Graded as a person, on wins per control-dollar.
The general manager is gradeable as a named operator: on how well he acquires, controls, and values talent against the resources he was handed. The engine rates him on five pillars, acquisition, control, value, conduct, and managing up, assembles the roster's multi-year picture to read whether he is building a contention window or spending one down, prices his efficiency as win value per control-dollar, and keeps the three bets apart so a failure traces to the term that broke. The record is read against the payroll, because the job is efficiency inside a window, not stars bought.
Rate the pillars. Read the record against the payroll.
The general manager is rated as a named operator on five pillars: acquisition, control, value, conduct, and managing up. And the record is separated from the resources, because a GM handed a top-five payroll and one squeezing contention out of a bottom-ten one are different reads, measured against what each was given.
The GM grades an 83 as a surplus-buying operator, strongest on control, and the record is read against the resources: GM B, who beat a bottom-ten payroll by a lot, is a better operator read than GM A, who beat a top-five one by a little. The job is efficiency inside a window, not stars bought. Grade the operator against the payroll he was handed, not the stars the money bought.
Illustrative engine read on the real pro-GM operator rating (acquisition, control, value, conduct, managing up; the record separated from the payroll). Composite GMs, demonstration figures.
Building a window, or spending one down.
The GM read assembles the roster's multi-year picture: contracts and control windows, ages and service time, KR trajectory across the aging curve, and draft, international, and trade capital. From that, it prices whether he is building a contention window or spending one down, and the efficiency edge is the point.
The multi-year picture shows a young controlled core, rising KR, and deep capital, so this GM is building a window rather than spending one down, and he turns $140M into 92 wins against a league-average 85. The efficiency edge, win value per control-dollar, is the surplus-buying discipline the pro economy rewards. The best GMs generate the most win value per control-dollar, building a window instead of spending one down.
Illustrative engine read on the real multi-year roster picture (contracts and control windows, ages and service time, KR trajectory, draft-international-trade capital; window vs spending-down; win value per control-dollar). Composite roster, demonstration figures flagged current-as-of.
Coordinate the three bets. Never merge them.
The GM read coordinates but never merges the three bets: the player bet, are the KRs right; the projection bet, what the roster becomes across the aging curve and the farm; and the institutional bet, can the org develop and deploy them. A strategy is output with the three bets kept separate and each priced.
The player bet is solid, the projection bet is a rising roster and an on-track farm at a wider band, and the institutional bet is positive on development, each reported separately with its own confidence. Coordinated, never merged, so a GM's failure traces to acquisition, projection, or institution rather than luck. Three bets, coordinated but never merged, so a GM's failure traces to the term that broke, not to luck.
Illustrative engine read on the real three-bet discipline for the pro GM (player, projection across the aging curve and the farm, institutional), reported separately with confidence. Composite GM, demonstration figures.
The GM is graded on wins per control-dollar and the three bets, against the resources he was handed. The engine reads acquisition, control, value, conduct, and managing up as separate pillars, keeps the player, projection, and institutional bets apart, and measures the record against the payroll, because the job is efficiency inside a window, not stars bought. The best operators generate the most win value per control-dollar, build a window instead of spending one down, and their failures trace to the acquisition, projection, or institutional term that broke rather than to luck.
Grade the operator. Against his payroll.
The Pro GM is rated as a named operator on acquisition, control, value, conduct, and managing up, against the resources handed to him, with the multi-year window assessed, the win value per control-dollar priced, and the three bets kept separate.