Rate the roster and you rate the money. Rate the general manager and you have to subtract it first.
The college general manager is the decision-maker who runs the roster operation, the portal and recruiting acquisitions, the retention calls, the budget discipline, and the roster architecture. The engine points inward and rates him, not the roster he inherited, on five quantified pillars. But a general manager builds inside an institution that sets his sandbox, so the engine separates the two and relates them through the owner-adjustment, crediting him for what he built with what he had rather than for the money he was handed. The rating carries a career-versus-current split and a confidence band gated on the sample of his decisions.
Five pillars, re-scoped for the portal-and-collective game.
A composite college general manager. The engine points inward and rates the builder, not the roster he inherited, on five quantified pillars re-scoped for the portal-and-collective game.
The identity triggers classify the kind of operation, a description of method and not a grade, because two general managers can reach the same rating through opposite identities. This is the evaluating-the-evaluators principle. Rate the builder on what he brings in, holds, funds, structures, and navigates.
Illustrative on the real Front Office pillar framework (the five pillars re-scoped, the roll-up to a rating on the KR scale against a legend, the identity triggers). Composite general manager, demonstration figures.
The institution sets the sandbox. Rate what he built in it.
A general manager does not build in a vacuum. The institution sets the budget, the stability, and the infrastructure, so a rating of the builder is meaningless until the sandbox is subtracted.
This is the same separation the head-coach read makes, a decision-maker is not the resources he commands, he is what he does with them. A general manager who builds a strong roster on a thin budget rates above one who buys a stronger roster on a bottomless one. Subtract the sandbox, and rate what is left.
Illustrative on the real front-office-versus-institution separation (the institution as the sandbox-setting owner-equivalent rated separately, the owner-adjustment, the residual after the resources are subtracted). Composite general manager and institution, demonstration figures.
A long track record is a tighter bet than two good cycles.
Because this is a rating of a person over time, it carries two separate objects and bands the whole thing by the evidence behind it.
Report the arc, report the regime, and band the whole thing by the sample of decisions behind it, so a decade-long track record is a tight bet and a first-year general manager is a wide one no matter how good his one class looked. A hot start is never sold as a career.
Illustrative on the real rating layer (the career-versus-current split, the confidence gated on the sample of decisions, the rating as a bet). Composite general managers, demonstration figures.
The engine that rates players and teams turns and rates the person who builds them, on the same scale and the same honesty, because the evaluator is as ratable as the evaluated. A general manager is scored on five pillars, what he acquires through the portal and recruiting, what he retains against the portal, how he builds against the revenue-share and NIL budget, how he structures the roster and its eligibility, and how he navigates the athletic department above him, and those roll up to a rating on the KR currency. But a rating of a builder is meaningless without subtracting what he was given to build with, so the engine rates the institution separately as the sandbox-setting owner-equivalent and relates the two through an adjustment that caps the strong manager under a weak program and amplifies him under a strong one. What is left after the subtraction is the residual, his own contribution, and it is the only honest thing to rate: a manager who builds a strong roster on a thin budget is worth more than one who buys a stronger one on a bottomless budget. And because it is a rating of a person over time, it carries the long arc and the current regime as two separate objects, and it bands the whole thing by the sample of decisions behind it, so a hot start is never sold as a career. Rate the builder, not the building, subtract the sandbox, and price what is left as the bet it is.
Rate the builder, subtract the sandbox. Band what is left.
College General Manager rates the decision-maker on five pillars, separates him from the institution that sets his sandbox through the owner-adjustment, and reports the residual as a bet with a career-versus-current split and a sample-gated band.