How to Lie with Statistics is one of those books that shatter the way you read information. You will never look at statistics the same way again. People have an inherent trust in statistics; this book will show them to be fools. Any number you need to back up your claims can be tweaked and proven using statistics. This book is the Dark Arts recipe book on statistics. If you want to spin, obscure, prove, confuse, or mislead your reader’s attention using statistics, this book will show you how.
This is not to say that you should use this book to obscure the truth. Far from it. Darrel Huff intended this book to be a guide for the everyman, who often falls prey to the spinsters, scam artists, and public relations experts that rely on the dubious nature of numbers and statistics. He exceeds his mission. You will learn more in an afternoon reading this book (it is an incredibly short and easy read) than you can learn in an entire semester’s worth of college statistics. (I know, I’ve taken the course.) There are three lessons that you really need to understand after reading this book; they aren’t lying to you, who benefits, and to be careful out there.
When I say that these scam artists aren’t lying to you. They aren’t. But they aren’t exactly telling you the truth either. Good numbers and statistical processes back up every fact they say. Oftentimes though, the numbers don’t mean a damn thing. Or don’t mean what you think they mean. “Statistics are used much like a drunk uses a lamppost, for support rather than illumination”, said Vin Scully. So every time you read a statistic really ask yourself what it is saying, and if it really helps the argument. Look at all the different ways of coming to a number, and decide, most often, to throw the number out. There are three different ways to calculate averages, and if the study or figure doesn’t tell you which one they are using, they are misleading you on purpose. Because they benefit.
Just like in courtrooms and politics it is often asked the immortal Latin phrase “Cui Bono.” Meaning who benefits?. If a company throws a statistic at you that claim their product is the best, well that number isn’t exactly the number to go on. Sure the numbers can back up the company, and they might be telling the truth with perfect statistical measures. But when errors of statistics are commonly skewed in the writer’s rather than the consumer’s favor, it is difficult not to call bullshit. So heed Huff’s warning.
It is a scary world of information out there. With the Internet age upon us, we are inundated with even more statistical information than Darrel Huff could have ever imagined. Remember to keep an open mind, with a skeptical bent on the truth. Just because someone uses statistics to prove themselves right, you don’t and probably shouldn’t listen to them. Armed with How to Lie with Statistics and a cynical mind, you too can walk in relative peace through the valley of dubious claims. Good luck out there, and remember, you are smarter than they think.
“There are lies, damned lies, and statistics.” – Mark Twain