Nate Silver, data guru returns to sport

PERSON IN THE NEWS

July 26, 2013 7:28 pm

FT

By Andrew Edgecliffe-Johnson

The political forecaster shows the new power of one-man brands, says Andrew Edgecliffe-Johnson

Early in the 2008 US presidential campaign, a blogger on the Daily Kos website started attracting attention. Writing under the pseudonym “Poblano”, he calculated that Hillary Clinton was not faring nearly as well in the primaries as pundits said. One by one, his forecasts embarrassed the pollsters and political reporters.

After months of finding flaws in others’ poll interpretations, Nate Silver outed himself. “You can’t name yourself after a chilli pepper and do much media stuff,” he said. His path from anonymity to stardom began with FiveThirtyEight, the politics site he created that year – named after the number of votes in the US electoral college. This week it culminated in a deal to move the blog from the New York Times, which has licensed it since 2010, to ESPN, Walt Disney’s sports network.

True to form, the 35-year-old made the decision only after analysing several variables: who could develop FiveThirtyEight into something more ambitious, and who offered editorial freedom, prestige, fun and money. The NYT tried to keep him but no newspaper is a match for a sports powerhouse with $9bn of annual revenues.

The Grey Lady was once a media pinnacle you left only in a box, and Mr Silver’s departure shocked many journalists. But it illustrated a growing tension between media’s traditional gatekeepers and the journalistic one-man brands who draw large audiences but have interests – and income – outside their editorial day job.

Like a star player in sport or on a trading floor, such signing can be hugely valuable but it can also cause friction. Individuals who grab so much of the limelight threaten institutions’ sense of control, wrote Andrew Sullivan, a friend of Mr Silver who took his blog from Newsweek Daily Beast to its own site. “The NYT needed Nate much more than he needed them.”

Mr Silver is an outsider and a late-comer to political journalism. He was struck by how vapid the reporting was while studying a 2006 Congressional attempt to ban internet poker, then an important source of his income. As the 2008 vote approached, he saw that most correspondents paid disproportionate attention to outlying national polls and not enough to swing states. This was built into a theory, articulated in his 2012 book The Signal and the Noise, that when we are swamped by information we block out the parts we do not like.

His model – 2,000 lines of computer code – predicted 49 of the 50 state races in 2008. Four years later, he got all 50 right and Gawker, a blog run by his friend Nick Denton, crowned him America’s chief wizard.

The statistical sorcerer’s move to the Mickey Mouse empire is a return to his roots as a sabermetrician – a cruncher of baseball data. As a boy growing up in Lansing, Michigan, the son of a political-science professor father and community activist mother supported the Detroit Tigers. His father would relax by analysing stadium attendances. By 14, Nate was building models to try to predict George HW Bush’s 1992 electoral fight with Ross Perot.

After studying economics at the University of Chicago, Mr Silver became a KPMG consultant. When the dry work on intra-company transfer pricing dragged, he began trying to predict ball players’ performance. His “Pecota” model earned him a following in baseball before he adapted his blend of sophisticated modelling and clear narratives to politics – a sport often likened to a horse race. But it gave rise to a popular critique, he says: “Nate’s a good baseball guy, but what does he know about politics?”

Some critics have been harsher. Shortly before the 2012 Obama-Romney election, Joe Scarborough, a former Republican congressman who hosts MSNBC’s Morning Joe show, called him a joke and an ideologue for thinking that the race was “anything but a toss-up”. Mr Silver, who has called conventional pundits “fundamentally useless”, bet Mr Scarborough $2000 that he would be proved right. Mr Scarborough paid up, sending a cheque to charity, but the NYT ombudswoman scolded Mr Silver, saying that by joining its newsroom he had lost “the right to act like a free agent”.

In his spectacles and brown suits, the nervy-eyed Mr Silver looks the part of the geek. But he likes hitting back, accusing dismissive editors at Politico, a Washington news site, of lacking curiosity for “the world outside of the bubble”, and shaming TV prognosticators who “may as well have been flipping coins”.

Mr Silver faced suspicion even from some NYT colleagues. “I don’t think Nate Silver ever really fit into the Times culture,” Margaret Sullivan, its public editor, wrote this week. “He was, in a word, disruptive.” Mr Silver smiles at this, saying: “We look at the word ‘disruptive’ as a positive.”

Yet he has jumped to another big media group, suggesting brands such as the NYT or ESPN still appeal to disruptive stars. Mr Silver notes that other big names found homes in large companies, such as Wonkbook, Ezra Klein’s policy blog for the Washington Post, and Grantland, Bill Simmons’ irreverent ESPN sports site.

The Signal and the Noise showed Mr Silver’s range, taking in subjects from weather forecasting to the global financial crisis. He has been courted by Hollywood and Wall Street , eager to source his wizardry, but he wants a larger audience.

At ESPN he will roam well beyond politics. “To some extent the point’s been proven in politics,” he notes. After a run of accurate election predictions, he adds, “to be frank [there’s] more downside risk than upside risk.”

Mr Silver plans to apply his model to subjects as diverse as education, the Oscars and economics. “We might start out by trying to forecast US payroll numbers,” he says, teasing that he is already working on creating economic indices. Some would like to see him go further. Mr Klein tweeted this week: “Nate Silver for Fed chairman”.

The writer is the FT’s media editor

Share this post