General Discussion
In reply to the discussion: The shy Trump voter. [View all]Drunken Irishman
(34,857 posts)That is all.
I am not talking specific polls. I am talking an average of polls. If every single poll was off, as you say, Hillary's support would've been off too - that is not the case. In these three states, when you average out the final polls, the polls were 100% correct on Hillary's support.
That means, on the whole, the polls were accurate. The only thing they couldn't account for were the undecided voters, which broke heavily for Trump. That has nothing to do with their poll sampling and everything to do with voters telling pollsters they weren't decided. I theorize many of these voters were, in fact, Trump supporters too shy to admit to pollsters they supported Trump.
I based this on several facts:
1. The average of all polls had an unusually large undecided population - especially on election day. Typically, undecideds narrow significantly the days before the election. It's rare that they remain at, or above, 10% in state polling averages - as they did with two state averages: Michigan (10%) and Wisconsin (13%).
2. The average of polls hit Hillary's support almost exact. This tells me the sampling was not off on the whole. Your sample can't be off dramatically and only effect one candidate. It would have impacted Hillary's numbers too. What you're suggesting would make sense if, say, Hillary's support in the averages went from 52 to 47 and Trump's went from 43 to 47 or whatever. But that isn't the case. Hillary's support didn't budge. They got her support correctly.
3. I posted evidence that undecideds broke heavily for Trump, which plays into my narrative about shy Trump voters.
I have backed up every one of my claims. You really haven't.
I feel comfortable saying I am right here and the evidence I posted pretty clearly lays out that fact.
Have a good night.