Rook and Bishop vs Rook: The Second-Rank Defense

Watch the method: press play to see the winning idea run, or step through it move by move.

Rook and bishop versus rook is the most common extra-piece ending in chess, and its reputation is schizophrenic: theoretically drawn from almost everywhere, practically lost by strong grandmasters every single year. The gap between those two facts is exactly one defensive setup.

The second-rank defense. The defending rook belongs on the rank directly in front of its king, the full rank, treated as a fence. From there it intercepts every check on the file, and the attacking king can never take the last step forward: each approach square is either covered or answered by a rook trade, and without rooks the bishop mates nobody. The attacker owns the rest of the board and it buys him nothing, because the fight is only ever about one rank.

What loses. A passive post is fine; an idle move is not. Wandering off the fence rank is fatal, and so is any neutral move: only Re7 and Rd7+ hold, and most of the alternatives allow Rf8 mate on the spot. Rook sorties for checks or counterplay give the attacker the tempo he has been begging for: the king slips to the sixth, the bishop seals a flight square, and the famous Philidor winning position appears by force. The corner is the other graveyard: driven to the edge files, the defense runs out of fence. In this drill even a plausible check loses on the spot, which is the lesson in one move.

The clock is a defender. Fifty moves is a short leash for an attacker who must reorganize repeatedly against a correct fence. Hold the rank, keep the king central, count in peace. Defenders who know this setup save a piece-down ending several times in a career; defenders who improvise usually get to watch the bishop close the box.