Replies

If MI is close why aren’t R/R killing in OH?

I don't care what anybody says, the polls this year are the strangest I've seen. It's not "omg McCain's gonna win" denial that I had in 2008, the polls are strange even when taken at face value. If O is winning MI by only 1.5%, then R is going to win OH, VA, FL, WI, CO, etc., period.

It does seem like that almost every state is showing a foundation for a blowout, except for a few battleground states. Have seen some competitive NJ + CT polls as well.

IMO, one of three explanations are possible:

1. For whatever reason, Obama is doing way better than Romney in the battlegrounds, enough to move about 10-15% of the electorate that would normally vote for Romney.
2. Polls in the battleground states are inaccurate because of pollster bias. Generally it's not NYT that's polling in second and third tier strategy states.
3. Voter exposure and fatigue, or lack thereof, disproportionately affects poll responses. I hear people are getting up to 10 calls from Obama + Romney in the battlegrounds. In that case, you can choose which are the more representative polls, the battlegrounds or the non-battlegrounds.

I agree that the polls are wackier than usual this year. How can a Republican be down 2 in MI but down 5 in OH? It's just not possible due to demographics alone. I think these battleground polls have a lot of special sauce in them. Oversampling Dems and over-inflating the undecided/prefer another candidate column. I have a feeling that anybody who says "I'm not crazy about Romney but I suppose I'll vote for him because it sure won't be Obama" are getting put into that column, taking votes away from Romney and making it look better for Obama.

I've seen outlier polls have that Romney within 5 in NJ and I'd like to see a reliable poll for MA. I think that due to the Scott Brown senate race and the hometown factor for Romney, Massachusetts is going to be much closer for Romney than people would think.

The voter turnout in 2008 was D+7 and the polls we are seeing are weighted that way. Fact is, the 2012 turnout is going to be more like R+2 (based on 2010) and that is a 9 point swing towards Romney that is not being reflected in these bogus polls.