Advanced Stats Won Obama Re-Election, Can Work for Baseball Traditionalists Too!

Dan LevyNational Lead WriterNovember 7, 2012

Nate Silver is the second-happiest man in America today. Silver is the brains behind the FiveThirtyEight blog, part of the New York Times network, who successfully called the race for president of the United States in favor of Barack Obama for the second time in as many elections.

In 2008, Silver correctly projected all but one state during Obama's historic victory, yet still people maligned his methods during this election cycle, questioning his numbers and mocking his assuredness that Obama would win again. 

Too many political pundits on TV and radio and even the Internet thought the race was too close to call between Obama and Mitt Romney. They were either looking at the wrong numbers or looking at the numbers wrong.

Silver had Obama winning re-election at a statistical probability of more than 3-to-1. The only thing Silver got wrong, based on FiveThirtyEight's latest pre-election projections, was that he might not have called the Obama victory over Romney a big enough blowout in the Electoral College. Even Silver may have thought it was going to be closer than it ended up being.

So what does this have to do with baseball?

Before becoming this nation's most spot-on political soothsayer, Silver was a leader in the Society for American Baseball Research (SABR) community, imploring traditional baseball pundits—and fans—to see a host of more advanced (read: better) ways to look at statistics that help determine which player or team has a better chance of winning.

In simpler terms, numbers are good, if you're looking at the best numbers. 

Whether it's politics or sports, numbers don't change, but the way we look at numbers should never stop evolving. Silver projected a relative landslide victory for Obama because he looked in the right places on the map.

Heck, I don't know the first thing about political tracking—what's the voting equivalent to xFIP, anyway?—but it wasn't hard to see that Obama would have had to lose six of the eight battleground states for Romney to win the election. If Obama were to retain Florida, Romney needed to win seven of the remaining seven battleground states.

How was that 50/50, like many national pundits were suggesting? The TV pundits who needed this election to stay close so their ratings would benefit all the way into November were calling this race a dead heat. So were they lying, or were they using the old way of looking at the numbers?

Maybe it was a little of both. Or maybe they were afraid of putting their reputation on the line for a bunch of statistics they can't understand. (Sound familiar, baseball fans?) The bottom line is that whether it's politics or sports, everyone pays more attention to a tie game. Nobody likes tuning in for a blowout, so the national numbers telling pundits it was a dead heat made them happy to report it was exactly that, even if it wasn't.

The election proved to be a bit of a blowout—a four-game sweep, if you will—and Silver called it.

It's important to remember that Silver wasn't right because of his gut. He was right because he looked at the evolution of the numbers. He paid attention to the numbers that mattered when his detractors were worried about the polls measuring popular vote.

As a friend suggested before the election, using the popular vote to determine how close the presidential race will be is like using RBI to determine which power hitter is more productive at the plate. Statistics have evolved, and the traditionalists need to evolve, too. 

During NBC's wall-to-wall coverage of the election, Chuck Todd had been preaching that swing states may come down to specific counties or even specific municipalities. On election night, Todd explained in specific detail how the Obama campaign targeted these specific locations to get people out to vote, buoying his support in crucial areas that could have leaned either way.

Todd called the Obama campaign the Moneyball model of winning an election. Todd told viewers that Obama advertised on the Big Ten Network on Saturdays nearly two months in advance of Romney to find any advantage to get his message out to the specific swing voters—in that case, working-class white males—to help him win key battleground states like Ohio, Wisconsin and Iowa.

Early in the night, the Obama campaign saw the returns from those specific regions his campaign had targeted were falling their way, and they knew, far before the American people knew, re-election was all but certain.

David Plouffe ran the Obama campaign to re-election because his team figured out a way to break the statistics down to the square mile. Todd illustrated that the Democrats knew which municipalities had gained more voters in the last four years and the likelihood those new voters could help him retain—or retake—the areas, a snowball effect that could help him take state after state that "from the gut" pundits thought would be too close to call.

Mike Trout of the Los Angeles Angels needed Plouffe to run his American League MVP campaign.

The problem in baseball—in all sports, really—is that generations of historical precedent have made many traditionalists see the game a certain way, and they are too old or crotchety or intolerant of change to admit there's a better way than what they've always known.

Political pundits were just as old-minded and pig-headed as some sportswriters (and fans) this election cycle, refusing to look at the numbers in the way that would give them the most likely outcome. Even Silver admitted that Romney could have won the election, but that outcome was statistically improbable. 

A team can come from behind in a five-game series to win three-straight games on the road en route to a second World Series in three years, but it still makes it statistically improbable. 

That, as the saying goes, is why they play the game. And it's why there are so many people hanging on every chad, er vote, on election night.

In baseball, more and more teams are relying on advanced metrics to help determine everything from which players to sign to mega-contracts to which reliever to put in the game with a runner in scoring position and less than two outs.

If the information exists, why not use it to your advantage? That's what the other side is doing.

This model isn't new, by the way. My friend and GOP political strategist Phillip Stutts worked on the Get out the Vote campaign for the 2004 Bush-Cheney re-election, and he said they were required to read the book Moneyball, basing the model for the campaign after that.

Stutts admits the Obama campaign took it to another level this cycle, but the model is not new. Smart behind-the-scenes people have been using this hyper-localized, in-depth statistical model for years in politics, just like in sports, and still there are people who publicly deride those who use it.

There's nothing to deride anymore. When Peter Gammons is re-tweeting political writers praising Silver's "50 for 50" state projections, it's time to admit the advanced statistical way of looking at something can—and does—give you a 100 percent more informed, and therefore better, chance of making the right prediction than simply using your gut.

One Hundred Percent. Feel free to check the math on that one. I'll be trying to figure out if the Phillies can hire Plouffe as a special assistant to the general manager. After all, the guy is a huge Phillies fan

Unless Silver is available…

The evidence is pretty clear that there is a different way to build a champion. More and more teams are looking at redefining the winning model by using advanced statistics and numbers-driven observations they had never imagined a few decades ago. 

If it has worked in baseball, it has worked for politics, too. Now, if Obama and Plouffe and Silver are showing everyone how it works, not only to project an election but also to win one, maybe the hangers-on to the old way of doing things—in both politics and in sports—will start to realize there's a better way than they have been willing to acknowledge.

Now that the election is over, the only thing left to decide is who will play Silver and who will play Plouffe in the movie about this election. Someone call Brad Pitt right now.