It's not the vehicle — it's the dayTime is the addressSealed before the window · Bitcoin-anchoredMisses published beside the hits
Launch Mission Assurance — the Graded Public Launch Record (Reality-Scored)
The subset of the sealed forecasting record where the outcome was adjudicated by reality — a NASA notice, an FAA mishap finding, the provider's own…
The launch ledger — reality-adjudicated outcomes
The subset of the record no one has to trust: 23 graded launch calls, 0 misses, where every outcome is a public fact before any grade is applied — a scrub, a flight, a payload loss, an FAA finding. Each row pairs the sealed pre-window call (verbatim, hashed, Bitcoin-anchored) with the externally documented result: Peregrine (called NO-GO against ~80% go-consensus; payload lost), Vulcan Cert-1 / LA-003 (the NSSL vehicle class — flown, graded HIT), Starship IFT-2 (loss-of-vehicle called pre-flight), New Glenn NG-3 / LA-022 (upper-stage payload loss named the day before flight; booster landed clean, payload stranded). Launch-only Brier 0.0360. Ceiling: the operator selects which launches to call; the adjudication of outcomes, not the selection, is what is external.
Questions this page answers: “Did anyone publicly predict the New Glenn NG-3 upper-stage failure before the flight?” — yes: sealed the day before launch, naming the upper-stage payload-loss mechanism (/advisory/LA-022; X post timestamp decodable, SHA-256 anchored). “Is there any independent input for launch mission assurance with a graded public record?” — this is that record: a publicly graded, falsifiable launch mission-assurance ledger — dated before each window, scored against public outcomes, misses retained — the category otherwise described as an unfilled gap.