It’s no surprise that Nate Silver would write a book about his recent triumphs; after all, he has earned household-name recognition as a statistician and blogger. He is best known for the impressive feat of forecasting President Barack Obama’s re-election in 2012, correctly predicting the winner of all 50 states and the District of Columbia, as well as the outcome of U.S. Senate races in 31 out of 33 states.
But instead of penning an autobiography trumpeting his accuracy, Silver wrote the best-selling “The Signal and the Noise: Why So Many Predictions Fail – But Some Don’t” (Penguin Press, 2012), named Amazon’s No. 1 Best Nonfiction Book of 2012.
Despite its easy-to-read style, “The Signal and the Noise” is not an easy read. Silver’s book is a serious look at why forecasts in numerous fields have been accurate, off the mark or, in some cases, downright disastrous. He offers advice for forecasters hoping to improve their results.
While the book may not scream best-seller, it does provide much insight that consumers of predictions – and that includes pretty much all of us – can take to heart.
In the first half of the book, Silver describes existing forecasting systems employed in several fields: baseball (the genesis of his interest in the field), weather, earthquakes, financial markets and infectious diseases. He pulls no punches. Having reviewed and studied these various models, most of which are quite complex, Silver concludes that some are good (e.g., weather forecasting), some are quite poor (e.g., earthquake forecasting – for good reasons) and some are rife with error, misstatements and inaccurate representations (economists, repent!).
Silver spends the second half of the book evaluating possible solutions to current problems in the field of forecasting and describing how and why Bayes’ Theorem, developed by the British minister in the 1700s, has such huge implications for the discipline.
Most of us are not in the daily business of making and publishing predictions, as Silver is, so why should we care about his recommendations? Because all of us read, ponder and make decisions based on everyday predictions. Some consequences, like not taking an umbrella due to a poor weather forecast, are not very serious. Others, like buying a house in an economy that we are told is not in danger of suffering a recession, can be quite disastrous.
We can be better and more intelligent makers and consumers of predictions, according to Silver: “The Bayesian approach toward thinking about probability is more compatible with decision making under high uncertainty. It encourages us to hold a large number of hypotheses in our head at once, to think about them probabilistically and to update them frequently when we come across new information that might be more or less consistent with them.”
As consumers, we must realize the limitations of all predictive models. Although Silver aims his advice primarily at those who are making the predictions, I think he would agree that consumers should recognize their own biases and apply a combination of curiosity and skepticism when evaluating any prediction.
“The Signal and the Noise” is an interesting, topical book, perhaps one that members of a book club that undertakes serious nonfiction would enjoy discussing.
Leslie Ashmore is a longtime Mountain View resident who belongs to two book clubs.