Barack Obama may have been officially announced as the winner of Tuesday night’s presidential election, but a statistical model could be crowned the victor in certain circles.
Nate Silver, a political blogger for The New York Times, runs the website fivethirtyeight.com, which accurately predicted 100 percent of the 50 state results on Tuesday night. His site is aptly named for the 538 electoral votes up for grabs to win the presidential election.
What sets Silver apart from most political junkies is that he does not just look at one poll and make projections, he looks at all of the polls.
Using a statistical algorithm, Silver inputs the data from every political poll conducted along the course of the campaign and lets the numbers speak for themselves. And if Tuesday night’s performance was any indication, the numbers do not lie.
Before throwing his hat in the political ring, Silver used his statistical prowess to create a mathematical model used to evaluate major league baseball players, called PECOTA.
Think of Brad Pitt playing Billy Beane in Moneyball – a movie that chronicles, and dramatizes, the Oakland Athletics use of a theory called sabermetrics to draft baseball players. Based off of the acronym SABR (Society for American Baseball Research), the system uses objective evidence, such as player statistics, to measure in-game activity.
Sabermetrics seeks to answer the question “who contributed the most to the Atlanta Braves this season?” or in the case of Moneyball, “which players should we draft to create a team that will take us to the World Series?”
This theory provided the basis for Silver’s PECOTA system. Fast forward to 2008. Silver’s new model, now trained to interpret political polls, accurately predicted 48 out of 50 states in the presidential election and four years later predicted 100 percent correctly.
Now, that's a pretty good batting average.