Against a decent candidate, I think Portman could have been a bit more vulnerable this year. But, they'd have had to be damn near perfect. With our bench being what it is, there wasn't anyone who was going to make a good run of it, to be honest. Strickland was probably the best option. Sittenfeld would have done as well, I think. I mean, Strickland is a total "hold your nose and vote for me" candidate. His gun stance is shit, and that's nearly a deal breaker for me. Strickland needed to go on the offense almost immediately...and he didn't. Because he's terrible.
Also, I just want to point this out again:
Those IPSOS things aren't "real" polls. They are real in that they exist, someone did them, and they aren't just pulled out of a hat. They're models based on breakdowns of their national tracking poll. Then, they add their own weights to it.
We talked about this the other day. For example, they assume that 70% of white men will vote. This will not happen. 70% of eligible white men aren't even registered to vote, number one, and, number two, that number has never, ever been higher than 65% (and in the last three or four elections has been in the low 60s). They also assume that only 60% of black women will vote, which would be a 10% decrease from 2012, and the lowest participation rate among black women since 1984. I also *think* they assume that only 32% of eligible Hispanic voters will turn out, and that the electorate will only be 13% people under 30.
None of these are good assumptions. Their whole thing is a version of Nate's model, really, if the only polls Nate used was IPSOS.
This is not unskewing, btw. It's just t.
