Time-pressure blunder. He was down to 1 second on his clock at one point.
10.Bxd6. A better move to keep the tension was 10.Nc3-e2 – thought of this after the game.
23.Rxe3. 23.fxe3 seemed more accurate, but I was in time-pressure here and couldn’t make a good decision.
24…e5?! There was no hurry to make this move, as he could sit on this position with 24…Kf8, for example.
25…e4. Probably not a wise decision. Again, 25…Kf8 is reasonable here.
26…g5. I could win a pawn now, and saw this relatively quickly analyzing this part of the game for the first time afterward, just now as I plugged in the game-score. I always like to make the ironic comment “The strongest move is the one that goes unrefuted”, and this is a great example of that as in the coming endgame this is the perfect square for that pawn.
27.b4? Just a quick move (in time-pressure, sickeningly enough). 27.dxc bxc, 28.RxR RxR, 29.RxR NxR, 30.Bxe4 is winning. If 27…RxR??, 28.RxR RxR, 29.c7! wins. I saw this effortlessly now, but was super-nervous OTB due to my standing on the clock, which I used horribly in this game, but still had enough time to find this move, certainly.
29.d6? White should have fair winning chances with 24.Rd4 Rxd5, 25.RxR RxR, 26.RxR NxR, 27.Bxe4 Nf4, 28.Bxb7 Nxh3+, 29.Kf1.
31.Bxe4. My intuition told me that 31.Rxe4 was correct, but I didn’t know why. As it turns out, I needed to keep a pair of rooks on the board and should be fine, but after I made my move I figured out what I was in for.
33.Rd5? Ridiculous. I already knew that it was desirable, if not necessary to keep a pair of rooks on the board. With that in mind, simply playing 33.Rc1! fits the bill.
After this, I knew I was lost and offered draws, and then he finally offered me a draw in the final position. It’s strange because in the post-mortem, I let him win both my a and b-pawns, but won comfortably taking his g and f pawns. However, there are two reasons for this: 1.He played it feebly, and 2.When I played f4, he took on f4, but if he doesn’t take on f4, then I have to play fxg5, followed by h4, and am simply lost, since he did manage to sac his knight for one of the pawns in the post-mortem. I almost want to say that there are three types of lies: 1.Lies, 2.Damned lies, and 3.Post-mortems. 😉