Reply #41: I understood that just fine [View All]

[dlaliberte]

I started out by agreeing that some adjustment of the exit poll numbers is required based on the population being sampled. So I don't disagree that some adjustment is necessary.

What I objected to is the adjustment based on the actual vote. Since we don't trust the actual vote, we can't trust the exit poll numbers either after they have been adjusted based on the actual vote.

Why should we trust the "raw" exit poll numbers? Because they have been adjusted based on turnout and other demographics independent of the actual vote adjustment. I would expect they should be able to adjust the exit polls in the same way they adjust the pre-election poll numbers. My question was why do anything more?

We also want the per-precinct exit poll numbers, so we can hopefully tell which precincts more likely had fraud. A random sample within a precinct should match up with the actual vote in that precinct pretty well.