Expected value per card. Every card's graded expected value uses its real PSA population split by grade — not a model. EV = P(10)×PSA-10 price + P(9)×PSA-9 price + … down through PSA 5, where P(grade) is that grade's share of the card's actual PSA population. Outcomes below PSA 5 are valued at the card's raw (ungraded) price. Half grades (9.5, 8.5, …) are excluded and probabilities renormalized over whole grades.
Prices. Each grade is priced at its most recent recorded sale within the last 12 months (last sold, not an average), from aggregated recent sales of graded cards. When no sale falls in that window, we fall back to the most recent recorded sale on file. Every graded price is floored at the card's raw price — a slab never counts for less than the ungraded card. Raw prices are current market prices for ungraded near-mint singles.
Grading decision. A pulled card is graded only when its expected graded sale minus the grading fee beats selling it raw: contract value = max(raw, graded EV − fee). The fee slider re-computes everything live; $80 reflects PSA's standard per-card rate. Cards whose expected value exceeds $1,500 incur higher declared-value fees, which this model does not add — treat top-end contracts as slightly optimistic.
Pull odds. Set-specific community pull-rate data — not era-wide templates — assigned by card number block. Source: set pull rates. These are community estimates, not manufacturer-published odds.
Pack EV. Pack EV = bulk value (commons, uncommons, non-holo rare and reverse slots at current market prices) + Σ over hits of pull odds × contract value. Return % divides by the pack's current market street price.
Caveats. Population grade rates reflect cards collectors chose to submit — pack-fresh grading rates likely run lower (survivor bias). Grades with no recent sales are flagged in each card's breakdown. Populations and prices are snapshots as of the date in the footer.
