12 min · 11 chaptersNarrated in your voice
What it's about
David Spiegelhalter argues that statistics is not a set of mechanical procedures but a disciplined way of reasoning about an uncertain world. Using real cases—serial killer detection, courtroom miscarriages, cancer screening, machine learning—he shows how careful statistical thinking distinguishes sound evidence from persuasive-sounding numbers. The book mattered because it reached general readers just as public debates over data, algorithms, and misinformation were intensifying, giving them tools to question numbers rather than simply accept or dismiss them.
Read a little every day
15 min a day
1 days · done Sat, Oct 31
Read together
Start a circle and share the code — everyone sees everyone's honest place in the book. Accountability, not leaderboards.
Who it's for
- A curious reader who distrusts statistics in the news but wants to know which numbers to trust and why. - A student or early-career analyst who knows the formulas but not the judgment calls behind them. - A policy or health professional who must explain risk and uncertainty to a public audience.
What’s inside
- 1Statistics as a Way of Reasoning1 min
Statistics is best understood as a cycle of inquiry—defining a question, planning, gathering data, analyzing, and concluding—not a toolbox of formulas applied after the fact.
- 2What a Single Number Hides1 min
A summary statistic like a mean or median can obscure more about a dataset than it reveals, especially when the shape of variation matters more than the center.
- 3Correlation Is Not the Crime, Causation Is the Charge1 min
A correlation, however strong, only becomes evidence of causation when it survives a structured set of independent checks, not through statistical strength alone.
- 4Probability Has More Than One Meaning1 min
Public confusion about risk often stems not from bad arithmetic but from mixing incompatible definitions of probability without realizing it.
- 5The Trouble With Significance1 min
The p-value, treated by much of science as the arbiter of a real effect, answers a narrow technical question that is routinely mistaken for the broader question of whether a finding is true.
- 6When the Law Ignores the Base Rate1 min
Courts and juries repeatedly convict or exonerate based on probability reasoning that ignores how rare an event is in the general population, turning a small chance into false certainty.
- 7Prediction Is Not Explanation1 min
Algorithms optimized purely to predict outcomes can perform well without offering any understanding of why, and mistaking predictive accuracy for causal insight leads institutions to deploy systems they cannot justify or correct.
- 8How Risk Gets Distorted in Translation1 min
Presenting risk in relative rather than absolute terms is one of the most common ways statistics mislead without a single number being technically false.
About David Spiegelhalter
David Spiegelhalter is a British statistician, Emeritus Professor of Statistics at the University of Cambridge, and former president of the Royal Statistical Society. Knighted for his services to statistics, he has spent decades applying statistical reasoning to medicine, criminal justice, and public risk communication, including work on the Harold Shipman inquiry and NHS clinical outcomes.
