A program can buy a season. Only the coach carries the record.
Wins are confounded by conference, prestige, money, and inherited roster, so the engine never grades raw wins. It grades what beats the baseline (develops players, beats its own talent, wins in March, brings talent in, lasts), then runs the one test a program cannot fake, portability, and does it for the whole staff, not just the man in the chair.
Raw wins belong to the money and the roster. The engine grades the coach.
A win is confounded: conference strength, budget, prestige, the talent already in the building. The engine strips all of it and grades only what a coach adds on top of the baseline, across five pillars, then adjusts for integrity. This is a composite head coach.
Two coaches with the same record can be twenty points apart here, because one inherited a loaded roster in a soft league and the other developed his own in a gauntlet. The rating answers the only question that matters when you hire: what does this coach add that the program did not already have.
Illustrative engine read on the real Coach Intelligence pillar structure (development, overperformance, March, acquisition, staying power, integrity-adjusted). Composite coach, demonstration figures.
The one test a program cannot fake: did the impact travel?
A coach is a portable actor. The engine follows him across every stop, and asks the natural-experiment question: when he changed schools, did the pillars travel with him, or stay behind with the building? Same composite coach, three stops.
Wins can be inherited; a signal that survives three changes of address cannot. Portability is the closest thing in sports to a controlled experiment, and it is the heart of the coach rating.
Illustrative engine read on the real portability structure (the coach followed across stops, pillars compared). Composite coach and programs, demonstration figures.
Not just the head coach. The whole staff, and what the rating drives.
The engine follows the assistants too, each a portable actor with a specialty that travels. It rates the staff as a unit and by chair, then feeds the decisions that hang on it: who develops a given prospect, and who to hire.
A head-coach name on a marquee is not a staff. The engine reads the whole operation, values each chair by what it carries, and hands that straight to the two decisions it drives: how a prospect will grow here, and whether this is the group to bet a program on.
Illustrative engine read on the real staff-breakout and feed structure (assistants as portable actors, DEV conditioning development projections, hire pricing). Composite staff, demonstration figures.
Prestige, money, and an inherited roster can lift a win total without a coach doing much, which is why raw wins are the worst way to judge one. The engine grades what a coach adds over the baseline, tests whether it survives a change of address, and reads the whole staff behind him. A season can be bought; a signal that travels from a mid-major to a blue-blood and back cannot. That is the record, and it belongs to the people, not the building.
Grade the people, not the building.
Coaches reads the staff as portable actors and feeds the decisions that hang on them.