liveBook · Manning

Base rate fallacy

As explained in Timeless Algorithms, The Foundational Ideas

The base rate fallacy is the error of neglecting an event's underlying prevalence when interpreting evidence about it.

The base rate fallacy is the tendency to overvalue new evidence and undervalue an event’s underlying prevalence. A positive result can still imply a low probability of the event when it is rare and false positives are common. Bayesian inference combines prevalence with evidence to estimate that probability. section 2.2.2

Positive tests and rare events

A medical test with 99% sensitivity and a 1% false-positive rate yields a 9% probability of disease after a positive result when prevalence is 0.1%. In a surveillance example, a 95%-accurate offender match with a 5% false-positive rate yields only a 0.19% probability that a flagged person is the offender when prevalence is 1 in 10,000. section 2.2.2

How Timeless Algorithms, The Foundational Ideas covers it

How deep it goes
Works through posterior probabilities for medical testing and surveillance, including prevalence and false-positive rates.
Examples
Uses a rare-disease test and an offender-matching surveillance algorithm to illustrate low posterior probabilities.
What you can do after
Evaluate positive test results by combining evidence with the event's underlying prevalence.
How it connects
Connects Bayesian inference, Bayes' theorem, prior probability, and false positives.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Timeless Algorithms, The Foundational Ideas. The section links point to the parts of the book it draws on.