Played Anthea again as White, our seventh encounter (hardly seems that many times).
It seemed as if this game would be about the opening or the middlegame, but it turned out to be about the endgame where I threw away the win with a rather cocky move in time-trouble. I played Kf3, too much bravado, as I didn’t even anticipate the completely obvious …h5! This is what time-trouble does, missed something that a blind person would have seen.
I had around five minutes for the ending, but that clearly wasn’t enough. I didn’t see how to make progress in that limited time.
Once my king got stuck on f3, she could have exchanged her bishop and rook for my queen and pawn, if she wanted to, but I couldn’t see how it was forced, or even who would be winning then. The interesting thing about that endgame, should she have allowed it, is that if Black plays …b6 or ..b5 it’s a draw, but if Black were to move her king, then I could play a5, fixing the backward pawns, and then win a tempo with the king, and win the pawn-race by one tempo, winning.
Kh3 instead of Kf3 was easily winning, but on f3, the king is immediately checked by dxe+, whereas if the king were on h3, then that is not a check and there is no need to immediately recapture the pawn. Actually, it’s only “easily” winning, if Black had followed up Kh3 with ….h5?, whereas …Rg7 would have held out for a long time to come and I didn’t have enough time to play that out correctly – it actually goes 80 moves, the king has to get in there, queen trades for rook, and then bishop has to sac itself for a pawn while the Black king had previously been pushed out of play by the White queen.
Earlier in the game, when I played Kh1, that was based on mating possibilities that I had, but needed her queen to be one tempo further away from defense of her king. For example, after Nf5, Crafty first sees the line …Bd7, but then Nxg7 and the knight can’t be taken or it’s mate. Of course, I was also wondering how I could keep up materially and would it compensate for ditching my queenside, but apparently it does work because of Qg3. In the game, my Qg5 was a mistake. Qg3 followed by Nxd6 (threatening Nxf7+ winning an exchanges!) is much stronger. That was a quick move, and I wasn’t sure which line was better. In the game, Black had a defense, ….Qe2. This move takes away Qh5 so that the White queen won’t be able to check on f7, and Black is okay or better.
In the Qg3 line, I had realized that …f6 was Black’s best defense, and it is supposedly equal, but with Crafty I looked at a line where White gives up the d3 pawn for the exchange, which Crafty first rated as .27, but I knew it was somehow winning. What is surprising is that I could not win that endgame without Crafty’s help because there were so many pitfalls against sloppy play by White, leading to draws. But I did realize that a forced trade of rooks, and Crafty helped me with the rest.
Sadly, in the future I will need to save even more time for endgames against strong players. There is the school of thought that it’s better to study endgames for this reason, but I think I simply need to play quicker middlegames, more decisively. There was no reason that I needed to spend so much time earlier other than I didn’t think I was going to need it later, but I did.
Well, this is what happens when you don’t have that second time-control to fall back on and have to save all that time on the clock up-front. It was an interesting and exciting game that got fully played out (given that not enough time was saved to play it out well).
Going back to the game, her move 15…Bd7 instead of 15…Be6 (what I had expected), was a blunder. It’s even more interesting to realize why she played this move, it was to set up the …Ng4 fork. However 16.Bb6 and eventually Bc7 will round up the d6 pawn once again, in a matter of moves.