Reply #18: What formula are you referring to? The weights? The probability? The MOE?

[TruthIsAll]

1) The weights are correct, if you believe the exit poll.

2) There were 17% new and 4% non-Bush, non-Gore voters, if you believe the exit poll.

3) The assumption that 79% voted as they did in 2000 is a fair one, if you believe the exit poll.

4) The probability is ABSOLUTELY CORRECT, if you believe the exit poll weights are correct, because then the deviation from exit poll to vote tally must be accurate.

5) The MOE is correct, if you believe the formula is correct: Actually, MOE= 0.875%= 1/sqrt(N), where N= 13,074= exit poll sample size.

Mitofsky states that the MOE is 1.0%. Is he applying the "Cluster Effect" or is he just rounding up to the nearest percent?

I was conservative and used his MOE of 1.0% in the probability calculation. If I had used an MOE of 0.875%, the probability would have been as close to ABSOLUTE ZERO as you can get.

So, it comes to this: Do you believe the exit poll?

If you do, game over.
If not, why not?