How to Lie with Statistics

Darrell Huff catalogs the everyday tricks that advertisers, pollsters, and journalists use to make numbers say whatever they want them to say, from small unrepresentative samples dressed up as authoritative surveys to graphs with truncated axes that exaggerate tiny differences into dramatic-looking trends. He works through each device methodically — the misleading average, the biased sample, the semi-attached figure that answers a different question than the one implied, the correlation mistaken for causation — showing that most statistical deception doesn't require outright lying, just selective presentation of technically true numbers.

The book matters because numerical claims carry an aura of objectivity that makes people less skeptical of them than of plain assertions, even though numbers can be arranged, cherry-picked, and framed just as persuasively as words. Huff's larger point is that basic statistical literacy — asking who collected this data, how, and what's being left out — is a practical form of self-defense in a world saturated with numerical claims from advertisers, politicians, and media.

Who should read it

Anyone who regularly encounters statistics in the news, in advertising, or at work, and wants a quick, practical toolkit for spotting when numbers are being used to mislead rather than inform, will find this short book immediately useful.

About the author

Darrell Huff was an American writer and journalist, not a professional statistician, who wrote this bestselling book as an accessible, skeptical guide for general readers rather than a technical statistics text.