Data and Methodology Wins
Steven L. Taylor
·
Wednesday, November 7, 2012
·
13 comments
I will, I expect, write more about this at some point, but I would like to note that the way the night has played out has been in line with the polling and the models of the stats geeks.
The unskewed movement has lost (as have those who make guesses based on their gut or hunches).
This matters, and not because of who won the election.
About Steven L. Taylor
Steven L. Taylor is a Professor of Political Science and a College of Arts and Sciences Dean. His main areas of expertise include parties, elections, and the institutional design of democracies. His most recent book is the co-authored
A Different Democracy: American Government in a 31-Country Perspective. He earned his Ph.D. from the University of Texas and his BA from the University of California, Irvine. He has been blogging since 2003 (originally at the now defunct Poliblog).
Follow Steven on
Twitter
A good night for Nate Silver and anybody else who believes in math and science.
I’ll draw a weird analogy. At the beginning of WW1 the French believed in elan and esprit. The Germans believed in machine guns. Machine guns kicked elan’s ass. Sooner or later reality trumps fantasy.
This matters, and not because of who won the election.
Indeed, but it will only make a difference if it means we stop paying attention to those who get paid to listen to their guts, if it means we can talk less about the horse race and more about the real issues that face the nation.
I doubt it means any of that. 24 cable news stations need things to talk about, and people to do the talking. They don’t want to invest what it takes to fill that kind of time will real reporting and analysis. It’s much easier to employ bullshit artists. There’s no overhead in pulling things out of your ass.
This means nothing until the right sees that math and science are real and decide to believe in evolution and global warming. I will not hold my breath.
What the numbers really show is that the Republicans are no longer a relevant political party at the national level. With Romney being barely competitive in places like Virgnia and New Hampshire, it should be obivous that no Republican candidate and no Republican campaign is going to be winning the president.
The real conclusion should be that GW Bush will be the last Republican president and that the U.S. continues down the road to being a one party state. For all of the discussion about money in poliitics, the elections shows that demogprahics trumps everything else.
Isn’t that like saying that if I tell you about half the spins on a roulette wheel will be red that my “roulette methodology” has been proven correct?
Look back at some of those polls. Take Ohio. The week of the election SurveyUSA, NBC/Marist and PPP all came out with surveys which said Obama + 5-6. Those polls obviously were bunk. They obviously were skewed in favor of Obama. They clearly were not going to be correct. For all of the reasons that some of us Quixotically were pointing out, e.g., oversampling Obama voters from ’08, oversampling Democrats, etc. And sure enough the actual vote was Obama + 2. So that “validates” the “science” of polling? Really?
Then Florida ironically enough presented the flip side of that coin. The week of the election two newspapers, of all sources, came out with surveys that said Romney + 5-6. That patently seemed far fetched if not impossible. How could Romney win by 5-6 points in a state in which non-whites were going to make up such a large percentage of the electorate? Sure enough the actual vote was Obama + 1. So now the “science” of polling further has been “validated?”
I look forward to SmoothJazz’s next contribution to this site on how everything Nate Silver says is wrong and not to be trusted.
Likewise I look forward to Jan’s defense of how Michael Barone and George Will’s “deep understanding” of the internal indicators in the polls is some how more legitimate and accurate than Silver’s model. I’m sure she will also have a good reason why anyone should pay attention to anything Bob Krumm says about polling and problems with Nate Silver’s model considering Krumm’s predictions have been spectacular wrong for two presidential cycles now.
Beyond that, I’d love a good explanation for how any of the people who have supported the unskewing efforts can seriously believe that they have enough awareness and understanding of data and methodology to form adult opinions on data and methodology heavy issues like climate science.
@Tsar Nicholas:
The entire point of Silver and other metapollsters is not to treat any specific poll as true or necessarily representative of the entire race. They work across polls.
So choosing individual single polls as a method of debunking the entire process is a crappy argument, even by your usual standards.
@mattb: Yes, can we now get George Will banned from pontificating on anything for the next five years?
Hell, if a CEO had made similar predictions about his company and had been similarly wrong, the BOD would have fired him in a nano-second.
Math and numbers are actually the best way to measure things. Who would’ve guessed? 😉
I’m hoping Will, Noonan, Scarborough, and Barrone get fired. They look like the same kind of idiots that opposed Galileo because the idea that the earth revolving around the sun just FELT wrong.
@Tsar Nicholas:
Wow, you are painfully misinformed about how statistics work and why Nate Silver’s methodology works well. You’re the exact same as the Unskewed guy: “I don’t know how it works, therefore it doesn’t.”
Also, Tsar, we don’t yet know what the final Ohio count is. As of this writing, 90% of the votes are counted and Obama’s up by 2. I don’t know if that figure includes the provisionals which won’t be counted for another 9 days. It could very well be that Obama extends his lead to get closer to the polls you cherry-picked.