Reply #59: Which analysis? [View All]

[Febble]

I agree that the final numbers represent a re-weighting.

Even if I didn't know they did, it is obvious from the math. If they represented new samples the between sample variance would swamp the within sample variance, and the MoE would go through the roof. But I think this is what Tunesmith was saying.

I also agree that discrepancy between the exit polls and the count was massive and signifant (and massively significant).