This open thread is to discuss all of the day's polls -- what they tell us about the election, their methodological strengths and weaknesses, notable findings others have missed or whatever else you want to talk about. Each day's open thread will appear in the morning and remain open for 24 hours. We also encourage you to use the "favorite" button to identify the most interesting or insightful comments.
Wednesday featured comments
"I love the dueling Indiana senate internal polls. Mourdock's shows him +1. Donnelly's shows him +7. It would be really nice to get a real polling outfit in there to get an independent assessment, but Indiana makes that tough with their calling rules. Nevertheless, common sense would seem to indicate that if your internal poll only gets you to +1, you are probably losing." - Symanski
"Taking a look at the whole 'the incumbent automatically loses if he isn't at 50%' rule, using only RCP averages, it's quite clear that Obama is in no worse position now in terms of vote share in the battleground states than Dubya was in 2004. On election day 2004, Bush was on 48.2% in Florida, 48.8% in Ohio, 47.4% in Iowa, 49.5% in Missouri and 47.8% in New Mexico, all of which he went on to win. At the moment, Obama is on 50% in Nevada, 49.8% in Wisconsin, 49.4% in Pennsylvania, 49.3% in Minnesota, 49% in Michigan (only 3 polls in the averages for MN and MI I should add), 48.9% in Ohio, 48.7% in Iowa and 48.7% in New Hampshire. With those states plus the ones that are actually safe for Obama, that would give him 281 electoral votes, 5 shy of what Bush achieved in 2004, and in 2004 Bush carried three states in which he was polling a lower share of the vote than Obama is in Iowa and New Hampshire where he doesn't even need to win, provided he wins all those where he has a bigger lead. And that's even assuming Romney can win in NC (he's on 49.8%), FL (49%), VA (47.8%) and CO (47.8%)." - cwpiper