Thursday, November 1st, 2012
20

Now Nate Silver Is Just Laughing At You


The Nate Silver Wars are still going—but it's embarrassing to even refer to it as a proper battle, since the weirdo pundits who think he's a LIBERAL MOUTHPIECE are too busy breathing through their own mouths to be understood. If you were busy "being without power" or "helping out your neighbors" or otherwise having a life and/or suffering in the hurricane, perhaps you missed the most hilarious intellectual breakdown of the election yet. Poor Dylan Byers—author of today's instant Politico classic "Media stumped by 2012 outcome" (yup)—took on Times stats boy and America's boyfriend Nate Silver the other day, with HILARIOUS RESULTS for all of us. It's still hard to pick which sentence was the dumbest in the Politico piece. Maybe this one: "For all the confidence Silver puts in his predictions, he often gives the impression of hedging." Hee. Welcome to math. Anyway, now Nate can just have fun.

As regards to "hedging," what is actually going on is that Silver tries to, you know, inform people. His latest, on state polling, is notable:

Bias, in a statistical sense, means missing consistently in one direction — for example, overrating the Republican’s performance across a number of different examples, or the Democrat’s. It is to be distinguished from the term accuracy, which refers to how close you come to the outcome in either direction. If our forecasts miss high on Mr. Obama’s vote share by 10 percentage points in Nevada, but miss low on it by 10 percentage points in Iowa, our forecasts won’t have been very accurate, but they also won’t have been biased since the misses were in opposite directions (they’ll just have been bad).

But that's a lot of words.

Meanwhile, Dylan Byers is out there treading the public intellectual trail, tweeting out some smart thoughts for everyone.

20 Comments / Post A Comment

stuffisthings (#1,352)

If I flip this coin and it comes up heads, all the people who said it had a 50% of coming up tails are going to look preeeeetty stupid.

stuffisthings (#1,352)

@stuffisthings I'm hoping that Politico will interview the guy who picks what animals go on the state quarters though because that would be an equally useful and valid perspective.

hershmire (#233,671)

Lies, damn lies, and…

riotnrrd (#840)

@hershmire ..people who don't understand statistics

mishaps (#5,779)

Hey, on behalf of everyone with deviated septums, don't associate us with the people dissing Nate Silver.

Mr. B (#10,093)

Every day I come to The Awl hoping for a Choire post with the #OH GIRL tag. (Please don't do it more often though, that would ruin it.)

Van Buren Boy (#1,233)

@Van Buren Boy I'm going to have to insist you pick that up and put it out with the rest of the trash.

Ralph Haygood (#13,154)

"…pundits are useless." YES. This isn't a statistical inference from a mere sample of pundits, it's a factual statement about the entire population of pundits. (Apparent exceptions, e.g., Paul Krugman, will invariably be found to be just moonlighting as pundits while devoting most of their time to legitimate work, e.g., Krugman is mainly an economist.)

Vulpes (#946)

I don't know if Nate Silver is gay (though I wish, so we could be geeky together), but that was some SHADE he threw right there! Werk!

hockeymom (#143)

FWIW, Joe Scarborough also does not like Nate Silver one bit. Joe's "gut" tells him Nate is wrong.

Maths>Joe's gut

Ham Snadwich (#11,842)

@hockeymom – Why would anyone trust their guts? They're known to be full of crap.

wb (#2,214)

If America's Boyfriend said Obama was losing, we'd probably all be freaking out and/or sobbing, not saying that he was wrong.

Antennapedia (#161,290)

@wb This. Exactly this. Say what you want for Nate Silver, he does not oversell what he's saying. He's a very good example of scientific research, ie: THE DATA ARE THE DATA. They can be interpreted, they can have trends, and it can sometimes be misleading.

riotnrrd (#840)

@Antennapedia "THE DATA ARE THE DATA" Bless you for understanding "data" is plural.

I think what's really hard for Nate Silver is that his work and words mean different things in a political punditry milleu vs. a math milleu. In the math world saying something is 75% likely to happen means that there's a quite signficant chance that it will not happen. In fact 1 out of every 4 times it doesn't. But in politics when you say "I give Obama a 75% chance of re-election" that translates to "I'm putting my entire reputation on Obama winning and if he doesn't I'm a huge jackass."

joshc (#442)

@Jeff Crystoff@twitter yes, I also get the sense that most of these idiots believe that his claim about a 70% chance of winning means that Silver is predicting that he'll get 70% of the vote. Because numbers are scary.

I can't believe the Giants put Sandoval in the line-up against the Tigers. Don't they realize the miserable loser only hit the ball 31% of the time against St. Louis?

john smith (#245,014)

I think what's really hard for Nate Silver is that his work and words mean different things in a political punditry milleu vs. a math milleu. In the math world saying something is 75% likely to happen means that there's a quite signficant chance that it will not happen.
mushroom

Post a Comment