This was the last time ever that I play a rapid tournament with 15'+10''. It was pure hell waiting for every next round. There is always two idiots with some interesting endgame, and 98 other players have to stand around and wait. One good example today was a rook vs knight endgame that the rook party eventually won.

Well, this is the position against Idaczek (1894), when after 24... Nd3 25. Rxe8 Rxe8 26. Nce4 Bxe4 27. Nxe4 Qb6+ 28. Kf1 d5 29. Qxd3 dxe4, I suddenly realised that 30. fxe4 Rf8+ 31. Ke2 Qf2# 0-1 is a problem. The problem is that he wins, and not I.

So, I had to think, which, according to Jan Gustafsson, is something you might be forced to do if you do not want to lose instead. Anyway, in the end I found 30. Qd7 Kf8 31. Qf5+ Kg8 32. Rd7 and as I was happy to pick up the central pawn without being mated, Black gifted me with a nice mate instead, 32... Qe3 10. Qf7+ Kh8 33. Qxe8# 1-0.

This is another position against Götze (1960). Since my team had put a low-ranked player on board one, Frank was unhappy to see me sit down at board four. First, one must note that Black can counter with b5 anytime he likes. The game proceeded 22... Bf6 (intending Bh4 and Ng3+) 23. g4 Ng7 24. g5 Bd4 25. Bxg6 (winning one pawn) hxg6 26. Qxd4 Nf5 27. Qg4 Kg7 28. h4 Rh8 29. Bg3 Rh5 30. Kg2 Qh8 31. Rh1 Kg8 32. Bf2 Qc3 33. Qe4 Kf8 34. Qc2 Qh8 35. Qb2 (giving the pawn up by blunder) Nxh4+ 36. Bxh4 Rxh4.

I am almost 100% sure that Black played 33... Kf8, so that 37. Qxh8+ was possible, which I must have missed. Instead I went 37. Rxh4 and managed to use the h-file with my rook to make my adversary use too much time.