Monday, March 10, 2008

How to track pundit prescience?

At the beginning of February, Steve Sailer mentioned, in pointing out the media infatuation with political momentum, that he rarely makes exciting predictions for a fear of being wrong.

That's a prudent route, but it might be overly so, especially given the novelty existent in his work. A year ago, for example, he might've predicted that Obama would prove unelectable in the general election (if he made it that far) when the revelations about his race-obsessive, African-nationalist-church-membership past were mainstreamed by forces on the right. That prescience would have drawn more attention than the revelations alone did.

"Exciting predictions" basically pertain to situations where more than one outcome seems plausible in the minds of most people--Presidential elections, sports games, etc. Because the premium on certainty is so high, people seek out those predictions as a way of cutting through the fog of uncertainty and being in the know. Problem is, that high level of uncertainty frequently means those making the predictions turn out to be wrong.

That's not a big deal for those who play augur, though. Fox News doesn't profile the predictions of its individual panelists the day after a big vote to show the public how well their prognostications panned out. If it did, those panelists would be hesitant to make predictions in the first place. And it's the predictions themselves, not their credibility, that draw people in, as anyone who has watched Dick Morris tell the future knows!

After reading Steve's post, I started outlining one of my own to tout the accuracy of my sparse predictions. Just make a few, whenever there is a big disconnect between popular perception and callous reality, and presto! I had the Duke rape case from the get-go on the grounds that white guys simply don't gang rape black women. Randall Parker had just quoted an economist who summed up the housing crisis as I'd imagined it come about a couple of years ago. I'd just have to wait a day for the Patriots to cleanup the Giants, thereby fulfilling my prophecy made early in the season of a New England Superbowl victory, and then one more day after that for Hillary to regain the acclaimed 'frontrunner' status on her road to the Democratic nomination, as I'd foreseen her doing in the midst of all those crazy Obama predictions.

Uh, anyway, if you're a regular reader, you didn't miss that prospective post.

How standard is that 50% (possibly 75%?) "exciting prediction" rate? Seems to me that a database on the performance of predictions made by Important People would be quite useful. Gauging what people and predictions deserve to be tracked could be solved by open sourcing the site, along the lines of wikipedia. I haven't found anything close to a resource like that, but this BusinessWeek piece hints at how much fun it would be if there was one.

No comments: