Book Review: How to Lie with Statistics, by Darrell Huff

If you can’t prove what you want to prove, demonstrate something else and pretend that they are the same thing. In the daze that follows the collision of statistics with the human mind, hardly anybody will notice the difference. Darrell Huff

Who would believe that Amazon’s current #1 Best Seller in Business Statistics is a slim volume, written in 1954 by Darrell Huff, who was not a statistician, but a prolific author who once was the managing editor of Better Homes and Gardens.


Huff wrote many books, such as How to Take a Chance, Score: The Strategy of Taking Tests, How to Figure the Odds on Everything, and The Complete How to Figure It: Using Math in Everyday Life. His expansive authorship also included books on home buying, home workshop projects, working with concrete and masonry, home improvement, and personal finances.


In How to Lie with Statistics, Huff exposes statistical “tricks” that can be used to deceive a reader. Why would Huff’s almost 70-year-old book on statistics remain a bestseller?

J. Michael Steele sums up the success of the book in his paper “Darrell Huff and Fifty Years of How to Lie with Statistics” (2005). He cites four key elements: the title, the illustrations – and illustrator, the style, and the content.


Huff drew on his experience as a magazine editor and writer by developing a catchy title that prevented the volume from being relegated to the dustbin of history. Steele observes that if the book had been titled An Introduction to Statistics, it would likely have sold a “few hundred copies for a year or two.”


The book is filled with cartoons and graphics designed by Irving Geis, giving it a 1950s-ish, light-hearted feel. The illustrations, on almost every other page, break up the text and illustrate the author’s key ideas.


When it comes to Huff’s writing, Steele notes, “In a word, Huff’s style was – breezy. A statistically trained reader may even find it to be breezy to a fault, but such a person never was part of Huff’s intended audience.”

Most importantly, Huff delivers the content. He introduces basic statistics, data visualization, cause and effect, and critical thinking … in a mere 144 pages. Writing clearly and concisely, he avoids sophisticated math while providing illuminating examples.


Reading How to Lie with Statistics today may cause conflicting reactions. On the one hand, the concepts in the book are for the most part still relevant. On the other hand, the examples are dated, and while interesting, are difficult to relate to.


Huff introduces the concepts of mean, median, and mode, explaining that the choice of a measure greatly affects its perception. He skillfully uses the example of significant differences in values regarding corporate salaries. In his example, the mode salary is quite low, as lower-paid workers are most common in the organization. The median salary is slightly higher, but the mean salary is significantly higher because the company owner has a much larger salary. Huff’s message is that the choice of a specific measure can be used to support a writer’s point of view and influence the reader’s perception.


The examples, specific to the time of writing, are less relevant today. In the mean, median, and mode example, the mode annual salary is $2,000, the median yearly salary is $3,000, and the mean yearly salary is $5,700. Not only are the salaries different from today, but the examples reflect a 1950s, white, male, North American perspective that may at times be jarring and offensive to the modern reader.


Huff’s insightful discussion of graphics could be part of any modern data visualization discourse. Using the example of “The Crescive Cow,” Huff shows how an attempt to show the increase in milk cows can go awry. The figure is intended to show a doubling of the cattle population, but the illustrator doubled both the height and width of the 1860 cow, resulting in the 1936 cow being four times as large, creating the erroneous impression for the reader that the population has quadrupled. This kind of mistake is easily made by an inexperienced designer or an intentionally deceptive one.


The Crescive Cow. On the outside chance you are not familiar with the word "crescive," according to Merriam-Webster, it is an adjective meaning “marked by gradual spontaneous development.”


Huff’s analysis is especially interesting: “But the effect on the hasty scanner [reader] of the page may be even stranger: He may easily come away with the idea that cows are bigger now than they used to be.” This observation foreshadows current research on the unintended interpretations of visualizations, such as misunderstandings of the meaning of the hurricane cone of uncertainty.

Huff’s final chapter, “How to Talk Back to a Statistic,” provides common-sense advice for critical thinking. When facing a statistic, he says that readers should ask the questions:

  • Who Says So?

  • How Does He Know?

  • What’s Missing?