The fact ledger uses two tools together: a verdict label that says what kind of finding a check reached, and the Veracity Meter, a five-step bar that shows how well a claim stands up. This page explains both so that readers can judge for themselves whether a score is fair.
The five-step Veracity Meter
The meter has five segments. The number of filled segments is the score.
- 1 of 5 - Contradicted. The claim is contradicted by primary evidence, or the supporting material is fabricated.
- 2 of 5 - Mostly unsupported. A small true element exists, but the claim as stated is misleading or lacks the context that changes its meaning.
- 3 of 5 - Partly supported. The claim holds in a limited setting but is not true in general.
- 4 of 5 - Largely supported. Evidence supports the claim with minor open questions.
- 5 of 5 - Fully supported. Primary evidence confirms the claim and no significant caveat remains.
Verdict labels
- Verified True: primary evidence confirms the claim. Usually scores 4 or 5.
- Context Required: the claim is accurate only within a narrow setting. Usually scores 2 or 3.
- Misleading Context: a real fact is framed in a way that leads to a wrong conclusion. Usually scores 2.
- Debunked: evidence shows the claim is false. Scores 1.
- Synthetic AI Origin: the material was generated or heavily manipulated. Scores 1.
How a score is assigned
A claim is stated in a single sentence, in the words of the person or post that made it. Two checkers then work separately: each identifies the type of primary evidence that would settle it, such as an inspection record, a legislative register or a dataset, and compares the claim against it. If their scores differ by more than one step, a senior editor reviews the disagreement and the entry notes that it was contested.
What the score is not
The meter does not measure how important a claim is, or whether the person who made it meant to deceive. It does not predict the future. A claim about something that may happen is checked against what the evidence supports now. Nor is it a precise measurement; the steps are editorial judgments and the written evidence note is the real substance of every entry.
The sample entries are fictional
The six entries currently in the ledger are illustrative samples created for the launch edition. They show the format and the reasoning but do not describe real claims, real organizations or real events. No real-world accuracy figure is attached to this scale, and none will be claimed until a track record exists.
Last updated 2026-10-04