Sunday, November 14, 2004

250 Million to One: A Must Read

Yeah, about those wildly "wrong" exit polls in Ohio, Pennsylvania, and Florida (among others) ... apparently, not so wrong. Or "wrong" insofar as there is a 250 million to one chance they're wrong. "Wrong" insofar as Dick Morris, Republican consultant, says this:
Exit polls are almost never wrong.... So reliable are the surveys that actually tap voters as they leave the polling places that they are used as guides to the relative honesty of elections in Third World countries.

Don't even get me started on the whole banana republic aspect of our current despotcracy. Is it any wonder that international election monitors were not particularly impressed with the way the oldest democracy did things? Hmmm. Apparently, Kazakhstan, Georgia (Republic of), and Venezuela did it better.

Oh, there's plenty to chew on. Let statistician Steven F. Freeman explain The Unexplained Exit Poll Discrepancy.

As dumbyass has stressed many times before, I am not a statistician. But I am a person who can take the giggle -- or, in this case -- blind fury test. That is, it does not take a numbers-crunchin' genius to figure out that the statistical probability that every single indicator used to predict election outcomes of the last 50 years would have had to be wrong for this "election" to have turned in dumbyass's favor is, oh, about eleventy jillion to one. That is, we would have to believe that his re-elect numbers and approval rating numbers both being below 50% did not matter; that a massive nationwide GOTV campaign, which always favors Dems, did not benefit Kerry; that younger people voting in record numbers did not help Kerry; that there are more registered Dems than Repugs did not translate for Kerry; that people voting in record numbers, which generally benefits Dems, did not benefit Kerry; that exit polls (mm hmm) which said without a doubt that the trend was in favor of Kerry were all wrong; that John Zogby, a wholly reliable and reputable pollster, would have called it early and been so very wrong, &c. and &c. All of these are counterintuitive. Why?