Endgame Riddle: Unzicker vs Botvinnik, Amsterdam 1954

MickyMaus90: Very good. Many thanks!

So is 49.Ke2 better or easier than 49.g3?
I think this a matter of opinion. After 49.g3 Kf8 50.Rd7 e5 first White must correct his move order by 51.h5! Kg8 52.Kg2. Now we are back in the game and after 52...Ra5, White should use either 53.Kh3 Ra2 54.f4 or 53.f4 with an active use of his king.
Personally I consider this last line the most systematic at the moment, but 49.Ke2 and some other ideas are very convincing wins as well.

JeanCatalan: 49.Ke2 is interesting. In my analysis I ignored that move, because I liked 49.g3. Giving White a second option to bring in his king by Kf1-g2, besides Kf1-e2-f3. Indeed the advantage of 49.Ke2 is to keep the spare tempo g2-g3 and highly surprisingly White can use this in the rook ending.
One of the critical positions occured after Botvinnik's 56...Rd5. Here the game move 57.g4 wasn't best (still winning), while Botvinnink's old line with 57.h7 is good. Another attempt is to hand over the move to Black by 57.Ra7, with the idea of 57....Kg8 58.Rg7+ Kh8 59.Re7, so a rook triangulation. Admittedly Black can try 57.Ra7 Rd6 with a bit of couterplay.
Back to 49.Ke2. The critical line featuring Botvinnik's defence is: 49...Kf8 50.h5 Kg8 51.Re7 e5 52.h6 Kh8 53.Kf3 Rd5 and now White has the elegant 54.g3, achieving the critical position with Black to move. If now 54...Kg8, then 55.g4 is possible, 55...e4+ 56.dxe4 fxg4+ 57.Ke2 d3+ 58.Kd2 Rd4 59.Rg7+ with check and Rg7xg4 next.

Poiuy Trewq: In your line 49.h5 f4 50.h6 Rh5 White can win on the spot by 51.h7, because White can transit to a winning pawn ending, if he wants to. Like 51...Kf8 52.h8Q+ Rxh8 53.Ra8+ Kg7 54.Rxh8 Kxh8 55.Ke2 Kg7 56.Kf3 e5 57.g3 fxg3 58.fxg3+-.
The reason 49.h5 is dubious is 49...Rb1+ 50.Ke2 Rh1, as this forces 51.Rh7, so the rook in front of the pawn. White is still winning, but many would consider this bad technique. For more details see my contribution in the solution to come.

In 2) it should of course be "53.f4 wins" not h4.

albitex:
1) Yes 51.Kg2?= is a mistake as 51.h5 wins.
2) Yes 53.h6? is a mistake as 53.h4 wins.
3) 61.Re8+? is a mistake as 61.e5 wins as already Botvinnik had proved.
But there are more mistakes...

53. h6? (53. f4 +-)
-
61. Re8+?? This error leads to a draw a won game: (61. e5 and Black can only resign)

From a quick preliminary analysis, I would say that 51. Kg2? is a bad mistake. This move wastes time unnecessarily, better 51. h5 is White has a clear advantage.

Poiny Trewq: The game move 49.g3 wins. Your suggestion 49.h5 wins as well. But 52.g4? is a mistake, after which it is drawn (52.Ke2 wins) due to 52...fxg3 e.p. 53.fxg3 Kg8=.

Looking at the adjourned position (and not reading the comments) my initial impression is that Botvinnik's sealed move is a loser, for it allows White to launch the h-pawn immediately: 49. h5 f4 50. h6 Rh5 51. Rh7 Kf8 52. g4 Rh4 53. f3 Kg8 54. Rg7+ Kh8 (54 ... Kf8 55. g5 Rh5 56.Ke2! Now with the K-side pawns secure, White's King can venture to the Queenside to attack Black's marooned pawns) 55. Rg6 Kh7 56. Rxe6 Rxh6 57. Re4, giving White a two-pawn advantage and most likely the win. [full disclosure: analysis not checked by silicon, as I do not own a chess program]. Thoughts?

brian8871: Yes the game move 51.Kg2? throws away the win and 51. h5 wins.
But there are more mistakes...

51. Kg2 is not best. 51. h5 is stronger.

JeanCatalan:
1) 49.Ke2 indeed wins, but the game move 49.g3 wins as well.
2) 53.Kh3! +- is indeed better than the game move 53.h6?=.
3) The game move 61.Re8+?= is indeed a mistake and 61.e5! wins as Botvinnik has proved.

But there are more mistakes to be found...

White had two better attempts:

49.Ke2 Rb2+ 50.Kf3 Rd2 51.h5 Kf8 52.Rd7 e5 53.Rd5 Rxd3+ 54.Ke2
and
53.Kh3 Ra2 54.f4 exf4 55.gxf4 Ra1 56.Kh4 Rg1 57.h6 Rg4+ 58.Kh5 Rxf4 59.Kg6 Rg4+ 60.Kxf5 Rh4 61.Kg5 Rh2 62.Rxd4
finally clear win instead of 61.Re8+? 61.e5 Rf4 62.Kxd3 Rxf2 63.Rg7 and win se table bases.

calvinamari: How?

White wins