Grading Economics
On July 28, 2026, a Baltimore collector named Nicholas Funk filed a proposed class action against PSA and its parent, Collectors Universe, in Maryland federal court. The complaint's central claim is that PSA sells grading as an objective, standardized service while actually applying a subjective standard. It is one of three active cases, and Collectors has since responded publicly, though no court response has been filed in the Funk case. None of the allegations have been tested in court.
Set the legal outcome aside, because you cannot act on it. The useful question for a flipper is narrower: if grading really is subjective, does that break the math you use to decide whether to submit a card? The answer is no, and the reason is worth understanding, because it also tells you which of your assumptions actually is fragile.
According to the filing and reporting on it, the complaint alleges:
Two details worth knowing before you assume this is about you. The proposed class covers people who beneficially owned cards submitted through an intermediary (a bulk or group submitter) in the four years before filing, and direct PSA account holders are excluded. And the named plaintiff's own claim is small, about $300 in fees across seven cards. This is a structural case, not a big-ticket personal grievance.
Everything above is an allegation, a motion to dismiss is widely expected, and this post is not a prediction about the case.
On August 3, Collectors gave a statement to The Athletic. The line that matters for your decisions:
A grade is an expert opinion. Its value to the hobby rests on the fact that specific grades cannot be bought, pressured, or litigated into existence.
Collectors also said PSA would "never compromise our grading independence," and characterized a second suit as a dispute over grades a collector wanted and PSA declined to assign.
Read that carefully, because it is the company's own framing and it settles the practical question. PSA's defense is that a grade is an opinion. That is not a concession wrung out of them, it is their position, and it happens to be the same premise this post started from. If the grader says a grade is an expert judgment rather than a measurement, then any process that treated a grade as a predictable output was always mispriced, lawsuit or no lawsuit.
There is a concrete consequence. If a grade is an opinion by design, then a 9 you believe should have been a 10 is not an error to be corrected, and resubmitting on the theory that you will eventually receive the "right" grade is a bet on variance, not a claim on justice. Price resubmissions and cross-grade attempts as fresh draws with fresh fees, because that is what they are.
Three, as of early August 2026, and they allege different things:
| Case | Filed | Core allegation |
|---|---|---|
| Funk v. Collectors Universe (D. Md.) | July 28, 2026 | Grading marketed as objective while applying subjective criteria; production quotas; profiting from inconsistency via repeat submissions |
| Lichtman v. Collectors Universe (C.D. Cal.) | July 21, 2026 | Undergrading; refusal to cross competitor-graded cards; "population control"; conflicts involving the CEO's personal collection |
| Rasmussen v. Collectors Holdings (C.D. Cal.) | April 2026, amendment proposed August 3 | Antitrust, over the SGC and Beckett acquisitions, reduced SGC capacity, PSA price increases and service suspensions |
The Lichtman case is the one to watch on the pop-report question, since "population control" would speak to whether published population data can be taken at face value. That is a live input to grading decisions, unlike the fee-refund theory in Funk. It remains an allegation.
Here is the part most coverage will miss. A decision model that assumed grading was deterministic would not need expected value at all. If you could know the grade in advance, you would simply look it up, compute the profit, and be done. The only reason expected value exists in this hobby is that the grade is uncertain when you make the decision.
So the model already encodes exactly what the complaint alleges. Concretely, in the engine behind The Report Card:
| What the complaint alleges | What an EV model already does |
|---|---|
| The same card can grade differently | Assigns a probability to each grade, never a single outcome |
| No fixed, measurable criteria | Defaults to an 8% chance of a 10 when you have no data on the card |
| Grading is inconsistent | Applies a haircut to top-grade odds rather than taking them at face value |
| Reliability is overstated | Refuses to call any raw-to-graded outcome high-confidence, because this card's grade is unknown until it is graded |
That last row is the one to sit with. The engine has a hard rule: no matter how good your comps and population data are, a raw card's grade is never treated as a high-certainty input. Not because of a lawsuit, but because it never was knowable. If your process is "the pop report says 43% of these gem, so this one probably gems," subjectivity is a real threat to your plan. If your process is "assign odds, weigh the outcomes, require a cushion before buying," you already built the variance in.
Two things deserve to move, and neither is the arithmetic.
We have made this point in the break-even guide and the adverse-selection framing: a published gem rate reflects the cards people chose to submit, not a random raw copy, so it is a ceiling on your odds rather than an estimate of them. If the standard also drifts by grader or by day, that is one more reason to treat a category-level gem rate as a loose prior and discount it for your specific copy's centering and surface.
The practical version: when you plug a gem rate into any grading decision, ask what would happen if the true number were several points worse. If a few points of gem rate flips your call from grade to pass, the call was never strong enough to act on.
The vertical-integration allegation is the genuinely new consideration, and it is not a math input. If one corporate family assigns the grade, publishes prices, holds the card in a vault, runs resale, and lends against graded cards, then the entity determining your card's grade also has interests in the venues where you discover its value. You do not need to believe anything improper happened to notice that arrangement deserves a discount on unquestioning trust.
What you can do about it is boring and effective: verify prices against sold comps you pull yourself rather than any single guide, and treat grader choice as a decision with resale consequences, which we break down in the grader comparison.
The uncomfortable, honest version: this complaint is describing something experienced flippers already price in. A grade is a probabilistic outcome produced by human judgment, and the only protection was ever a model that expects to be wrong a predictable share of the time. That is why we publish our calls before outcomes exist, wins and losses both. A guide, not gospel.
The Report Card turns an ask price, your comps, and the pop report into expected value, a max-buy price, and an honest confidence range. Every call we make with it is logged publicly, timestamped, before the outcome exists.
Join the waitlist See the public track record →