This is not the narrative you are hearing in many places in the MSM. For a while, we've been hearing about Romney's alleged momentum. In the last few days there has finally been enough push back that many have now finally moved on a little from it-though there are still those holding on.
For those who want to insist on Romentum, a new Washington Post/ABC tracking poll seems to offer some hope. In the last few days the poll has gone from Obama leading 49-48 on Sunday to trailing 50-47 yesterday. There were also lots of claims in the report that Romney has moved ahead on who is favored on the economy by 9 points: 52-43.
Interestingly, this 50-47 margin is now the same that Gallup is showing for Romney among likely voters; Rasmussen in it's latest showed the same margin. But other polls like Rand and Reuters show Obama leading and cumulatively, the national polls are basically a wash.
Meanwhile, the state polls mostly are not showing good news for Romney. He has not lead an Ohio poll in 2 weeks. He has never show a lead in Wisconsin; he's even trailing in Virginia now according to Talking Points Memo's Poll Tracker. Indeed, according to PT, he's even trailing slightly in Florida-which even Nate Silver wrote a rash poll suggesting that Obama pull out.
Speaking of pulling out, a week ago, the Romney team feigned that they were considering pulling out of North Carolina. Yet today, PPP shows a dead heat between Obama and Romney. NC is the swing state that was considered most safe for Romney. Yet even here his lead is small.
A lot of the problem is context. You can look at polls all day, but if you look at the wrong ones. misread them, or cherry pick you get a totally false impression.
If you cherry pick polls like this recent Wapo poll or Gallup when it showed Romney up by 7-since dropped to 4-you can fall for this illusion that Romney has momentum-in truth his momentum from Denver has been over for two weeks. Since then you've seen Obama's chance of electoral college victory rise from just over 60% according to Nate Silver's model, to over 73% last night. Before the Denver debate, Obama's chances reached a high of 85%, so he's gained back roughly half of what he lost after Denver at this point.
Indeed, while some believe that Nate has a nefarious plan to boost Obama by falsely claiming he leads by a lot, if anything he's a lot more conservative on Obama's chances than other poll analysts are.
Here's what Sam Wang has to say about the claim that Obama is slightly more likely to win:
"This is false – he’s a lot more likely to win."
"In a race today, President Obama would win with about 90% probability. The true probability is even higher, since the Meta-Analysis does not correct for individual pollster errors. We could – but the political blowback from unskewing polls is too large."
http://election.princeton.edu/2012/10/25/do-you-understand-polls-as-well-as-david-brooks/#more-7922
He also, for the record, predicts that the Democrats will pick up significant seats in the House: we haven't heard much about the House races elsewhere. Most polls show the Democrats will hold onto the Senate, and may pick up a seat or two, even: which would be a major accomplishment as so many more Democratic seats were in play than for the GOP.
The question is how many they pick up. His model says there is as much as a 33% chance that the Dems retake the House-so not the most likely but still possible outcome.
On average the model is showing 208 seats for the Dems in the next House. It will certainly by a tighter body than it is now, whoever wins.
UPDATE: Just in case you are still hyperventilating about the Wapo poll-though really you shouldn't-today's Wapo poll shows Romney's lead back down to 1, hopefully underscoring the folly of worrying so much over one cherrypicked poll on a particular day.
http://livewire.talkingpointsmemo.com/entry/obama-cuts-romneys-lead-to-1-in-abc
No comments:
Post a Comment