Football has three phases. So it has three coordinators, and the engine rates all three.
The offensive, defensive, and special-teams coordinators each own a phase, and each is rated on the pillars a head coach is: his scheme identity read from the data, his play-calling as an in-game actor, his unit's performance against the talent it was handed, and his development of his position groups. Because special teams is a co-equal third phase, the special-teams coordinator is a real rated actor, not an afterthought. And because a phase is a joint output of the coordinator, the head coach who may call the plays, and the roster, the engine splits the credit, and its cleanest read is what travels with a coordinator when he changes buildings.
One rated actor per phase, and special teams is not an afterthought.
A composite staff, three coordinators, one per phase. Each is read against the scheme identity the engine infers from the data, at the richest tier available, and rated on the same pillars a head coach is, scoped to his phase.
A coordinator is read against his inferred scheme, and the two engines apply: a college coordinator carries the added families and the heavier option and RPO volume. Three phases, three identities, three rated actors.
Illustrative on the real coordinator framework (the three-phase Scheme Inference identities, the per-phase pillar rating with archetype and regime-and-roster separation, special teams co-equal). Composite coordinators, demonstration figures.
His identity is read from the data, and his play-calling is the actor.
A coordinator is not a label. He is his scheme, inferred from the tendency signature: the offensive family and its personnel mix, the defensive shell and package, the special-teams approach. The rating reads him inside the identity he actually runs.
The play-caller tendencies are the same tendencies the scouting engine builds an opponent advance from, the coordinator read and the game-plan read are one system seen from two sides. The head coach owns the game-management, the coordinator owns the play-calling. A coordinator is judged on the scheme he chose and the calls he made, not a name.
Illustrative on the real scheme-and-situational pillar (the inferred identity, the play-calling tendencies conditioned on the matchup, the play-caller tells from charted history). Composite coordinator, demonstration figures.
A phase belongs to three parties. The engine splits it, and watches him move.
A phase belongs to three parties. The engine splits the credit in the open, then applies the one test a roster and a head coach cannot fake: it watches what happens to the unit when the coordinator changes buildings.
The portable-actor read is the cleanest isolation, because a unit that over-performs its talent wherever he goes and regresses the moment he leaves is a contribution no roster or head coach can explain away. The phase residual and the position-group development residual round it out. Credit the phase to the parties who produced it, and the coordinator to what moves with him.
Illustrative on the real credit and isolation mechanics (the transparent credit split, the play-calling-autonomy swing, the portable-actor across-stops read, the phase residual, the never-touches-a-current-KR rule). Composite coordinator and units, demonstration figures.
A coordinator owns a phase, but a phase's numbers belong to three parties: the coordinator who designs and calls it, the head coach who may be the real play-caller, and the roster that runs it. So the engine refuses to read a coordinator off his unit's raw output, and instead splits the credit in the open, asks who actually calls the plays rather than what the title says, and prices the unit against the talent it was handed. It rates all three coordinators, because football has three phases and special teams is co-equal, and it grades each on the scheme he chose and the calls he made, read from the data as an identity and not a label. And then it applies the one test a roster and a head coach cannot fake: it watches what happens when the coordinator changes buildings. A unit that over-performs its talent wherever he goes, and regresses the moment he leaves, is a contribution that has separated itself from every roster and every boss he has worked under. That is his, and the engine credits it to him, on the sample it has, at the confidence that sample earns, and never by moving a single player's own number. Follow the man across the stops, and the scheme stops being anonymous.
Rate all three phases. Follow the coordinator across the stops.
Coordinator Intelligence rates the offensive, defensive, and special-teams coordinators on their inferred schemes and their play-calling, splits the phase credit between the coordinator and the head coach and the roster, and isolates the coordinator by what travels with him, never touching a player's own number.