A statistician steps into the spotlight
BY NEIL STEINBERG firstname.lastname@example.org November 8, 2012 5:40PM
Nate Silver | Wikipedia photo
Updated: December 10, 2012 6:20AM
As satisfying as a Barack Obama victory is, skating past jeering mobs who held him in contempt — I’m looking forward to someone asking Mitt Romney how it feels to lose to the WORST PRESIDENT IN HISTORY — there’s a second God bless America, bad-guys-end-up-on-a-skewer-while-the-credits-roll tale of triumph in the face of derision that may be even more delicious than Obama’s: the Nate Silver saga.
Silver, a plucky blogger (his blog, “FiveThirtyEight” is named for the number of electoral votes) had his breakout moment Tuesday by precisely predicting the outcome of the 2012 race, calling each of 50 states. He did it not by magic (though one satiric headline wondered if he is a witch) but by math.
A graduate of the University of Chicago — with a B.A. in economics — Silver, 34, first drew notice analyzing baseball stats, with an amazingly accurate model for predicting outcomes, the Player Empirical Comparison and Optimization Test Algorithm, or PECOTA.
In 2008, he tried his hand at politics, starting up his blog and calling the winner in 49 of the 50 states. His blog was picked up by the New York Times. The trouble began this year when Silver put his chips on Obama early and left them there, while the pundits pontificated about horse races and conjured an electorate whose opinions seemed to shift with the breeze. Silver, on the other hand, kept upping the odds of Obama winning, topping a 90 percent chance of victory by election day, predicting that Obama would take 313 electoral votes to Romney’s 225.
This did not go down well in Fox TVland. Republicans, who are stuck in a particularly fraught relationship with facts as it is, reacted with lordly disdain. MSNBC talk show host Joe Scarborough quivered with disgust while dismissing Silver. “Both sides understand it is close and it could go either way,” he pronounced. “And anybody that thinks that this race is anything but a tossup right now is such an ideologue . . . they’re jokes.”
Other right-wing pundits zeroed in on the obvious flaw in Silver’s analysis — he’s gay — one indicting him for being “thin” and “effeminate” (the non-logic at work seems to be that besides having marriages that infect the institution for everyone else, gays also must have trouble with numbers, a slur of particularly staggering ignorance and idiocy — igniocy? — since probably the greatest computer analyst of all, Alan Turing, was gay, and was also persecuted by the torch-wavers of his time).
In the days before the election, nervous Obama supporters clutched at Silver’s analysis like a blankie. “KEEP CALM AND TRUST NATE SILVER” placards were passed around Facebook. Conservatives meanwhile put up a “Nate Silver is Wrong” website, heaping contempt upon him.“This website is dedicated to exposing a newly created media God by the name of ‘Nate Silver,’” it snidely announced.
When the election results lined up exactly with Silver’s analysis, some worried that the practice of paying oracles to reach into their bag of biases and confusion and make predictions about future elections would fall by the wayside, replaced by computer models.
“Has Nate Silver destroyed punditry?” the Christian Science Monitor asked, noting that conservative lion George Will predicted Romney would win with 321 electoral votes.
After the election, Silver wisely chose not to leap into the mosh pit of the media, ignoring most requests for interviews from me and everyone else, apparently, even though he has a new book to flog, The Signal and the Noise: Why So Many Predictions Fail — but Some Don’t.
So hats off to the Chicago Humanities Festival, for scoring one of the better “gets” of the year, when the elusive Silver returns here Friday for a lecture at the University of Chicago, “Nate Silver on Baseball and Politics: The Numbers Don’t Lie.” A happy coincidence?
“It’s not a coincidence,” said Matti Bunzl, the festival artistic director. “We have been following him for years. The way the election was shaping up, it was so clear it would be very close, so much of the discourse would turn on polling, all issues. The fact that he has completely exploded, that we could not foresee But we assumed that he would play a pivotal role in the election, and it played out exactly as we anticipated. We assumed that by now he would be a household name.” Bingo.
And no, you can’t attend: It’s sold out.
In late October, Silver sat with Charlie Rose and explained how he does what he does.
“Any one poll is not a good indicator, but if you take them all together . . .” he said. “Our model averages the polls and simulates the electoral colleges.”
He also illuminated why parts of the media find him so confounding.
“Sometimes pundits like to create the impression things are wildly changing every minute and states are coming into play and dropping into play,” he said. “When usually most voters are basing their votes on fundamental things. I frustrate narratives told by pundits and journalists who don’t have much grounding in objective reality.”