Category Archives: Media Criticism

Bayesian Analysis

An explanation, perhaps, of why so many people were rubes when it came to Obama:

…in 2008 nobody had direct experience with Barack Obama. He was the man from nowhere. We had no a posteriori way of judging him. But in 2008 a very large number of people saw Barack Obama stumping on the stage. Interestingly, many saw him exactly as Peretz, Brooks, and Noonan did: a fairly competent politician who might be left of center but whose policies and likely actions would fall well within the mainstream and bounds of rational behavior. But others saw him right off as a huckster. Neither group could quite explain to the other why what they saw was the “true Obama.” Why did two different sets of people see the same Obama images and come to different conclusions?

The reason I think is prior collateral information, or experience. It would be interesting to study whether the same group of people who tended to view Barack Obama unfavorably in 2008 also saw John Edwards as a sharper. There was something about Edwards’ hair, the unnatural emphasis with which he delivered his messages, some oleaginous quality that hung about him that, like the burglar example in the Amazon review, stirred memories of something unpleasant in the viewer.

But these unpleasant memories were largely absent in middle class, college educated, white America. These were nice people. They didn’t routinely associate with the con-men, hucksters, pawnshop brokers, and street corner grifters. To them the perfect hair, the nice suit, and the emphatic speech were simply proof of good personal grooming and culture. But to others these very same things were too clever by half. And just as the sight of a man climbing out of the window with a bag at night would arouse no suspicions in persons unfamiliar with burglars, neither would Obama’s papered over resume ring any alarm bells in people prepared to think the best of everyone.

I think it’s more than that. I don’t routinely associate with con men, etc., but I do have an intrinsic distrust of politicians in general that I think Noonan, Brooks et al don’t, or at least didn’t. And I actually did some research on this particular con man. The press refused to vet him properly, but a few people, like David Freddoso dug into his past, and what he found, from Reverend Wright and the socialist New Party to the general Chicago sleaze, wasn’t pretty. The notion that someone who voted to the left of avowed socialist Bernie Sanders would govern as a “centrist” was always lunacy, but most simply averted their eyes. Perhaps it was Bush fatigue, but there was something about the rubes that desperately wanted to be conned, wanted to believe in the unicorns and rainbows, wanted to be impressed by Reinhold Niebuhr and well-creased pants.

I predicted that he could never be elected, not because he was black (that was the main thing that I saw going for him), but because I didn’t realize that others weren’t seeing him for the obvious charlatan that he appeared to be to me. I’m relieved, though, that the scales have finally fallen from their eyes, and that we’re on the same page finally. It’s almost like the scene in LOTR when Wormtongue’s spell is broken.

[Update a few minutes later]

Wretchard has a follow-up comment:

Back in the days when whiteboards were new we fooled around with decision matrices. The decision criteria went down the left hand side of the board the weights were ranged in adjacent columns. Some criterion were ‘critical’. You had to have them. Uncertainty in their outcomes was not to be tolerated.

In electing a President you can argue that one of the ‘critical’ criterion is having a good handle on who he is; on what he will do based on a good empirical database. You cannot wing this criterion because it is a critical one. You can get the “want” objectives wrong, but the “must” objective is absolute.

The substitution of the “of course he’s brilliant” assertion (because it is an assertion) in place of solid data is nothing short of negligence. You would never select a pilot on that basis to fly your plane. Who would choose a man with no known flight experience, “a blank slate,” but with a wonderful tailored uniform, sharp trouser creases and magniloquent manner? But that is more or less what some voters did in 2008. Someone called it the “Dancing With the Stars” system of election.

So they chose the pilot on that basis.

Now some of the passengers are uneasy. The pilot is hitting buttons at random, whirling around the tabs, twisting the control column this way or that like a man possessed, yelling at the tower. That’s the Bayesian. That’s the a posteriori which they observe. Now the doubts begin, but a little too late. The question is, what now?

In retrospect every candidate ought to come to the nomination a fairly known quantity, someone with a measurable history of policy and executive history. Because this is all the nice people can see. That is evidence in their eyes. And maybe they are right. The hunches and heebie-jeebies of other people may or may not yield valid results. They would not be inclined to accept the gut feelings and judgments of others. But the existence of a measurable track record is something on which everyone should be able to agree.

But we’re still not allowed to see his transcripts.