LSU's roster costs twice the cap. Both numbers are real.

LSU's football roster reportedly cost $40–50 million for the 2026 season. The revenue-share cap created by the House settlement is roughly $20.5 million per school, across every sport combined.

Both figures are accurate. Understanding why requires reading what the settlement actually caps, which is narrower than the coverage suggests.

The first thing to correct is the number itself. The settlement does not state a dollar cap anywhere. Article 3, § 1(e) of the injunctive relief agreement sets the annual Benefits Pool at 22% of Average Shared Revenue, and fixes it there for the term. The $20.5 million everyone quotes is that formula's reported output in its first year — a figure that appears in coverage, not in the document.

That distinction matters in two directions. The pool tracks the underlying revenue rather than a negotiated number, so it moves without anyone reopening the agreement. And a school's compliance question is not whether it stayed under a dollar figure but whether it stayed under a percentage of a revenue base that is itself calculated.

What the pool binds. Article 3 § 1(a) frames the pool as institutional payments and benefits to student-athletes. It is a ceiling on what the school itself provides, applied across every sport it fields rather than siloed by program.

What sits outside it, precisely. Article 3 § 3(c) provides that genuine third-party payments are not counted against the pool. This is the structural point, and it is worth stating carefully because the mechanism is routinely described wrong — including in an earlier version of this piece.

Three distinct things are often collapsed into one:

- A reporting threshold. Article 2 § 4 requires third-party NIL agreements above $600 to be reported. That is a disclosure obligation, not a valuation test. Reporting a deal is not submitting it for approval. - A valid-business-purpose and compensation standard, requiring compensation commensurate with that paid to similarly situated individuals who are not athletes. This applies to Associated Entities or Individuals — the category capturing collectives and school-affiliated boosters — rather than to all third-party NIL above some dollar figure. - Ordinary third-party endorsements, which are outside the pool and outside the associated-entity standard entirely.

So the structure is narrower and sharper than the shorthand suggests. It is not that all outside money above a threshold gets valued. It is that money from entities associated with the school is held to a market standard, and everything else is genuinely uncapped.

Where the pressure lands. That places enormous weight on the definition of an Associated Entity, because the entire enforcement question is whether a given collective falls inside it. A market standard applied only to associated entities is a standard whose reach is determined by a classification rather than by a dollar amount.

The counterargument worth stating. The framework is coherent on its own terms: cap what the school pays, require disclosure of what others pay, and hold school-affiliated money to a market test. The objection is not that the design is incoherent but that the associated-entity boundary is new, and comparator sets for athlete endorsement value largely do not exist yet. That is an enforcement problem rather than a drafting one, and enforcement problems get solved by the first contested determinations.

None of the reported roster figures come from an audited disclosure. College football has no centralized payroll reporting, no equivalent of a league-published cap sheet, and no requirement that a collective open its books.

The circulating numbers come from a broad but unattributed sample. CBS Sports reported its 2026 roster spending rankings after polling more than fifty sources across Power Four athletic directors, general managers, personnel executives, salary-cap analysts, agents, and boosters. That is a wider base than the "rival general managers" framing this piece previously used, and the correction cuts both ways — the estimates are better sourced than that characterisation suggested, and still not audited.

Brian Kelly has said LSU spent roughly $18 million in 2025. The reported 2026 figure is roughly double that. Neither can be checked against a filing.

That is not a footnote. The number shaping a recruit's expectations, a rival program's budget, and a booster's sense of what is required is a number nobody has to stand behind. In a market, a price that cannot be verified tends to rise — each participant bids against an estimate rather than a disclosure.

The programs at the top of the second track, and the gap compounds.

A cap that binds institutional payments and leaves third-party money open does not restrain spending. It relocates it — from a channel every school can access equally to one where access depends entirely on the depth of the local collective. The pool is the floor every school must reach to stay competitive. The collective is a ceiling only some can build toward.

The absence of reporting sharpens it further. A program that cannot match the spending still has to recruit against the perception of it, and perception in an unreported market is set by whoever talks to reporters. There is no disclosure to correct an inflated estimate and no incentive for the school benefiting from it to try.

The athletes gain, materially and unambiguously. But the distribution among them tracks collective capacity rather than performance, which is a different allocation than either the old model or the settlement's stated design.

The pool replaced one compensation ceiling with another, and the last one did not survive contact with antitrust law.

NCAA v. Alston was unanimous, delivered by Justice Gorsuch. Its holding was narrow — education-related benefits. Justice Kavanaugh's concurrence went further, and it is worth quoting rather than paraphrasing, because it is routinely overstated: he wrote that there are serious questions whether the NCAA's remaining compensation rules can pass muster under ordinary rule-of-reason scrutiny, that it is highly questionable whether the NCAA can justify not paying athletes, and that if that justification is unavailing it is not clear how the remaining rules can be legally defended.

That is an invitation, not a holding, and the difference matters. An earlier version of this piece said the concurrence stated the remaining rules would not survive. It said no such thing — it said the questions are serious and the answers unclear. A concurrence flagging doubt is a roadmap for a future plaintiff; it is not a ruling anyone can cite as controlling.

A court-approved settlement is a stronger position than a rule an association wrote for itself. It has judicial approval, a defined class, and released claims. But approved as a resolution between these parties and immune from challenge by someone who was not a party are different propositions.

Two things to watch. First, whether an associated-entity determination goes against a major collective — that is the moment the outside track acquires a boundary, and the moment someone has standing to test where it sits. Second, whether Congress acts. A federal standard is the only mechanism that would end state-by-state variation and give the pool a statutory basis rather than a contractual one.