The world is not enough

Round 5 at the Kingston Invitational gave me the increasingly predictable assignment of playing against IM Peter Large (2391). After my gut-wrenching loss to him at Kidlington in February I had mixed feelings ahead of this game: although I had to some extent cracked the code, and knew that putting him under pressure in a chaotic position gave me chances to confuse him, I likewise play unreliably in such positions and my history of squandering good positions against him created a feeling of inevitability. Also, in a slight change to the usual formula, he had the white pieces this time, and that is always a stern test of opening erudition even for me.

This game is perhaps most noteworthy for its psychological content, as the actual chess was rather poor. I duly lost the opening battle, however a few careless moves on Large's part enabled me to regain my 'sacrificed' pawn whilst building a strong attack on his king. From there things started to get really strange:

[Event "Kingston Invitational"] [Site "?"] [Date "2024.08.14"] [Round "5.14"] [White "IM Large, Peter G"] [Black "Young, Daniel J"] [Result "1/2-1/2"] [ECO "C68"] [WhiteElo "2391"] [BlackElo "2017"] [Annotator "Young,Daniel J"] [PlyCount "56"] [EventDate "2024.??.??"] [SourceVersionDate "2024.08.18"] {After consecutive wins I was, naturally, feeling pretty good. This pairing was a downfloat, and though I have a bad score against Large it belies the positions I've had: I shoot for the stars against everyone, and he doesn't like being put under pressure!} 1. e4 {(1:30)} e5 {(1:30)} 2. Nf3 Nc6 3. Bb5 a6 4. Bxc6 {Although there was no time to prepare between rounds, I was at least expecting this Large speciality. The next few moves are standard theory.} dxc6 5. O-O Bg4 6. h3 h5 {(Of course, the bishop is untouchable, though Black has also tried the pawn sacrifice 6... Bh5!?)} 7. d3 Qf6 8. Nbd2 g5 {This turned out to be my last move in book. Since White has castled in front of a compromised pawn shield, Black immediately begins a pawn storm, intending ...Be6 and ...g4. This idea is so strong that White usually plays 9. Nc4 to provoke a queenless middlegame, but I was wrong to imagine that he is forced to.} 9. b4 $5 {White anticipates queenside castling by Black and prepares to fianchetto his dark-squared bishop, with pressure on e5. Large has played me enough to know that embracing a complex struggle where I'm forced to calculate is usually the right policy, and so he sows the wind... but on this occasion he reaps the whirlwind!} (9. Nc4 {All told I still think this is the most sensible move, though Black is holding his own here. The threats against g5 and e5 force simplification on f3.} Bxf3 10. Qxf3 Qxf3 11. gxf3 f6 {White's main try is Fischer's} 12. h4 gxh4 13. f4 {, sacrificing a pawn to break Black's bind on the kingside and activate the c1 bishop, however} b5 14. Na5 exf4 15. Bxf4 Kd7 $11 {seems to cover everything; Black continues with ...Bd6 and ...Ne7, comfortably completing the development of his pieces.}) 9... Be6 $6 {I rejected 9... Bxb4 at once in view of 10. Rb1, but the move played also hands the initiative to White. This is a natural reaction if you've learnt that 9. Nc4 is essentially forced, however in making it I overlooked White's 11th move.} (9... Qg7 $142 {is the approved continuation. The queen stays in touch with the h-file, ensuring Black's bishop remains untouchable, will not be attacked after a later Bxg5, and frees the f-pawn to support its neighbours on e5 and g5. If Black is given time to implement this idea then White will be in trouble, so he has to act quickly, however this time after} 10. Nc4 {Black is ready to play the surprising} Bc8 $1 {, retreating the bishop to a square where it isn't attacked after Nxg5(e5) and offering a sham pawn sacrifice for active (or perhaps 'hacktive') play on the kingside. White can now temporarily win a pawn in three different ways, but none of them is really satisfactory. Best seems to be} 11. Bxg5 (11. Nxg5 $6 f6 12. Nf3 Bxh3 $15 {is similar to 11. Bxg5 except that White's bishop remains undeveloped on c1}) (11. Nfxe5 $2 {is definitely out since} g4 12. h4 f6 $17 {unceremoniously traps the knight in the middle of the board.}) 11... f6 12. Bd2 {, when White at least gets a tempo out of the trade of g5 and h3 pawns. After} Bxh3 13. Nh4 O-O-O 14. a4 Ne7 15. Ne3 Bd7 $11 {the engine gives a cold-blooded assessment of equality, but I'm not surprised that Black has racked up a big score in practice since he is ahead in the race to prise open the enemy king position. Large doesn't like being attacked (cf. the game!) so he would not have enjoyed this.}) 10. Bb2 {(1:27)} Bd6 {(1:29) Now Black would stand excellently if not for} 11. d4 {, and I cannot take because of the skewer on the long diagonal. So White is bound to win material, and after two independent moves I already find myself fishing for counterplay.} Qh6 $2 {Plausible but insufficient. I played this anticipating the game continuation - though both sides missed a big improvement for White - and after spending some time satisfying myself that 11... g4 didn't work - which it does.} (11... Qg7 $2 {was played in the high-level blitz game Naiditsch-Anton Guijarro, Biel 2020, the only previous game to reach this position. This doesn't even have the hit on d2 to recommend it over my move, and after} 12. Nxe5 Bxe5 13. dxe5 Ne7 14. Nb3 Rd8 15. Qc1 $18 {Black was denied ...Ng6, he remains a pawn down and White's knight is heading for c5.}) (11... g4 $142 $146 {is therefore essential. After} 12. dxe5 {I only considered 12... Qg6 which is indeed hopeless for Black, but the tactics on the long diagonal work both ways since White's bishop is undefended, so} Qg7 13. exd6 Qxb2 {Interestingly Stockfish now wants to change the story and make it a sacrifice of a piece for three pawns by White, and continues} 14. dxc7 $5 (14. Ng5 O-O-O 15. Nxe6 fxe6 $44 {will lead to Black regaining his pawn with easy counterplay on the kingside, since} 16. dxc7 $6 Kxc7 {is rather too greedy; 17... Qc3 is threatened and White has no time to prevent the opening of the kingside.}) 14... gxf3 15. Nxf3 Qf6 $11 {This at least has the benefit of clarity for White, and also prevents Black's king scurrying away to safety on the queenside. Even so, an extra piece is worth some trouble in the middlegame, and this is certainly better than how the game could have turned out for me.}) 12. Nxe5 $6 {What could be more obvious than winning a pawn and getting out of the g-pawn's firing line? Still, I was happy to see this as now I at least to get to play a game in spite of my shaky start.} (12. d5 $142 $1 {I had a vague sense of unease about this move at the board, though I couldn't understand why. Still, opening the centre is logical when Black has neglected his development.} cxd5 13. exd5 Bxd5 {the point is to follow up with} 14. c4 {, regaining the initiative thanks to this rolling pawn. Black has to try and mix things up with} g4 (14... Bxf3 {to win time for} 15. Qxf3 O-O-O {is still too slow, as} 16. c5 Be7 17. Bxe5 $18 {regains the pawn with a powerful initiative whilst Black's kingside pieces are out of play:} Rxd2 {doesn't work because of} 18. Qc3) 15. Nxe5 f6 {, however} 16. cxd5 fxe5 17. h4 Nf6 18. Nc4 $16 {gives White a big positional advantage. He has managed to keep the kingside shut, his pieces are active and the e5 pawn is none too healthy.}) 12... Bxe5 13. dxe5 O-O-O {Now there is a pin on the newly opened d-file and Black has succeeded in generating some counterplay for the lost pawn. And immediately my opponent's discomfort starts to show.} 14. Qe2 $6 {White understandably seeks to unpin, however this allows me to get my next move in with tempo, and he starts to lose control of the position.} (14. Bc3 $142 {instead defends the knight in advance, so that he can meet} g4 {with} 15. h4 {At first the engine evaluation is heavily in White's favour, but I don't think Black needs to despair just yet. Soon White will have to weaken the light squares around his king further with g3, otherwise Black's knight will land on f4 via e7 and g6. The doubled e-pawns also don't carry full value and work to blunt the long diagonal for his bishops.} Ne7 16. g3 Ng6 17. Qe2 {The most active square for the queen, connecting the rooks. Now if Black does nothing White will go over to the attack with 18. a4 followed by b4-b5, so the only hope of muddying the waters is} Nxe5 $5 {, sacrificing the exchange but gaining room to manoeuvre for Black's pieces.} 18. Bxe5 Rxd2 19. Qe1 {Threatening 20. Bf4, so Black can't save both of his rooks.} Rxc2 20. Bxh8 Qxh8 $14 {Black only has one pawn for the exchange, but his queen, rook and bishop are all active, and it's hard for White's king to feel entirely safe with an enemy light-squared bishop on the board.}) 14... g4 15. Rad1 $6 {(1:18) An obvious developing move, however the computer suggests a more active defensive try.} (15. f4 $142 {I considered this, but assumed that White should not create even more weaknesses in front of his king. However the central pawn mass controls some useful squares and White can fight for the initiative this way, rather than face one-way traffic as in the game. The point is that} gxh3 16. Rf2 {defends the knight anew, and Black has to deal with the threat of 17. f5. Therefore} f5 17. exf5 Bxf5 18. Nf3 Ne7 19. Ng5 $11 {, gaining a tempo with the threat of Nf7. In the world of Stockfish this is balanced, although in such a sharp position that evaluation feels meaningless in a game between humans. I would still be happy as Black here, for the simple reason that my king is safer.}) 15... gxh3 {(1:15)} 16. g3 Ne7 {I was starting to get excited about the position here, but I did the disciplined thing and made sure I was attacking with all my pieces. White can't stop me playing ...h4 in any case, so there is no need to rush.} 17. Kh2 h4 $17 18. Nb3 Rdg8 {White would be delighted to simplify, so naturally I preferred to transfer the rook to the main theatre instead. Conceding the d-file is immaterial.} 19. Qd2 $2 {This move, again borne out of the natural desire to chop wood, allows a combination which I saw and then proceeded to not play!} (19. Rg1 $142 {was what I expected, reinforcing the tender g3 square, however Black can insist on its weakening with} Bg4 20. f3 hxg3+ 21. Rxg3 Be6 22. Rdg1 {In this way White does achieve his cherished trades, but} Rxg3 23. Rxg3 Bxb3 {(preventing White relieving the weakened dark squares with Bc1)} 24. axb3 Qf4 $17 {ties up his kingside completely. Even if White does somehow limp into an ending, it's hard to imagine it being a pleasant one for him given the two sets of doubled pawns.}) 19... hxg3+ 20. fxg3 {(1:06) I have never claimed to be a tactical genius, but even so I spotted 20... Rxg3! almost immediately after White's 19th move. And not just me - I saw it, Large saw it, Stephen Moss and John Saunders who were blogging the tournament saw it, and anyone watching at home with an engine certainly saw it. Why then did I play} Qh5 $2 {(57) instead? Simply put, this was where all the winning positions I've had against Large in the past started to flash before my eyes...} (20... Rxg3 $142 $1 {I satisfied myself without much analysis that White couldn't take the rook and live to tell the tale, but that wasn't why I rejected this move.} 21. Kxg3 (21. Qxh6 {was what gave me pause. After} Rg2+ 22. Kh1 Rxh6 $19 {Black has won a pawn, but nevertheless I feared I was selling my advantage too cheaply; after all the disappointments in the past I really wanted a direct kill and no room for ambiguity. However, if you look at this position on its own merits, rather than being influenced by what came before, it's obvious that Black is completely dominating. As well as the extra pawn, he has a rook on the second, White's pieces are passive and the weak pawns on c2, e4 and e5 are all targets. Certainly against someone of my own strength this would have been no impediment to playing 20... Rxg3.}) 21... Rg8+ {White cannot keep his king under cover, and with most of his pieces sidelined from the defence the end is not long in coming. The relatively best hope seems to be} 22. Kh2 (22. Kf3 {going for a walk ends about as well as one would expect:} Bg4+ 23. Kf2 Qh4+ 24. Ke3 Qg3+ 25. Kd4 Rd8+ 26. Kc5 b6+ 27. Kc4 Be6+ $19) 22... Rg2+ 23. Kh1 (23. Qxg2 hxg2+ 24. Kxg2 {With all of the lines open, White is swiftly overwhelmed:} Qg5+ 25. Kf2 Qf4+ 26. Ke2 Bc4+ 27. Rd3 {(27. Ke1 allows an epaulette mate on e3)} Qxe4+ 28. Kd1 Bxd3 29. cxd3 Qxd3+ $19 {Rook and bishop are no match for queen and two pawns.}) 23... Rxd2 24. Nxd2 {(or 24. Rxd2 Qe3, winning the e4 pawn)} Qe3 25. Rg1 Ng6 $19 {On pure numbers White is in the game, with RR vs QP, but he is completely tied down and his king remains in jeopardy. White's minor pieces are on the wrong side of the board.}) 21. Qe3 {Of course, without a queen trade I would have snapped at a second opportunity for ...Rxg3, but I didn't really expect it to happen. Instead I was planning on ...Ng6xe5, so continued as advertised with} Ng6 $2 {, but ignoring White's right to create threats was very unprofessional. At this moment I thought his next move was unplayable, but whilst he was thinking I realised it wasn't the case. I should know not to trust my calculation, and it really would have been simpler to prevent it outright.} (21... Kb8 {was the move I regretted not playing afterwards, but White still gets some chances with} 22. Nc5 {. Therefore the best solution is}) (21... b6 $142 {, both shutting the g1-a7 diagonal and controlling the c5 square. After} 22. Nd4 Bc4 23. Rg1 Rd8 24. Rde1 Rhg8 $15 {Black still retains the better chances thanks to the ugly e-pawns and his more secure king.}) 22. Qa7 Qe2+ $8 23. Qf2 $4 {A big relief, especially as Large had spent 34 minutes on his previous move and another nine on this one. In that time I'd realised that now it should be White's turn to sacrifice a rook.} (23. Rf2 $142 $1 {I thought this should be played to seize the d-file, but the real motive is just to win a tempo! After} Qxd1 {White should continue with the quiet} 24. Nc5 $1 {, sealing off the d7 square. The only way for Black to stop Qa8# is to flee back towards the centre, when White secures a perpetual check without much trouble.} (24. Rd2 $143 $2 {was what I expected. This has similar ideas, but with the important difference that Black can return material and consolidate his king position.} Qxd2+ 25. Nxd2 Rd8 26. Qa8+ Kd7 $17 {Black is quite safe here since White is attacking with a single piece. An important point is that} 27. Qxb7 $4 {loses to} Ra8 {followed by ...Rfb8 and the queen is trapped.}) 24... Kd8 $8 25. Qb8+ Ke7 26. Qxc7+ Ke8 {Black can't step onto the f-file because of the pin from White's rook, so the king must remain in the centre. Equally, White is a whole rook down, so the most prudent is to immediately force a draw with} 27. Nxe6 fxe6 28. Qc8+ Qd8 29. Qxe6+ Qe7 30. Qc8+ Qd8 31. Qe6+ $11) 23... Qxe4 {Large said in the postmortem that he'd overlooked this move. I'm not sure if he meant it literally, or that he'd missed what it allows next turn.} 24. Rfe1 Qg2+ {This is almost forced as otherwise the enemy queen is going straight back to a7, but in any case trading queens and returning my extra pawn is a price well worth paying to finally open the h-file.} 25. Qxg2 {(22)} hxg2+ {(47)} 26. Kxg2 Nf4+ {Black's main problem here is that he has so many good moves to choose from.} (26... Rh3 {is the computer preference. The rook takes the lead on the h-file before Black plays ...Nh4+ next move, after which White can't defend both the g-pawn and the entry square on h2 with his king.} 27. Rg1 {At least saving the pawn, but} Nh4+ 28. Kf2 Rh2+ 29. Ke3 Ng2+ 30. Kf2 Nf4+ 31. Ke3 Nd5+ $19 {leaves White without a good king move, and nothing better than giving up the exchange on d5.}) (26... Bd5+ {was the move I first intended, but I talked myself out of this for the nonsensical fear of} 27. Rxd5 (27. Kf1 {is Stockfish's attempt, but} Ne7 $19 {soon breaks through on the kingside files}) 27... cxd5 $6 (27... Nf4+ 28. Kf3 Nxd5 $19 {was pointed out immediately by Large in the postmortem}) 28. e6 {, as, after all, my opponents' passed pawns always behave differently. In truth, here too the pin on the g-file allows Black to untangle easily.} Rh7 29. e7 Nf4+ 30. Kf3 Ne6 $19) 27. Kf3 Nd5 28. Rd4 Rh2 $19 {Here, for reasons that I later came to bitterly regret, I offered a draw, which Large accepted after some thought. I believed that I stood slightly better, but that essentially the moment had passed and the momentum had shifted. Great was my surprise, and my disappointment, when I discovered that Stockfish evaluates this position as better for me than at any other point in the game!} (28... Rh2 {Clearly, my misevaluation of this position is of a piece with the same faulty judgement of 20... Rxg3 21. Qxh6: I had too little confidence in my technical ability, or more accurately too much faith in my opponent's ability to slip away yet again. Obviously Black should play on, if for no other reason than that he can hardly lose. But let us consider some moves. The c-pawn is attacked, and I expected White to defend it with} 29. Rc1 {For the moment White is grimly defending everything, but he has too many weaknesses and after} (29. Re2 {drops the g-pawn to either order of captures, most precisely} Rxg3+ 30. Kxg3 Rxe2 $19 {when the white king is driven away and Black retains his actively-stationed rook}) (29. Rd2 Rh3 30. Rg1 Nxb4 $19 {similarly wins a pawn whilst remaining overwhelmingly more active}) 29... Rg5 30. Rh4 Rf5+ $1 31. Ke4 Rxh4+ 32. gxh4 Rf4+ 33. Kd3 Nxb4+ 34. Ke3 Rxh4 $19 {his position collapses. It is all so easy when you can step away from the heat of the moment. It's a shame since I have largely cured the urge to offer draws which used to plague me so much, but maybe I was too desperate to get something out of this pairing for a change. In any case, it abruptly snapped the momentum I was starting to build, and I certainly paid for my cowardice over the next two rounds.}) 1/2-1/2

Was it all decided by a moment of madness? Perhaps. It is indeed hard to imagine me being as merciful against someone of my own strength, either in the final position or on move 20. Who else could have seen 20... Rxg3!, calculated it accurately and not played it?! Once upon a time I used to offer all sorts of unfathomable draws, but this was the first time that my old illness had shown itself for many years. I think this game was heavily coloured by the Young-Large games of the past, especially its immediate predecessor: I believed I had to find lines where I could kill him on the spot with no conceivable wriggle room to be able to win the game, and that gradually converting a clear advantage was insufficient. Hence in both cases I managed to convince myself that the clear advantage I had was no advantage at all. It need hardly be said that this line of thinking is utter nonsense, especially from someone who has beaten three IMs in classical games and two former top 10 players in less serious formats.

Turning to the game in itself, my poor reaction to 9. b4!? supports the idea that knowing a little theory is worse than knowing none, as I played on autopilot for a kingside attack and didn't respect White's possibilities. Nevertheless that move was, in my opinion, a poor practical decision by Large, because you have to be conscious of your own weaknesses as well as your opponent's. Besides, in principle randomising the game advantages the lower-rated player because it creates more risk. Much better to play 9. Nc4, idly torture me in an ending and let my nerves crack that way. The momentum shifted once I was able to castle to safety, and 14. Qe2?!, 15. Rad1?! and 19. Qd2? are all nervy moves. The consequences of that momentum shift continued to be felt later on. I could easily imagine him finding 23. Rf2! Qxd1 24. Nc5! if he hadn't been under pressure for so long, and then I would have been the one forced to think about how to save the game.

Speaking of momentum, giving away a cheap draw instantly snapped the positive energy I'd built up by winning the last two games. Karma for this act of cowardice was not long in coming, as both of my games the following day were unmitigated disasters. I'm going to cover them jointly in one short entry, as in my mind they are really one conjoined game.