Weekly Endgame Column #5

White to move and win.

This week the subject is knight + pawn versus bishop. We shall have a look at positions where the shortest diagonal in front of the pawn is 4 or 5 squares.

1. The bishop can defend against the pawn on its own

The pawn is being stopped at b5, and thus the shortest diagonal in front of the pawn is 5 squares long: from a4 to e8. When the shortest diagonal is more than 4 squares long, then the bishop can handle the pawn on its own – in the above position, the black king is not needed.

The reason for more than 4 squares is that the knight is able to cover 2 squares of a diagonal, same goes for the king. That gives a total of 4 squares at the maximum. I have written about this in Sharp Endgames too.

White can try:

1.Nc3 Be8 2.Nb5 Bd7 3.Nd4

The idea is to block the diagonal with Nd4-c6, but Black plays:


The Black king is only needed to make waiting moves: if White for instance plays 4.Nc6, then Black is not able to move the bishop.

So what if the Black king is not able to make waiting moves? These are the few exceptions where a 5-square diagonal is still not sufficient for a draw. We are entering the world of studies!

2. When 5 squares are not enough

M. Dimentberg 1949: White to move and win.

The pawn has to be stopped on d7, which makes the shortest diagonal in front of the pawn 5 squares long (a4-e8). So, in theory the Black king is not needed, and it’s even close to pawn, taking an active part in the defense! However, there are zugzwang motives in the position:

1.Kd7! Kd5

The only way to get the bishop onto the a4-e8 diagonal. The knight on g5 covers the squares e6, e4, f3  and h3, making a shift to the h3-c8 diagonal impossible.

2.Kc7 Lc6 3.Ne4!

Black is in a deadly zugzwang as the knight indirectly covers the remaining 4 squares on the a4-e8 diagonal due to knight forks.

Horwitz & Kling 1851: White to move and win.

In the above position, the diagonal a4-e8 should suffice for Black to draw – but the placement of the Black king in the a8-corner creates zugzwang motives:


Also possible is 1.Na2!

1…Ka7 2.Nb4!

The Black king is kept in the corner. If now 2…Ba4/2…Bb5, White plays 3.Nc6+- and the pawn promotes.

2…Ka8 3.Nc6!

The Black king is trapped in the a8-corner, and thus his majesty has no waiting moves. Unable to make useful bishop moves, Black is in a deadly zugzwang.

The study composer A. Kalinin has built on this idea for the following position. We shall see more to Mr. Kalinin in a later Endgame Column with 3 or 4 squares for the shortest bishop diagonal, he has made some valuable contribution.

A. Kalinin 1975. White to move and win.

1.Ne6! Bc6

This example is more complex and we are approaching moves and maneuvers from endgame theory, not just the (flashy) world of studies. I am a fan of this mix.

1…Bg4 fails to 2.Kd7 (threatening Kd7-e7 and d6-d7) Bd1 3.Kc7! Ba4 4.Nd8 Be8 5.Nc6!+- with the final position from Horwitz & Kling.


It is vital to get the knight to d4. If 2.Kc7? Be8= and Black is in time to get the king out of the a8-corner.


If 2…Bg2, White replies 3.Kd7! It is curious that this type of king move – in front of the pawn on the square of the bishop – keeps turning up in these endgames. At this point, it is the only move to win. The main idea is to keep the flexibility to go to both c7 and e7 depending on the circumstances 3…Bh3 4.Ne6 Bf1 5.Kc6+- or even 5.Ke7+- as 5…Bb5 6.Nc7+ wins the bishop.

3.Kd8! Ba4 4.Kc7! Be8 5.Nc6+-

Again the final position from Horwitz & Kling.

3. Starting out with 4 squares

V. Kosek: White to move and win.

Another brilliant composer in these types of endgames. As the king+knight can cover a maximum of 4 squares on a diagonal, such positions are interesting in connection with (mutual) zugzwang. The basic endgame theory is filled with complex examples, and I will treat some of them in a later blog post – but for now, the above position is a good starting point.


The only move to win – White has to force Black to show his hand before Ke1 moves. Should Black go to the long (a4-e8) or the short (e8-h5) diagonal?


From the main continuation, it will be clear what White should reply to 1…Bf7.

If Black stays on the long diagonal and plays 1…Ba4, White is able to show that the king has optimal flexibility on e1:

2.Kd2! Aiming for d6 over b4 and c5, gaining a few tempos on the way. 2…Kg2 3.Kc3 Kf3 4.Kb4! Bc6 5.Kc5! Ba4 6.Kd6+- White will block the diagonal with Nf6-d7 next.


It’s all about mutual zugzwang later.

2…Kh2 3.Kf2! Kh3 4.Kf3! Kh4 5.Kf4!

This is the key position to aim for, and White wins because Black is to move. White to move would not have access to either f5 or g5, but Black to move has to make a concession.


5…Bf7 6.Kf5+- leads to yet another zugzwang position for Black.

6.Kg5 Bf7 7.Kh6 Kh4 8.Kg7+-

The knight covers e8/h5, and Kg7 the squares f7/g6.

It is now clear that 1…Bf7 at the starting position should be answered with 2.Kf2!

…and Black will end up in the same zugzwang positions.

As I said, a later Endgame Column will treat more complicated positions with 4 squares on the shortest diagonal. Later, we shall also have a deeper look into 3 squares on the shortest diagonal.


%d bloggers like this: