Comment: perfect

(See in situ)



I think the most telling point in this document is that Romney and Paul's lines diverge from the expected path at nearly identical spots. If one started losing votes at say 10,000 and the other at say 5,000 I would question assumptions. But since both diverge at the same place its CRAZY creepy! The flipping idea is soooo perfect because it keeps the overall count the same. NO ONE would be able to notice by skimming the surface. I think that flipping votes is a PERFECTLY legitimate theory about what happened.

It only takes a random sample size of 30 or so to get a good representative prediction line of the larger group. That assumes that the smaller sample is truly a random and evenly distributed sample of course.

07 Daytona 675
Don't Tread on Me!