1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 d6 5.h3 Be7 6.O-O Nf6 7.Re1 O-O 8.c3 b5 9.Bc2 b4 10.d4 exd4 11.Nxd4 { A somewhat strange idea, but for some reason I was trying to use the pawn tension to provoke an exchange of knights. } 11...Nxd4 12.cxd4 c5 { The downside of my concept. Now black gets this break. } 13.Bf4 Qb6 14.Nd2 { Ignoring the pawn, as white will be able to regain it fairly easily anyway. } 14...Bb7 15.Nc4 Qc7 16.b3 Rad8 17.Rc1 Rfe8 18.Bd3 Qd7 19.Qc2 d5 20.Ne5 Qc8 21.exd5 Rxd5 22.Bc4 Rxd4 23.Bxf7+ Kf8 24.Be6 { 24.Bxe8 is best, of course, but I found myself unable to resist this shot. } 24...Qxe6 { The point of 24.Be6 is that if black plays a move like 24...Qa8, white wins with 25.Qg6! with mate next move whatever black plays. For example, 25...hxg6 26.Nxg6. } 25.Ng6+ hxg6 26.Rxe6 Rxf4 27.Qxg6 Be4 28.Qg3 Rf5 29.Re1 Bd5 30.Rxe7 { A bit of a gamble, but I thought the resulting endgame with the three connected passed pawns would offer good practical chances. } 30...Rxe7 31.Rxe7 Kxe7 32.Qxg7+ Ke6 { I had expected 32...Bf7, which seems a bit more solid. } 33.g4 Re5? { The move Elliot and I both missed here was 33...Rf3!, blockading the white f-pawn, attacking h3, and maintaining defence of the f6 knight. Black's pieces coordinate nicely here to defend each other. I still think the position offers reasonable chances for a white win, but it certainly would've been much more difficult. However, after 33...Re5?, black loses the knight by force. } 34.f4 Re3 35.f5+ Ke5 36.Qe7+ Kf4 { The king can't keep up defence of the knight. } 37.Qxf6 Kf3 38.Qa1 Kg3 39.Qf1 Rf3 40.Qg2+ Kh4 41.Kh2 { White just has to avoid falling into 41...Rg3, which would have been embarrassing. } 41...Rd3 42.Qf2+ Kg5 43.f6 Bf7 44.Qf5+ { Winning the rook. } 44...Kh6 45.Qxd3 1-0