Kenilworth Kibitzer

To answer the last question first, if the pawn were on the a-file, this is a dead draw, simply because Black's rook is placed excellently behind the passed pawn.  Black's rook can check White's king endlessly from this position because it will have nowhere to hide except in front of the pawn, which impedes its progress.  If the opposite were true and White's rook were behind the pawn, the position becomes winning for White because Black's rook is forced to be entirely passive, guarding the a-file while White's king runs rampant.  See for example Alekhine-Capablanca 34th game WC match (1927) 1-0.  Alternatively with the Black rook behi

nd the pawn see: Federowicz-Yermolinsky USA Ch (1997) 1/2-1/2.  

So what happens on the b-file?  It turns out to be quite tricky.  In fact, in this case, former world champion Petrosian lost the game, even though he had an overnight adjournment with helper monkeys to solve the puzzle.  First let's see what has since been determined to be Black's best approach to a draw (according to Kantorovich):

47....Kf6! 48. Ke3 Ke6 49. b5 b5 Ke5 50. b6 Ke6

It seems that White has made all the progress and Black

 is merely marking time.  However, after: 51. Kd4 Rb1 52. Rb8 Kd6 53. b7

It turns out the Black is completely safe with the paradoxical 53...Ke7!  

With the White rook on b8, this position is functionally identical to one in which the pawn is on the a-file, since Black's rook can check the White king at will, which has no shelter and no way to advance the pawn.  As long as Black's king stays away from cheap tricks (54...Ke6?? 55. Re8+ wins), there is no progress to be made.  Likewise, had White advanced his pawns earlier in this variation it makes no difference, all Black has to do is keep White's king away from his pawns.

47....Kh6 ?! (Kasparov) ! (Emms)......!? (Mangion)

Until the above defense was found Black's idea seemed the most compelling - he intends to advance his pawns and create a weakness in White's position that can provide counterplay and shelter for his king.  There is a long series of analyses of the strength or weakness of this move in On My Great Predecessors Vol. 4 which I'll leave to the hard-core.  I like Black's idea because I think it would succeed against most opponents and is easiest to understand as well.