DIAG-10 · Diagnosis · Invariant: holds by construction, no expiry
Verdicts are asymmetric under partial coverage
The law
A directly witnessed, locally decidable fault is established by one observation regardless of unmeasured dimensions; an unqualified all-clear is a universal claim requiring evidence on every dimension. Signal absence under partial coverage licenses neither verdict.
Scope
Composite health/trust/audit verdicts. The fault side requires the defect be positively observed, a score depressed by the coverage gap itself is not a witnessed fault.
What this licenses you to do
Attach coverage metadata to every composite score; gate all-clear verdicts on full coverage; before passing a low score through, verify the fault was directly witnessed rather than inferred from absent data.
Evidence
Observed building a tracking-quality scoring system: when coverage of the underlying signal is asymmetric, a low score can be trusted (it's confirmed by the data present) while a high score cannot (it may only reflect what wasn't checked yet), exactly the asymmetry this law describes.
Riibon's own analysis is held to these laws
The laws on this site are a public subset of the rulebook Riibon's analysis works to. Before a metric movement is called a market event, the order of checks they set has to rule out the measurement first.
Ask the Riibon app why a number in your account moved, and see which checks it ran first. Access starts with a short call: we look at your ads together, then run them for you or set the app up with you.