The Obama Election: Analytics Makes the Call

Nate Silver, the New York Times‘ FiveThirtyEight blogger, used a database of polling statistics to accurately predict the winner of all 50 states on the night of the U.S. election.

Image courtesy of JD Lasica, Socialmedia.biz.

Even though President Barack Obama won the closely called race against Mitt Romney, it seems like the real star in last week’s presidential election is Nate Silver.

Silver, the New York Times’ FiveThirtyEight blogger, utilized a tremendous database of polling data, both historic and current, to accurately predict the winner of all 50 states on the night of the election. He also gave Obama a 91% chance of winning, going against the media tide calling for a very close race (or a Romney win). And it wasn’t a fluke. In 2008, Silver correctly predicted the winner of 49 out of 50 states.

What’s the big deal? His results (and others like his) could have ripple effects on the way political campaigns are covered — and more importantly, run — in the future. According to a Huffington Post blog:

By disregarding the industry wisdom that gut feelings, day-to-day poll movements, and talking head commentary are what matter, [Silver] has decimated the value of these superficial judgments.

But it’s not just the baseball-statistician-turned-political-analyst Silver who is changing the way elections are conducted. The Obama election campaign itself has used data analytics to understand voter behavior — and adjust strategy accordingly.

In a Time article published on Nov. 7, Obama campaign manager Jim Messina explained his tactics during the run up to the 2008 and 2012 elections: “Measure every single thing in the campaign.”

While much was written about the campaign’s use of data analytics, little was known about the actual details of that work until a group of senior campaign advisors agreed to share details with Time — as long as the information wasn’t published until after the new president was announced. According to the Nov.