Knight Endgames – a Crash Course

Knight endgames are supposed to be like pawn endgames. That is true for the most part, but they are only similar. They are not the same, and in this article I would like to point out a few important differences. As usual, there will be puzzles at the end, so keep reading!

In general, king activity is extremely important, as is knight activity. That’s fairly logical and goes without saying. For instance, take a look at this position from one of my old games:

Walton, John (1856 USCF) – Brodsky, David (2100 USCF) Marshall U2300 September 2013


Black to move

Black has a powerful knight on d3, while the white knight is very passive on d1. Black is simply much better here, and I went on to convert my advantage. That should be fairly intuitive and logical. Now for a more complex example from another game of mine:

Katz, Gabriel (1942 USCF) – Brodsky, David (2032 USCF) New York State Scholastic 2013

G. Katz 1

White to move

Black appears to be a clean pawn up, but the white king comes and ruins black’s party. The game went 35.Ke2 Kg8 36.Kd3 Kf8 37.Kc4 a6 38.Kc5

G. Katz 2

White to move

Boy, is white’s king more active! Black has to be careful not to lose the a6-pawn outright. I played 38… Nc7 and after 39.Na7 Ke7 40.Nc6+ Kd7 41.Ne5+ Ke8 42.Nc6 we agreed to a draw as anything other than repeating with Kd7 would have gotten me into trouble.

What’s the overall conclusion? Just like in pawn endgames, king activity can be really important. In the previous position game, king activity was more or less the equivalent of an extra pawn.

In knight endgames, calculation is also important. Big time. Calculation in knight endgames can be more complicated than in pawn endgames, as there are generally more branches of the calculation “tree”. Knights can hop, fork and gobble loose pawns. Things can get messy, and intuition alone won’t guide you. Of course, there’s the “technical” side of knight endgames, where one side has everything under control, methodically improves their position, and strikes at the right moment. In my experience, however, those don’t appear very often.

With pawns on the same side of the board, positions that are easily winning in pawn endgames aren’t winning in knight endgames. Don’t despair, however, as in many of those positions your winning chances in a knight endgame are much higher than in, for instance, a rook endgame. Also, not all positions that are knight + pawn vs. knight are draws, as a lot depends on where the pieces are. Outside passed pawns can be very powerful as long as they’re supported. They can be excellent distractions, but they aren’t of much use if they drop immediately!

A complex example

This game combined outside passed pawns, active knights and kings, grabbing pawns, calculation, and many other aspects of knight endgames.

Brodsky, David (2314 USCF) – King, Alex (2365 USCF) Marshall Grand Prix March 2015

King 1

White to move

In this game, my two connected passed pawns on the queenside ended up serving as distractions. They won’t queen, but they keep the black king and knight occupied. It’s time to go for the kingside pawns. I played 37.Ne4! with the threat of Ng5 forking the f- and h-pawns. After 37… Kxb4 38.Ng5 Kc5 I was at the crossroads.

King 2

White to move

I had to choose what pawn to take. Black’s next two moves will most likely be Kb6xa7, while white should grab as many pawns as he can and bring his king into action. 39.Nxh7! was stronger, because after 39… Kb6 40.Ng5 Kxa7 41.Nxf7 white’s knight is much better placed on f7 then on h7, and white is just winning. Black can avoid this by playing 40… f5, but white can decisively run his king up with Kf2-e3-d4-e5. Instead, in the game, I made my life harder than necessary (seems to be a hobby of mine) by playing 39.Nxf7?! which still (barely) wins. The game went 39… Kb6 40.Ng5 Kxa7 41.Nxf7 Ne6! 42.Kf2 Kb6 43.Kf3 Kc6 44.Kg4 Kd6

Puzzle 1

King 3

White to move

The semi-forced moves are over, and now it’s up to you to find how white wins! The win isn’t really forced; you need to find the first couple of moves, and then the rest falls into place.

Now, I have a couple more puzzles to keep you guys busy. As usual, I’ll publish the answer on Sunday in the comments. Enjoy!

Puzzle 2

Karthik P

White to move

Is 27.Re7 a good idea?

Puzzle 3

Rohde 1

White to move

Calculation time! What move(s) is/are winning for white? Give yourself a few minutes to think and decide what you would play in an actual game.

Until next time!

One thought on “Knight Endgames – a Crash Course

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s