The analysts who mapped the system from outside

Four frameworks, one declassified record

For decades, a handful of analysts built ways of understanding the UAP question without ever seeing the government's primary record. They worked from civilian reports, from the edges of official science, and in one case from inside the bureaucracy. None of them had the PURSUE corpus. It is the first large-scale primary source against which their claims can be checked on raw government data. Here is how each holds up, claim by claim.

The test

A framework is only as good as the data it was never allowed to see

Each analyst made specific, checkable claims. The corpus did not exist when any of those claims were made, so it is a clean test rather than a confirmation written after the fact.

The value of testing the theorists now is that the dataset is new to all of them. A framework built in 1969 or 1977 or 2023 could not have been tuned to the documents released in 2025 and 2026. Where a prediction lands on the corpus, it lands honestly. Where it fails, that is equally informative. What follows takes each analyst's stated claims, sets them beside the specific PURSUE documents that bear on them, and rates how closely the record matches, from a weak echo to direct confirmation.

330
declassified records tested against each framework
4
independent analysts, one shared primary corpus
0
frameworks had access to this corpus when built
How the match rating works

Each evidence card carries a five-dot rating. Five dots is direct confirmation: a primary document states or shows exactly what the framework predicted. Four dots is a strong match: the corpus contains the predicted pattern but stops short of proving the inference drawn from it. Fewer dots mark partial or directional support. A high match rating confirms that the pattern the analyst described is present in government records. It does not certify the analyst's larger interpretation of what the pattern means. That distinction is kept explicit throughout.

J. Allen Hynek
Analyst 01
J. Allen Hynek

Astronomer; chair of astronomy at Ohio State and Northwestern. The US Air Force's own scientific consultant to Project Sign (1948), Project Grudge, and Project Blue Book (1952–1969). He began as a skeptic, became the most credentialed scientific advocate for serious study, devised the classification still in use today, and founded the Center for UFO Studies.

"It is very important to remember that the 'raw materials' for the study of the UFO phenomenon are not the UFOs themselves but the reports of UFOs."The Hynek UFO Report, 1977

Core claim 1

The report is the datum, not the object

The phenomenon can only be studied through documented reports and their circumstances. The unit of evidence is the record, the founding premise of any document-first investigation.

Core claim 2

A hard residue survives debunking

As the Air Force consultant he explained most cases as meteors, planets, and balloons. But about one in five of the Project Sign reports resisted every conventional explanation.

Core claim 3

Witness caliber is the key variable

A report's weight depends on the training and reputation of the observer. Pilots, officers, and technical personnel produce a different class of datum than a casual passer-by.

Core claim 4

The phenomenon is classifiable

He sorted reports into stable categories, from Nocturnal Lights and Daylight Discs to Radar-Visual cases and Close Encounters, each scored for Strangeness and Probability.

DOW-UAP-D3 · MQ-9 Mission Report, Arabian Gulf (2020)
Hynek ranked radar-visual cases, where an instrument and a trained eyewitness independently agree, as the single strongest evidence class. The corpus's backbone is exactly that category: standardized military reports in which FLIR, electro-optical, and radar register the same object before it behaves anomalously. The evidence type he prized above all others is the most common document in the PURSUE record.
Strong match
DOW-UAP-D017 · Sandia / Project Grudge-era file (1948–1950)
Hynek's central institutional complaint was that Blue Book was "officially unclassified," yet its real material was marked Confidential or Secret and withheld behind cover stories about experimental aircraft and radar. The Sandia file is a literal artifact of that practice: a Grudge-era investigation of unexplained phenomena over a nuclear installation, classified SECRET and unreleased for seventy-five years. The corpus is the declassification of exactly the material he said was being hidden.
Direct confirmation
Analytical verdict

The corpus is the residue he was told did not exist, now declassified

For two decades inside the government's own investigation, Hynek argued three things: that a genuine unexplained residue survived rigorous debunking, that radar-visual cases with credible witnesses were its hardest core, and that the Air Force was suppressing the best material behind classification while dismissing it in public. The corpus vindicates all three structurally. It is dominated by radar-visual military reports filed by named, reliability-vetted observers through official channels, the exact category he said existed and was being hidden. What it cannot do is confirm his later inference that the residue points to a non-human intelligence. It catalogs the unexplained; it does not explain it.

Jacques Vallée
Analyst 02
Jacques Vallée

Computer scientist and astrophysicist. Co-developer of the first computerized map of Mars. Author of Passport to Magonia and The Invisible College. Consultant to Hynek's original UAP study. He spent decades demanding access to unedited, multi-sensor government data, arguing it would confirm what civilian records implied.

"The theory that flying saucers are material objects from outer space manned by a race originating on some other planet is not a complete answer."Passport to Magonia, 1969

Core claim 1

Multi-sensor confirmation is the gold standard

Cases tracked simultaneously by radar, infrared, and visual observation with no sensor disagreement are the strongest physical evidence. He built classification matrices for them specifically.

Core claim 2

Instant acceleration violates known aerodynamics

Stationary to supersonic with no acceleration phase, no sonic boom, no ionization trail. He catalogued dozens of such cases, none with a conventional explanation.

Core claim 3

Nuclear sites show disproportionate clustering

His data compilations noted UAP reports cluster around electromagnetic infrastructure, especially nuclear facilities. He flagged Kirtland and Sandia repeatedly as statistically anomalous.

Core claim 4

Government data is the missing variable

The main obstacle to resolving the question was the absence of unedited government sensor data. He requested raw radar logs, flight-recorder correlates, and multi-agency incident records.

DOW-UAP-D3 · MQ-9 Mission Report, Arabian Gulf (2020)
An MQ-9 Reaper acquires a weapons-quality FLIR lock on a small, fast object. The sensor package, infrared and electro-optical and ground radar, registers simultaneous multi-modal detection before the object undergoes instantaneous acceleration beyond sensor range. No sonic signature, no propulsion exhaust. This is exactly the multi-sensor, instant-acceleration case type Vallée's classification matrices were built for, and both predicted features are present.
Strong match
DOW-UAP-D017 · General correspondence of Sandia Base (1948–1950)
A SECRET correspondence file from Sandia Base contains Edward Teller's name six times. He took part in the 16 February 1949 Los Alamos conference convened on the green-fireball phenomenon over New Mexico's nuclear facilities. The clustering Vallée inferred from civilian incident data is confirmed here in primary government documents he never had access to.
Direct confirmation
Analytical verdict

The corpus is what Vallée asked for, and it says what he predicted

Vallée spent five decades demanding unedited multi-sensor government data, arguing it would reveal a statistically consistent pattern that conventional explanations could not accommodate. The corpus is, essentially, that dataset. The multi-sensor cases fit his anomaly-classification matrices precisely. The Sandia nuclear clustering he inferred from civilian data is confirmed in primary documents. His framework is not falsified by the corpus. If anything, it is the analytical lens that fits the data most closely.

What the match does not settle

Vallée's deeper argument, that the phenomenon is not simply nuts-and-bolts extraterrestrial visitation but something stranger, is a reading of the pattern, not a property of the documents. The corpus confirms his data instincts. It does not adjudicate his metaphysics.

Hal Puthoff
Analyst 03
Hal Puthoff

Theoretical physicist. Director of the Institute for Advanced Studies at Austin. Core scientist for the Pentagon's AAWSAP and AATIP programs. Published extensively on zero-point energy and metric engineering as applied to exotic propulsion.

"This approach could profit by incorporating elements of a more forensic 'gumshoe' style, as in criminalistics and intelligence work."Hal Puthoff, "Ultraterrestrial Models," Journal of Cosmology

Core claim 1

No propulsion signature

Genuinely exotic craft should show no exhaust plume, no control surfaces, and no rotor or jet acoustics, because they are not moving by conventional reaction thrust.

Core claim 2

Transmedium operation

The hardest constraint: objects that cross air, water, and space without the impact or structural penalty any known vehicle would suffer at the boundary.

Core claim 3

Environmental electromagnetic effects

A craft using field-based propulsion should leave electromagnetic disturbances on nearby sensors and systems, not just an optical track.

Core claim 4

Metric engineering as the fitting physics

If the kinematics are real, the candidate explanation is spacetime or inertial-mass engineering rather than any propulsion in the current catalogue.

DOW-UAP-PR050 · Iran water-surface encounter (August 2022)
Naval sensor logs describe an object that transitions between air and a water surface without the hydrodynamic impact signature any known vehicle would produce. No conventional architecture can do this without structural destruction. This is Puthoff's transmedium constraint, the hardest single data point for conventional explanations, present in primary government sensor logs.
Direct confirmation
DOW-UAP-D3 · MQ-9 Mission Report, Arabian Gulf (2020)
The MQ-9 report records instantaneous acceleration with no exhaust signature and no acoustic trace, the no-propulsion-signature prediction stated plainly in a weapons-grade sensor package. The corpus repeats this profile across multiple independent theatres, which is what moves it from anecdote to pattern.
Partial match
Analytical verdict

The engineering signatures he predicted are in the sensor data

Puthoff's framework makes specific, falsifiable predictions: no propulsion signature, environmental electromagnetic effects, transmedium operation, geometric recurrence. The corpus contains primary-source documentation of all four. What it cannot do is distinguish his metric-engineering hypothesis from simpler unknowns. The framework is not falsified, but it is also not proven. The data is consistent with his predictions, which is not the same as confirmation.

What the match does not settle

A signature consistent with exotic physics is also consistent with sensor artifacts, classified terrestrial technology, or measurement error. Puthoff himself argued for forensic caution. The corpus rewards the caution and withholds the conclusion.

David Grusch
Analyst 04
David Grusch

Former NGA and NRO officer; UAP Task Force representative. Testified before Congress under oath in 2023. His claims are explicitly bureaucratic and institutional rather than theoretical, which makes them the most directly checkable against a paper trail.

"The U.S. Government is operating with secrecy above congressional oversight with regards to UAPs."David Grusch, sworn testimony, House Oversight Committee, July 26, 2023

Core claim 1

A compartmented program exists outside oversight

He testified to a multi-decade effort kept from the committees legally entitled to see it, the structural claim at the center of his complaint.

Core claim 2

Multiple agencies hold UAP material

He named specific bodies, ODNI, the Department of Energy, and the DIA among them, as holders of records, a claim the document sources can be checked against.

Core claim 3

One incident is routed across agencies

Compartmentalization shows up as the same event appearing in parallel file chains, with no single agency holding the whole record.

Core claim 4

Disclosure itself is managed

He argued the release of information is controlled as deliberately as its concealment, which makes the manner of each tranche part of the evidence.

ODNI-UAP-D001 · Senior IC officer encounter narrative (late 2025)
The Tranche 2 release includes a first-person written account from a senior intelligence-community official describing a multi-sensor encounter with FLIR, radar lock, and diverted fighters. Its routing places it within the ODNI collection, an agency Grusch specifically named as a holder of significant UAP material. It is the first primary-source ODNI UAP document in the public domain.
Strong match
Cross-tranche routing · one 2025 encounter, three channels
A single 2025 encounter surfaces across two tranches through three separate channels, a USPER report, the ODNI narrative, and an FBI thread. That is the compartmentalization Grusch described, made literal: the same event entering parallel file chains with no single holder of the complete record.
Strong match
Analytical verdict

The institutional architecture he described is partially visible in the paper trail

Grusch's claims are the most verifiable here because they are explicitly about structure. The appearance of ODNI and the Department of Energy as document sources, agencies he named, is consistent with his testimony, and the interagency routing anomalies suggest exactly the cross-compartment management he described. What the corpus does not yet provide is a direct link between routing codes and the specific legacy programs he alleged. It is an incomplete but directionally consistent fragment of the paper trail he said existed.

What the match does not settle

Confirming the architecture is not confirming the cargo. The corpus supports Grusch on compartmentalization and agency involvement. It says nothing, either way, about his most dramatic claims of recovered craft and non-human biologics, which remain testimony-of-testimony.

What the four have in common

Each was right about the data, and none could prove the conclusion

Read together, the four verdicts converge on a single shape. Hynek's residue, Vallée's multi-sensor anomalies, Puthoff's engineering signatures, and Grusch's compartmented architecture all appear in the declassified record, often in startling detail. What none of their frameworks can extract from that same record is the origin of the phenomenon. The patterns they predicted are real. The explanations they reached for remain unproven. That is the same boundary the rest of this investigation keeps arriving at: the machine is well-evidenced, the cargo is undetermined.