Too Good to Be True¶
Gunn, L. J., Chapeau-Blondeau, F., McDonnell, M. D., Davis, B. R., Allison, A., & Abbott, D. (2016). Too Good to Be True: When Overwhelming Evidence Fails to Convince. Proceedings of the Royal Society A, 472(2187).
Cited by¶
3 citations across 3 artifacts.
Each citation links to the sentence it supports in the citing article.
Primes¶
- Paradox of Unanimity
- In forensic and jury deliberation, the ancient Sanhedrin rule acquitting a defendant convicted unanimously is the classical statement, with modern descendants in the scrutiny of too-perfect lineup identification.
This sourceFormalizes the non-monotonic posterior and the unanimity sign-flip, using witness identification and the Sanhedrin acquittal rule as motivating cases.
- In forensic and jury deliberation, the ancient Sanhedrin rule acquitting a defendant convicted unanimously is the classical statement, with modern descendants in the scrutiny of too-perfect lineup identification.
Mechanisms¶
- Blind Pattern Comparison Round
- That inversion is the point: in genuinely noisy independent observation, perfect unanimity is less likely than modest disagreement, so flawless agreement often signals a hidden common cause rather than a strong signal.
This sourceShows probabilistically that, when observations are subject to even rare systemic failure, sufficiently unanimous results can be less credible than modest disagreement and can indicate a hidden failure state.
- That inversion is the point: in genuinely noisy independent observation, perfect unanimity is less likely than modest disagreement, so flawless agreement often signals a hidden common cause rather than a strong signal.
- Corroboration Ladder
- Sometimes suspiciously unanimous agreement should lower confidence, not raise it, because it signals a common cause or systematic error rather than independent convergence: the paradox of unanimity.
This sourceShows that unanimous evidence can reduce confidence when observations may share a systemic failure rather than being independent.
- Sometimes suspiciously unanimous agreement should lower confidence, not raise it, because it signals a common cause or systematic error rather than independent convergence: the paradox of unanimity.
Verification¶
This reference passed the adversarial substantiation pipeline: it was checked to exist and to support the claim it is attached to. See how references were verified.
Registry ID ref:5070e6bdfdfd · see in the full table