The Doomsday Calculation

April 8, 2019
Poundstone (How to Predict the Unpredictable), who studied physics at MIT, provides an intriguing, if less than revelatory, look at Bayes’ theorem as a useful way of predicting the probability of future events. Poundstone explains the theorem, the creation of 18th-century mathematician and clergyman Thomas Bayes, as a way of “assigning a probability to something that has never happened” and applies it to a host of questions, ranging from the mundane (how long will one’s relationships last?) to the cosmic (are there other universes?) and the existential (are humans inhabitants of another civilization’s digital world?). He also applies it to his central question: when will civilization end? His litany of ways the world might end is impressively varied and creative, and includes the human race being rendered sterile by mutated salmonella, and an errant experiment at the CERN supercollider creating a quantum condition that destroys not just life on earth but the entire universe as well. Readers concerned with the big questions Poundstone explores will find much of interest in this enjoyable mathematics primer, even if they are likely to remain unconvinced the equation is as intellectually transformative as he claims. Agent: John Brockman, Brockman Inc.

Starred review from April 15, 2019
An amusing, entertaining effort to answer the unanswerable. In his latest thought-provoking journey, journalist Poundstone (Head in the Cloud: Why Knowing Things Still Matters When Facts Are so Easy to Look up, 2016, etc.) writes about how modern scientists have applied the mathematical theorem of obscure 18th-century British clergyman Thomas Bayes to a host of important questions and come up with unsettling answers. "By applying Thomas Bayes' rules to the technique of self-sampling," writes the author, "we can address cosmic mysteries. Was life on Earth probable or a rare accident? Why don't we see any evidence of extraterrestrials? Is the world we see real or a simulation? Is the universe we observe all there is?" There is perhaps less than meets the eye because the results are always in probabilities, but many defy common sense, and the enthusiastic Poundstone delivers a steady stream of delicious jolts. Bayes' theorem finds the probability of something if one knows other probabilities. Sound boring? Here's an example: A woman learns that her mammogram is positive. What are the odds that she has breast cancer? The known probabilities are that 1 percent of women have breast cancer and that mammograms are 90 percent accurate. The startling answer: The odds are only about 1 in 9. Since the test is only 90 percent accurate, 10 percent of the 99 healthy women will also be positive. Poundstone turns up brilliant scientists who have applied Bayes' theorem to unanswerable questions and then delivered answers. One determined that, with a 95 percent confidence level, the human race will survive at least 5,100 but not more than 7.8 million years. Another charming example of the Malcolm Gladwell school of writing: making an implausible statement and then producing evidence that it's true--maybe. Poundstone's examples mix statistics and serious philosophical arguments, and readers who pay close attention will be rewarded.

COPYRIGHT(2019) Kirkus Reviews, ALL RIGHTS RESERVED.