Evidence you were shown, not sampled

What it looks like. "Legal already signed off." "Everyone on the team is doing it this way." An approach that arrives exactly when it is needed, from someone who turns out to share your background, with a track record you can personally verify.

The familiar case. The stock-picker scheme. Ten thousand people are told a named stock will rise and ten thousand that it will fall; the wrong half are dropped and the exercise repeats. After six rounds, a hundred and fifty-six people have personally verified six consecutive correct calls and are invited to invest. Their evidence is real and their reasoning is sound over the sample they can see.

Why it fails. Apparent consensus is cheap to fabricate and rarely traced to primary sources. Apparent coincidence is cheap to manufacture by trying many times where the failures are never seen — nothing about the instance in front of you is false, and you have no way to observe the attempts that missed.

Instead. Require cited consensus to trace to primary sources; treat unverifiable approval citations as unconfirmed. Treat unsolicited improbable fit as a negative signal rather than a neutral one, and never let a track record you were shown substitute for one you sampled independently. Detection here belongs to whoever can see the whole attempt population, not to the individual target.