The move is the answer. Not the player's grade in isolation.
A roster move is priced by what it does to the team. Add, swap, or cut a college player, and the engine re-runs the additive Team KR and reads the resulting delta against the scholarship-and-package cost, all under the 34-man fully-funded cap and the eligibility clock. The delta is the marginal team-KR movement, so a high-KR bat at a position of strength can move the team less than a lesser glove that fills a hole, and the output is value added per dollar with the constraint check, not a raw grade. The roster layer aggregates and reconciles; it never re-evaluates a player the engine already read.
Re-run the team, read the marginal delta.
Add a recruit or portal target to the roster and re-run Team KR through the additive core: the delta is the answer, the marginal team-KR movement he produces given the current roster, park, and needs. Then read the delta against his cost, the scholarship-and-package number, so the output is value added per dollar, not a raw grade.
Adding the 81-KR second baseman moves Team KR from 86.4 to 88.5, a +2.1 delta, read against his $210K cost for value added per dollar. The move is the answer, and the same-cost 85-KR corner bat at a position of strength would move the team a third as much. The delta is the answer: marginal team-KR movement per dollar, not the player's grade in isolation.
Illustrative engine read on the real add-and-re-run (the marginal Team KR delta through the additive core, read against the scholarship-and-package cost as value per dollar). Composite player, demonstration figures flagged v0 and current-as-of.
Under the cap, every add is a swap.
Every move is checked against the amateur constraints: the 34-man fully-funded roster cap, so adding one means the headcount and the funding both bind; the eligibility clock, so the player has to have the seasons; and the three-phase coverage, so the roster still covers the defensive spectrum, the platoon, and the arms. To bring one in under the cap, someone comes off.
Under the 34-man cap the add is really a swap: the 81-KR glove in, the 58-KR redundant bat out, a +1.6 net move, with the cap, the clock, and the three-phase coverage all checked and holding. To fit one in, someone comes off, and the net is the answer. Under the cap, every add is a swap, and the net team-KR delta is what matters.
Illustrative engine read on the real move-under-constraints (a swap under the 34-man cap, the eligibility-clock and three-phase-coverage checks, the net team-KR delta). Composite players, demonstration figures flagged v0 and current-as-of.
The construction margin, where a roster is built well or badly.
Beyond the marginal add, the engine reads the construction margin: depth and injury resilience over the season, platoon and handedness balance, and the bullpen-and-lineup shape. It is the thin layer on top of the additive core where a college roster is actually built well or badly, and the output is a construction plan with the constraint check, not a single number.
The construction margin is where the roster is built well or badly over a full season: this one is deep at the top and thin behind it, left-handed heavy, and set at the back of the bullpen, and the output is a plan with the constraint check, not a lone Team KR. The additive core is the answer; the margin is how you build it well. The margin is the thin layer on top of the additive core where a roster is actually built well or badly.
Illustrative engine read on the real construction margin (depth and injury resilience, platoon and handedness, bullpen and lineup shape) as a construction plan with the constraint check. Composite roster, demonstration figures.
The move is the answer. The engine adds the player, re-runs the additive Team KR, and reads the delta against the cost under the 34-man cap and the clock, because a roster is not a pile of grades, it is what the team becomes when you make the change, priced against what the change costs. Under the cap every add is a swap, the net delta is what matters, the constraint check confirms the headcount and the clock and the coverage all hold, and the construction margin is the thin layer on top where the season is actually built well or badly. The roster layer aggregates and reconciles, and never re-evaluates a player.
Make the move. Read the delta.
College Roster prices a move by what it does to the team: add, swap, or cut, re-run the additive Team KR, read the delta against the scholarship-and-NIL cost under the 34-man cap and the clock, with the three-phase coverage check and the construction margin on top.