Grade the fingerprint that travels, not the record the roster handed him.
The manager and the coordinators are graded as named actors on what beats the baseline, never on raw wins. The engine separates the person from the situation: it rates the manager on his pillars and holds his record against the roster baseline, splits every coordinator's credit transparently between the coach, the player, and the org's development machine, and runs the portability test as the honest check, because a real fingerprint shows up in the next building and a borrowed one does not. In baseball the pitching-development read moves more value than any other staff read, so it is the centerpiece.
Rate the pillars. Measure the margin over the baseline.
The manager is rated across his pillars, including an in-game and bullpen-management pillar and a development pillar, and assigned an archetype. His career record is tracked while the regime and roster context are separated, because a manager who won with a loaded roster and one who beat the run differential with a thin one are very different reads.
The manager grades an 82 as a bullpen-leverage tactician, strongest on the in-game pillar, and the career read is the margin over the roster baseline, not the win column. The 84-win manager who beat a thin baseline by 7 is a better read than the 96-win manager who beat a loaded one by 2. Grade the margin over the baseline, not the wins the roster would have produced anyway.
Illustrative engine read on the real manager rating (the pillars including in-game and bullpen management and development, the archetype, the record measured against the roster baseline with regime and roster context separated). Composite managers, demonstration figures.
Split the credit. Then test whether it travels.
The pitching coach is rated with independent career tracking and a transparent credit split between the coach, the player, and the org's development machine, then run through the portability test: did the staffs he moved between carry the gain with him. This is the high-leverage case, because in baseball the pitching-development read moves value more than any other staff read.
An 86-rated pitch-design developer whose staff gains split 45% to the coach, 35% to the player, and 20% to the lab, and whose gain traveled across three organizations. The portability is the proof: the fingerprint shows up in the next building, so the credit is his to carry. The pitching-development read is the highest-leverage staff read, and portability is the honest check on it.
Illustrative engine read on the real pitching-coach rating (independent career tracking, the transparent credit split between coach, player, and org lab, the portability test across organizations). Composite pitching coach, demonstration figures.
A gifted lineup is not a coaching record.
The hitting coach is rated on the gain his hitters made over their projected curves, separated from the quality of the bats he inherited. A coach handed three ascending prospects did not necessarily coach the ascent, and the credit split says so, then the portability test asks whether the fingerprint shows up across rosters.
The hitters beat their curves by a hair, net, so the coach carries just 22% of the credit and the ascent that was already projected carries most of it. A gifted lineup is not a coaching record, and the credit split and the faint portability say so. Inheriting three ascending prospects is not coaching the ascent, and the engine refuses to score it as one.
Illustrative engine read on the real hitting-coach rating (the gain over the projected curves separated from inherited talent, the credit split, the portability test). Composite hitting coach, demonstration figures.
A coach is graded on what beats the baseline and travels, not on the record his roster produced. The engine separates the person from the situation, splits the credit between the actor, the player, and the machine, and holds the portability test as the honest check: a real fingerprint shows up in the next building, and a borrowed one does not. The manager is measured by the margin over his roster baseline, the pitching coach by a gain that traveled across three organizations, and the hitting coach by how little his hitters beat the curves that were already pointing up.
Rate the actor, not the roster.
Coaches grades the manager and the coordinators on what beats the baseline, splits the credit transparently between the coach, the player, and the machine, and runs the portability test, with the pitching-development read as the highest-leverage one in the sport.