Project Blue Book: Special Report No. 14

In 1951 the US Air Force contracted the Battelle Memorial Institute in Columbus, Ohio to bring statistical method to the case files accumulated by Project Grudge and its successor, Project Blue Book. The study ran for several years and was published in 1955 as Project Blue Book Special Report No. 14. It remains the largest formal analysis of unidentified aerial reports the United States has produced.

It is also among the most misrepresented documents in the subject. It is quoted as proof that the unknowns were merely poor reports, and quoted with equal confidence as proof that they were not. Both readings survive because most people citing it have read the Air Force press summary rather than the report.

What Battelle actually did

Analysts sorted 3,201 cases into knowns, unknowns, and insufficient information. A case was placed in the unknown category only when all four evaluators independently agreed that the available data ruled out every conventional explanation they could identify. Roughly one case in five survived that filter.

Each report was also graded for quality — the experience of the witness, the number of observers, the internal consistency of the account, the availability of instrument data. This grading is what makes the study interesting, and it is the part that is usually left out.

The comparison the report is built on

Battelle then compared the knowns and the unknowns across characteristics including colour, number of objects, duration, brightness, speed and shape, testing whether the two groups behaved like samples drawn from a single population. The report concluded that the probability of that was low.

This is a narrower finding than it is often made to carry. It says the unknowns do not look statistically like misidentified aircraft, balloons and astronomical objects. It does not say what they were, and the report does not claim to.

The Air Force summary released alongside it emphasised that no evidence had been found of anything beyond conventional explanation — a defensible statement about evidence, frequently repeated as though it were a statement about the data. The two are not the same, and the distance between them is where sixty years of argument has taken place.