1. ## Blitz pawn endgame

Yesterday I played an interesting pawn endgame in Blitz (internet). In a queen endgame I had managed to throw a two-pawn lead away, and was hanging in the ropes with 53.Ke3!:

 FEN Viewer

We were in time trouble both of us, I had 8 seconds left and he had 28 seconds. It was a 3-minute 2-second increment Blitz game (as is currently the official standard to be used for the FIDE World Blitz Championship, which does not use 5-minute 0-second increment). The game continued 53...b3 54.Kd3 a5 55.Kc3! a4 56.h5! and my opponent suddenly realized that there was no way to stop the h-pawn from queening, and resigned.

I think the position is quite amusing in the twist and turns that happens at this point. Analyzing it might make you a better pawn endgame player, at least there are some instructive points. The question is whether to defend or attack. What were the mistakes in this sequence, and how would the game have ended in each instance (draw, White wins, Black wins)?

2. As Black, I might have considered 1...Kd5 so as to stay within the square of the h-pawn (given that its Blitz format and one is presumably not calculating but playing on instinct).

3. Originally Posted by The Snail King
As Black, I might have considered 1...Kd5 so as to stay within the square of the h-pawn (given that its Blitz format and one is presumably not calculating but playing on instinct).
In fact 53...Kd5! is the only move that earns a draw, to keep the king in the square of the h-pawn. I had hoped for a bit more analysis than a 1-mover though
White has to think about defense because black's pawns are further advanced. However, I could have played (with king on f3) 53.h5?!,b3 54.h6,b2 55.h7,b1=Q 56.h8=Q Qf5+ but it is black that starts the checks so at least a psychological advantage for black - but I think it should be a draw.
It is more difficult to see why 53.Ke3!,b3?? actually loses for Black. How can White win?

4. Originally Posted by Jesper Norgaard
In fact 53...Kd5! is the only move that earns a draw, to keep the king in the square of the h-pawn. I had hoped for a bit more analysis than a 1-mover though
But SK's whole point is that knowledge of the square rule makes a lot of analysis unnecessary, in the sense of specific calculations. 53... Kd6 would also draw for the same reason.

5. Originally Posted by Jesper Norgaard
It is more difficult to see why 53.Ke3!,b3?? actually loses for Black. How can White win?
Kd2. The b-pawn falls in two moves if black fails to protect it. Each side has now used one move compared to the position after 53.Ke3 but black is no longer able to get inside the square of the pawn without losing the b-pawn. So he has to try to promote his own pawns but ...Kb4, Kc1 Ka3, Kb1 is too slow; white promotes first in time to win.

I first tried Kd3 but this actually loses after ...Kb4 forcing Kd2 (the horror!); the loss of that critical tempo results in white getting mated by the a-pawn the move after promoting.

6. Originally Posted by Jono
But SK's whole point is that knowledge of the square rule makes a lot of analysis unnecessary, in the sense of specific calculations. 53... Kd6 would also draw for the same reason.
Yes I was not precise Jono, 53...Kd5 is the only idea that draws, but is not the only move, because 53...Kd6 works the same way. Checking the square of the pawn is useful here to avoid some calculations, but not to determine whether running for the square or attacking with own pawns is better.

Originally Posted by Kevin Bonham
I first tried Kd3 but this actually loses after ...Kb4 forcing Kd2 (the horror!); the loss of that critical tempo results in white getting mated by the a-pawn the move after promoting.
Exactly so when I could have won with 54.Kd2! instead I got a losing position after 54.Kd3??

I guess the variaton you have lined up - Kevin - is 54.Kd3?? Kb4!-+ 55.Kd2 Ka3 56.Kc1 b2+ 57.Kb1 Kb3 58.h5 a5 59.h6 a4 60.h7 a3 61.h8=Q a2# which is cute, but much easier is 56...Ka2! 57.h5 b2+ 58.Kd2 b1=Q

The final sequence of errors was therefore 53...b3??+- (53...Kd5(Kd6)=) 54.Kd3??-+ (54.Kd2!+-) 54...a5??+- (54...Kb4!-+) 55.Kc3! a4 56.h5!+- 1-0
Wouldn't have happened with more time on the clock for both players, of course! Blitz is Blitz.