Nate Silver began his career analyzing baseball statistics. In the 2008 Presidential Election, he applied his skill set to politics and predicted the election outcome with uncanny accuracy. Now working with The New York Times, Tuesday will be a day of reckoning for Silver and his FiveThirtyEight blog. If his predictions fare as they did in 2008, his work may change the landscape of political journalism.
 |
| Come Wednesday morning, all eyes will be on Nate Silver and the election predictions published in his FiveThirtyEight blog, licensed by The New York Times. (Photo Credit: Melissa Ann Pinney, nymag.com) |
With less than 48 hours until the 2012 Presidential Election on November 6, magazines are rolling out cover stories on the choice Americans will face on Tuesday. Depending on where one lands on the political spectrum, Wednesday morning could bring anything from a dark episode of oligarchy and Gilded Age inspired government policies under Mr. Romney to a reinvigorated economy and a return to prosperity; from gun snatching and oppression of religious freedom under President Obama to staying the course on a path of hope and recovery. Regardless of which candidate wins Tuesday’s election, many political insiders and lay readers alike will be checking one website: Nate Silver’s FiveThirtyEight blog, a site licensed by The New York Times and centered on an election forecasting model. Friends and foes of Silver will gather around computer screens, click refresh and look on with glee or grief as the final tallies appear. How close will Nate Silver’s predictions come to the actual outcome? Many have argued that America’s future is at stake this November. For Silver, his career may hang in the balance.
To be accurate, FiveThirtyEight does not make predictions as much as it assigns the probability, in the form of percentages, each candidate has of winning. In the case of the 2012 Presidential Election, Nate Silver’s model focuses on the chances both candidates have in winning the Electoral College, the vote counting system that is used to determine the outcome of the presidential election.
Nate Silver was not always in the business of politics. He began his statistical career in baseball, where he created PECOTA, a system which is used to forecast the performance of MLB players. In 2008 he migrated to politics and founded FiveThirtyEight.com. The results were stunning: according to Forbes, his model accurately predicted the results of the 2008 Presidential election for 49 out of 50 states plus Washington D.C., came within less than 1% of the actual popular vote and successfully forecasted the winner in every Senate race. In 2010, The New York Times licensed FiveThirtyEight and Nate Silver’s articles regularly appear on the newspaper’s website as well as in its print edition. His forecasts do not ring of partisan damage control: the one state his model failed to call accurately in 2008, Indiana, went to Barack Obama. Earlier this year, when Democrats were claiming momentum in their effort to recall Wisconsin Governor Scott Walker, Silver’s model repeatedly showed the opposite was true. Governor Walker would go on to survive the recall effort.
While spurning the title of wizard, Silver does have his secrets. FiveThirtyEight is built around a proprietary model and the algorithm Silver uses to make his forecasts, which some pundits have called his “secret sauce”, is a closely guarded secret. What is known is that the FiveThirtyEight model aggregates both national polls and state polls, provides for non-poll related aspects of the race like economics with an “economic index”, measures polling reliability by examining how accurate polling has been compared to “real-world conditions” since 1968 and has strict criteria for which polls get more weight in the model and why. For example, a poll that consistently leans Democrat or Republican will get less weight in the model than a poll that has produced more even results. Polls that deviate from the norm get less weight in the model as well. Newer polls replace older polls in the model and all polls used must conform to industry standards. The strength of the FiveThirtyEight model is that it draws upon numerous state and national polls rather than just relying upon one or two national tracking polls, as most news outlets do when reporting on the state of the race. This allows for strength of consensus.
During Silver’s journey from baseball to The New York Times, he acquired professional respect and notoriety for his objective and thorough approach to political statistics and showed a great propensity for holding partisan brush strokes at bay in a line of work that is often colored by ideological fervor. Silver has been a featured speaker at TED and SXSW; he has appeared as a commentator on CNN and MSNBC and has published articles in The New York Times Magazine. In 2009, he was listed in Time magazine’s 100 Most Influential People in the World and this year he was featured in the magazine’s Time 100 Roundtable discussion. Silver is currently promoting his book, The Signal and the Noise, which is being published by Penguin and is slated to speak November 9th at the Chicago Humanities Festival as part of the annual Karla Scherer Endowed Lecture Series for the University of Chicago.
Silver (right) was a keynote speaker at SXSW in 2009. He is pictured here being interviewed on stage by Business Week writer Stephen Baker (Photo Credit: Daniel Terdiman/CNET)
As in other professions, with visibility comes criticism and in the midst of a hotly contested election cycle, Silver’s criticism has come in family size servings. Most notably, Joe Scarborough, host of the popular MSNBC morning talk show “Morning Joe”, reviled Nate Silver on air, saying, “Anybody who thinks that this race is anything but a tossup right now is such an ideologue, they should be kept away from typewriters, computers, laptops and microphones for the next 10 days because they’re jokes.”
Scarborough, who was a Republican congressman before joining MSNBC, was taking offense with the probability of an Electoral College victory FiveThirtyEight had been assigning to President Obama, currently at 85.5%. Before Gallup suspended their national tracking poll on October 29th, citing Hurricane Sandy and reasoning that the disaster had “compromised the ability of a national survey to provide a nationally representative assessment of the nation’s voting population,” the poll had President Obama and Mr. Romney tied at 48% among registered voters.
According to the Gallup national tracking poll, Scarborough’s statement seems plausible. But the FiveThirtyEight model does not make its forecasts based on the popular vote, rather it bases its numbers on the probability that either candidate has of attaining the 270 Electoral College votes necessary for a win. Because President Obama has maintained leads in states, that when combined, give him 270 votes or more, FiveThirtyEight has the president heavily favored—much to the ire of conservative pundits like Scarborough. Silver responded to Scarborough’s criticism by offering a friendly wager: if President Obama won reelection, Scarborough would donate 1,000$ to charity, likewise if Mr. Romney won, Silver would donate 1,000$. This offer, made on Twitter, caught the attention of Margaret Sullivan, public editor of The New York Times. As public editor, Sullivan’s job is to act as a liaison between the Times’ staff and its readers, refereeing conflicts and addressing controversies that arise within the newspaper in order to protect its integrity. Writing in the Public Editor’s Journal, Sullivan stated that “…the wager offer is a bad idea—giving ammunition to critics who want to paint Mr. Silver as a partisan who is trying to sway the outcome [of the election].” Sullivan also wrote that as a New York Times employee, Silver’s actions reflected upon the newspaper and that she believed his offer to wager was “inappropriate for a Times journalist, which is how Mr. Silver is seen by the public even though he’s not a regular staff member."
In a follow up post, Sullivan stated emphatically that she has observed no partisan bias in Silver’s work and while she believes there was no conflict of interest inherent in the offer to wager a bet, “…there is an appearance of one. And appearance matters—if affects credibility, which is at the heart of good journalism.” While reprimanded by the Times’ public editor, many came to Silver’s defense. As proof of his good standing in the professional community, the American Journalism Review, the Washington bureau chief for The New York Times, the Columbia Journalism Review, and many others, showed support for FiveThirtyEight and Nate Silver. Rem Rieder, AJR’s editor and senior vice president, responded to Sullivan’s reprimand by asking, “Why is backing up your prediction with cash any more partisan than making the prediction in the first place? This doesn’t suggest to me Silver is a true believer trying to sway an election. What it says is that he’s a poll maven with no shortage of belief in his own handiwork who is willing if not eager to stand up to a TV pundit calling him names.” David Leonhardt, the New York Times’ Washington bureau chief, praised Silver, writing in the Public Editor’s Journal, “There is no pollster, no political scientist and no other writer who has a better public record of analyzing elections data than Nate. It’s not perfect, as he himself tries to convey with his emphasis on uncertainty and humility. But it’s serious, impressive and nonpartisan, and the same is not true of many of his critics’ claims.”
Criticisms of FiveThrityEight’s projections are only partly rooted in partisan loyalty. It is a changing of the guard, a failure to nod at orthodoxy that the mainstream political pundit class fears and this fear forms the underbelly of Nate Silver’s blowback. “Who will win the presidential election next Tuesday,” asks Brendan Nyhan of the Columbia Journalism Review, “Until recently, the market for analysis of questions like these has been dominated by mainstream political reporters and commentators. Their style leans heavy on qualitative impressions and hazy narratives. But as the audience for quantitative analysis of politics has grown, the establishment analysts have become increasingly defensive about their status.”
Nate Silver’s approach provides this emerging information market with the quantitative data it seeks. When a poll is released, the cable networks discuss its results vigorously, dispatching their commentators to shed light on its meaning and the campaigns send in their surrogates to lay claim to the high ground of momentum. The political static spikes to unbearable levels. Meanwhile Silver examines the same poll, combines its results with other polls that were released the same day, and with arithmetic, arrives at an objective conclusion. This scares the establishment, notes Ezra Klein of the Washington Post. “Silver’s work,” writes Klein, “poses a threat to the more traditional—and, in particular, to more excitable—forms of political punditry and horse-race journalism.” As the President and Mr. Romney spar over who will better protect the middle class, FiveThirtyEight has become the epicenter in a class war that pits the vanguards of opinion against the vanguards of percentages.
Nate Silver’s approach provides this emerging information market with the quantitative data it seeks. When a poll is released, the cable networks discuss its results vigorously, dispatching their commentators to shed light on its meaning and the campaigns send in their surrogates to lay claim to the high ground of momentum. The political static spikes to unbearable levels. Meanwhile Silver examines the same poll, combines its results with other polls that were released the same day, and with arithmetic, arrives at an objective conclusion. This scares the establishment, notes Ezra Klein of the Washington Post. “Silver’s work,” writes Klein, “poses a threat to the more traditional—and, in particular, to more excitable—forms of political punditry and horse-race journalism.” As the President and Mr. Romney spar over who will better protect the middle class, FiveThirtyEight has become the epicenter in a class war that pits the vanguards of opinion against the vanguards of percentages.
According to Silver’s Saturday post, of the 21 swing state polls released yesterday, 16 showed President Obama in the lead, compared to 2 that showed Mr. Romney in the lead and 3 that showed a tie. In his most recent report, posted tonight at 9:23 PM EST, Silver reported that of the 12 national polls released in the last 24 hours, President Obama lead in all of them by an average of 1.3%. After incorporating this data into the FiveThirtyEight model, Silver is projecting that the incumbent has an 85.5% chance of winning the Electoral College on Tuesday. FiveThirtyEight is also projecting that President Obama will win 306.4 of the 538 Electoral College votes and 50.5% of the popular vote.
Fair or not, in less than 48 hours, Nate Silver will have his reckoning. While he does not have as much at stake as the candidates come Tuesday night, he acknowledges that he is feeling the pressure of possibly having his reputation evaporate Wednesday morning. “I’m sure that I have a lot riding on the outcome,” Silver told BuzzFeed recently, “I’m also sure I’ll get too much credit if the prediction is right and too much blame if it is wrong.”
Isaac Matson can be reached for comment at isaac.matson@outlook.com
0 Comments