Back to school, part 1

It is said that time heals all wounds, but almost a decade after graduation I still have rather mixed emotions about my three years in Bristol as a student. I mostly remember being overdrawn the entire time, and a lot of hard work trying to get a first not being rewarded, but a spartan existence on my own was a valuable life lesson, and history remains a fascinating subject to me. And as much as I like to poke fun at the weather (they don't call it Brizzle for nothing), it is a nice city to which I've returned a few times since. Last month was my first trip in nearly seven years though, and since the reader is seeing this it was of course for chess - the Open section of the Bristol Centenary Congress, doubling as the West of England Championship.

I entered the fray from the second round on Saturday morning, and disappointingly the pairings didn't go up until just beforehand. I was just on the mid-line of the scoregroup so would either play one of the highest or lowest seeds - in the end I got the black pieces against IM Conor Murphy (2515). I last played him in the 2018 British Championship when he was 'just' an FM, and lost a hard-fought game; he has made a lot more use of six years than I have and now has two GM norms to his name, including an outstanding performance on top board for Ireland in the last Olympiad.

This game sees the debut on the blog of my opening project for the season: driven to despair by no-one letting me play the Najdorf any more, I've been learning 1. e4 e5 for Black. I already used it once in a blue moon when specially prepared, and had played it in a few rapidplay encounters more recently, but this was its first appearance in a serious tournament game. A very sharp battle soon arose, belying the supposedly quiet variation, but fittingly for the location of the game it eventually turned into a bit of a lesson:

[Event "Bristol Open"] [Site "?"] [Date "2024.04.27"] [Round "2.11"] [White "IM Murphy, Conor E"] [Black "Young, Daniel J"] [Result "1-0"] [ECO "C55"] [WhiteElo "2515"] [BlackElo "2024"] [Annotator "Young,Daniel J"] [PlyCount "75"] [EventDate "2024.??.??"] [SourceVersionDate "2024.04.29"] {This was my first visit to the city where I was a student since 2017. It was cold and rainy (of course) but I could appreciate the surroundings more than when I was a miserable and impoverished undergraduate; Bristol is really quite a nice place and the Great Hall at Bristol Grammar School, at the top of Clifton and right next to the university, was a grandiose place for a chess tournament. I took a bye on the Friday night and began on Saturday morning against a very strong IM.} 1. e4 {(1:30)} e5 {(1:30) I tend to take on one meaty opening project per season, and this time it's 1... e5. I'd had spotty results in the anti-Sicilians for some time, without the Najdorfs to compensate. It stands to reason that the 'anti' lines are easier with a pawn in the centre and straightforward development, whilst my longstanding fear that White can get a dull game if he wants one have turned out to be unfounded.} 2. Nf3 Nc6 3. Bc4 {No longer the preserve of junior tournaments and old romantics, the Italian Game has enjoyed a renaissance of late. Admittedly this owes a lot to White's crisis of confidence in the Ruy Lopez, however a lot of new ideas have been discovered too. Now Black faces an important choice.} Nf6 {Since I can't improve on Short's exposition (in 'Winning') I will instead quote it at length: "The world can be divided into those that choose the dependable 3... Bc5 and wilder, more reckless types, who prefer the text. They are both developing moves, but there is a fundamental difference of philosophy: after 3... Bc5 the position can become very sharp, but only if White is prepared to take risks [...] With the Two Knights Defence, however, Black is the one who can be compelled to sacrifice and be drawn into wild complications [...] I consider both main continuations to be objectively equal in merit, but one must understand well what one is doing when one forms a repertoire."} 4. d3 {In the 21st century you should expect to see this most of the time; it is not the modern way to give up the initiative as White by embracing murky complications so early, and slow systems like this are relatively proof against being analysed out by engines. Of course, in former days White was not nearly so modest, and the older tries are still brought out of the museum sometimes.} (4. Nc3 $6 {is played frequently at low levels but probably does White more harm than good because of the 'fork trick'} Nxe4 5. Nxe4 d5 6. Bd3 (6. Bxd5 $143 $2 {is the most-played move, but only because of the volume of beginner games in the database; this simply accelerates Black's development and} Qxd5 7. Nc3 Qd6 8. O-O Bg4 $17 {, with queenside castling and ...f5 to follow, promises Black a strong attack before White can really get going.}) 6... dxe4 7. Bxe4 Bd6 8. O-O O-O $11 {Black certainly has no problems, and with a freer pawn centre he has scored superbly in practice.}) (4. d4 {leads to a position that can also arise from the Scotch Gambit, and is recommended by Gawain Jones in his 'coffeehouse repertoire'.} exd4 {Here there are two main tries:} 5. e5 (5. O-O {this apparently swashbuckling move really just leads to dull liquidation after the well-known sequence} Nxe4 6. Re1 d5 7. Bxd5 Qxd5 8. Nc3 Qa5 {(8... Qh5 is also played)} 9. Nxe4 {. That said, the computer is sceptical of the automatic} Be6 {, preferring 9... Be7. The reason for this is at the end of the long main line} 10. Neg5 O-O-O 11. Nxe6 fxe6 12. Rxe6 Bd6 13. Bg5 Rde8 14. Qe2 Kd7 15. Re1 Qxe1+ 16. Qxe1 Rxe6 {it proposes the new move} 17. Qf1 $5 $146 {, intending Qc4 followed by taking the d-pawn. I have a pet theory that, contrary to received wisdom, a queen is usually better than two rooks... maybe there's something in it?}) 5... d5 6. Bb5 Ne4 7. Nxd4 Bd7 {(7... Bc5 is an interesting move order but often just transposes)} 8. Bxc6 bxc6 9. O-O Bc5 10. f3 Ng5 11. f4 Ne4 12. Be3 {An interesting and imbalanced position. Black has easy development and the half-open b-file, in return for which White has a mobile pawn front and more space. The engine and the database both suggest chances are even.}) (4. Ng5 {is of course the critical option, essentially forcing Black to sacrifice a pawn. However Black is currently considered to have few problems here, and since White is running much more risk with this anti-positional pawn grab there seems little motive for him to go in for it.} d5 {(the maddening complications of the Traxler Variation, 4... Bc5!?, have still not really been worked out)} 5. exd5 Na5 6. Bb5+ c6 7. dxc6 bxc6 {I played 3. Bc4 as a beginner, and back then my attitude to Black's gambit (encouraged by the engines of the time) was complete incomprehension: White is a pawn up with no weaknesses, Black's queenside structure is destroyed, he is no better developed than White and the knight on a5 is stupid. Now that I understand dynamic play much better, I know matters are hardly so simple: Black's position is indeed ruined in static terms, but he has open lines for his pieces whilst White has to waste even more time retreating from b5 and g5.} 8. Bd3 {This clumsy-looking move is now considered the most testing, keeping e4 under control. On the flip side, the bishop is exposed here and Black is able to shift gear to playing for positional compensation by an exchange on d3.} (8. Be2 {is the traditional approach.} h6 9. Nf3 {(Steinitz and Fischer favoured 9. Nh3)} e4 10. Ne5 Bd6 11. d4 exd3 12. Nxd3 O-O 13. O-O {At first sight it 'only' remains for White to develop his queenside, but this turns out not to be very easy! After} Qc7 {(Black should delay this move until after castling, provoking the reply)} 14. h3 c5 $44 {White has actually scored appallingly. Black's play is very easy: ...Bf5, ...Nc6, ...Rad8 and ...Rfe8. White simply has to defend.}) (8. Qf3 {Back in those carefree days of innocence I mentioned earlier, I thought that Boguljubov's move, winning a second pawn, is a refutation of Black's play. In reality this is much too greedy. Black has various good moves, but just for fun I'll mention that} cxb5 $5 9. Qxa8 Qc7 10. Qf3 Nc6 11. c3 $8 Bg4 12. Qg3 Nd4 $1 $44 {leaves Black with excellent compensation for an exchange and a pawn, with White being almost entirely undeveloped.}) 8... Nd5 9. Nf3 Bd6 10. O-O Nf4 (10... O-O {leads to a second pawn sacrifice and even sharper play after} 11. Re1 f5 $5 12. Nxe5 Qf6 13. Nf3 g5 $13 {, straightforwardly aiming to blow White away on the kingside}) 11. Re1 (11. Be2 $6 {is the obvious attempt to defend without making any concessions, but it fails to} Nxg2 $1 12. Kxg2 e4 $17 {recovering the piece, as if} 13. Nd4 $2 Qg5+ 14. Kh1 Qf4 $19 {mates}) 11... Nxd3 12. cxd3 {By comparison with 8. Be2, White is relatively safe against a lightning attack on his king, however the d-pawns are a sorry sight which further complicate the development of his queenside and Black has a static trump of his own in the shape of the two bishops. The main continuation is} O-O 13. Nc3 Re8 14. Ne4 c5 15. b3 Bf8 16. Ba3 $44 {, and here Black has made a plus score with both ...Nc6 and ...f5.}) 4... Be7 {A big breakthrough for me in taking up 1... e5 as a full repertoire choice was realising that these d3 lines, which I'd always dismissed as lifeless stodge, actually have plenty of life. In contrast to, say, the London System, where White simply strongpoints d4 and just dares Black to do something, both sides keep various pawn breaks in reserve - the action is delayed relative to many other opening systems, but it's still coming. I played this unassuming move in accordance with the repertoire book I'd brought with me to Bristol, though there is a wide range of options.} (4... d5 {is a direct equalising attempt which has been popular recently, but on principle I'm a bit suspicious: White's extra tempo feels more useful if the game opens up early on.} 5. exd5 Nxd5 6. O-O Be7 7. Re1 f6 8. d4 Nb6 {Here White has amassed a score of over 70 percent with} 9. Bb3 $14 {(there have been some recent games with 9. Bb5 too), the main line continuing} Nxd4 10. Nxd4 Qxd4 11. Qxd4 exd4 12. Bf4 {. No doubt a computer can defuse White's nagging initiative, but it looks a joyless chore to me.}) (4... h6 5. O-O d6 6. c3 g5 {is much more like it! In these closed-centre positions White should always assess the consequences of a quick ...g5 when he castles quickly, though it works best for Black if White has also volunteered h2-h3 early on.} 7. a4 Bg7 8. a5 {Here I agree with Harry Grieve in a recent ChessPublishing column, who advocates} Ne7 {to economise on the normal 8... a6; the knight is heading to g6 and perhaps f4. The engine still doesn't seem very impressed but this strikes me as a good imbalancing choice for a must-win game.}) (4... Bc5 {is the most natural and popular move. I usually play this in blitz games but, since Murphy knows the Italian lines very well and has an excellent score with them (including a win against Gawain Jones in the 4NCL last year), I preferred to play differently this time.} 5. c3 d6 {It's hard to follow all the transpositions, but the main sequence in the database is} 6. O-O O-O (6... a6 {A rule of thumb in this opening is to not combine short castling with an h-pawn move until the opponent also castles kingside, so this move order has its adherents. However I'm not sure it has much independent significance after} 7. Bb3 Ba7 8. Nbd2 {since Black's best move is probably 8... O-O, transposing back into the main line. However some players have tried} h6 $5 9. Re1 g5 10. Nf1 g4 11. N3d2 h5 {It would be nice if such fairy-tale moves could work, but alas this is the real world! White's kingside is unblemished and Black has not completed his development, so this feels very rash to me. That said, White has scored remarkably poorly in the line} 12. Nc4 h4 13. Be3 {, so maybe I don't know what I'm talking about...}) 7. Bb3 a6 8. Nbd2 Ba7 {(pre-empting a later b2-b4 with tempo)} 9. h3 {- only now that Black has castled! -} h6 10. Re1 {There are three main moves here - 10... Nh5, 10... Re8 and 10... Be6 - all of which score about 55 percent for White.}) 5. O-O O-O 6. Re1 {Established as White's best move order, ruling out a quick ...d5 break. Compare this with my next Black game 24 hours later!} d6 {Time was short after the pairings went up, so I was forced to rely on my aforementioned book, which covered the two main moves 7. c3 and (particularly) 7. a4. But there are other moves!} 7. a3 $5 {An obscure move, and I can't find any record of Murphy playing it before, so perhaps it was improvised to avoid preparation - this was my first use of 1... e5 in a published game, which should set an alarm off for any practical player. White's priority in this position is to save his light-squared bishop from being exchanged off by ...Na5, however whether by accident or design it also works to discourage the plan I had brought to the board. Now I had to come up with something different, which I finally did after 18 minutes:} Nd7 {is a rare choice but actually the best-scoring move. This knight is heading to b6, reinforcing d5, and then Black will continue with ...Kh8 and ...f5. In any case, we were now both on our own.} (7... Na5 8. Ba2 c5 {was what I intended against 7. a4, when the knight returns to c6 and we have a Chigorin Ruy Lopez for all intents and purposes. Unfortunately, White has an irritating extra option here in} 9. b4 Nc6 (9... cxb4 $6 10. axb4 Nc6 11. b5 $16 {drives the knight offside since the desirable} Nb4 {leaves Black without a good move after} 12. Bb3 {followed by 13. c3.}) 10. bxc5 dxc5 $14 {- I didn't want to play this structure, where the outpost on d5 is exposed and I've lost my central pawn break, and Stockfish agrees.}) (7... Be6 {is the usual retort to White's last move.} 8. Nbd2 Qd7 9. c3 {But Stockfish wants to vary here.} Bxc4 (9... Rfe8 {is almost always played, however it believes that} 10. Bxe6 Qxe6 11. d4 exd4 12. cxd4 d5 13. e5 Nd7 14. Nf1 {(Bauer-Stevic, European Team Ch. 2019) favours White, presumably on account of his extra space. My experience of the Canal Sicilian (3. Bb5+) makes me sceptical; I'd be more worried if White still had his light-squared bishop to complement the central pawn constellation. Is} f6 $146 15. exf6 Qxf6 16. Bg5 Qd6 17. Bxe7 Rxe7 {really that impressive for White?}) 10. Nxc4 {In the one game to reach this position Black chose 10... b5; the oracle offers the more patient} h6 $146 {instead, preventing Bg5 and not committing on the queenside yet. The obvious attempt to take advantage of Black's leisurely play is} 11. d4 {, though Black is fine after} exd4 12. cxd4 d5 {since e4-e5 can only be achieved by further simplification.} 13. Nce5 Qd6 14. Nxc6 Qxc6 15. e5 Ne4 $11) 8. Nc3 (8. c3 {White might also have tried breaking out in the centre, which is what I thought stronger.} Nb6 9. Bb3 Kh8 {However, the follow-up is definitely from the engine:} 10. a4 $1 {, intending} f5 11. exf5 Bxf5 12. a5 Nc8 13. d4 $14 {and with the point that} e4 $2 14. a6 $1 {finally reveals what Stockfish is up to.} bxa6 15. Bd5 {Computer witchcraft perhaps, but I suppose there is some logic in White making strides on the light squares after Black lashes out with ...f5.}) 8... Nb6 9. Ba2 Kh8 10. h3 {(1:17) I thought this move was merely weakening, however White will have to take the plunge soon and he can make use of the h2 square as we will see.} f5 {(1:11)} 11. Nd5 $6 {Played after fully 27 minutes, but perhaps this is a bit too direct. It is a big theme for the rest of the game that White wants to open the centre and Black very much does not, and this move gave me a chance to keep the game closed for a long time. On that basis, 11. exf5 is the most logical.} f4 $6 $146 {Extremely ambitious of course, but this seemed quite natural to me and I played it quickly. Black's argument is that the centre is closed and therefore he can indulge in throwing the kingside pawns forward without jeopardising his king, hoping to make use of the hook created by White's 10th move. If the position does become open then Black will have much to regret, so this is a move that only works if it works, but deferential play against strong opposition never got me anywhere.} (11... fxe4 {A previous game continued with the straightforward double trade} 12. dxe4 Nxd5 13. exd5 Nb8 {, leading to a structure more often seen in the King's Indian. In general White's chances are better if he can control e4 with pieces unless Black has d4 in return, since White can press ahead on the queenside whereas Black struggles to generate counterplay on the kingside (his dark-squared bishop has little to do if the e-pawn can't advance). Therefore I feel Black has already gone slightly wrong, and sure enough after} 14. c4 Bf5 (14... Bf6 {If Black tries to play by analogy with the variation below, the open centre soon starts to cause problems for him.} 15. Bb1 {Preparing to arrange a battery on the b1-h7 diagonal, a plan that would make no sense with a pawn still on d3!} Qe8 16. c5 a5 17. Bc2 Na6 18. cxd6 cxd6 19. Be3 $16 {Qd3 follows. Black's minor pieces are out of it and with the c-file open White has a route to invade the enemy queenside.}) 15. Be3 Nd7 16. Bb1 {, removing Black's only active minor piece,} Bxb1 17. Rxb1 $14 {the engine starts to wake up to this. Black's cause certainly wasn't helped by} c6 $2 18. c5 $1 {(Raghunandan-Visakh, World U-20 Ch. 2019), and the black centre is smashed.}) (11... Nxd5 $142 $146 {is a significant improvement, keeping a mobile pawn centre and making it harder for White to open the position.} 12. exd5 Nb8 13. c4 Bf6 {Black can seriously contemplate ...g5, and White's queenside play is not so threatening -} 14. c5 a5 15. Be3 Na6 {Stockfish offers the slightly desperate-looking} 16. c6 (16. cxd6 cxd6 17. Qb3 {, preparing Bb6, is a more conventional way to play but then comes} g5 $15 {and Black's threats start to mount.}) 16... bxc6 17. dxc6 Qe8 $15 {, but the c6 pawn looks like a liability to me, the newly opened a2-g8 diagonal is easily neutralised by ...Be6, and Black's kingside ideas have not gone anywhere.}) 12. c3 g5 $1 {Of course! Black's play is nothing if not consistent. Here for the first time I sensed a flicker of uncertainty from the other side the board.} 13. Nh2 Bd7 $6 {As Botvinnik wrote (in the notes to his game with Sokolsky, Leningrad 1938, which illustrates the point very well) "there is no 'better' way of ending up in a cramped position than by aiming merely for simple development." His point was that development needs to be purposeful, whereas a move like the text is merely development for its own sake. It felt offensive to keep the bishop on its starting square indefinitely, however the immediate 13... Be6?? fails to 14. Nxb6 and so I settled for this non-move - the bishop achieves no more here than on its starting square, spending a valuable tempo in the process.} (13... Bf6 $142 {This move serves two purposes: it reinforces d4 and it clears e7 as a retreat square, thereby renewing the idea of ...Nxd5. I understood this, but overestimated the importance of getting the other bishop going and connecting the heavy pieces. Now if White simplistically ploughs on with his queenside offensive,} 14. b4 $6 {, then Black quickly creates counterplay, precisely with the newly-activated dark-squared bishop.} (14. Nxf6 $142 {, as in the game, should be preferred, but a simple comparison shows this is an improved version for Black because} Qxf6 15. b4 Be6 {comes in one move. The exchange of light-squared bishops is not as ruinous as it would be in similar positions from the Classical King's Indian, since White's queenside play is much slower and Black has that bit more time to build up to the big ...g4 breakthrough.} 16. Bxe6 Qxe6 17. d4 Qg6 18. a4 h5 $132) 14... Nxd5 15. exd5 Ne7 {If Black had played the immediate 13... Nxd5?!, then in the equivalent position with the knight on b8 and bishop on e7 White could break open the centre to his advantage with d3-d4. Here that's off the menu, and after} 16. c4 {the nature of the game abruptly changes with} e4 17. Rb1 e3 $1 18. fxe3 Nf5 $44 {and confusion reigns as Black's pieces spring to life. Not, for instance,} 19. exf4 $4 Bd4+ 20. Kh1 Ng3#) 14. b4 Bf6 15. Nxf6 $1 {(29) Good judgement. In static terms it may seem odd to trade an excellent knight for a bishop shut in by its own central pawns, however White removes a defender of d4, and once the game starts to open up the bishop pair will come into its own.} Qxf6 {(55)} 16. d4 $14 {Exploiting the potential pin on the long diagonal, White makes a big strategic achievement with this move. Now I have to juggle kingside counterplay with holding my queenside together, and with my minor pieces clumsily placed it ultimately turns out that I just don't have enough time. The lost tempo with 13... Bd7 makes a significant difference.} Qg6 17. f3 Rae8 {Intending ...Re7-g7, lending further support to ...g5-g4, but it's clumsy. Such is the price to be paid for giving White unchallenged control of the a2-g8 diagonal; now that he has managed to fit d3-d4 in I can't challenge with ...Be6.} 18. Bb2 Re7 $6 {I didn't consider a very computery attempt to fish for counterplay, and following a normal course leaves Black doomed strategically.} (18... h5 $142 {Nothing impressive yet, of course - this was the first move I considered. But I thought it was a dead end, since if White carries on with his queenside play (let's say} 19. Rc1 {to mirror the game)} g4 $1 20. hxg4 hxg4 21. fxg4 {and I couldn't see a follow-up. Not so, says Stockfish, which offers} (21. Nxg4 $143 $2 Bxg4 22. fxg4 Nd7 $17 {, followed by ...Nf6, is not what the doctor ordered; the g4 pawn will fall and White may find himself embarrassed with his bishops stranded far from the action around his king}) 21... Kg7 {and if} 22. b5 {the resolute} Rh8 $1 {, leaving the knight to its fate.} (22... Nd8 $143 23. c4 $18 {brings the b2 bishop to life, and chess gravity rapidly reasserts itself if Black tries to press the attack in the face of his crumbling centre:} Rh8 24. dxe5 dxe5 25. Nf3 {, with the twin threats of Bxe5+ and c4-c5.}) 23. bxc6 bxc6 {Now Black threatens 24... Rxh2! followed by ...Bxg4, ...Qh6 and ...Rh8 - with three attackers staring down a lone king it appears that this would offer at least a draw. White can avoid all that with} 24. Nf3 Bxg4 25. dxe5 dxe5 $13 {, but Black will continue with ...Nd7-f6, and with the white bishops out of play his threats to the kingside are not to be dismissed even if the engine thinks White can keep a plus with accurate play. This was far too creative for me, and I assessed the position after 21. fxg4 by intuition rather than calculation, but it is interesting to see what resources such a miserable position can contain.}) 19. Rc1 {Careful prophylaxis - in earlier games against Murphy I'd detected a tendency to ignore his opponent's ideas, but he is a much stronger player than six years ago.} (19. b5 $143 $2 {would suddenly offer Black counterplay from an unexpected quarter - his knights come to life after} Na5 {, taking aim at c4, and if} 20. Qe2 {to control that square} Na4 $15 {deprives White of his precious dark-squared bishop, greatly lessening the danger to Black's king on the long diagonal and so buying time to continue the attack.}) 19... h5 $6 {Fatalistically ploughing on. The unstereotyped 19... a6 20. a4 Ra8!? would at least have slowed White's next move down, though again I did not consider it, preferring to follow the natural course and see where it led.} 20. b5 {(14)} Nd8 {(32) (Now 20... Na5 21. c4 covers everything; c4 is protected and White has a1 as a retreat for the bishop)} 21. c4 Nf7 $2 {From here the knight renews its defence of e5, and in a fantasy world it might find the time to reach h6 and support ...g4. However it has a tactical flaw and offers White a chance to wrap up the game quickly.} (21... c5 {is offered by the computer as a better try, however} 22. bxc6 {(the only move to keep even equality, but not a hard one to find)} Nxc6 23. c5 dxc5 24. Rxc5 $16 {leaves Black with his centre dissolving in any case.}) 22. c5 Nc8 23. a4 $6 {A slight reprieve.} (23. c6 $142 {was what I expected, but I didn't notice the killing second punch} bxc6 24. Bxf7 $1 {, trading off a wonderful bishop but winning the fight for e5 in the process.} Qxf7 25. bxc6 Be8 26. dxe5 dxe5 27. Rc5 Qg7 28. Qa1 $18 {The pawn is lost and Black will not survive for long once White draws a bead on his king.}) 23... exd4 $2 {When I first committed to this King's Indian-esque game, way back on move 11, I was banking on the idea that once White inevitably got d3-d4 in I could strongpoint e5 indefinitely, or at least long enough for my kingside play to land. By now I'd come to the conclusion that this approach was untenable, and so reluctantly decided to anchor the long diagonal with a knight instead to at least deprive White of his d-pawn as a can opener. Unfortunately, the opening of lines makes this new edifice too rickety to hold for more than a few moves - now Black is definitely lost.} (23... Be6 $142 {holds out more hope, neutralising the Bxf7 threat seen in the previous variation. Black is happy to provoke} 24. d5 {, closing the centre, but this is a prelude to opening the queenside instead. Now} Bd7 25. c6 {compels} bxc6 26. bxc6 Be8 {, when} 27. Bc4 Nh6 28. Ba6 Bf7 29. Rb1 $16 {prepares to invade either on b7 or b8.}) 24. Qxd4+ Ne5 25. cxd6 {(10)} cxd6 {(21)} 26. Rc7 $18 Kh7 {A useful defensive move, but it only underscores that tempi have not been on Black's side in this game since I desperately need to make something happen on the kingside.} 27. Rf1 $5 {White can happily take on b7, but Murphy is again very careful - a sensible approach with only a few minutes left. Black is strategically busted and has been for some time, yet even so it would be easy to make a couple of careless moves and lose control, allowing Black genuine counterplay.} g4 $5 {White's last move was clearly aimed against this, but what else?} 28. fxg4 hxg4 29. hxg4 Bxg4 {(or 29... Nxg4 30. Rxf4, and the rook can't recapture because of mate on h8)} 30. Rxe7+ {(3)} Nxe7 {(9)} 31. Nxg4 Qxg4 {Taking with the knight would again invite Rxf4, however any hope of holding my position together now disappears along with my d-pawn.} 32. Qxd6 f3 {A last chance - 33. Qxe7+?? Rf7 - but of course White is having none of it.} 33. Rf2 $8 N5g6 34. Qh2+ Nh4 {In playing this move I realised that I was likely going to resign very shortly, however the alternative was no better.} (34... Qh4 $144 {I avoided this by elimination, since White can simply take on h4 with a trivially won ending, but in fact he can do better with the incisive} 35. g3 $1 Qh6 36. Bc1 {, insisting on breaking down the barrier on the h-file.} Qxh2+ 37. Rxh2+ Kg7 38. Bb2+ $18) 35. Qe5 {(2) The cleanest kill, jumping onto the long diagonal now that Black can no longer obstruct it. My time trouble reply loses instantly but there is no longer a defence.} Nc8 $2 {(0)} 36. Be6 Re8 37. Bxg4 Rxe5 38. Bxe5 {I must confess that I believed straight after the game that I'd had some chances, and certainly the fact that we were both reduced to little more than increment suggests it wasn't completely straightforward. However, analysing it dispassionately reveals that I was comprehensively outplayed. My strategic conception was sound, however it required very accurate play to justify such risks and I was not up to the task I set myself. I was too superficial at key moments.} 1-0

I find it's easier to be creative when an opening is still fresh and your imagination has not yet been stifled by pattern recognition, but evidently even something as outwardly tedious as a 4. d3 Italian Game can produce excitement. Strategically this game has a lot in common with the King's Indian, which speaks to my tastes. Just as in that opening, resolving to launch a kingside attack with ...f4 and ...g5 is a legitimate route to counterplay, but has the potential to make you look very silly if it goes wrong. Besides the danger of my queenside melting I also had to reckon with White busting open the centre here, since he has not committed to d4-d5. On the flip side he isn't quite so fast, and I believe I lost because I was subsequently outplayed rather than because my chosen plan was inherently wrong.

The main lesson for me from this game is that 13... Bd7?! was not a good move. The bishop doesn't contribute anything more from there than it did on its starting square, and in a closed position I didn't need to be in such a hurry to connect the rooks. Compare the 13... Bf6 variation to the game and it can be seen how much of a difference is made by the wasted tempo. In a cramped position I was forced to spend more time on ...h5 and the artificial ...Re7-g7 manoeuvre, and by move 18 I was effectively lost. Murphy was able to make progress in the centre and on the queenside more quickly than I had foreseen. Against a lesser opponent I don't think my highwire act would have been the right choice given the level of risk, but against an IM one can swing for the fences with a clear conscience.

In the afternoon I had another 1. e4 e5 game, this time as White in a Ruy Lopez; my opponent knew the line well and equalised comfortably, and I was relieved to steer to a draw. Sunday morning's game turned it into a thematic tournament.