Chicago Sun-Times
Staff reports on all things politics - from City Hall to Springfield to Washington, D.C.

Nate Silver's mic drop

| 3 Comments | No TrackBacks

Nate Silver of fivethirtyeight (Randy Stewart / Wikipedia)

The below images are electoral forecast maps, the first one posted this past weekend by Nate Silver of fivethirtyeight and, following that, posted by Unskewed Polls on October 28, 2012. Unskewed is, essentially, a conservative-bias imitator (and a pale one, at that) of Silver's blog.



Unskewed Polls


Bit of a difference, no? Unskewed's Dean Chambers gained notoriety for not only criticizing Silver a few days before he posted the above forecast, but for stooping to homophobia to do so.

Nate Silver is a man of very small stature, a thin and effeminate man with a soft-sounding voice that sounds almost exactly like the "Mr. New Castrati" voice used by Rush Limbaugh on his program. In fact, Silver could easily be the poster child for the New Castrati in both image and sound.

But here we are, the day after the presidential election and, according to the most recent returns, Silver not only correctly predicted the 50%-48% popular vote split for Obama, but correctly picked all 51 Electoral College votes (all 50 states plus Washington, D.C.), including the ultra-close Florida vote. So, yeah, he's good at what he does. And he's $2,000 richer - or at least the Red Cross is - after a bet with political loudmouth pundit Joe Scarborough over who would be right about the election (a bet that caused much hand-wringing only because people have to worry about something). Chambers, meanwhile, picked Romney to take 51% of the popular vote and 275 Electoral College votes. The only discrepancy in Silver's predictions are his total Electoral votes projections which don't match his "chances of winning a state" math, but something that relates to the way Silver weighs different polls and those outcomes are weighed differently.

In the harsh light of a new day, even Chambers is having to admit: Silver was right, I was wrong. And he's not alone. Politico posted a list of which polls were most accurate in projecting the outcome and oft-cited polls like Rasmussen and Gallup were at the bottom while PPP was at the top.

But none were perfect like Silver was. It's unfair to call this Silver's Redemption because the backlash that hit Silver in the final days weeks so wrong-headed and born out of a conservative distaste that he had Obama in the lead. If anything, it was a victory lap for someone who understands the math and models necessary to make accurate prediction. In the last two presidential elections, Silver has accurately predicted 101 out of 102 Electoral votes (99 states and D.C. twice) and twice nailed the popular vote split. His work speaks for itself and the vitriol directed towards him is simply the voice of denial from the right, the same right that wouldn't hesitate to back him if he had correctly picked a Romney win (and one has to assume that, while not based in homophobia, dems would have similarly dismissed Silver's numbers the way they dismissed Chambers).

Silver has earned our respect and earned the right to gloat. But it's best that he didn't; towards the end of the campaign, Silver did let his cool demeanor slip just a bit in defending himself against criticism. Instead, in a tweet that went out shortly after multiple media outlets had called the election for Obama, Silver simply posted a tongue-in-cheek tweet (posted below), perhaps the best mic-drop moment he could have hoped for and one he deserved.

No TrackBacks

TrackBack URL:


When people like Dean Chambers suggest someone else might be gay...........there is usually some self loathing involved.

Nate Silver is so reasonable in his analysis. Most of the people attacking him were jealous, or partisan or both. Pundits thrive in the "nothing is knowable, reality is what we say it is" world, and fight to defend it against interlopers with data.

Why does one "have to assume" that Dems would dismiss Silver had his numbers pointed towards a Romney win? I've been reading Silver's blog since before the 2008 election, long before he was working for the NYT. He also predicted the 2010 midterms; while he wasn't quite as astoundingly accurate as in the Presidential years (he freely admits his model doesn't work as well for Senate and, especially, House races), he still was pretty successful calling the Republican landslide. No one on the left, that I recall anyways, had anything bad to say about his predictions.

Presidential years are a little different, of course, but its not like the left has had any particular sort of aversion to data. Let's not force a false equivalency.

This assumption that liberals would have dismissed Silver for predicting a Romney win just as they did Unskewed is intellectually dishonest. Unskewed was dismissed by everyone that wasn't hoping against hope that the actual numbers were wrong.

If the polls legitimately showed Romney leading 90% of the important polls over the last several weeks, as they did in 2004 with Bush, those of us who understand statistics/probability would tend to believe them. But ultimately Nate Silver is simply a guy with a fantastic track record in the 2008 and 2010 races who make fairly transparent adjustments, in both directions when needed, and that for the most part simply takes the data as it is.

Chambers entire system likely consisted of an excel spreadsheet that any college freshman could have crapped together after Intro to Computing. And rather thank working with complex algorithims he just altered the turnout models to whatever level he needed to maintain a Romney lead. In case nobody noticed he originally claimed to be using the Rasmussen model, but when Rasmussen began showing a near tie race his predictions never changed. He basically found the most favorable possible numbers and stuck with them... the only thing I really wonder is if he did it because he believed it or to try and manipulate the race by injecting some hope into the right wing activists.

Leave a comment