THE EVIDENCE.

Our question is simple: does the available documentary record support what an American presidential administration says? We check specific, testable claims, not political loyalties or whether we like a policy.

What the scores mean

A 95% Fact score is our editorial confidence that the precise claim is supported. A 95% Fiction score is our confidence that the available record contradicts that claim. This number is not a measure of whether we like the speaker or a scientifically validated probability. Both sides require at least 85% editorial confidence to appear.

Where evidence comes from

We seek the president’s or administration’s original words, executive orders, federal data, court documents, scientific findings and independent reporting. Each article links to source material, presents a narrowly worded claim, documents caveats and explains plausible consequences.

What if the claim is partly true?

We distinguish exact statements from overstatements, missing context, predictions and opinion. If the record is ambiguous, we should not force the statement into either column merely to fill space. The verdict must be appropriately narrow.

How corrections work

Claims are reassessed as new evidence becomes available. The publication’s content is stored in GitHub source files rather than Ghost. Corrections need reviewed source references and a new commit. Retrospective editions identify their edition dates separately from the time they were actually written.

Our confidence scores are editorial judgments, not yet empirically calibrated. Return to the claims ↗