Shout Out and Alaskan Absurdity Analysis

Especially since leaving for Israel (though I think also before) I became obsessively interested in the election, most likely I imagine as a way of not feeling like I'm "missing" a lot by spending the year away from the US. (More on this in another post.) The last time I was here for a year, I definitely felt that, perhaps only highlighted by the then necessary tactics of my father sending me massive shipments of VHS tapes with our favorite television shows (including a brand new show that had caught his attention, The West Wing) and me watching the Rams' improbably playoff triumphs via a slowly connected desktop in the basement of the old Beit Nativ (זצ"ל).

This obsession has picked up over the last few weeks, and did not disappear in spite of the formidable lead Obama had in the polls. In fact, I had profound fears that the polls would be wrong, that the Bradley effect might actually exist (more on this in a second), and that the right-wing anti-science base would now be able to include statistician/pollsters in with evolutionary biologists, geneticists, and astrophysicists as practitioners of, as it were, patently false crafts. Nothing has been a greater relief to me (and will, hopefully, assuage future fears or disillusioned hopes, depending on polls' predictions in future elections) than to see that the polls were, with almost no exceptions, pretty dead-on.

So I wanted to take a moment to "publically" thank the website that (in partnership with Slate) brought me through the last few months, Pollster.com, and two other sites I became aware of (thanks to JAR and Jason Rubinstein, respectively) only in the last ten days before the election: realclearpolitics.com (claimed by JAR to be even-keeled but revealed by another site to be oddly in favor of polls that showed McCain was doing better, not worse) and fivethirtyeight.com. I will check them all regularly in the future and look forward to continuing to benefit from their insightful reporting and statistical awesomeness.

On that note, 538 reports today about one of the only polls that did not match up with the outcome, in Alaska, and sheds fascinating light on a variety of things, including the likely 10s of 1000s of outstanding ballots, McCain dominance far outside how he was polling, the strange (and, possibly and hopefully, soon to be rectified) lead by the incumbent convicted felon in the state Senate race, a suggestion that "convicted felons are the new Black" vis-a-vis a possible converse-Bradley effect, and the quite real ramifications of Alaska being so far away from the rest of the country. Great website; great post.