- What is the most economic position in a pawn endgame where Black queens first, but only draws, White pawn(s) not higher than rank 5 on time of Black queening? (This subsumes ranks 6 and 7)
An answer (maybe unique) is (seen from Black)
[FEN "8/4K3/8/5P2/8/2k5/3p4/8 b - - 0 1"]
where after 1...d1Q 2.f6 it's the usual draw of queen against bishop pawn. The point is that the black King stands very unfortunate: 2...Qe1+ 3.Kf8 and 3...Qb4+ isn't possible (thus f7 can't be impeded and the bK is out of the winning zone either).
This end position BTW is the heart of the following brilliant study:
[FEN "5K2/2p5/8/2k5/8/2P5/4P3/8 w - - 0 0"]
Mikhail Zinar, Shakhmaty v SSSR 1981, 1st special prize, 1.Kg7!!
What?? Why not 1.Ke7 or 1.Kf7? 1.Ke7? Kc4 is obvious: either Black gets both pawns or again Q vs bishop pawn arises, this time with black K already on b2. And after 1.Kf7? Kd5! we have mutual zugzwang, White's only try is 2.Kf6 Kc4 3.e4 Kxc3 and so on, where White queens even another move earlier, but now Black plays the K to d2 and we have the above position (with mirrored colors)! Thus 1.Kg7! Kd5 (Kc4 is easy, as above, again White queens another move earlier and Qg5+ wins in the end, unimpeded by a Kf6) 2.Kf7! and Black is in zugzwang. (Running to c3 is, as seen, one move short, running to Pe2 hopeless either.)
In case we want White to be winning and not just drawing, we need more pawns. This is probably not the most economical solution:
[FEN "8/5K1k/6p1/6PP/8/5p2/4p2/8 w - - 0 1"]
1...e1Q leads to an immediate mate: 2.hg6 Kh8 3.g7 Kh7 4.g8Q#
After 1...gh5 2.g6 Kh6 3.g7, Black has a choice:
- They still can promote first, but after 3...e1Q 4.g8Q they will get mated by 5.Qg7# or 5.Qg6#
- Or they can run away with 3...Kg5, but then they won't promote at all, e.g. 4.g8Q Kf4 6.Qg1+-