Transparency

State of the Record

Every figure here is counted live from the 307 cases currently on the record. Nothing is hand-picked. A fact-checking archive should be as auditable as the claims it checks.

307 casessince 29 Jun 202684.4% reached a primary source

Verdicts

Where cases land on the calibrated scale.

Mostly accurate115 / 37.5%

Source exists but framing is misleading86 / 28%

Partially accurate but misleading52 / 16.9%

Confidence

How sure the evidence allowed the verdict to be.

High169 / 55%

Medium122 / 39.7%

Low5 / 1.6%

Claim types

The only category dimension in the archive.

Distortions observed

How often each pattern shows up. Cases can carry more than one.

Exaggeration155 / 50.5%

Omitted qualifier136 / 44.3%

Species extrapolation61 / 19.9%

Fabrication59 / 19.2%

Causal overreach56 / 18.2%

Subgroup generalization48 / 15.6%

Temporal overreach34 / 11.1%

Marketing as evidence30 / 9.8%

Date or context mismatch16 / 5.2%

Quote manipulation2 / 0.7%

Visual manipulation1 / 0.3%

Where claims came from

The origin platform or input type.

Instagram181 / 59%

Submitted image85 / 27.7%

Submitted text35 / 11.4%

Submitted audio3 / 1%

TikTok1 / 0.3%

YouTube1 / 0.3%

Submitted video1 / 0.3%

Cases per month

How the record has grown.

126-06

6426-07

16726-08

7526-09

Percentages are share of the whole archive and are recomputed on every visit. Cases can carry several distortions, so those bars do not sum to 100 percent. The archive reflects what has been submitted for checking, not a representative sample of everything circulating.