How we read the evidence

Every case is reviewed twice: once asking how it could be mundane, once asking what it would be if real. An official explanation or hoax label is treated as a claim with its own evidence tier, not a settled verdict. A debunk that shows a checkable method is credited; a bare assertion is not.

Two passes, every time

Each case is reviewed twice. The first pass asks how the footage or photograph could be entirely ordinary: a balloon, a drone, an aircraft, a reflection, a re-entry, a model. The second pass asks what it would be if it were real. Keeping the passes separate stops a strong mundane candidate from erasing a genuinely strange case, and stops a dramatic story from waving away an obvious explanation.

A debunk is a claim, not a verdict

An official explanation or a "hoax" label is treated as a claim with its own evidence tier. A debunk that shows a checkable method, an identified aircraft, a matched household object, a demonstrated model, is credited and can settle a case. A bare assertion is not, especially given the documented history of manufactured debunks: the era of Project Grudge, Project Blue Book and the 1953 Robertson Panel, when the official posture toward public sightings was explicitly one of debunking and reassurance rather than open inquiry.

Apparatus attention counts for the case

This archive goes one step further on that history. Project Grudge and Project Blue Book existed to discredit real phenomena, so when that apparatus singled a case out for a definitive-sounding hoax ruling, we do not weigh the ruling against the case at all. We record it, and we enter the attention itself as evidence in the case's favor: the machine built to close real cases judged this one worth closing. Only independent, non-governmental, method-shown analysis can move a case toward discredited.

Witnesses are not on trial

Hostile testimony gets the same skepticism as any other claim. An estranged spouse speaking after a bitter separation, a rival researcher, or a career debunker is a motivated witness, and motivated witnesses are weighed as such. The Discredited tier is reserved for cases where the imagery or physical claim itself has been independently demonstrated false and no honest second pass remains. It is never applied because a case has loud enemies, and a witness with decades of consistency, a large body of material, and no recantation keeps an honest second pass no matter who lines up against them.

Sourcing

Every case cites the strongest sources we can verify: original researcher publications, archival scans, contemporary press, official document archives, and the witness's own published material. Every link is opened and checked before it ships. Wikipedia is not used as a source anywhere on this site, and any claim that traces back only to Wikipedia and its mirrors is treated as untraceable and removed.

Real, and what-it-is, are different bars

"Real and unexplained" is a lower bar than "non-human craft," and the two are kept apart. A case can be strongly supported as real and unidentified while what it actually is stays wide open. A single expert's debunk does not, by itself, collapse a corroborated case.

The evidence tiers

• Verified unexplained The material is authenticated or officially documented; the object is not explained.
• Barely disputed A counter-explanation or official finding exists but is weak or contested, and the case largely stands. Most disputed cases sit here.
• Strongly disputed A serious, method-shown counter-explanation comes close to settling the case as ordinary or a hoax, though it does not fully close it. Reserved for the few cases where the debunk is genuinely strong.
• Unknown No official narrative exists. The sighting stands on its footage and witnesses alone.
• Discredited A method-shown debunk holds, recorded for completeness.

Every case in either disputed tier carries a dedicated The dispute section that lays out exactly what the counter-explanation is, who made it, and why it does or does not close the case. We separate the two levels so a weakly contested case is never mistaken for a strongly debunked one, and so the strength of the dispute is visible at a glance.

Who applies this method, and where the archive's data comes from, is covered on the about page.