According to FiveThirtyEight, Quinnipiac as a systemic bias in their polling of R+0.7
Key stat too being an incumbent polling at less than 50%.
Not that I wouldn’t love to dance on the grave of the GOP majority, but the incumbent being under 50% in a poll is meaningless. For proof, Obama never topped 50% in the RCP average for the entire 2012 campaign. And yet he managed to win in an electoral landslide.
I seem to remember them being pretty good before, but they’ve been awful recently. They just had a couple of polls that showed Clinton up 8 in Florida while tied in Pennsylvania, which makes no sense.
But the 2 most accurate pollsters had it at 50-47.
Apples and oranges. You’re also talking about national polling vs state polling. And I don’t know that it’s the same when the challenger is a recent former governor who was quite popular. Either way, Portman has been a pretty mild, fairly inoffensive, establishment pol. The fact that he’s in any trouble speaks volumes.
This is June. When is that election? November? Do you think we might be a fixated on polls a little too early? What are Strickland and Portman doing and saying? How about some reporting that isn’t tied to a poll TMP? You can do better.
I prefer to take Larry Sabato’s take on this election
http://www.centerforpolitics.org/crystalball/articles/as-deadline-approaches-rubio-ponders/
The Quinnipiac is too squishy. Took a look at the underlying press release and they don’t say whether they’re reporting likely or registered voters. That’s a big deal. Plus, the difference in these polls is the algorithm they use to predict the demographics of the actual electorate. I can’t see that so I don’t know whether it makes any sense or not.
The Dems don’t have 45 plus Sanders (I) Vermont. They have 44 plus Sanders (I) Vermont and Angus King (I) Maine.