Bayesian Suspicion Machine

Start with how common the claim should be, add evidence one piece at a time, and see exactly how each assumption moves the odds.

The precision is only as good as the likelihood ratios you choose.

Set the prior

Prior probability

Percent

Before considering this evidence, how often would a claim like this be true?

Garbage assumptions still produce garbage precision. Likelihood ratios are judgment calls. The useful part is forcing those calls into the open.