Thanks for the diagram
an ordinary chessplayer wrote on 02/06/18 at 19:00:08:
I just did a quick pass through the moves. After 53.Rb6+, 53...Ke5 seems wrong. Instead 53...Ke7 54.Rxa6 Kd8 and for the moment white cannot trade rooks, later black might draw even without the h-pawn.
Re 53...Ke7 – no, it makes no difference at that point where the king goes, since 54 Rxa6 is a winning endgame according to Lomonosov. Apparently after 55...Kd8, it's mate in 30! But otherwise, yes, a drawn R+a+c v R is what I was playing for in a lot of cases. I just missed that with 52 d7! he could force one that is winning for White.
Stefan Buecker wrote on 02/06/18 at 20:28:39:
A hard fight and instructive ending - thanks for the link, Jonathan. I fear 53...Ke7 is losing, but 29...Ke7 comes into consideration: 30.Rh6 f3 31.c5 Rf1+ 32.Kc2 f2 33.Re4+ Kf7 34.Rf4+ Ke7 35.Rxh5 Rd8 36.Rhf5 Ke6 37.Rf6+ Ke7 38.Rxf2 Rxf2 39.Rxf2 Rd5. I don't have access to the Lomonosov tablebases though...
Re 29...Ke7 – maybe I didn't look at that as closely as I should, since I was deliberately aiming for the endgame in the main line. But in general I thought Black had more chances with the h-pawn than the f-pawn. For instance, after something like 30 Rh6 f3 31 Rxh5 Rf8 32 Re5+ Kf6 33 Rdd5 Rh8 34 Rf5+ Ke6 35 Kc2 Rhh2 36 Rf4 Rf1 37 Rd3 f2 38 Rdf3, I'm not sure whether it's possible just to sit on this, as White can make progress with Kd4, d2-d3, edging slowly forwards, and getting the a-pawn for the f-pawn isn't enough for Black here. Those endgames seem always to be lost.