ANALYTIC STANDARD
Evidence before mythology.
The archive preserves unusual ideas without granting credibility they have not earned. The same rules are applied whether a claim concerns propulsion, cognition, UAP, quantum physics, biology, or nonhuman technology.
HSARPA Evidence Scale
H0Pure HypothesisConceptually describable, but no persuasive empirical basis is established.
H1Theoretical BasisRelated mechanisms exist in accepted theory, but the proposed system itself is unverified.
H2Indirect EvidenceSuggestive observations or statistical results exist but conventional explanations remain viable.
H3Anomalous EvidenceA documented observation remains difficult to explain after serious conventional analysis.
H4Experimental ClaimA claimed effect has been produced, but robust independent replication is absent.
H5Reproducible AnomalyAn effect is independently reproducible while its mechanism remains unresolved.
MEASUREMENT
Direct measurement is not inference.
A UAP video can measure angular motion without measuring range. A force balance can measure deflection without proving a new propulsion mechanism. A statistical hit rate can differ from chance without identifying an information channel. Every dossier records what was directly observed before calculating what it might mean.
PHYSICS
Mathematical possibility is not engineering feasibility.
Einstein’s equations can admit unusual metrics; quantum theory can admit constrained negative-energy states; metamaterials can exhibit extreme effective parameters. None of those facts automatically supplies stable matter, manufacturable geometry, usable energy, or a device.
CONTROLS
Conventional explanations are part of the test.
Thermal expansion, magnetic coupling, ion wind, parallax, glare, contamination, publication bias, postselection and hidden energy are not dismissive excuses. They are competing hypotheses that must be measured or eliminated.
REPLICATION
Independent reproduction outranks narrative confidence.
Strong evidence survives different laboratories, different apparatus, blinded protocols, fixed analysis plans, matched nulls and documented chain of custody. Extraordinary interpretations become weaker when they depend on one instrument, one analyst, or unavailable raw data.