Domination

Domination is the endgame pattern in which a piece is not attacked, not pinned, not forked — and lost anyway, because every square it can move to is covered. No tactic ever lands; the trapped piece simply waits while the enemy king walks over to collect it. Recognizing these cages, and steering enemy pieces into them, converts material advantages that "should" be unwinnable and wins material in endings that look dead level.

What Domination Means

The word comes from the study composers — the Armenian grandmaster Genrikh Kasparyan built a famous collection around the theme — but the pattern is intensely practical. A dominated piece differs from a merely trapped piece in one respect: there is no immediate threat at all. The piece has moves; all of them lose it. Its owner is reduced to watching the position while the opponent improves at leisure — domination is zugzwang's quieter cousin, aimed at one piece instead of the whole army.

Two ingredients appear in virtually every example, and they are worth memorizing as a checklist:

• The rim. A piece on the edge has already lost half its squares to the geometry of the board. Knights suffer most — a corner knight has two moves total — but bishops and even rooks can be caged against the edge.
• Separation from the king. A piece standing next to its king is almost never dominated, because the king defends its escape squares and shelters it. Domination happens to stragglers — pieces that grabbed a distant pawn, blockaded a passer, or got shouldered away during a race.

When both ingredients are present, start looking for the cage. The two boards below are the patterns I teach first, because between them they cover the geometry of most practical cases — and every position on this page is tablebase-verified.

One Rook, One Frozen Knight

Rook versus knight with no pawns is normally a routine draw — but not when the knight is cornered far from its king. White: king c3, rook g2. Black: king a7, knight h1. The knight in the corner has exactly two exit squares, f2 and g3 — and the rook on g2 covers both at once, f2 along the second rank and g3 down the g-file. One piece, one square, a complete cage:

Now watch how little White needs to do. 1.Kd2 Ka6 2.Ke3 — the king simply walks toward the corner, adding a second guard to f2. Black's king sprints back — 2...Kb5 — but the arithmetic was settled on move one: the knight cannot move, so Black's king must cross the entire board alone, and it will not arrive. 3.Rg1 keeps the cage while White regroups, and the knight can do nothing but choose its moment: 3...Nf2 4.Kxf2, and with Black's king still on b5, the position is the basic rook checkmate. The knight was lost the moment it touched h1 with its king cut away — everything after that was bookkeeping.

The pattern to store: a rook covers a corner knight's both exits from the adjacent knight-file square — g2 against h1, b2 against a1, g7 against h8, b7 against a8. When a knight grabs a rook-pawn in a race and lands in the corner, this cage is usually waiting.

The Crossing Diagonals: Bishop Dominates Knight

A bishop covers only one color — so it seems absurd that it could cage a knight, whose every move changes square color. The rim makes it possible. White: king c3, bishop d5, pawn a3. Black: king f8, knight a5. The knight on the rim has four exits — b7, c6, c4, b3 — and all four are light squares. Better still for White, they lie in pairs on two diagonals, a8-h1 and a2-g8, and those diagonals cross on d5. From that one square the bishop covers the knight's entire world:

1.Kb4 attacks the frozen knight, and the game is effectively over. The desperado 1...Nc6+ jumps into a covered square with check and changes nothing after 2.Bxc6; 1...Nb3 meets 2.Kxb3 (the bishop covers b3 too, through c4). Black's king is a spectator: 1...Ke7 2.Kxa5, and the extra piece plus the a-pawn win at leisure — note that the pawn queens on a8, a light square, so even the wrong-bishop drawing resource is absent.

The geometry generalizes beautifully: a knight on the rim always exits onto squares of one color, and those squares always cluster on a pair of crossing diagonals. Find the crossing square, and one bishop does the work of a whole army. Knights on the rim are vulnerable to their own kind too — the proverb about the knight on the rim being dim is really a warning about this entire family of cages, one reason knight endgames punish careless pawn-grabbing expeditions so brutally.

The Antidote: Keep Pieces Near the King

Domination has a simple vaccine, and seeing it clarifies the whole concept. Same material as the first example — rook against knight — but now the knight stands beside its king. White: king e1, rook h8. Black: king d4, knight d5. The tablebase's verdict: dead draw.

The rook can check, probe, and circle forever; the king defends the knight's squares, the knight shelters behind the king, and no cage exists. This is the general truth about the rook vs knight endgame — drawn while the knight stays in touch with its king, lost when it is cornered or cut off — and the practical rule for defenders follows immediately: when simplification approaches, walk your short-range pieces home. A knight is a team player; the moment it operates solo across the board, every one of its stops is a potential h1.

For the attacking side, the same sentence reversed becomes a plan: separate, then drive to the rim, then cage. Cut the defender's piece from its king — a rook on an intervening file does this best — push it edgeward with king and threats, and only then look for the covering square. This is target-picture thinking, not calculation: you are steering toward a known cage the way schematic thinking steers toward any winning setup. In races, remember that a piece that "wins" a distant pawn at the cost of its coordination has often lost the game — check the cage patterns before you cash the pawn, whether the grabbing piece is the opponent's or your own. The concept even has its own entry in the tactical vocabulary — domination — but in the endgame it is less a tactic than a slow fate: a piece walks to the rim, its king cannot follow, and the board itself closes around it.

Frequently Asked Questions

What is domination in chess?

Domination is a situation where a piece is lost without being attacked: every square it could move to is controlled by the opponent, so it must wait to be captured. It differs from an ordinary trap in that no immediate threat exists — the piece's own mobility is the illusion.

How can a rook trap a knight all by itself?

Against a corner knight the geometry is exact: a knight on h1 has only f2 and g3, and a rook on g2 covers both — one along the rank, one down the file. The same cage works in all four corners, and the attacking king then walks over to win the frozen knight.

Can a bishop really dominate a knight?

Yes, on the rim. A knight on the edge exits only onto squares of one color, and those squares lie on two diagonals that cross at a single point. A bishop posted on the crossing square — d5 against a knight on a5 — covers every exit at once.

How do I avoid getting a piece dominated?

Keep short-range pieces, especially knights, within reach of your king as material comes off. Most dominated pieces are stragglers: they grabbed a far pawn or blockaded alone on the rim. Before sending a knight on a solo errand, check the return route, not just the destination.

Is rook versus knight a win or a draw?

With no pawns it is a draw whenever the knight stays near its king — the defense is easy. The losing cases are exactly the domination cases: knight cut off from its king, or caught on the rim or in a corner, where the rook can cover its few exits and the attacking king collects it.