Originally posted by heinzkat
I would say black can easily "keep" the endgame by 24...Ra8 but don't take my word for it
I agree with this analysis, after fiddling with it a bit. Black needs to play to draw, since the salient features (more remote passed pawn, more advanced passed pawn, slightly better development) all favor white. Blockading the pawn with the idea of applying pressure to it sounds like the best way. Attempting to hold on to material equality instead of seeking counterplay and stifling white's biggest threat would probably do more harm than good- as the game result tends to verify.
An interesting side note to share as I was studying this:
In Beliavsky and Mikhalchishin's "Winning Endgame Technique", they devote a chapter to rook endings with the same number of pawns on one side, and one side having a passed pawn on the other flank. They make the following observation:
"The Moscow chess player Kantorovich invented a very interesting statistical system to help evaluate this type of ending. The stronger side has the best winning chances with the knight's pawn, which he gives 0.7-0.8, i.e. from ten games one would score 7-8 points. The bishop's pawn is evaluated at 0.65, a centre pawn at 0.7, and a rook's pawn at around 0.6. It should be noted that these evaluations are averaged out, and the result of any concrete position depends in the first instance on the activity of the pieces."
An awesome book, and highly recommended.