The last week of the 2012 election cycle has seen a last gasp effort by Republican whores to claim that projections of an Obama win are based on flimsy evidence, with the additional claim that adjustments to these biases would show a solid Romney lead. The first claim is true. I bet Nate Silver $1000 that his bullshit method has no scientifically valid basis–and the only way he could achieve “validity” is through the admonishments of his equally tainted “public opinion” peers. Heaping shit on top of shit gives you a load of shit. On average, if you heap together shit, it doesn’t equal filet mignon. The Central Limit Theorem is not something to be fucked with, nor can we reinterpret it to act as if bullshit polls conducted by fly-by-night beltway bandits (with surly minimum wage workers asking the questions, no less) somehow amount to random samples of the American voting population. They are not. The BEST polling organizations like Gallup and Pew now admit they are polling fewer than 9% of targeted respondents. The crap Nate Silver is stewing together includes polls which don’t even bother keeping track of such things. They aren’t even probability samples, since they have no denominator to calculate the likelihood of being in the sample.
It is good that some are questioning the quality of data from these whore polls, but there is minimal reflection on why these bullshit polls became profitable, much less how they have had a negative impact on real social scientific research. Bullshit research makes real research even more difficult. And, the noise about what is happening TODAY creates an ADHD mentality that privileges new information over good information. That plays into the hands of people like Silver, who never could cut the mark in real academics. But, now he doesn’t have to. He just makes shit up, and gets paid.
But, while little Nate is a complete whore, he’s right about his projection. Obama will win, and win big. It has nothing to do with any of the poll voodoo bullshit. Instead, the best track is to look at the highest quality polls (probably Gallup) and then think about where the biases fall. Who, in this day and age, answers their fucking phone for a stranger? The answer is almost uniformly “old people and nutcases.” So, these polls may be a random sample of the elderly and nutcases, who are about split on Obama versus Romney. Young people, Brown People, people with professional jobs and higher education, they don’t answer their phones unless you have the special ring. Last election is the best predictor of this election, and I think the Republicans will do even worse with Romney Ryan than they did with McCain Palin.