METHOD · ALIEN PROOF INDEX 2,880 cases scored · highest 53 · median 3

The Alien Proof Index

Every UAP case in this archive carries a score from 0 to 100. It answers one narrow question: how far does this event go toward showing that non-human intelligence has actually been found? Not how strange it was, not how famous it is — how much it would move a careful person who did not already believe.

The honest answer, almost always, is: not far. Of the 2,880 cases scored, the median is 3 and the highest reaches 53. That is not a bug in the scoring. It is the finding. The best-evidenced UAP encounters in the historical record — the 2004 Nimitz encounter, the Belgian wave, the Minot Air Force Base incident of 1968 — are multi-sensor, multi-witness events that remain officially unresolved, and they still sit in the forties, because being unexplained is not the same as being extraterrestrial. A light in the sky that nobody can identify is a light in the sky that nobody can identify.

To reach 100 an event would need physical evidence that survived analysis, independent scientific replication, and scientific consensus. Nothing in this archive has any of the three, and it is worth being clear that nothing in the wider world does either. The scale runs to 100 so that the distance still to travel is visible.

What the score is made of

The index is deterministic. It is computed from each case's own record — the witness count, the kind of evidence, the language of the government's own summary — and not from anyone's opinion about the case. That means you can check it, and it means two cases with the same evidence get the same score whether or not one of them is famous. Every event page shows its own arithmetic.

Two ideas do the work. The first is evidence quality: how good is the evidence that something unexplained happened at all? The second is probative value: if the account is accurate, how much does it actually indicate non-human intelligence? A case needs both, because perfect footage of a blurry dot is still a blurry dot, and a vivid story about a landed craft with occupants is still a story. These components add:

Component	Max	What earns it
Instrumented evidence	+14	Radar corroborated by imaging scores highest. A single sensor scores less. A photograph alone scores least. Nothing scores nothing.
Witnesses	+10	Scaled by number and by whether the observers were trained — pilots, aircrew, astronauts, officers. Capped below instrumentation on purpose.
Independent corroboration	+8	Separate observers at separate vantage points, or — rarest and most valuable — a witness account and an instrument record of the same event.
Physical evidence	+12	Recovered material, a landing trace, electromagnetic or physiological effects. Almost nothing in the archive scores here at all.
Official status	+4	Whether an agency investigated and still carries the case as unresolved. Worth little: unresolved is a fact about the investigation, not about the object.
Anomalous performance	+6	Behaviour described as beyond known aircraft capability. Discounted, because "impossible" speed is usually inferred from a sensor at unknown range.
Structured craft	+5	A solid, structured object rather than an ambiguous light or a blob of contrast.
Occupants or artifact	+5	Occupants seen, or an artifact recovered. If this ever scored honestly and survived scrutiny, it would be the story of the century.

What takes it away

A scoring system that can only add is a scoring system that calls solved things mysteries. These subtract, and they subtract hard enough to zero a case out:

Penalty	Max	What triggers it
Conventional explanation	-25	If the phenomenon is affirmatively solved — a known, documented effect — the score collapses. The Apollo light-flash tapes score zero for exactly this reason.
Provenance	-10	A broken chain of custody, an anonymous source, or media that was altered before it reached the investigators.
Testimony delay	-6	An account recorded long after the event. Memory at thirty months is not a measurement, however precise it sounds.

There is one further rule. A case with no instrument record and nothing physical left behind is held at a ceiling of 12, however many people saw it and however credible they were. This is the hardest rule in the index and the one most likely to annoy people, so it is worth defending plainly: testimony is how nearly every case enters the record, and it is still not a measurement. A crowd can be sincere, trained, unanimous, and wrong about what an unfamiliar light was. Holding anecdote below the recorded cases is the single biggest thing that keeps the index honest.

What it is not

It is not a probability. A case scoring 40 is not "40% likely to be aliens" — it means the case is unusually well documented and still a long way from proof. It is not a measure of how interesting a case is; some of the most historically important documents in the archive score zero, because the government explained them. And it is a heuristic built from the text of the record, so it will occasionally be wrong in both directions. Where it is wrong, the breakdown on the case page will show you exactly why, which is more than most scores will do.

The highest-scoring cases

Sorted by score. Note how far the top of the table still is from the top of the scale, and note that of the 2,880 cases scored, only 37 reach the highest band at all.

The full archive is at /disclosure/archive/, and the PURSUE disclosure releases are indexed at /pursue/.