Everything Is Predictable: How Bayesian Statistics Explain Our World

This beguiling mathematical romp from science writer Chivers (The Rationalist’s Guide to the Galaxy) surveys the far-reaching applications of the statistics theorem elaborated by the 18th-century English minister Thomas Bayes, who showed how to estimate the probability that a hypothesis is true by considering new data alongside “prior” assessments of the hypothesis’s accuracy. (For instance, the theorem might determine the probability that a middle-aged woman has Covid by considering a positive test result alongside the virus’s prevalence rate among middle-aged women generally.) Bayes’s theorem produces startling insights that can upend conventional wisdom, Chivers writes, noting that the equation explains why “a cancer test can be 99 percent accurate even though 99 percent of the people it says have cancer don’t.” Examining the theorem in a raft of offbeat contexts, the author suggests its focus on evaluating new information in the context of previous beliefs sheds light on why vaccine skeptics are unmoved by evidence demonstrating vaccines’ safety and efficacy, and why contestants guessing which door hides a prize on Let’s Make a Deal should always switch their pick after the host reveals one of the losing doors. Chivers’s dive into probability theory is heady but lucid, and conveys arcane concepts in commonsensical prose. The result is a stimulating take on making sense of a murky, uncertain reality. Photos. Agent: Melissa Flashman, Janklow & Nesbit Assoc. (May)