What are they loosely basing the +10 oversample on? Or are they even trying to come up with a reason?
I wonder what the actual voter breakdown by party ID in Ohio is.
It’s more conservative than PA, but the econ is better than nationally there so... It sure is NOT +10 Dems. They have a Repub gov for one thing...
I know your first thought is always of the samples, so here you go: Im listing the WSJ/NBC poll partisan split first and then the 2008 exit poll split (in parenthesis) for each state for comparison. And yes, this is likely voters:
Virginia: 31D/26R/43I (39/33/27)
Ohio: 38D/28R/32I (39/31/30)
Florida: 35D/33R/30I (37/34/29)
http://hotair.com/archives/2012/09/13/wsjnbc-poll-obama-now-leads-by-five-in-virginia-and-florida-by-seven-in-ohio/