[Lysandru]

You missed my point. Why aren’t the polling organizations polling people who have voted?

[cgbg]

Why aren’t the polling organizations polling people who have voted?

Excellent question.

What I would want to do is use that data to _validate_ polls.

How would you do that?

Let us say 90 voters already voted early in State X, 30 registered Republicans, 30 registered Democrats, 30 registered Independents.

You ask the question: "Did you vote early?"

You ask the question: "Are you registered as Republican Democrat or Independent?"

You ask the question: "Who did you vote for in the Presidential election?

Then, you _weight_ your poll so that the "yes" answers match the _known_ 30_30_30 turnout.

Then you can be sure your survey accurate reflects the actual early voting.

Most pollsters are way too sloppy to validate their polls.

[Flick Lives]

You missed my point. Why aren’t the polling organizations polling people who have voted?

—

It’s hard enough to get people to answer a call from a pollster. Harder still to conduct a poll limited to one State. Now you would have to poll a subset of that with people who have submitted EVs. The probability of a meaningful statistical sample approaches zero.