Stephen Baker

The Numerati
Home - Viewing one post

Why Nate Silver is never wrong

November 8, 2012News

Election day was a big one for the Numerati. Rayid Ghani, who I profiled in the "Shopper" chapter, was President Obama's chief data scientist. He moved from modeling people's preferences in food to their politics, and apparently pulled it off brilliantly.

Then there's Nate Silver. I interviewed him onstage at South by Southwest in 2009. He represented the rise of statistical methods in politics. Now everyone knows about it, and him. Actually, he was a huge star at that 2009 event. There were a couple thousand people in the auditorium, and hundreds of others in spill-over rooms watching it on TV. But then he was a more of a star for politicos and geeks. Now he's mainstream.

As I mention in the headline, he's never wrong. And this is something that the gut-level pundits can't understand. He's never wrong, because he doesn't call elections, he calculates probabilities. He wrote on the Election Day eve that Romney had a 10% chance of winning. In other words, if he ran his statistical model through 100 scenarios, Romney would come out on top in 10 of them.

Silver made his first big splash in baseball analysis, where this type of thinking is older than the spitball. Romney, in essense, was coming to the plate with the chances of a miserable hitter, someone like Sandy Koufax. The Dodger lefty was one of the greatest pitchers ever, but a sorry hitter. His lifetime batting average was .097. An easy out. But still, 10% of the time his grounders found a way through the infield, his bloops fell in. Twice in his career, once in 1962 and again a year later, he actually hit a home run. So Romney on Tuesday was essentially Koufax at bat. Likely to lose, but not certain to. If Romney had won, Nate Silver would not have been wrong.

People are going to have to get used to this type of statistical thinking, because we're going to be besieged by probabilities. In medicine, for example, we're going to hear that a certain condition has a 10% or 14% chance of killing us within 10 years or 15 years. It's not mortal, but it could be. Before, we didn't have the numbers to play the odds. Now we will.

The rise of the Numerati threatens pundits and ruminators in the worlds of politics and sports. They make money by sitting around and speculating about events that could go either way. Some, like Joe Scarborough, feel threatened by the statisticians. He should, because statisticians like Silver make it clear just how empty his talk really is.


©2023 Stephen Baker Media, All rights reserved.     Site by Infinet Design

Kirkus Reviews -

LibraryJournal - Library Journal

Booklist Reviews - David Pitt

Locus - Paul di Filippo

read more reviews

Prequel to The Boost: Dark Site
- December 3, 2014

The Boost: an excerpt
- April 15, 2014

My horrible Superbowl weekend, in perspective
- February 3, 2014

My coming novel: Boosting human cognition
- May 30, 2013

Why Nate Silver is never wrong
- November 8, 2012

The psychology behind bankers' hatred for Obama
- September 10, 2012

"Corporations are People": an op-ed
- August 16, 2011

Wall Street Journal excerpt: Final Jeopardy
- February 4, 2011

Why IBM's Watson is Smarter than Google
- January 9, 2011

Rethinking books
- October 3, 2010

The coming privacy boom
- August 17, 2010

The appeal of virtual
- May 18, 2010