Sign up FAST! Login

Nerds Rush to Nate Silver's Defense


Stashed in: Politics!, Best PandaWhale Posts, America!, Nate Silver

To save this post, select a stash from drop-down menu or type in a new one:

The great war between people who write about politics for a living is not between liberals and conservatives, but between humanities majors and math nerds, and their battleground is currently the validity of Nate Silver's FiveThirtyEight election prediction model.The English majors have been writing that the presidential race is a total tossup-down-to-the-wire-nail-biter, and the math nerds, most prominently Silver, have been writing that actually, the race has been pretty stable and President Obama has a solid chance of winning. A 72.9 percent chance, in Silver's estimation. The week before Election Day, the English majors launched a major offensive against the math nerds, who pose a threat to fierce urgency of their headlines. "Nate Silver: One-term celebrity?" Politico's Dylan Byers asked Monday, chronicling the complaints of David Brooks, Joe Scarborough, and others whose livelihoods depend on people caring about their subjective feelings about elections saying they think Silver's overrated. On Tuesday, nerds rallied to the defense of their own.

key phrase there: "people who write about politics for a living"

pollsters are math nerds too.

Heh.

Jason, in your opinion is it the pro-Silver vs anti-Silver that leads to such differences in believed outcomes?

what do you mean 'believed outcomes' ?

Silver believes it's very likely Obama will win.

Others believe the outcome will be very, very close.

Is that because they're anti-Silver or because they have conducted their own research?

Political operatives across the spectrum are well aware that public polls skew Democrat and have been so consistently since Watergate. With the exception of 2000 and 2008, the Republican candidate outperformed the polls by about 5pts (iirc, the range is 3-9pts with the 80s being the most egregious period). The major MSM promoted pollsters are among the worst skewed pro-Democrat this cycle: i.e. the D+8 group. These outfits base their models upon a repeat, or near repeat, of the demographic turnout rates of 2008. Any operative who is honest about his craft knows this is nonsense, imo, and any pollster who knows his craft knows full well that you can analytically determine preferences by demo tab, but you can't accurately predict who's going to actually show up: it's a third tier analysis with WAY too high MoE. It's all wild guess, which is why most pollsters base their demo-turnout percentages on the previous equivalent election.

Nate Silver is doing largely the same types of analysis on the public results that the pollsters use to create them. The political math geeks know this, tell the political writers that it's a GIGO situation, who then write scathing articles which back up their preferred narrative. There's also probably a bit of "magician's creed" in there too.

Pollsters know quite well how white working class men break down, and how college educated women from the south break down, etc. and all the other demographic scales break down, and when you commission a national or state wide poll, THAT is the data you'll show to the strategists, not the aggregate numbers you see on a tv screen. What pollsters can't accurately predict is the human element of turnout, in what ratios will these crosstab demographics actually show up on election day.

Case in point: my buddy is running the GOP GOTV for Southern NJ, the area hardest hit by Sandy. Since the most Republican towns were washed out to sea, it's thrown 3 congressional districts into the lurch and he's trying to game out just how bad the turnout will be affected... He's not looking at preference polls, he's looking at the district demographics to see which portion of his guy's 15 point lead is now gone.

Wow, so an "Act of God" aka "October Surprise" really could affect the outcomes of some elections.

Thank you for taking the time to explain that.

October Surprises definitely can. If you recall, BushII was up in the polls leading into election weekend, until the hit piece on his Maine DUI... and "Act of God" are the worst, because we simply can't trust ANY polling between now and the election. There just isn't enough time to digest the events.

Things could go either way: Obama's swift response could give him a boost (tho I think Christie & Bloomberg will get the credit), but as the situation deteriorates in the affected areas (and it already is), people could start blaming him instead. People blamed Bush for Katrina despite Nagin's incompetence and Blanco's mental breakdown (tho having a dipshit at FEMA didn't help much, when FEMA exists specifically to step up when local resources faceplant).

It certainly won't affect the outcome in any of the coastal states: they're voting Obama no matter what. The only directly affected area that's in play is SE PA.. and Oh! the coincidence that Romney is going to Philly on Saturday!

Some quick analysis of that: Romney has huge leads in the vast majority of the state, except for the Lehigh Valley, the Mainline, and Philadelphia (there's a good chance he'll even win Pittsburgh). The LV is an unknown due to the huge population shifts there from the oil/gas fracking boom, but the Mainline counties are a notorious swing area. I'm absolutely certain that the calculus there is that since there are still 1.2M people without power and a good amount of storm damage, that Obama will get the blame due to negligence (with NJ and NYC gettign all the attention) and provide enough of a protest to push Romney (and Senate Candidate Tom Smith) over the top both from a support shift in the Mainline, and from turnout depression in Philly.

source?

for which?

PA analysis is what I was curious about?

PA GOP internal polling, via my friend in Hburg.

The much more troubling concern is a look at that map.

This is technically, but not figuratively true any longer (if it ever was):

"I pledge allegiance to the Flag of the United States of America, and to the Republic for which it stands, one Nation under God, indivisible, with liberty and justice for all."

We are not one nation, and we are not indivisible.

That map represented why this country is moving sidewards instead of heading forward; we have to find a way to come together. You basically have two countries: the western and northeastern states, and the rest of the nation.

You May Also Like: