A blunt pen-knife

"And when personal computers appeared, I was the first player to include machine analysis in my system of preparation and to make systematic use of playing programs and databases. Soon I discovered how weak some of my earlier preparations had been. A useful, sobering discovery! Under the microscope of powerful computer programs it transpired that at times I had gone along to a game not with some kind of Excalibur, but with a blunt pen-knife."

Garry Kasparov, in the foreword to Garry Kasparov on Garry Kasparov: Part I 1973-1985

I first read those words some years ago, and though they've stuck with me I have to admit that in the naivety of youth I thought that such a sensation belonged to the stagecoach era. After all, everybody uses engines nowadays, and they are so fantastically strong that further advances can only amount to tinkering at the edges, right?

I'd always planned to blog today's game, against the 2065-rated junior Ruqayyah Rida, just because I very rarely score opening knockouts any more, and I'm not immune to nostalgia for bygone days. Though I do get irritated by some adult players complaining that juniors are so well prepared it's impossible to play against them, as this just isn't true - in round 4 I was inexplicably allowed to put the famous ...d5!! equaliser in the Canal Attack on the database for a second time - the general standard of preparation has come a long way, and I'm no longer so obsessive about playing sharp main lines. However, when I came to annotate it at home I was able to look at my pre-game notes on a desktop computer with the latest Stockfish, rather than a creaking old laptop. It soon dawned on me that my old optimism about the end of history was misplaced - I had gone to battle with a very blunt pen-knife indeed:

[Event "British Championship Major Open"] [Site "?"] [Date "2023.07.28"] [Round "7.18"] [White "Young, Daniel J"] [Black "Rida, Ruqayyah"] [Result "1-0"] [ECO "C78"] [WhiteElo "2007"] [BlackElo "2065"] [Annotator "Young,Daniel J"] [PlyCount "113"] [EventDate "2023.??.??"] [SourceVersionDate "2023.07.30"] {As my rating isn't (yet) as spectacularly deflated as most of the people in this tournament, I was usually being paired with the lower-rated people in each score group - so it was with a groan of disappointment that I realised I was actually outrated on the ECF scale for this game. Nevertheless, having gone through the games of my young opponent, I concluded that she was not as tactically sharp as most juniors are nowadays, and that allowed me to hope to at least play a game. And after several rounds where I couldn't prepare at all, I could also bring the full weight of my opening knowledge to bear this time.} 1. e4 {(1:30) In case the reader is wondering why I stuck with 1. e4 despite knowing a line I dislike was coming, a flank opening would have led to a King's Indian which I like even less - I try to follow Botvinnik's advice and avoid playing against myself in the opening.} e5 {(1:30)} 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O b5 6. Bb3 Bc5 {Some years ago I had a number of battles in the Neo-Archangelsk, and found myself hopping between sidelines since I could never get a workable advantage. Then the line languished for a while, but now it seems to be back - Grandmaster Sethuraman advocates it in a Chessable course, and another female junior who I faced in last year's Major Open also plays it - both of them are part of the SPTW group run by Lorin D'Costa, so perhaps he is teaching it. At any rate he took a close interest in this game in the opening. Last year my opponent was scared of my preparation and avoided it, but I didn't expect to get lucky twice and spent a long time the night before figuring out which line to choose this time.} 7. a4 Rb8 8. Nxe5 {I can't honestly say that White gets much playing this way, but the same goes for everything else (except for maybe accepting the challenge with 8. c3 d6 9. d4 Bb6 10. Na3 etc., but I find those lines too difficult to play). This variation is at least challenging, I find the positions more comfortable, and Black can easily get muddled up if she tries to play by analogy with the equivalent lines following 7. Nxe5, which is ten times as common. As a bonus, the game in which I already tried this (Young-Ahluwalia, e2e4 Gatwick Challengers 2016) seems to have somehow eluded the database.} Nxe5 9. d4 Bxd4 10. Qxd4 {(1:32)} d6 {(1:26)} 11. f4 c5 $6 {Immediate vindication! This is how Black plays after 7. Nxe5 (see for instance Young-J. Golding, Colin Crouch Masters 2016) however it is less effective here because the b3 bishop has an escape square. Then again, as I cautioned in my notes to the Ahluwalia game above, it probably isn't quite as bad as its reputation.} (11... Nc6 {is the safer response, and what Ahluwalia played - see that game for more info. I will only add that I checked my old notes whilst preparing for this game and still feel that} 12. Qd3 {is a better response than the more usual 12. Qc3 - if anything, driving the knight back to e7 only helps Black, since ...c5 is available to fight for the weakened d4 square.}) 12. Qc3 {Nothing ages as badly as old analysis - I was happy to have my notes from 2016 to fall back on, but modern engines favour the more-played 12. Qd1. Though I did go over the old material on the day of the game, my laptop is ancient now and can't run a current engine, unlike my home PC. As seen below, the difference in results is striking!} (12. Qd1 $142 {From here the queen is safe from the knights' attentions.} Neg4 (12... Nc6 {Since ...Nd4 no longer threatens anything, White is in time for} 13. e5 dxe5 14. Qxd8+ Nxd8 15. fxe5 Ng4 16. Bd5 Be6 17. Be4 $14 {(Messinger-Marimoutou, French Junior Ch. 2014) with a stable edge thanks to the two bishops and Black's clumsily placed knights. If} Nxe5 $6 18. axb5 axb5 19. Be3 c4 20. Bc5 {with terrible pressure against Black's under-developed forces; Bd6 is immediately threatened.}) 13. c4 $1 $146 {This is an important new move - Black actually gains an advantage against everything that White has tried to date. The point is to clear c2 for the bishop - that is the prelate's best diagonal since Black is committed to castling kingside.} (13. c3 $146 {is the initial Stockfish suggestion, but it doesn't seem quite energetic enough - Black can just about get away with} Nxe4 14. Bd5 (14. Bxf7+ $2 {recovers the pawn after} Kxf7 15. Qd5+ Be6 16. Qxe4 Nf6 $17 {, but loses the initiative. White needs his queenside pieces!}) 14... Nef6 15. axb5 (15. Bc6+ Kf8 16. h3 h5 $1 $13 {is a mess, but Black is the only one having fun here.}) 15... Nxd5 16. Qxd5 Rxb5 {, not ceding the a-file to White. Now the black king gets displaced after} 17. Re1+ Kf8 18. Na3 Rb7 19. Nc4 {, however} Rd7 20. Qa8 g6 $44 {leaves White with compensation for his pawn, but no more - it is difficult to attack the backward pawn on d6 again, and the black rooks will connect once the king reaches g7.}) 13... bxc4 (13... Nxe4 $6 {should be met a la Morphy, just developing as rapidly as possible.} 14. Nc3 Nxc3 15. Re1+ Kf8 16. bxc3 bxa4 17. Bxa4 Nf6 18. Be3 Be6 19. Qd3 $16 {Whilst White's pawn structure is a mess (even with one fewer pawn to worry about!) this position is much more favourable for him than the 13. c3 variation above: he has two bishops, and crucially the black king is unable to free itself this time -} g6 20. f5 $1 Bxf5 21. Bh6+ Kg8 22. Qd2 {, when the rook on h8 is buried alive.}) 14. Bc2 $1 {Protecting White's strongest pawn.} (14. Bxc4 Nxe4 15. Nc3 Ngf6 16. Re1 Bb7 {seems to cover everything. White can actually win two pawns here with} 17. Bxa6 Bxa6 18. Nxe4 O-O 19. Qxd6 Qxd6 20. Nxd6 {, however} Rfd8 21. Nf5 Kf8 $44 {is not so bad for Black with opposite-coloured bishops and the one on c1 stuck defending b2.}) 14... O-O 15. h3 Nh6 16. f5 {Making it obvious which side of the board White is playing on. Those black knights make a sorry impression...} Re8 17. Qf3 $14 Bb7 18. Bxh6 gxh6 19. Nd2 $14 {Despite being a pawn up, Black seems to be struggling because of her split queenside pawns and the compromised king position. Note that she cannot try to free herself with} d5 $6 {in view of} 20. e5 $1 {exploiting the exposed king.}) 12... Neg4 {Still following the other line, however 12... Nc6 is also possible.} (12... Nc6 {In my old analysis I gave} 13. axb5 {and now 13... axb5 14. e5, concluding that White gets the better of it. However Black can improve with} Nd4 $1 $146 {, centralising with tempo as ...Ne2+ is threatened.} 14. Re1 axb5 15. e5 dxe5 16. Qxc5 (16. fxe5 Ng4 17. h3 c4 {causes White problems since 18. Ba2 would be strongly met by 18... Qb6. That leaves} 18. hxg4 cxb3 $13 {and White loses control of the position. For example, Stockfish gives as its mainline} 19. e6 bxc2 20. exf7+ Kxf7 21. Be3 cxb1=Q 22. Raxb1 Ne6 23. Ba7 Rb7 24. Rxe6 Bxe6 25. Qf3+ Kg8 26. Qxb7 {and White clings on, which I'm not even going to try to explain. Good luck finding all this over the board!}) 16... Ng4 {Apparently the most likely outcome here is Black managing to force a draw over the next few moves - maybe not so surprising in the face of White's slumbering minor pieces. A sample line is} 17. Nc3 Qh4 18. Rxe5+ Be6 19. Qxd4 Qxh2+ 20. Kf1 Qh1+ 21. Qg1 Nh2+ 22. Kf2 Ng4+ 23. Kf1 Nh2+ $11) 13. h3 (13. e5 $143 $2 {As I mentioned in the Ahluwalia game, White has to stop following the 7. Nxe5 line first since this runs into} Ne4 14. Qf3 c4 15. Qxe4 Qb6+ 16. Kh1 Nf2+ 17. Rxf2 Qxf2 {- in the equivalent, Black's rook would be hanging on a8.} 18. exd6+ Be6 {White can at least improve over my old analysis with} 19. d7+ Kxd7 20. h3 cxb3 21. Nc3 Rhc8 $15 {, but it is an uphill struggle to show he has enough for the exchange.}) 13... c4 14. hxg4 cxb3 $6 {Understandably, Black did not hesitate to recapture the piece, but even here the engine has an improvement!} (14... Nxg4 $142 $1 {Just one knight and the queen are enough to cause problems for the airy white king. I think the best try for White is the unexpected} 15. f5 $5 {, interrupting the defence of g4 and clearing f4 for his light-squared bishop.} (15. Qxg7 $143 $4 {loses immediately to} Qh4 $1 16. Qxh8+ Kd7 {and the only way to prevent mate is} 17. Rf3 Qh2+ 18. Kf1 Qh1+ 19. Ke2 Qxg2+ 20. Ke1 Qxf3 $19 {and White is left hopelessly uncoordinated.}) (15. Ba2 {White cannot preserve his extra piece this way because of} b4 $1 16. Qf3 {(so as to meet ...Qh4 with Qh3)} b3 17. cxb3 Qb6+ 18. Kh1 cxb3 {and it is trapped nonetheless.} 19. Bxb3 Qxb3 20. Ra3 Qb6 21. f5 h5 $11 {This is an imbalanced position, but the engine's assessment of dynamic equality is understandable - White's space advantage and the target on d6 are offset by the weakness of b2 and e5, as well as the insecure position of his king.}) 15... Qb6+ 16. Kh1 Nf2+ 17. Kh2 {White's king has to walk the plank, since giving up the exchange at once results in an instant repetition.} (17. Rxf2 Qxf2 18. Qxg7 Qh4+ $11) 17... Ng4+ 18. Kh3 Nf2+ 19. Rxf2 {Only now! This time h4 is under White's control.} Qxf2 20. Qxg7 {To stay in the game Black has to find the unexpected blow} Bxf5+ $1 (20... Rf8 $143 $2 {Black has no time for consolidating moves, since} 21. Nc3 $1 cxb3 22. Qf6 $18 {followed by Bg5 gives White a crushing attack on the dark squares. For once it is White who benefits from completing his development in good time here.}) 21. exf5 Qxf5+ 22. Qg4 Qxg4+ 23. Kxg4 cxb3 24. cxb3 $14 {after which, finally, White will emerge with two pieces against rook and pawn. I feel that he preserves some slight chances here since he still has his other rook. Nevertheless this is better than Black ought to get after the text.}) 15. g5 {(1:27) So, after sifting through the memory banks - and even that wasn't so easy - everything was in order and my target position was on the board. For all the unwitting holes in my preparation, I had correctly evaluated this as better for White since the immediate threats to my king are warded off - I am effectively using the initiative as a shield here. Therefore, when Rida quickly played} Qb6+ $2 {(1:11) , a move which I didn't remember, it set an alarm off. I immediately saw the combination that she was relying on, but now I was prompted to look deeper and spotted the catch.} (15... b4 $142 {Miraculously, one line at least has survived Stockfish's attentions intact - as I gave in my earlier notes,} 16. Qd4 Nh5 17. cxb3 O-O 18. Be3 $16 {favours White, less because of his extra pawn and more because of the kingside juggernaut and Black's weak queenside.}) 16. Be3 Nxe4 {Really, it was a bit naive of Black to play like this - if your opponent is clearly in preparation, it's unlikely that a two-move tactic which seems to win out of hand is going to work. Therefore I was very happy to uncork} 17. Qxc8+ $1 {. D'Costa didn't look at the game again after this.} Rxc8 18. Bxb6 bxc2 19. Nc3 $8 {The last bit of accurate calculation needed in the game - I played this having seen up to my 25th move, and rightly concluded that the game would then effectively be over.} Nxc3 20. bxc3 {(1:21)} Rxc3 {(1:02)} 21. axb5 axb5 22. Ra8+ Kd7 23. Rxh8 c1=Q 24. Rxc1 Rxc1+ 25. Kf2 $18 {(1:20) Black looked absolutely distraught, as she had done ever since my 17th move, and would keep on doing until she resigned 40 moves later. Perhaps she too was hoping to sneak into the Championship qualifying places and watching the chance slip away.} b4 {(1:02)} 26. Bd4 {Two split pawns are no match for the bishop here, especially as Black's king is distant. Though I was a little bit nervous of the b-pawn (guess how I lost in round 1!) my basic plan was to grab all of the kingside pawns, give up the bishop for the passers and then easily win the resulting rook ending. Far-fetched perhaps, but...} Rc2+ 27. Kf3 Ke6 {27... g6 or 27... h6 would have slowed me down a little, but without changing the result of the game.} 28. Bxg7 Kd5 29. Rxh7 b3 30. Rh8 {(52) Now I don't even need to sacrifice my bishop, as if 30... b2 31. Rb8 wins the b-pawn outright.} Rc6 {(22)} 31. Rb8 Kc4 32. f5 d5 33. Ke3 {Blockading the dark squares.} Rc7 34. Bf6 d4+ 35. Bxd4 $5 {(43) I should clarify that I deliberately fell into the 'trap' rather than blundering a piece - I'd seen her last move coming.} Re7+ {(19)} 36. Kf4 Kxd4 37. Rxb3 {Now my plan outlined on move 26 has been realised - whilst some would have preferred to keep the extra piece, I would rather play a position in which my opponent has no counterplay at all. It's a matter of taste.} Re4+ 38. Kf3 Re7 39. Rb4+ Kd5 40. Kf4 {(41)} Rc7 {(13) Allowing me to cut off the black king from the pawns, making things even simpler.} 41. Re4 Kd6 42. g6 fxg6 43. Re6+ {I'd originally planned the simple 43. fxg6, keeping the enemy king out, but I was eventually drawn in by the neater choice. With hindsight I'd probably take with the pawn, which is easier to play, but by this stage White is not going to be denied in any case.} Kd7 44. Rxg6 Ke8 45. g4 {(1:05)} Kf7 {(40)} 46. Kg5 Rc2 47. Ra6 Rc7 48. Rh6 {I'd never studied this ending, but I knew that the only time it gets fiddly is with N+R pawns. From there I realised that outflanking the enemy king is the way to the goal.} Rc4 49. Rh7+ Kg8 50. Ra7 {(1:04)} Rc6 {(28)} 51. f6 Rc4 52. Kf5 Rd4 53. g5 Rd8 54. Kg6 (54. g6 {To be pedantic, this wins more quickly, straight out mating Black instead of merely escorting one of the pawns through.} Rd5+ 55. Kf4 Rd8 56. Kg5 Rb8 57. Kh6 Rc8 58. f7+ Kf8 59. g7# {Still, you can only win the game once.}) 54... Rb8 55. Rg7+ {(1:02)} Kh8 {(25)} 56. Rh7+ Kg8 57. f7+ {Black had finally seen enough, and scarpered from the hall having flatly refused my offer of a postmortem. This was the first time in a long while that I beat someone of my own strength purely with preparation - it was a weapon that I lost by cutting back on 1. e4. However, going over my pre-game notes with a more powerful engine is a sobering experience.} 1-0

Kasparov continues the foreword with which I began this post by reflecting that there is a kind of mythical link between preparation and competitive success; the belief that you are carrying a 'deadly weapon' gives you confidence even if it turns out you really aren't. That is a good summation of this game. For some time I've largely gotten by on the old analytical fat that I generated around 2016-2018, and it's chilling to think how error-strewn much of it is likely to be. Even in an era of 3600-rated monsters, we haven't yet reached the heat death of opening theory.

I will limit myself to two other points about the game, just emphasising comments from the notes. Firstly, though 15... Qb6+? might be a forgivable tactical lapse in isolation, under the circumstances it was foolhardy to play a move which loses instantly if the underpinning two-move tactic doesn't work. True, I wasn't blitzing my moves (I couldn't remember them that quickly!) but I think it was obvious I was still in preparation, and even 2016-vintage engines can see that deep. This was one of the very few cases where you really do get a "White to play and win" klaxon at the board! Secondly, my technique was for once perfectly assured, and that was because I was careful to keep the counterplay under control. The engines initially react with horror to 35. Bxd4!?, but it was a good practical move because it showed that I understood myself.

By this point I had finally arisen from the slumber at the start of the tournament and recovered to a positive score, with 4/7 to my name. I expected 6/9 to be the target score to finish in the top 10 and thereby qualify for the 2024 Championship, as it proved to be, and the dream was suddenly alive. But not for long - the eighth round was a case of diamond cut diamond, as I played 25 moves of Najdorf preparation... but so did my opponent, and I had no winning chances. I therefore contented myself with a six-move draw in the last round, and the very next day went off to the job interview that led to me writing this from a quiet Oxfordshire village.

August and September were quiet months after the action of July, but before my annual spreadsheet-assisted review of the year I'm going to pick up one more game. Here I showed off an opening triumph, so for the sake of balance let's look at an opening disaster...