Rook and Pawn Versus Rook — The Endgame That Decides Most Games

There is one statistic that repeats in any game database: rook endings are the most frequent endgames in chess, by a distance. They show up in roughly half of all endgames played, for a structural reason: rooks are the last pieces to come into play and the last to be traded.

And there is a second statistic, less known and more useful: they are also where the most results get overturned. Won games are lost and lost games are saved at a rate that resembles no other type of endgame.

The good news is that almost everything is solved by two positions.

Why they are so resilient

The club saying is "all rook endings are drawn". It is an exaggeration with a real basis, and understanding that basis changes how you play them.

The rook is an exceptionally good defensive piece, for three reasons:

It checks from a distance indefinitely. A king trying to advance gets checked from the other side of the board, and every check costs it a move. That happens with no other piece.

It attacks the pawn from behind. A rook behind the enemy pawn stops it at no cost: the pawn cannot advance without being lost.

It can cut. A rook on a file or rank cuts the enemy king off completely, and that function — cutting — decides most of these endings.

From which comes the first practical rule, and the most important of all:

The rook goes behind the passed pawn. Yours, to push it; theirs, to stop it.

It is an old rule — usually attributed to Tarrasch — and it is right in the overwhelming majority of positions.

The Philidor position: how to defend

It is the most important defensive position in chess and it takes twenty minutes to learn.

The situation: your opponent has a rook and one pawn; you have a rook. Your king is in front of the pawn, on the promotion square or beside it.

The technique, in two stages:

Stage 1: the rook on the third rank. Put your rook on the third rank counting from your side of the board — the sixth, seen from the attacker's side. From there, the rook stops the attacking king from crossing that rank, because crossing would mean moving into check.

And there you stay, sliding the rook along that rank and doing nothing else. The attacker cannot make progress: their king cannot come forward and their pawn alone, without the king's support, is worth nothing.

Stage 2: when the pawn advances to that rank. At some point the attacker pushes the pawn onto that same rank, to dislodge your rook. That is the key moment: your rook drops to the back rank — the furthest one — and starts checking from behind.

Now the attacking king has nowhere to hide from the checks: its own pawn, having advanced, no longer serves as shelter. It is a draw.

Why it works: the advanced pawn took away the shield the attacking king needed, and from a distance the rook can check forever.

The mistake that ruins Philidor: starting to check too early, from the back rank, while the pawn is still behind. There the attacking king shelters in front of its own pawn, the checks run out, and the defence collapses.

The Lucena position: how to win

The mirror image: what to do when you are the one with the pawn.

The situation: your pawn has reached the seventh rank, your king is in front of it — on the promotion square — and cannot get out because the enemy rook is waiting. Your rook is free.

It looks drawn and it is winning. The technique is called building a bridge:

Step 1. Put your rook on the fourth rank — counting from your side — on a file away from the pawn.

Step 2. Step your king out of the promotion square. Your opponent starts checking from the side.

Step 3. Walk your king towards those checks, closing in on the enemy rook.

Step 4. When the check comes from the side and your king has nowhere left to go, interpose your rook: that is why it was on the fourth rank. That block is the "bridge". The checks end, the king comes out, and the pawn promotes.

It is five or six moves and they have to be known: without the bridge, the side checks never end and the position is a draw.

The five practical rules

Beyond those two positions, five principles settle most decisions in rook endings:

1. Rook behind the pawn. Already stated and the most important.

2. Activity before material. In rook endings, an active rook is worth more than a pawn. Giving up a pawn to activate your rook is often a good deal; sitting passively defending a pawn usually ends badly. It is the hardest rule to accept because it contradicts instinct.

3. Cut the enemy king off. A rook on the file separating the enemy king from the pawn is worth an enormous amount. The further away the king is cut off, the better: each file of distance is worth roughly a pawn.

4. Defending king in front of the pawn, draw; behind it, loss. It is the rule that decides whether to enter a rook ending or avoid it, and it has to be evaluated before the pieces come off.

5. Side checks need distance. To check from the side, your rook needs to be three files or more from the attacking king. Any closer and the king walks over and attacks it.

How to study them in a week

• Days 1 and 2: Philidor. Set the position up on a board and play it against yourself from both sides, until the defence comes out without thinking.
• Days 3 and 4: Lucena. Same. Pay attention to putting the rook on the fourth rank: that is the step people forget.
• Day 5: the five rules, with examples.
• Days 6 and 7: find the rook endings in your own games and review what happened.

Five or six hours in total.

The third position worth knowing: Vancura

Philidor and Lucena cover the vast majority, but there is one case that escapes them and comes up more often than it seems: the rook's pawn, with the defending rook trapped in front of it.

The situation is this: your opponent has a fairly advanced pawn on the rook's file, your rook is boxed in on the promotion square holding it, and your king is far away. Philidor does not apply — your king did not get back in time — and the position looks lost.

The Vancura defence saves it. The idea is the reverse of everything else: instead of putting the rook behind the pawn, you put it to the side, attacking the pawn from a distant rank. From there it does two things at once: it attacks the pawn, which then needs defending, and it keeps the option of side checks.

The attacking king then cannot advance without abandoning the pawn, and their rook is tied to defending it. It is a draw.

It takes twenty minutes to learn and it turns up precisely in the ending everybody assumes is automatically lost, which is where a known defence is worth most.

The commonest conversion errors

From the side with the pawn, three concrete errors that throw wins away:

Pushing the pawn too early. It is the mirror of the error that ruins Philidor. A pawn that reaches the sixth rank before the king is ready stops serving as a shield, and the defence appears by itself.

Letting the defending king get in front. The rule is single and decides everything: with the defending king in front of the pawn it is a draw, behind it is lost. So the manoeuvre to make before any other is cutting the enemy king off with your rook, on the file or the rank that separates it from the pawn.

Trading rooks at the wrong moment. A rook ending that looks complicated can turn into a lost king and pawn ending with a single exchange. Before trading the rook, the question is always the same: is the resulting pawn ending winning? If you do not know, do not trade.

And one rule that organises all three: the advantage is built before pushing. First the cut, then the king, and the pawn last. The other way round, most of these positions are drawn.

How they show up in practice

It is worth saying where you will meet this, because it is not in a study position with five pieces:

It shows up as a rook ending with several pawns that gradually simplifies. Nobody arrives at Philidor directly; you get there after twenty moves of pawns coming off, and most people do not recognise the position when it finally appears because they studied it with a clean board.

So it is worth studying them with extra pawns too. A rook ending with three pawns against three on the same side is almost always drawn, and it is the position that actually gets played. Knowing that decides whether to enter it, twenty moves earlier.

And that is why the clock matters so much. These endings are played at the end of the game, when little time is left, and they are precisely the ones that demand precision. An ending you know how to play and play with two minutes is an ending you can still lose.

From which comes the practical recommendation that closes everything: when the game is heading for a rook ending, save time. Five minutes saved in the middlegame are worth, in that phase, more than a pawn.

Why this pays so much

Because of a very concrete asymmetry: most of your opponents know neither Philidor nor Lucena. They are positions known for four hundred years that almost nobody below 1800 has studied, because endgames are not entertaining.

That means that in the endgame that comes up most often in chess, you will be the side that knows what to do while the other improvises. Nowhere else in the game does so little study buy so much.

And there is one more thing, which shows up over time: knowing these endings changes your middlegame. Once you know that a rook ending with the king in front is drawn, you start handling exchanges differently twenty moves earlier. That is the part that does not appear on the syllabus and is worth the most.