, and ...b5 is back on, and yet the loss of time neutralises my last move. Things would have been a lot less clear after either 19... Bxb2 or 19... Ne5.} 20. Ne4 $1 {(50) Seymour said later that he felt very confident at this point, and it rubbed off on me for I made this move with some scepticism, but my intuition did not betray me: White must fight for the initiative, and material considerations are of no great importance. This is the keen dynamic judgement of a 1. e4 player, illustrating yet again that it's essential to experience a wide range of positions to make effective decisions.} Bxb2 {(25)} 21. Rb1 $4 {Typical of my games: good ideas, but shaky implementation driven by weak tactical vision. I thought that if Black was given a move's respite he had chances to organise a defence, but of course my rook has nothing to do on the b-file. Meanwhile, Black's supposed defence to 21. Rad1 turns out to be illusory.} (21. Rad1 $142 Nf6 {was what I feared, but I missed that after} 22. Qc2 $1 Nxe4 23. Qxb2 $18 {Black is totally uncoordinated and has no good defence to the simple attack on h8. In particular, after} f6 {, seemingly covering everything, White has} 24. Bb8 {winning a rook. An omission caused by failing to look that little bit deeper; I didn't like having to exchange pieces and spend two tempi with my queen, and made an emotive evaluation.}) 21... b5 22. Qc2 Be5 $4 {Black lost this game by underestimating the dynamics, and in particular because he wasted too many tempi. Exchanging pieces is useful in isolation, and it allows him to finally dream of castling, however White is allowed to double rooks on the open file with tempo, after which Black's position is cut in half and his prospects immediately become hopeless.} (22... Qxe4 $142 {looks similar to 21. Rad1 above, however the stray rook on b1 (as well as the knight's sounder position on d7) makes a big difference because after} 23. Qxb2 f6 {the Bb8 resource is no longer available. White can only satisfy himself with a repetition of moves after} 24. Re3 Qd5 25. Rd3 Qe4 26. Re3 $11) 23. Bxe5 Nxe5 24. Rd6 $18 Qc8 25. Qc3 {(37) Driving the knight away from its central position, and taking control of the long diagonal. Black is still feeling the pain of his 16th move and is reduced to thrashing about.} Ng6 {(5)} 26. Rbd1 f5 27. Nxg5 Re7 {Seymour was also in desperate time trouble, and it was hard to believe he could last more than a few moves, but try as I might I couldn't find a clincher. Another of my calling cards...} 28. Qxc5 $6 {A spectacular move (28... Qxc5 29. Rd8#) but not one that I really wanted to play, since allowing Black to finally castle allows him to prolong his resistance.} (28. Qf6 $142 {was obvious, coming with the baleful threat of 29. Nxe6, but I couldn't see what to do next after the forced} Rf8 {. If only I had pushed through the adrenaline and slowed down (and I still had half an hour in reserve) I would have realised that no next move is required once castling has been ruled out - as Seymour rightly observed in the postmortem, it's sufficient to retreat the queen and ask Black to make a move with his flag hanging.} 29. Qc3 $18 {Now Qxc5 will be outright terminal, and if Black stops it with} c4 {then} 30. Qd4 {mates. This would have been an appropriate end to the game. The game continuation is also winning for White, just a lot more slowly.}) 28... O-O 29. Rc6 Qb7 30. Nxe6 {(21) Black has caught up in development, but White's pieces dominate the open central files and this is already the third sickly pawn he has harvested in the last few moves, with more to come.} Rfe8 {(0)} 31. Nd4 $6 (31. Nd8 $142 {I sensed this could decide matters, and rightly so, but couldn't see anything conclusive so preferred to keep the knight centralised.} Qb8 {The way to win is} 32. Qd5+ Kg7 33. Rxa6 $18 {, and the b-pawn goes too since Nc6 is threatened. Even worse for Black is} Rxe2 $2 34. Qf7+ Kh8 35. Rd7) 31... Re5 32. Qc2 bxa4 33. Qxa4 a5 34. Qc4+ (34. Nxf5 $1 {wins another pawn. I saw the main point (34... Rxf5 35. Rxg6+), but not that} Rxe2 35. Qc4+ Kh8 36. Qd4+ Ne5 37. Re6 $1 {leads to a win based on the precarious position of the black king. However White has to see all the way to} Qb8 38. Rxe8+ Qxe8 39. Qd8 Ng4 40. Qxa5 $18 {, when there is no defence to the threat of 41. Rd8. There's no need to rely on calculating this deeply when the position is already so favourable.}) 34... Qf7 35. Qxf7+ $5 {(11) The decision of a practical player. White is willing to settle for a highly favourable ending in which he is already one pawn up and the straggler on a5 will soon be picked off. Black is thus permitted to play another 43 moves, but in a winning position why take any risk?} Kxf7 {(1)} 36. Ra6 Rd5 37. e3 {Now that the centre is stabilised I can turn my attention to the a5 pawn. Black can't organise a defence in time because his knight is too far away - even in the endgame the impact of his sluggish development continues to be felt.} Ne5 $6 (37... f4 {was what I expected, since in the endgame the defender should aim to exchange pawns. This also allows the black knight to become active, however White can keep his counterplay in check without too much trouble.} 38. gxf4 Nxf4 39. Ra1 Nh3+ 40. Kf1 Rg5 {Black is close to generating threats against f2, which White can snuff out with} 41. Ra7+ Kg8 (41... Kg6 {leaves the king more exposed than active in a double rook ending. White can now play} 42. R1xa5 {since} Rg1+ 43. Ke2 Rg2 44. R5a6+ $18 {followed by a rook check on the seventh rank wins material.}) 42. Nf3 Rf5 43. Kg2 Kh8 44. Kxh3 Rxf3+ 45. Kg2 Ref8 46. Ra2 $18 {, reaching a won rook ending once the a-pawn is captured.}) 38. Ra1 Nc4 39. Ra4 {My tactical vision was on point in the endgame, partly redeeming my earlier play. By attacking the knight White begins a combination which leads to the win of the a-pawn.} Rc8 40. Rc6 $1 {(8)} Nb2 {(1) Allowing a desperado sacrifice, but the a-pawn was doomed anyway.} (40... Rxc6 $144 41. Nxc6 {could have led to a decoy sacrifice instead, since I was hoping for} Rc5 42. Rxc4 $1 $18 {winning a piece.} Rxc4 43. Ne5+) 41. Rxa5 Rxa5 {(if 41... Rxc6 42. Nxc6 and the other rook is now defended)} 42. Rxc8 Nd3 43. Rc3 {The last difficult move in the game, expelling the knight before Black can threaten f2. With four sound pawns against two weak ones, the outcome is no longer in doubt even with play limited to one side of the board.} Ne5 44. Kg2 Ng4 45. h3 {(5)} Nf6 {(0)} 46. Rc2 {White is playing to force the exchange of rooks, cooling the dying embers of Black's counterplay. But first he takes control of the second rank, just in case. There is no need to hurry.} Kg6 47. Rb2 h5 48. Rb5 Rxb5 {Otherwise the f5 pawn is lost, but a material advantage is felt very strongly in pure knight endings.} 49. Nxb5 Ne4 50. Nd4 {(6)} Nd2 {(1)} 51. f3 Nc4 52. Kf2 Kf6 53. Ke2 Nb2 54. h4 {Fixing a weakness on h5. Now it only remains to bring my knight to f4, driving the black king all the way to h6 to defend the pawn, and then escort the central pawns up the board with my king.} Nc4 55. Kf2 {(7)} Ne5 {(2)} 56. Kg2 Nc4 57. Kf2 Ne5 58. Ne2 Kg6 (58... Nd3+ $144 {would have temporarily pushed my king to the back rank (this was why I repeated once) but} 59. Kf1 Ne5 60. Kg2 $18 {and White will not be denied.}) 59. Nf4+ Kh6 60. Ke2 {(6)} Nc4 {(3)} 61. e4 {The final stage of the winning procedure: White creates a passed pawn in the centre.} fxe4 62. fxe4 {To be honest I expected Black to resign here, but since I played 40 superfluous moves the morning before I can't really complain about this.} Ne5 63. Ke3 Nc6 64. Kd3 Nb4+ 65. Kd4 {(7)} Nc2+ {(1)} 66. Kd5 Ne3+ 67. Kd6 Nf1 68. Ne2 {By now White can even afford to release the black king from its prison.} (68. e5 $144 Nxg3 69. e6 Nf5+ 70. Kd7 Kh7 71. e7 Nxe7 72. Kxe7 $18 {was another way to win. I was trying to convert without even letting Black give up his knight for the e-pawn, but in the end had to allow him that small triumph.}) 68... Kg6 69. e5 Ne3 70. Nf4+ {(6)} Kf5 {(3)} 71. e6 {Played after calculating to the end. 71. Nxh5 is also good enough to win.} Kg4 72. e7 Nf5+ 73. Ke6 Ng7+ 74. Kf7 Nf5 75. e8=Q {(4)} Nd6+ {(5)} 76. Kf8 Nxe8 77. Kxe8 Kxg3 78. Ng6 {Black loses his final pawn to zugzwang. Though it was hardly perfect, I think Botvinnik would have appreciated my play in this game. I prepared an effective opening idea right through to a middlegame plan, showed good feel for the initiative and converted the endgame with aplomb. Even the string of missed wins in the middlegame was very Botvinnik-esque. As such I consider this one of my best fighting games.} 1-0

One of the most (in)famous games in the Botvinnik canon is this one against Larsen. After Black misplays the opening everything is poised for a sacrificial onslaught against his denuded king, but after missing a series of forced wins White made the controversial decision to abandon the attack and take things to an ending which he eventually won with the neat breakthrough 44. g4! after the adjournment. His exasperation comes through clearly in the notes ("really it was a comedy of errors [...] on six separate occasions I missed a forced win [...] who knows how many more might still be discovered?") but he argued strongly that the endgame was the right practical decision.

This game ended up as a tribute act to Botvinnik-Larsen, right down to the opening moves; even the ...Qa8 manoeuvre which got Black into trouble is identical. I suffered from the same tactical short-sightedness and the same feeling that the game was somewhat grubby as a result; thankfully I also imitated the good endgame technique called for by making the same practical decision. And this was above all a good practical game, even if its aesthetic value is compromised by not ending in mate around 50 moves earlier.

It is a shame to have a drawn ending as the cornerstone of your preparation as White, but there are 0.00s and there are 0.00s. My decision to target this one in particular was guided by my overnight diagnosis, and sure enough a tense position where White has good general compensation for a pawn but no actual threats for the time being proved to be very uncomfortable for him. 16... g5? is a nervy, weakening move not in keeping with the position when Black should be hurrying to catch up in development, and both 19... Qc6? and 22... Be5?? cost too much time. Simply put, I handled the dynamics better, and it was no accident that I made a dynamic position appear on the board. I made my fair share of oversights in this game, especially when concrete analysis was called for (I misevaluated 21. Rad1 and 28. Qf6 because of an inability to slow my calculation down) though I was happy with my endgame play for a change.

Just like that, back to 50%. Now I was definitely warmed up! The afternoon round would be an even stricter test, as I got downfloated to play an IM with Black and with history definitely not on my side.