Method
How this archive decides
A site that assigns verdicts has to show its working. This page states the standard each verdict is held to, what counts as a source, which subjects are excluded outright, and how to tell us we got something wrong.
The three verdicts
Every entry carries one, and every entry carries a one-sentence written justification for it, displayed at the top of the page. There is no fourth option and no “you decide”. Refusing to reach a conclusion is a conclusion, and usually the wrong one.
CONFIRMED51 entries
Declassified records, sworn congressional testimony, court findings, or an admission by the responsible institution establish that a real conspiracy took place — coordinated action by multiple parties. A fabrication by a single author is DEBUNKED, however well documented: one person is not a conspiracy.
See all confirmed entries →UNRESOLVED9 entries
Qualified, independent experts examining the same evidence reach different conclusions, and no finding has settled the question. Not a hedge, and not a courtesy extended to weak claims.
See all unresolved entries →DEBUNKED105 entries
The central claim has been tested against physical evidence, documents, or investigation and fails. Where the claim is unfalsifiable by construction, that is stated as part of the refutation.
See all debunked entries →UNRESOLVED is not a hedge. It is the verdict most easily abused, and overusing it would destroy the point of the other two. It appears on 9 of 165 entries. A claim being unproven is not enough. A claim being popular is not enough. The test is whether qualified, independent people looking at the same evidence currently disagree — and in most of those cases the entry also names the specific document or dataset that would settle it.
Verdicts can change. When a Freedom of Information release, a court finding or a replication changes the evidence, the entry changes with it, and the change is noted rather than quietly applied.
The share of entries marked CONFIRMED reflects this archive, not the world. Roughly a third of what is catalogued here is documented fact, and that is a fact about sourcing rather than about how often conspiracy theories turn out to be true. Encyclopedic coverage exists reliably for cover-ups that produced congressional hearings, court judgments and declassified files; it does not exist for the enormous, forgettable tail of false claims that circulated for a week and left no record anyone thought worth keeping. An entry needs three sources, so the archive systematically oversamples the confirmed end. Read the ratio as a description of what is documented, never as a base rate.
Sourcing standards
Three sources, minimum
No entry ships with fewer. The build fails if one is missing — this is enforced in code, not by good intentions.
Primary documents first
Where a declassified file, court judgment, inquiry report or dataset exists, it is cited directly and listed above the journalism. CONFIRMED entries are built on these.
Written here, not copied
Wikipedia and other encyclopedias are research inputs, not the product. Every section is written for this archive. Where an encyclopedia article informed an entry it is cited as a source and its CC BY-SA licence is acknowledged in the footer.
Quotes are rare and attributed
At most one short quotation per source, in quotation marks, with the speaker or document named. If a famous line cannot be verified against the record, it does not appear.
The claim is steelmanned first
Every entry states what believers assert, plainly and at its strongest, before the evidence section answers it. A debunk that mocks the claim it is answering persuades nobody who needed persuading.
What is excluded, and why
Some material is left out on purpose, and the reasoning is stated rather than hidden.
Theories whose core claim is an ethnic or religious conspiracy narrative get no entry. Blood libel, material derived from the Protocols of the Elders of Zion, the Great Replacement and their descendants are not curiosities to be catalogued. Giving one a page, a card and a badge hands it a surface area that no amount of debunking offsets. Where such a narrative is unavoidable context for an entry that is included — the reptilian claim is the clearest case — it is named and labelled inside that entry’s prose instead.
Theories targeting living private individuals get no entry, and neither does anything active enough that a page here would function as a harassment vector. This archive is not a research aid for people looking to find someone.
No comments, accounts, votes or belief polls. Each of those turns an archive into a forum, and a conspiracy forum is a different and much worse product. There is nothing to join here.
Imagery
Every cover plate on this site is generated procedurally from the entry’s slug — halftone, redaction, contour or signal trace, chosen deterministically so the same entry always draws the same plate. Where a genuinely free-licensed photograph is used, its credit and licence are printed under it.
No photorealistic images of real events or real people are generated anywhere on this site. A site about fabricated evidence cannot ship fabricated evidence, even decoratively.
Corrections
Errors here are worse than errors elsewhere, because the whole proposition is that the verdicts are trustworthy. If something is wrong — a date, a name, a citation that does not support the sentence attached to it, a verdict the evidence no longer supports — it should be fixed.
The content of this archive lives in the repository as plain data files, one entry per record, with sources attached. A correction is a change to that file: cite the document, name the field, and the entry is updated. Substantive corrections that change a verdict are noted on the entry rather than applied silently.
The standard for changing a verdict is the same as the standard for assigning one. Volume of belief is not evidence. A new document is.