Spots on the sun

Today's post is about my most memorable game of the 2021-22 season, though not necessarily the best. This was another 4NCL game for the perennially outrated KJCA Kings team, with black on board four against Blackthorne. That was a flattering place for me by rating, but after all sometimes the stars can align for me...

On a high board against a team with first-division pedigree, I had opposition to match: IM Andrew Ledger, rated 2398. Not only was this a surprise strengthening of their team over the day before, but I could expect to face a 4. e3 Slav, a passion-killing line that I do not have a good score against - at home with the engine running it all looks so harmless, but at the board... 

So I made the short drive from Staverton to Daventry without much optimism for the game, but instead things turned out rather differently. Ledger tried to surprise me in the opening, but I beat him to the punch; it seems the psychological shock was decisive as a string of questionable decisions left him in a lost position as white within 20 moves:

[Event "4NCL Div. 1/2"] [Site "?"] [Date "2022.05.01"] [Round "8.4"] [White "IM Ledger, Andrew J"] [Black "Young, Daniel J"] [Result "0-1"] [ECO "B12"] [WhiteElo "2398"] [BlackElo "2004"] [Annotator "Young,Daniel J"] [PlyCount "80"] [EventDate "2022.??.??"] [WhiteTeam "Blackthorne"] [BlackTeam "KJCA Kings"] {After a comfortable draw with a 2250 the day before, I was on board 4 against Blackthorne, an established first division team. My opponent hadn't played the day before, so I only had 90 minutes' worth of preparation to go on. Time control: 40 moves in 100 minutes + 50 minutes to completion + 30 seconds/move. } 1. e4 {(1:40) After a ten-minute drive through the rain from my hotel, I arrived to the board slightly late and was very surprised to see that this move had appeared on it - Ledger almost always plays 1. d4 and his last Sicilian was in 2006! However, surprised does not mean unprepared...} c6 $1 { (1:39) I didn't want to find out what Ledger had cooked up, and as my Caro-Kann games are off-database this was sure to unsettle him. I'd settled on this before the game, and correctly guessed that he would fall back on what he'd played before if ambushed.} 2. d4 d5 3. e5 c5 4. dxc5 e6 5. Nf3 Bxc5 6. Bd3 Nc6 7. O-O {Here the main move is 7.. . Nge7, which Speelman beat Ledger with in 2003. But I remembered Dreev writing that he didn't understand why Black didn't play} f6 $5 {- so I thought I'd give it a try! I didn't look any further, since time didn't allow, but I felt I had won the psychological battle. Also the fire alarm went off at this point, just to reinforce any confusion in Ledger's mind.} 8. exf6 $6 {A safety-first move. White avoids any surprises, but having given up his pawn centre and accelerated Black's development he can hardly aspire to an opening advantage.} (8. Qe2 $142 { is played more often.} fxe5 9. Nxe5 Nxe5 10. Qxe5 Nf6 11. Bb5+ {This temporarily displaces Black's king (11... Bd7?? 12. Qxe6+) but he can safely castle by hand:} Kf7 12. Bf4 Rf8 13. Nd2 Kg8 {. Still, I think that if White can get an advantage against 7...f6, he should look here and particularly at} 14. Bg3 {, stabilising his control of the central dark squares.}) 8... Nxf6 9. a3 $6 $146 {This bears some similarity to the popular 5. a3 system, but is less logical with the e-pawn gone as d6 is available; now it merely wastes time driving the bishop to a better diagonal whilst behind in development. More promising would be 9. c4 followed by 10. Nc3, leading to an IQP structure, though Black should be fine with all the pieces still in play.} O-O 10. b4 { (1:33)} Bd6 {(1:33)} 11. c4 {The position now bears some resemblance to a 3... Nf6 Tarrasch French, and accordingly I was looking at ideas like ...Bd7-e8-h5, ...Qe8-h5, ...Ng4/...Rxf3 and so on. White's pawn standing on c4 rather than d4 changes things though; on the one hand I have more freedom in the centre, and my dark-square weaknesses are less pressing, but in return White has more pressure on the pawn centre. Still, I would certainly take Black purely on stylistic grounds: a strong centre and active pieces in return for static weaknesses is a trade-off I will gladly make.} Kh8 $6 {"A very professional move", as one of our top boards put it, neutralising the a2-g8 diagonal and preparing my next two moves. This is a good move, but Black can already take the fight to Black more directly.} (11... Ne5 $142 {forces White's hand rather than letting him keep the tension, with the result that I'm able to expand in the centre without any concessions. I didn't consider this move, possibly filtering it out as too fanciful before completing development.} 12. c5 (12. Nxe5 $5 {is actually preferred by Stockfish, which is willing to lose a tempo and hence a pawn in order to disrupt Black's idea.} Bxe5 13. Ra2 {Black now wins material with} dxc4 14. Bxc4 Bxh2+ 15. Kxh2 Qc7+ 16. Kg1 Qxc4 $15 { , when White can gain time on the enemy queen and the e6 pawn is admittedly not much of a material advantage. Still, a pawn is a pawn!}) 12... Nxf3+ 13. Qxf3 Bb8 14. Qe2 e5 $15 {Now without a pawn break, White will struggle to develop counterplay against Black's rolling centre. There could hardly be a better advert for my seventh move!}) 12. Bb2 dxc4 13. Bxc4 e5 {Revealing the point of my 11th move - none of the thematic Tarrasch moves seemed to work in the face of my loose central pawns, so I concluded my chances were connected to ...e5, for which I would need to release the tension. I was feeling quite happy here and received a nasty jolt when Ledger played} 14. Ng5 $6 {after a lot of thought - I had analysed this, but hallucinated that the bishop was still on d3, allowing ...e4 and then ...Bxh2+. Now it dawned on me that I can't stop White winning the exchange, as 14... Qe7 is thwarted by the loose bishop on d6. However, once I regathered myself I realised that White's play is rather crude - it would be counter to chess logic for him to saunter in and win material just like that when his queenside is under-developed and my pieces are pointed at his king.} (14. Nbd2 $142 {is more realistic.} Bf5 15. h3 {Here the computer offers several ideas, but I would undoubtedly have continued directly with} e4 (15... Nh5 $5 {is a wacky idea, aiming for f4 and intending} 16. g4 e4 17. gxf5 exf3 18. Nxf3 Rxf5 {, producing a messy position where White's weakened king offers counterplay against the bishop pair.} 19. Be6 Rf8 20. Qb3 Nf4 21. Rad1 $13) 16. Nh4 {and now Stockfish gives} (16. Nd4 $143 $2 {Staying centralised is more natural, but White removes kingside defenders at his peril; after} Nxd4 17. Bxd4 Rc8 18. Qe2 Bb8 $17 {followed by . ..Qd6, the storm clouds are gathering over the white king.}) 16... Bxh3 $1 17. gxh3 Bf4 {, regaining the piece.} 18. Nxe4 (18. Nb3 $143 $2 {saves one knight.. .} Qc8 19. Kg2 $8 Bg5 {... but not the other. Trying to cling onto it with} 20. Bxf6 Rxf6 21. Qh5 $2 {would only make matters worse because of} Bxh4 22. Qxh4 Rh6 $19 {with twin threats of ...Qxh3+ and ...Rg6.}) 18... Nxe4 19. Qg4 { The simplest solution here is} Be5 $1 20. Bxe5 Nxe5 21. Qxe4 Qxh4 {- the point -} 22. Qxh4 Nf3+ 23. Kh1 Nxh4 $11 {, and whilst the bishop should be a better piece with play on both sides, White's kingside structure is inferior so chances are balanced.}) 14... e4 $1 {The exclamation is for keeping calm - in truth this wasn't a difficult move to find. I can't stop Nf7+, so I should play my own moves and let him get on with it. I am playing for a direct attack, so gaining space and opening the b8-h2 diagonal is obvious. This gives rise to some fascinating variations, but I didn't bother to calculate anything yet as it wouldn't change my decision - and rightly so, as Ledger opted to duck them.} 15. h3 $2 {(49) Ledger spent a while on this move and started to look uncomfortable. Indeed, White now begins to face serious problems: whilst he takes control of g4 and rules out Greek Gift sacrifices, weakening the king's position and losing a tempo in such a sharp position is risky to say the least! } (15. Nf7+ $142 {is obviously critical, and likely best.} Rxf7 16. Bxf7 { I did no more at the board than identify 16... Bxh2+ and 16... Ng4 as candidate moves - but this is analysis, so let's dig deeper. This would be a good position to use for training purposes; despite producing a mass of variations below I feel that there are still secrets to be discovered here.} Ng4 {This seems to be best, calmly bringing another piece into the attack.} ( 16... Bxh2+ {Black seems to be able to force a draw with the more obvious move, but no more than that.} 17. Kxh2 Ng4+ 18. Kg1 (18. Kg3 $143 $4 {is often a serious challenge to Greek Gift attempts, but in this case Black can regain the sacrificed piece with} Qc7+ 19. f4 Qxf7 {, when he is hardly down material and White's king is stranded; White is unable to meet the twin threats of ... Qg6 and ...Ne3.} 20. f5 Ne3 21. Qc1 Nxf1+ 22. Qxf1 Bxf5 $19 {at the end of it all Black emerges two pawns up, to say nothing of the white king.}) 18... Qc7 $8 (18... Qh4 $143 $4 {The more thematic continuation unexpectedly fails to} 19. Qd6 {, covering h2}) 19. g3 Qxf7 20. Qe2 {My instinct was that Black should be able to find his way to the enemy king here, as White's light-squared bishop is sorely missed. However, White is apparently able to defend, mostly with a well-timed f2-f3 to liquidate the e-pawn.} Bf5 {Simply developing another piece, and defending e4. However, the weakness of this pawn gives White time to get his pieces into play.} (20... Qh5 $2 {illustrates the point well; after} 21. f3 $1 exf3 22. Nd2 $1 {, exploiting Black's back rank, White is just in time to scramble a defence of his king, and as the game opens up the extra exchange will tell.} Be6 23. Nxf3 Qh3 24. Rfe1 Qxg3+ 25. Qg2 Qxg2+ 26. Kxg2 $16) (20... Nce5 {So next I tried this move, locking down f3 and bringing in reinforcements. Naturally White should reduce the attacking force with} 21. Bxe5 {(the immediate 21. Nd2 may also be playable)} Nxe5 22. Nd2 (22. Qxe4 $4 {is poisoned -} Nf3+ 23. Kg2 Bh3+ $1 24. Kh1 {(or 24. Kxh3 Ng5+)} Qh5 $19 {mates}) 22... Bh3 {and now the simplest defence is to return material with } 23. Qxe4 Bxf1 24. Rxf1 Re8 $11 {when the position is level.}) 21. Nd2 Qg6 { (rather than the more obvious h5 square, for reasons that will soon become clear)} 22. f3 {Again this move, but now there is a new resource:} Ne3 $1 23. Qxe3 Qxg3+ {forcing a perpetual check. This seems to be the limit of what Black can achieve with 16... Bxh2+.} 24. Kh1 Qh3+ 25. Kg1 Qg3+ $11) 17. g3 $8 ( 17. h3 $143 $4 {loses the queen to} Bh2+ 18. Kh1 Nxf2+ $19) 17... e3 {Anyone who played over the 16... Bxh2+ variations would have seen that this pawn can easily become a liability on e4, so once White has had to weaken his kingside pawns it's tempting to use it as a battering ram instead.} (17... Nxh2 $143 $4 {At Epsom Chess Club the next day, Graeme Buckley and I analysed the game together and concluded that this led to another perpetual, but we missed} 18. Kxh2 Qh4+ 19. Kg1 Bxg3 20. Qh5 $1 $18 {when the attack is repulsed}) 18. f4 Nf2 {Stockfish suggests this to be the best, forcing White to return the extra material - precisely why I don't want to recommend it! However, Black can't improve his position enough by quiet moves.} (18... Qe7 19. Bc4 Bc7 {is an interesting idea, redeploying the bishop to the a7-g1 diagonal, but by returning his queen White is able to survive.} (19... Nf2 {is possible here too, and after} 20. Rxf2 exf2+ 21. Kxf2 Bh3 22. Qe2 Qd7 23. Nc3 Re8 24. Qd3 Bf5 {we end up with a curious draw by repetition: White's queen must stay on the d-file to stop ...Bc7-b6+, but ...Bc5+ is also in the air so it is restricted to the light squares, and therefore perpetual attack from the bishop.} 25. Qd5 Be6 26. Qd3 Bf5 $11) 20. Qd5 $1 {White is virtually busted without this move, forcing Black's light-squared bishop to a clumsy square to prevent 21. Qg8#.} Bd7 21. Nc3 e2 22. Bxe2 Bb6+ 23. Kh1 {Black now wins White's queen with} Nd4 24. Bxg4 Bc6 {, however White is finally able to mobilise his remaining forces. } 25. Rae1 Qf7 26. Re5 Rd8 27. Qe4 Bxe4+ 28. Nxe4 $11 {White has rook, bishop and pawn for the queen, which usually isn't quite enough, but crucially his king is safe (now White is the only one with the right-coloured bishop) and he should be active enough to hold his own.}) 19. Rxf2 (19. Qe2 $2 {Greedy play will be punished! If White tries to preserve his material advantage then} Bc7 20. Nc3 Bg4 {forces White to part with his queen.} 21. Qxe3 (21. Qc4 $2 { is simply too much, turning the kingside into a veritable hornets' nest of invading minor pieces after} Nd4 22. Qd5 Nf3+ 23. Kg2 Qe7 $19 {, followed by .. .Rd8 and a crushing attack}) 21... Bb6 22. Qxf2 Bxf2+ 23. Rxf2 Qd3 $17 { Compared with the 18... Qe7 variation above, Black has preserved his light-squared bishop and White is less active, so his survival chances are correspondingly bleaker.}) 19... exf2+ 20. Kg2 (20. Kxf2 $143 $4 {will not do because of} Bc5+ 21. Ke1 Bf2+ 22. Ke2 Bg4+ $19) 20... Bf5 21. Nd2 (21. Nc3 $2 { allows Black's queen to reach the kingside with decisive effect after} Qf6 22. Bc4 {and now the stunning} Be5 $3 {, forcing White to expose his king to avoid losing a piece.} 23. fxe5 (23. Qb3 $2 {loses at once to} Nd4 24. Qd1 Nc2 $19) ( 23. Qf3 Bd4 $19 {leaves White virtually immobilised, with ... Re8 to follow.}) 23... Nxe5 24. Be2 Rd8 25. Qa4 Bg4 26. Bxg4 Nxg4 $19 {According to Stockfish, Black has mate in twelve - although he doesn't have many pieces left, White's king has no defenders at all.}) 21... Bc7 {Using the momentary obstruction on the d-file to reposition the bishop.} 22. Nf3 Qe7 23. Bc4 Rd8 {White can provoke a queen trade with} 24. Qe2 Qxe2 25. Bxe2 {, however this is not the end of his problems; after} Bb6 26. Bc3 Re8 $15 {White's pieces lack co-ordination, and the f2 pawn is surprisingly tough to get rid of. My instinct is that White should be able to survive this, but I have to admit that the engine is rather bullish about Black's chances.}) 15... Qe7 {(49)} 16. Re1 $6 {Ledger had apparently decided that the e-pawn was his only hope of counterplay - we'll see where this leads him, but in fact I could have punished the idea straight away... loose pieces drop off!} (16. Nf7+ $143 $2 { no longer frightened me, and rightly so: I intended} Rxf7 17. Bxf7 Ne5 18. Bb3 {and now 18... Nd3, which is very strong, but in the evening my team was very keen for me to show the game, and one of our top boards quickly pointed out that} Bxh3 $1 {is simply devastating.} 19. Re1 {With such an overwhelming local superiority on the kingside, it's not surprising that Black has several ways to win here, but it's hard to resist a sacrificial finish:} (19. gxh3 Nf3+ 20. Kg2 Nh5 {White is a rook up, but his pieces are simply on the wrong side of the board, and he cannot defend.} 21. Rg1 $5 {An ingenious computer try, freeing f1 for the king, but insufficient.} (21. Bc1 Nh4+ 22. Kh1 Qe5 $19 {or}) (21. Qc1 Nf4+ 22. Kh1 Qh4 $19 {would both lose quickly.}) 21... Qg5+ {(the engine thinks 21... Qh4 is even stronger, but this looks simpler to me)} 22. Kh1 Qh4 23. Be6 Nxg1 {(23... Qxf2 also wins)} 24. Qxg1 e3 $1 $19 {, decisively opening the long diagonal and undermining the guard of g3.}) 19... Nfg4 $1 20. gxh3 Rf8 $1 {(instructively bringing all the pieces into the attack)} 21. hxg4 Qh4 22. Bd4 Qh3 $1 {This quiet move is the real killer, sealing the white king's tomb.} 23. Bxe5 Bxe5 24. f4 {(the only defence to the stock mate by 24.. . Bh2+ etc.)} Bc7 25. Rxe4 Bb6+ 26. Rd4 Rxf4 $19 {and mate in a few moves.}) ( 16. Nc3 {Stockfish recommends plain development. The point is that 16... Qe5 can be met by 17. f4, so Black does better to harass the floating knight on g5 with} Bf4 17. h4 h6 18. Nd5 Nxd5 19. Bxd5 {The knight is immune because of Qh5#, but} Bf5 20. Qh5 Be5 {leaves White's pieces tripping over each other. The only way to save them all is the desperate} 21. g4 {, after which the simplest is} (21. Bxe5 Nxe5 $19 {leaves White with no good answer to the threat of 22... Bg4, so the g5 knight is doomed}) (21. Rab1 {also does not avail because of} Bxb2 22. Bxc6 bxc6 23. Rxb2 Qf6 $19 {, hitting the rook and threatening ...Bg6.}) 21... Bxg4 22. Qxg4 Bxb2 23. Rae1 Ne5 24. Qh3 $17 { , and whilst White may get his pawn back it's hard to believe he can survive such an initiative with his king wide open.}) 16... Bf5 $2 {Routine development isn't bad in itself, but a far more punishing move was available... } (16... Bf4 $142 {essentially wins on the spot, as White has no good way to defend his knight.} 17. Nxe4 (17. Bc1 {doesn't help either because Black seizes the long diagonal, after which he just goes back to winning the knight - } Be5 18. Ra2 h6 19. Nxe4 {is again out of the question because of} Nxe4 20. Rxe4 Bf5 $19 {winning material.}) (17. h4 $2 {From memory, this was what put me off, but I didn't realise that further weakening of the kingside is fatal for White -} Ng4 {threatens to break in on f2, and if} 18. Rxe4 (18. g3 Bxg5 19. hxg5 Nxf2 20. Qc1 {- defending g5 -} Nh3+ 21. Kh1 Bg4 $19 {and White is defenceless}) 18... Bh2+ 19. Kf1 Qxe4 $1 20. Nxe4 Ne3+ 21. Ke2 Nxd1 22. Kxd1 Bf5 $19 {leaves Black material up with a more harmonious position.}) (17. Nf7+ {is 'playable', but giving up two pieces for a rook in the early middlegame is moreorless hopeless for White.} Rxf7 18. Bxf7 Qxf7 $19 {This is similar to how the game turns out.}) 17... Nxe4 18. Bd5 Bf5 19. f3 {This doesn't get the piece back because of} Qh4 20. Bxe4 Bxh3 $1 $19 {with a winning attack, as the bishop cannot be taken:} 21. gxh3 Qg3+ 22. Kf1 Qxh3+ 23. Ke2 Qg2+ 24. Kd3 Rad8+ ) 17. Nd2 Rad8 $15 {Despite the missed opportunity, I was still very optimistic. Not long before this I'd been looking at some Morphy games, and it occurred to me how rarely strong opposition actually allow me to develop all my pieces to good squares - and how good my results are when I can. We now have what might be called an IKP structure; much rarer than an IQP, but with the central pawn advanced to the fifth rank, all the pieces in play, and open lines along the f-file and the b8-h2 diagonal, surely a favourable one for Black.} 18. Qe2 Rde8 {After trying and failing to make more direct tries work, I went for this quiet shuffling move. I saw that I couldn't stop White winning a pawn if he really wanted to, but was sure he wouldn't dare try...} 19. Bxf6 $4 {Wow - if White wasn't playing anti-positionally before, he certainly is now. I didn't believe this for a moment, and apparently Ledger didn't either because he started to look very distressed once I recaptured. As is often the case, what looks outwardly to be a positional blunder is actually the result of a tactical oversight.} (19. Rad1 $142 {White too ought to develop his last piece, but even here the g5 knight continues to cause difficulties for him.} Bf4 20. Bxf6 {Stockfish suggests that White should give up two pieces for a rook and pawn - a recurring theme in this game, but in this case it's a reasonable idea because queens are exchanged in the process.} (20. Ndf3 $5 Nh5 21. h4 $15 {looks very suspicious, but there doesn't seem to be any immediate refutation...}) 20... Qxf6 21. Ndxe4 Bxg5 (21... Qxg5 $2 {looks clever but backfires after} 22. Nxg5 Rxe2 23. Rxe2 Bxg5 24. Rd5 $14 {, when g2-g4 will win material.}) 22. Nxf6 Rxe2 23. Rxe2 Bxf6 $15 {White isn't without chances thanks to his active rooks, but Black's bishop pair shouldn't be underestimated, and so long as he can preserve his remaining rook I think the second player is for choice.}) 19... Qxf6 20. Nf7+ {(12) From his expression, Ledger was clearly well aware that this isn't far short of resignation, but indeed it is not possible to take on e4.} (20. Ngxe4 $143 $4 {One of their other players speculated that Ledger had missed} Bxe4 (20... Qg6 $17 {In reality, I have to admit I didn't see this variation at all and was intending this move, which is positionally winning but not yet concretely! The reader may have noticed that my failure to spot forced wins is a recurring theme in this game, and that is an accurate reflection of my play - I am much better at sensing what should and shouldn't work than I am at producing the evidence. Maybe that's why I get disproportionately good results against strong players - they trust their intuition, whereas weaker opposition is more likely to demand that I prove it...}) 21. Nxe4 Rxe4 $1 22. Qxe4 Qxf2+ 23. Kh1 {, when White is winning after every move except} Qg3 $19 {which mates.}) 20... Rxf7 { (21)} 21. Bxf7 Qxf7 $19 22. Qc4 Qf6 $6 {Correctly refusing to trade queens: as mentioned earlier, the advantage of two pieces against a rook is most pronounced in the middlegame. I knew that I should continue playing actively, but this isn't the right way; I was aiming for a battery on the b8-h2 diagonal, and encouraged by the attack on a1 indirectly defending my e-pawn, however this proves to be a dead end for me.} (22... Qg6 $142 {, the other move I looked at, is much stronger. First of all Black gains a tempo by threatening .. .Bxh3, but the real point is that e5 belongs to a different piece - or, as John Emms put it more concisely to me, "followed by ...Ne5-d3 and White resigns". And indeed, after} 23. Kh1 Ne5 24. Qf1 Nd3 25. Re2 Rc8 $19 {White is completely tied up and there's no question of the counterplay that arises in the game.}) 23. Rad1 Qe5 $2 24. Nf3 Qf4 25. Re3 $6 {(10) 25. Qc1 would have forced an exchange of queens.} Bg6 {(11)} 26. g3 Qf6 $6 {Although most people who've seen this game have lauded me for playing excellently, I feel they are paying attention to the first half and then judging by result - to be honest I'm not entirely convinced, because this was one of many games in the season that were marred by time pressure. White had serious chances to get back into the game, but for once an opponent failed to make the most of them...} (26... Ne5 $142 27. Nxe5 (27. gxf4 $144 {Simplification isn't enough for White here; following} Nxc4 28. Rc3 exf3 29. Rxc4 b5 30. Rcd4 Bc7 $19 {a pure BB vs R scenario doesn't promise White anything.}) 27... Qxe5 $19 {avoids any counterplay.}) 27. Nh4 Bf7 28. Qb5 g5 $6 {Clearly a move induced by the clock, nervously creating luft in the most direct - and weakening - manner possible.} 29. Nf5 Be5 30. Qxb7 {(7)} Qxf5 $2 {(2)} (30... Rb8 {was my intention, then 31. Qd7 Rd8, but now I realised White has} 31. Qa6 Rb6 $2 32. Qc8+) (30... Bg6 $142 {would have broken through on the kingside however - if} 31. g4 Bxf5 32. gxf5 Rf8 $19 {and now ... Qxf5 will come with a threat to f2.}) 31. Qxc6 Bg7 32. Rd7 Bg6 33. g4 $6 Qe5 $2 (33... Qf4 $142 {is obvious - I wanted safety first in time trouble and felt sitting on the long diagonal (also covering my rook) was the best way to avoid any accidents. The truth is that Black should focus on his kingside counterplay.} 34. Rxa7 $4 {is not a threat owing to} Rf8 35. Re2 Be5) 34. Re1 $6 (34. Rxa7 $142 $1 $11 {The point is not really to grab a pawn, but to have Ra8 forcing the exchange of my rook - then his king becomes much safer.} Qa1+ 35. Kg2 Bd4 36. Ra8 Rxa8 37. Qxa8+ Kg7 38. Rxe4 {Simplest.} Bxe4+ 39. Qxe4 Bf6 $11 {A draw remains favourite with two exposed kings, but only White has any winning chances.}) 34... Rf8 35. Qb7 $4 {(3) Finally settling the outcome of the game. Now that my rook is in the right place, White can't defend his kingside with just one piece.} Qc3 {(0)} 36. Re3 Qa1+ 37. Kg2 Qf6 $2 {37... Qb2 would have crashed through on f2.} 38. f3 {38. Rd2 was the last chance, but should still be winning for Black. For once in my life I settled on the practical choice here of repeating moves to reach the time control -} Qb2+ 39. Kf1 Qc1+ 40. Kf2 {(4)} Qb2+ {(1), whereupon Ledger resigned - my first win against an IM since the 2018 British. The conclusion left a lot to be desired, and I was helped by my opponent playing far short of his best, but I was proud of my play in the first half of the game: I played good, logical moves and set a strong player problems he couldn't solve. For the team, this was arguably even more valuable: having fought for six draws from the seven other boards against far stronger opposition, my heroics pick-pocketed a 4-4 draw and a precious match point.} (40... Qb2+ {Once the adrenaline rush of the time scramble is over, White doesn't have any serious threats - ...g5 may have been rash, but it at least gave my light-squared bishop a useful post on g6 from which to defend h7. So Black is free to mobilise his rook and dark-squared bishop, quickly encircling White's king.} 41. Kf1 {(41. Ke1 is met in the same way)} Bd4 42. Rxd4 (42. Re2 Rxf3+ {mates}) 42... Qxd4 43. Qe7 Rc8 44. Re1 h6 $1 $19 {The final touch, freeing Black's rook from defence of the back rank and setting up a decisive invasion on c2.}) 0-1

The Caro-Kann strikes again! I play it very rarely because the Sicilian works well, but as a surprise weapon it has brought me excellent results, and this was certainly the most spectacular case. In the past I would have confidently played 1... c5 anyway and invited Ledger to do his worst, but I am more practical these days.

This was certainly my best result for some years, but just as there are spots on the sun it's important not to be dazzled by the outcome if one is to learn the lessons of this game. In fairness to my opponent, he was clearly not on form at the time and had a disappointing result at the English Seniors Championship shortly after; possibly external factors were involved, or maybe it was just one of those phases we all go through.

Although the first half of the game is striking, having analysed it I realise how little I actually saw at the board. I only allowed 14. Ng5 because of a simple oversight, and it was always the violent finishing moves (18... Bxh3! after 16. Nf7+?, misevaluation 17. h4 in reply to 16... Bf4!, 21... Rxe4! after 20. Ngxe4??, and so on) that I failed to see. On the other hand, my intuition was always in the ballpark; I understand this sort of position quite well, but attention to detail has never been my strong suit. Thankfully I wasn't required to prove my ideas; if Ledger was a weaker player he might have won this game!

As usual, the finishing stages were played patchily at best, since I had the twin handicaps of time trouble and a winning position. To me, my moves look nervy and quite passive - 28... g5?!, 30... Qxf5? and 33... Qe5? all make a poor impression. 22... Qf6?! may be the most insightful moment of all though, since it sets the tone for the rest of the game - the ...Qe5 idea is obviously not going anywhere, so it leads to a position where I wasn't really using my material advantage, just hoping that gravity would do the work for me. Contrast this with the 22... Qg6 variation, where Black wastes no time in putting his extra pieces to work and promptly suffocates White.

So all in all, a game of big conceptions but poor execution on both sides, decided by a little cunning and a lot of psychology. Yes, even strong players can be nervous! I wasn't entirely happy with it even at the time, but like my win against Snape it was of course a big confidence booster. I felt on great form at this point, and had real chances against an FM the next day before winning another game in the Epsom club championship to get to 4/4. That set up a showdown for the title against another IM...