King and knight

The square b2 is next to the corner a1. A king steps there in one move. A knight on a1 needs four moves to reach b2, and there is no faster way. By the middle of the chapter you will be able to explain why.

In the last chapter the rook, bishop and queen travelled along lines as far as they liked, but could not jump over anyone. Now for pieces that are built the other way round. The king moves only to a neighbouring square. The knight does not go far and does not follow a line, but it jumps over anything. Range has nothing to do with it, and a different question measures these pieces better: how many moves does it take to reach the target? For the king and the knight the answers are quite different, and the whole chapter rests on that.

The king: one step in any direction

The king moves one square in any direction: forward, back, sideways or diagonally. You could also say that the king moves like a queen, but only one square. From the middle of the board it has eight neighbouring squares.

Near the edge there are fewer neighbours, and in a corner very few. The king attacks all the neighbouring squares and defends its own pieces standing next to it: if the opponent captures such a piece, the king captures back. The king also has a special move together with the rook, castling, but it has a chapter of its own, chapter 6.

Where the king cannot go

You already know the main rule about the king from the first chapter: a king never steps onto a square that an enemy piece attacks. Every other piece may take risks, but the king may not. A move after which the king would still be in check is simply not allowed. So before you step with the king, check which enemy pieces are looking at that square.

One consequence of this rule surprises beginners: two kings never stand next to each other. Each attacks the neighbouring squares, so to stand right beside the enemy king is the same as stepping under attack. There is always at least one empty square between the kings.

Kings facing each other across one square, as in this question, get in each other's way: neither can enter the square between them. In endgames such a face-off decides whose king gets through, and it has a name of its own: we give it in First checkmates and study it properly in the chapter on pawn endings. For now remember the simple thing: kings do not stand next to each other.

The king captures the way it moves: a piece on a neighbouring square. The rule still applies here. If someone defends the enemy piece, the king cannot take it, because after the capture it would be in check. A king may capture only unprotected pieces.

The king counts in its own way

Now the question we began the chapter with: how many moves does the king need to reach a given square? On an empty board the answer is simple. In one move the king can shift by one file and, if it goes diagonally, by one rank at the same time. So the number of moves equals the larger of the two shifts, in files or in ranks. A king goes diagonally no slower than straight.

All strong players keep this property of the king in mind: in endgames it decides whether a king can catch the opponent's pawn or support its own. We come back to it in the chapter on pawn endings.

While there are many pieces on the board, the king hides in a corner behind its pawns. Walking forward into checks is dangerous for it. But if the opponent's pieces are tied up and cannot get at it, the king becomes a strong fighter. One of the most famous king walks in chess was played in Tilburg in the Netherlands in 1991, and it is remembered as Short's king walk. The Englishman Nigel Short had White against the Dutchman Jan Timman, and by move 31 White's queen and rook were already close to the black king.

The king crossed four squares, h2, g3, f4 and g5, and each time it stepped where Black's pieces could not reach. It never broke the rule "do not step under attack".

The knight: a jump in the shape of an L

The knight does not move along a line. Its move looks like the letter L: two squares along a file or a rank and one more to the side. You can also go the other way round, one square straight and two to the side, and you land on the same squares. The squares the knight passes over do not matter: whether they are empty or occupied by your pieces or the enemy's, it simply jumps. Only the landing square counts. The knight cannot land on a piece of its own side, and it captures an enemy piece like any other.

Look at the diagram more closely. The marked squares form a ring around the knight, but the neighbouring squares are not part of it. The knight attacks no square next to it and no square on its own file, rank or diagonal. Its attacks lie where neither a rook nor a bishop is looking, which is why they are so easy to miss.

The jump gives the knight what the line pieces lack. In the starting position the rooks, bishops and queens are shut in by their own pawns, but the knight comes out at once, straight over the wall of pawns. On the first move of a game only pawns and knights can move.

The closer a knight is to the edge, the fewer squares it has: some of its jumps lead off the board. From the centre a knight attacks eight squares, at the edge half of that or fewer. Players in different countries have sayings about this, and the English one even rhymes: "a knight on the rim is dim".

Compare the pieces of the last chapter. On an empty board it makes no difference to a rook where it stands: it attacks 14 squares from anywhere. A bishop in the corner attacks almost half as many squares as in the centre. A knight in the corner has four times fewer squares than in the centre. So the knight is usually put nearer the middle of the board: from there it reaches both wings in a couple of jumps, while from the edge it first has to come back.

The knight changes colour

The knight's jump has a property that makes its counting unusual. Look again at the diagram with the knight on d4: the square d4 is dark, and all eight squares it jumps to are light. It is always so. With every move the knight changes the colour of its square, from light to dark and from dark to light. The bishop, as you remember from the last chapter, lives on one colour all its life, but the knight cannot stay on its own colour for even one move.

From this comes a rule for counting. In an odd number of moves (1, 3, 5) the knight lands on a square of the other colour, and in an even number (2, 4, 6) on a square of the same colour as the one it started from. You have known the colours of the squares since the first chapter: a1 is dark, h1 is light.

Knight routes

Back to the puzzle from the start of the chapter. The knight is on a1, the target is the neighbouring square b2. Both squares are dark, so the number of moves must be even. Let us see whether two are enough.

How do you find such a path in your head? The easiest way is to think from the end. First ask from which squares the knight reaches the target in one jump: it comes to b2 from a4, c4, d3 or d1. Then find which of those squares is easiest to reach from the start. The chain is built from both ends towards each other. Try it yourself: the trainer names the square where the knight stands and the square it must be brought to, and you click the squares it jumps to.

When two moves become easy, go on to the levels with three and four moves. It is good training not just for your fingers: in a game you need to see where a knight can get to in a couple of moves before it has moved.

One last thing about counting. From d4 a knight on an empty board reaches any square in at most four moves. The longest trip for a knight is to the opposite corner: from a1 to h8 it needs six jumps. For a king it is seven steps, so the knight gets to the far corner faster. But to the neighbouring b2, as we saw, the king gets there four times faster than the knight.

Knight attacks

Knight attacks are easy to miss, because the eye is used to running along lines. The rook, bishop and queen aim along files, ranks and diagonals, and you can see that. The knight aims round the corner. When you check a square, it is not enough to draw eight rays from it, as we learned in the last chapter. You must also ask: is there a knight an L-shaped jump away?

Here is an example from the other side: this time White overlooked a knight. White wants to attack the pawn on b7 with the queen and chooses the square d5. From d5 the queen really does attack b7 along the diagonal. But who else is looking at d5?

The pawn on b7 could also have been attacked from a safe square, for example from b3, along the b-file. Before moving the queen, check not only the rays but also the jumps of the enemy knights.

A check from a knight is especially unpleasant. You remember from the first chapter that there are three ways to get out of check: move the king, block, or capture the checking piece. To block, you put a piece on the line between the attacker and the king. But the knight has no line: it attacks by a jump.

A knight check that also attacks another piece is the knight's favourite weapon. The king must move, and on the next move the knight takes what it attacked. A blow at two targets at once is called a fork, and chapter 11 is about it. For now, find a fork yourself.

King against knight

Now let us bring our two heroes together. You already know that a knight in the corner is poor: it has only two squares. The king uses this. See how a fork like the one in the problem above turned into a trap for the knight itself.

Notice that the king boldly stepped up to the knight. A knight does not attack the neighbouring squares, so a king beside a knight is safe. A knight can attack a king only from an L-shaped jump away; right up close it is harmless. Now trap a knight yourself.

Now the reverse story: the knight against the king. A king usually hides behind its own pieces, and that protects it. But against a knight such protection turns into a disaster. If all the squares next to the king are occupied by its own pieces, and a knight gives check, the king has nowhere to go and the check cannot be blocked. This mate is called smothered: the king is shut in by its own men. Other mating patterns are collected in chapter 18.

The last problem of the chapter takes two moves. Here the king has one free square left, g8. Think how to make Black occupy it themselves.

Both themes of the chapter worked together in this problem. The knight attacked squares you cannot see along any line: g8 and f7. And the black king, which has eight neighbours only in the centre, had three in the corner and lost all three.

What's next

The king and the knight do not go far, and the two of them cannot give mate. If the opponent has only a king left and you have a king and a knight, mate is impossible: there is not enough material, and the game ends at once in a draw. Such draws are covered in chapter 6.

You now know five pieces out of six. Pawns have already appeared on our diagrams, but only as targets and walls. The pawn is the slowest piece: it goes only forward, one square at a time (from its starting square it may go two), and it captures differently from how it moves. But when it reaches the last rank, a pawn can turn into a queen, and then the counting of moves that we learned from the king decides the game: will the king catch the pawn? How the pawn moves and what it can do is in the chapter "The pawn".