Nate Silver sits on the stairs at Allegro hotel in downtown Chicago, Friday, Nov. 9, 2012. The 34-year-old statistician, unabashed numbers geek, author and creator of the much-read FiveThirtyEight blog at The New York Times, correctly predicted the presidential winner in all 50 states, and almost all the Senate races. (AP Photo/Nam Y. Huh)
Updated: July 19, 2014 5:24PM
In the final installment of our three-part ‘‘Statistics Are Non-Nutritional and In Some Cases Will Kill You’’ series, Couch Slouch deconstructs the myth of statistical superstar Nate Silver.
I’m at war with Nate Silver. Well, not really ‘‘at war.’’ He’s a wonk; I’m a wuss.
But somehow, using a calculator, a pocket protector and a mouse pad, Silver has turned small numbers into big dollars, moving his revered FiveThirtyEight blog last year from the New York Times to ESPN.
◆ Note: Leaving the New York Times for ESPN is like leaving the Vatican for the Church of Scientology.
(I guess I should be jealous of him. After all, the New York Times never would want me to work in its hallowed halls, and ESPN only trusts me to do poker. Frankly, there are probably 200 news organizations that would hire Silver but not me and maybe two places — one of them is Krispy Kreme — that might hire me but not him.)
We’re told Silver is a singular talent, which is just a fancy term for being a one-trick pony. And what is that trick? He projects election results well.
To quote Jerry Seinfeld from a ‘‘Seinfeld’’ episode: ‘‘Big deal.’’
I’d bet Silver can’t even boil a three-minute egg.
All he does is build polling-aggregation models. The pollsters do the polling, then Silver listens to the polling and predicts who will win. In the 2012 presidential election, Silver was right on 49 of 50 states. My dog Sapphire, analyzing the same numbers, was right on 48 of 50 states.
Plus, there’s a difference between correctly projecting electoral winners by analyzing polling data, which is Silver’s ‘‘gift’’ and allows him to be an eHarmony superstar and a terrific ‘‘get’’ for any Goucher grad, and forecasting sports results through numbers analytics.
(By the way, don’t tell me Silver is getting all the great dames with his analytics charm. What, you think I wooed Toni, a k a She Is The One, with sabermetric sweet nothings? No. I won her over the old-fashioned way — with a medium-well steak at Ruth’s Chris.)
I’ll proclaim right here and now, in front of all of you people, that I could flip a coin and do as well as Silver in any sport against the point spread.
When Silver was looking at potential employees for his FiveThirtyEight team at ESPN, he judged them by a set of coordinate axes. Coordinate axes, coordinate schmaxes. To be FiveThirtyEight material, you had to land in the upper-left quadrant of the coordinate plane, needing to be ‘‘quantitatively inclined, rigorous and empirical.’’
(You know where I would land? At happy hour at Chili’s.)
I did apply to FiveThirtyEight on a whim but was rejected because two of my references only could be reached at the racetrack and the third is Fidel Castro.
My biggest problem with Silver is this: His data might be right, but I don’t grasp what he’s saying much of the time. Even when I comprehend his scribblings, they’re boring.
Here’s a recent FiveThirtyEight sampling of Silver’s prose:
‘‘I recalibrated the estimates to ensure that the whole matched the sum of the parts.’’
‘‘There’s a modest inverse correlation between teams’ performance against the point spread and their market size.’’
‘‘The Spurs’ point distribution appears to be relatively normal (both in the common-language and mathematical senses of the term) and relatively symmetrical about their average margin of victory.’’
‘‘We can check this by means of binomial distribution.’’
‘‘The best way to test probabilistic forecasts is to check their calibration and to compare them against alternative probabilistic estimates.’’
‘‘Singles increase in value relative to home runs when there are runners in scoring position.’’
I CAN’T TAKE IT ANYMORE.
Oh, what the heck: I figure if you can’t beat ’em, join ’em.
Sure, Silver won’t hire me. But I’ve decided to fight statistical fire with statistical fire. My new blog — FiveThirtyEightAndAHalf — debuts Sept. 1. Our slogan: ‘‘More Numbers, Less Nuanced, Free Cocktails.’’
Ask The Slouch
Q. When they televise the closing-bell ceremonies at the end of the business day on Wall Street, why is everyone applauding, even if the economy just tanked? (Hayden J. Vredenburg, East Greenbush, N.Y.)A.
It’s no different than NBA players slapping hands with a teammate after a missed free throw.
Q. How did LeBron James go from ‘‘The Decision’’ to SI.com? (Scott Palmer, Charleston, W.Va.)A.
By the way, it’s possible Sports Illustrated edited his piece so badly that he actually wrote he wasn’t going back to Cleveland.
Q. Striking out four times in one baseball game is known as the ‘‘golden sombrero.’’ What do you call it when four women reject you? (Yale Reardon, Indianapolis)
A. A typical Saturday night.
Q. Any chance of motorcycle cops using their own version of the speed limit to ticket umpires who have their own version of the strike zone? (Bob Cayne, Scottsdale, Ariz.)
A. Pay the man, Shirley.
You, too, can enter the $1.25 Ask The Slouch Cash Giveaway. Just e-mail firstname.lastname@example.org. I f your question is used, you win $1.25 in cash!