the evils of poll results

Here in America we take polls to heart, which is quite ridiculous. Everything is polled nowadays: how many people objected to Janet Jackson's "wardrobe malfunction", how many people think that Bush is a "man of God", who would you vote for if the election was held today, etc. No poll can be completely accurate - you'll note at the bottom of any poll graphic (most of the time in teeny lettering) a degree of accuracy/inaccuracy. So, say, 45% of Americans like turkey hot dogs, 50% dislike turkey hotdogs, and 5% don't care. Then this is with a +/- 3% degree of error. Well, with this 3% degree of error, it could really mean that 47% of Americans like turkey hot dogs, 51% don't and 5% don't care. Or any other variation.

Basically, polls are crap. When I hear that Bush is leading in the polls with "likely" voters, well, bully for him. As I learned in my media and politics class in school, that doesn't mean much at all. We all know who the media favors (and always has), so if they want their results to show that Bush is in the lead, they can take it however they like.

Given that Bush is only leading by 1% among registered voters in a USA Today poll, I don't think they've got much to cheer about. I'll take it with a grain of salt and remain ever-skeptical.

update: I felt I should qualify this post by noting that I am skeptical of election polls no matter the result. So even if they said Kerry was in the lead, I still wouldn't buy it until the votes were counted.