ELECTIONS with Nate Silver

Is it possible to predict the future? What is “data journalism”? Have polls become increasingly unreliable? To answer these questions, Pedro Pinto interviews Nate Silver in this episode of “It’s Not That Simple”, a podcast by the Francisco Manuel dos Santos Foundation. A statistician famous for correctly predicting the results of the 2008 US presidential election in 49 of the 50 states of the Union, Nate Silver is the founder of the website FiveThirtyEight.com, which uses “data journalism” to cover areas such as politics, science, health, and sports. Silver started out as a baseball statistical analyst (having written several books on the topic), before turning to politics and polls. In 2012, he published the book The Signal and The Noise: The Art and Science of Prediction, and has written for publications such as The New York Times, ESPN, Sports Illustrated and Slate. In 2009, he was named one of Time Magazine's 100 Most Influential People. In this episode, Silver explains how his election prediction model works, as well as what probabilistic thinking is and the difference between correctly predicting who wins an election and reliably calibrating the probability that a given candidate will win. Silver also discusses the value and usefulness of polls, as well as their limitations. Finally, Silver analyses the complexity of the American electoral system and the obstacles that this (and the relatively low frequency in which elections take place) poses to the accuracy of electoral forecasts, in a conversation well worth listening to. More on this topic • The Signal and The Noise: The Art and Science of Prediction, Nate Silver, 2012 • New York Magazine article on Nate Silver, 2008 • Newsweek article on Nate Silver, 2008 • Article about Nate Silver in The Atlantic, 2020 • Conference “How Far Can We See the Future?”, with Philip E. Tetlock at the Francisco Manuel dos Santos Foundation Other references in Portuguese • Essay of the Foundation “Sondagens, Eleições e Opinião Pública”, by Pedro Magalhães