Library › Study

Quantum Pings Library

Methods for Studying Coincidences

By Persi Diaconis and Frederick Mosteller · Study

Type
Study
By
Persi Diaconis and Frederick Mosteller
Journal
Journal of the American Statistical Association
Source
JASA ↗
Published
1989

The paper that gave coincidence a mathematical home: it catalogs the fallacies that make coincidences feel impossible and shows how to study them properly.

Persi Diaconis and Frederick Mosteller's "Methods for Studying Coincidences" (1989, Journal of the American Statistical Association) gave the topic a mathematical home. It catalogues the fallacies that make coincidences feel impossible — selection, multiple endpoints, the law of truly large numbers — and shows how to estimate the real probability of famous cases.

The paper is the foundation of the skeptical position, and a curious detail keeps it honest: the authors end by admitting that some coincidences resist easy explanation. That concession, from two of the century's great statisticians, is still quoted by the other side. It is the rare skeptical text that both camps respect.

The meaning

Even this skeptical classic concludes that some coincidences resist easy explanation — a concession often quoted by the other side.

The skeptic

The paper is the origin of the "law of truly large numbers" as a serious tool: most coincidences are explained by scale, selection, and hidden causes.

About the research

Two of the most influential statisticians of the 20th century, at Stanford and Harvard.

Key takeaways

researchstudyskeptic
← Back to the Library View source ↗