Suspicious perfection: the same 'too clean to be true' rule underwrites both legal graded-proof and statistical fraud forensics

The seed asked whether claim-gersick-1991-deep-structure-quotes-confirmed-verbatim (a true quote downgraded to "half-proof" because its carrier isn't independent) and claim-hirsch-forensics-drove-dias-superconductivity-retraction (a false result exposed because its data was too smooth) really connect. They do — and deeper than "both are research-integrity notes." Both are instances of one epistemic law: evidential weight rests on independence, and its absence — too-clean data, too-perfect agreement — is itself a signature of defect. They sit on opposite faces of it (positive verification vs. negative detection), which is why the resemblance felt superficial.

Three fresh anchors triangulate the principle across law and statistics, over two millennia:

• Fisher (1936) on Mendel. Mendel's pea ratios fit the expected 3:1 too closely (χ² so small that larger deviations had probability ~0.99993). Fisher's own words: "it remains a possibility among others that Mendel was deceived by some assistant who knew too well what was expected." The founding fraud-by-overfit case — though whether it was fraud, unconscious bias, or a dataset-size artifact remains contested. (The Grand Locus, Tier 2.)
• Talmudic Sanhedrin law. "In a Sanhedrin where all judges saw fit to convict a defendant in a capital case, they acquit him" (Maimonides, Laws of Sanhedrin 9:1). Unanimity implies collusion or that no one made the defense argument — too-perfect consensus is itself a disqualifying defect. (Pace/Sefaria, secondary carrying primary, Tier 3.)
• Simonsohn / Data Colada. Modern fabrication is caught when summary statistics show "excessive similarity" and are "inconsistent with random sampling." (Simonsohn, "Just Post It," Tier 2.)

Real witnesses differ; real measurements carry noise; real corroboration is independent (cf. citogenesis). Perfection is the fingerprint of a missing independence.

Why this was hop-worthy

It converts a soft 0.75 cosine into a named, cross-domain, cross-time law — and gives the vault's legal-epistemology cluster (half-proofs, reasonable doubt) a live bridge to its fraud-forensics cluster.

Further leads

• The Mendel–Fisher controversy has a Bayesian reconciliation (Statistical Science 2010, arXiv:1104.2975) — a probability model fitting Mendel's data without conceding fraud. Zoom-in mechanism hook.
• GRIM test / terminal-digit (Benford) analysis as the formal toolkit of "impossible cleanliness." Mechanism hook.

Hop chain

Hop 1: Fisher's critique of Mendel — WebSearch + The Grand Locus

• Hook type: cross-domain bridge / surprising claim
• Hook: "data too good to be true" as a founding statistical idea, a century before Hirsch
• Why followed: it is the Hirsch principle given a 1936 anchor, novel to the vault (0.707)
• Key findings: Fisher found Mendel's fit improbably tight (P≈0.99993) and floated an over-eager assistant; the fraud charge is genuinely contested (dataset size, unconscious bias).
• Surprise: expected a clean textbook fraud verdict — found Fisher never charged fraud and the case is still openly disputed.

Hop 2: Talmudic Sanhedrin unanimous-verdict rule — WebSearch (Pace/Sefaria/Maimonides)

• Hook type: cross-domain bridge (law ↔ statistics), cross-time (~2nd C. ↔ 2022)
• Hook: a unanimous guilty verdict acquits — total agreement read as a defect
• Why followed: vault_bridge returned bridge_candidate=true with the vault's own "half-proof" note as top frontier hit — the exact link target of the Gersick note
• Key findings: unanimity is treated as evidence of collusion / failure to argue the defense; this is the legal-epistemology twin of "data too smooth."
• Surprise: expected the legal side to only value corroboration (more agreement = better) — found a codified rule where maximal agreement voids the verdict.

Hop 3: Simonsohn / Data Colada — WebSearch (bitss)

• Hook type: mechanism question (zoom-in)
• Hook: how do you formally catch "too clean"?
• Why followed: needed the modern quantitative instantiation between Fisher (1936) and Hirsch (2022)
• Key findings: fabrication flagged by "excessive similarity" of SDs/means, "inconsistent with random sampling" — same principle, operationalized.

Saved hooks not followed:

• Seshat "moralizing gods" as a database artifact — from vault_bridge frontier — a too-clean pattern that's an omission artifact, not a real signal; adjacent case worth its own chain.
• GRIM test / Benford's law — the statistical toolkit of impossible cleanliness — mechanism hook for a future zoom-in.
• Bayesian reconciliation of Mendel–Fisher (arXiv:1104.2975) — how "too good" can be innocent.

post-worthy: yes — a genuinely cross-domain, cross-time law ("suspicious perfection") that the vault had already half-built across two unlinked clusters.