Random Rant about Conventional Poll Wisdom
I'm sick of hearing Democratic boosters regale us with their optimism using this nugget of statistic-based wisdom: on election day, undecideds break for the challenger! Yeah yeah maybe they have some numbers to support that theory, but all I know is this: I watched the damn polls in the leadup to the 2004 election with a stupid amount of attention. And if the undecideds had broken to the challenger we'd have President John freaking Kerry right now. I think this is a good time to employ some Colbert reasoning here. You keep your books and statistics and I'll go with my gut. I listened to all that cheerleading 2 years ago and all I got for it was a big lump in my throat.

So, maybe over the last 30 years undecideds have generally broken for the challenger. Big whoop. Here's a newsflash: exit polls used to work to. And it used to snow here. Things change. We can win in November but only through hard work and organization and being smart with good candidates. I gain no new hope from that other garbage. As far as I'm concerned, my gut tells me the rule is this: undecideds break for the Republicans. That's what I've seen, in 2004, 2002 and 2000. If we're going to win--and we must--it's by bucking the trends with concerted effort, not by trying to ride the coattails of some kind of election statistic magic.