The game was agreed drawn after move 26 as we were both down to seven minutes remaining and realized that the position was too complex to be played out. I had seen all the way to move 29, but wasn’t sure what he’d do there and it looked scary how he could reposition his knight (in our mutual time-trouble). I show what seems obvious to me, the 30.d5 push, and that is around +.6 after …Qc3 coming up, but actually 30.b5 is winning (..Na8 instead of ..Nd7 had been a blunder), but this is quite difficult to see at the board, and if you had seen it you wouldn’t have enought time to analyze or play it correctly most likely – a pity, really.
A great game, but the piece de resistance would have been on move 30, and it’s more of a combination, and not a tactic.
I am not even sure why Stockfish thought that 29.Na8 was +.9 or more. My main computer crashed (probably a bad data cable), and stockfish on an Android tablet, well these tablets are hard to navigate (I am on my laptop now).
Actually, I wasn’t worried about the 28..Qf6 variation during the game, I was worried about a 28..Nf6 variation, where the knight would like to go to e4 or g4 and it’s very difficult to account for knight like this in time-pressure, as I told him after the game. During the post-mortem, he even got in this ..Rxd4, …Rd2 followed by ..Qb2 or ..Qd4 with this mating battery against my king; this could happen for example if he simply sacs his Nc7, and then I lose instantly by playing Qxb7 after that instead of Qb6. I would have been surprised and most likely have lost if he uncorked this during the game, as one of the reasons that I accepted the draw was because I thought that my king was looser than his.
One thing I have to improve on, even in time-pressure, is simply to believe that he repositions an active piece to somewhere else that it is then attacking something else, yet undefending possibly an even more important point. I can give up d4 to take e6, which is more devastating for Black. If the knight moves back, then not only does it undefend d5 and b5, but is also pinnable to the king along the back-rank. The Nc7 sac variation would have been extremely challenging for me during time-pressure, and that is the one variation that I am still not clear on; undoubtedly, I am missing a resource which I will find, but would not have found in time-pressure.
I found the simple defenses to the Nc7 sac, and I am glad that I don’t have a computer to help me because I know that my opponent tonight for example, Rhett, would never miss these simple defenses. For example after …Rd2, Black can resign after Qc7-e5. If instead of ..Rd2, Black plays ..RxR, BxR Qd4+, Kh1 Qd2, Rf1 with Bf3 coming up if needed, and Black can resign again as White also has Qe5 to play if needed.
All this takes us back to the …Nf6 variation which I was afraid of, but mostly because of the time-pressure. After ..Nf6, Bc2, I thought that neither side had a breakthrough, and didn’t realize after all of his defensive play that he might choose to play it sharp here, instead. We both admitted that we didn’t have time to play the position out correctly, which is why we both were okay with a draw.
…Qf6 is a bad move. It is better to develop the knight instead. Intuitively, I fell this, but OTB I have to calculate refutations and often by this point simply do not have the time to do so so opponent’s can then get away with stuff, and unrefuted moves often turn into wins for the side playing them.