---
title: "R. A. Fisher"
type: "entity"
entity_kind: "person"
status: "hub"
canonical_name: "R. A. Fisher"
aliases: ["Ronald Fisher","Ronald Aylmer Fisher","R.A. Fisher"]
first_seen: "2026-07-11T00:00:00.000Z"
writer_model: "claude-opus-4-8"
connects_to: ["suspicious perfection","statistical fraud detection","Gregor Mendel","design of experiments","statistics"]
seek_code_commit: "a619c8a"
---


R. A. Fisher (1890–1962) was a British statistician and geneticist who built much of the foundation of modern statistics — analysis of variance, maximum-likelihood estimation, the randomized design of experiments, the modern synthesis linking Mendelian genetics to Darwinian selection. He was also, as Galton Professor of Eugenics at University College London, a central and lastingly controversial figure in early-twentieth-century eugenics.

In this vault Fisher anchors the founding case of the "suspicious perfection" cluster: his 1936 paper "Has Mendel's Work Been Rediscovered?" pooled chi-squared values across [[entity-gregor-mendel|Mendel's]] reported pea-segregation experiments and put the probability of a genuinely random dataset fitting the predictions at least this closely at P≈.99993 — too clean to be ordinary sampling luck. The move is the ancestor of modern data forensics: too *little* noise treated as evidence of a defect in provenance, not as confirmation.

## References

- [[claim-fisher-1936-flagged-mendels-pea-data-as-improbably-close-fit]] — the 1936 re-analysis and the "deceived by some assistant" sentence.
- [[observation-suspicious-perfection-independence-absence-signals-defect]] — the general law Fisher's verdict founds.
- [[moc-too-clean-to-be-real]] — the data-forensics map that cross-links Fisher as the method's lineage.