Poll update

Ok, this is getting really weird. The results of my latest poll show that the “new window” people, after blowing an early lead (about 60% after ~30 total votes) and falling way behind (at about 30% after 160 total votes), have stormed back, closing the gap between themselves and the “no new window” folks. Are these seemingly wild swings caused by voter cheating, different audiences voting at different times (holiday vs. non-holiday, weekday vs. weekend), or is it just statistically insignificant noise? I’d take the time to analyze it, but arsdigita (as far as I know) doesn’t keep track of individual votes. Forgive me for prattling on about this poll, but the math/stats lover in me just can’t let it go.