When I got back to Colorado, I got two hours sleep Tuesday night, then less than three hours sleep on Wednesday night, running around town partying with Alex. So by the time Thursday comes around and I find out there is a game, perhaps that does leave me with some excuse for what was to follow, Shirley’s nice quiet house and free coffee not withstanding (it is ideal tournament conditions playing there, BTW).
My feeling was that 1.d4 and a King’s Indian Defense would give me a solid advantage against Rhett, but I decided to play this sharp system instead, even knowing that it’s virtually walking right into his preparation.
My first suspect move was 10.Bd2. I wanted to play 10.Kh1, but figured he would at least equalize with ..exf, 11.Bxf Qh5. But now I missed that after 10.Bd2 exf, that my planned, 11.Nd5 does not work because of …Qc5+, picking up the knight.
Stunned by this, I play 11.Kg1? when a more interesting move may be 11.Ne2!? with a nice attack on Black’s queen, plus interesting recapture on f4 possibilities. Of course I was holding this knight in reserve to trap the queen in lines where Black might take on b2, and Na4 traps the queen. It’s interesting how my tendency to maximize caused me to swot lines. Earlier, it was Na4 lines which caused me to miss the …Qc5+ threat, and here it was this queen trap line which caused me to not realize that this was a fresh position where b2 pawn could be given up with an attack on his king. For instance. 11.Ne2!? Qb6, 12.Bxf4 Qxb2??, 13.Qd2, and actually 14.Rf1b2 would be trapping Black’s queen!
13.gxf3? I had so much time to think at this time-control that I went into this sketchy position for the allure and fun of it. Most of my time was spent calculating the serious 13.Rxf3 Ne5, 14.Bxf4 Bd6, and then what to do with the rook? 15.Rf1 would simply give up the d-pawn after …NxBd3, 16.cxN, Rxd3, but there is probably a lot more here than just that, and even 15.Rh3 or Re3 may be possible.
17.Na4? 17.b4 deserves the most consideration, although Black has such fancy continuations available as …Bc5, 18.bxc6 Be3, 19.c6xb7+ Kb8! or 17…c5, 18.b6 a6. White might still be losing, but at least is sort of forcing, if not tempting Black, to develop his remaining two pieces.
18.Kg1? 18.Rg1, which Rhett offered post-mortem, is a far better defense as …Nxf3, 19.Rg2, threatening 20.Rf2! appears playable, and indeed we got into a double rook endgame where I was only down a pawn yet still managed to lose with risky bad play.
19.h2xNg3?? Not good. By this point I was getting more short-sighted in time-pressure, and not able to think rationally. 19.Rf2 NxB, 20.RxN Nxf3+, 21.Kh1 was still a whole lot better than losing immediately.
This looks like a terrible loss, and is, even on paper, but it did give me a better sense of when not to sac a pawn and when to. If I had had a real attack, it may have been a different story. I thought that Anthea’s loss to Alex began when she decided to recapture Alex’s pawn instead of letting it stay there and ignoring it – different game, different position – whereas I think others looking at that game with me didn’t possess that sort of sense as much as I have developed by making dumb pawn sacs myself (I know a good one when I see one, or one that can be recaptured later).
It would be a little intriguing to know whether or not an engine agrees with my analysis here or finds anything that I didn’t find, and slightly wonder if it won’t, although surely it ought to.