My eyes lit up when Calvin played 17…f6??
Perhaps the biggest surprise of the game was the early resignation. After 24.Rh6, at first I intended 24…Qc3, 25.Rxh7+ RxRh7, 26.QxRh7+ Nf7, 27.BxNf7 QxNb3 and here I sort of stopped my analysis to see if there wasn’t a more direct mate. When Calvin resigned, I was still looking at 27.RxNf7+ BxRf7, 28.QxNf7+ Kd8, 29.Qf8+ Kc7, 30.Qe6+, so I showed him that line, and said “is that mate?” (it’s a 0.0 perpetual after 30…Kb7) We both thought it was mate, although he had thought I was going to play the line which actually does lead to mate, 27.BxNf7.
I believe that I would have actually played correctly by taking on f7 with the bishop, but I probably would have been down to 9 minutes by the time I would figure out the position and play it, whereas I was only down to maybe 18 minutes when he resigned. Winning a won position is one of the hardest things because there are many junctures where one can easily get sidetracked.
Even at this hypothetical 9-minute-point, the best I would have come up with would have been 27.BxNf7 (or BxBf7) QxNb3, 28.BxBe8+ KxBe8, 29.Rf7 which does win because Black has to sacrifice queen for rook to stop the mate, and then White will also have a passed g-pawn. It was as if he didn’t give up on the original line, calculating it ’til mate, whereas my analysis was to hop around a bit whenever I got stuck, which is more my style; sometimes I will calculate a line all the way through, but often during a move I will base the move on a patchwork of analysis, which is not the ideal way to play, but also lets me examine more lines.