Of course, I have no idea which polls are accurate. I'm hoping that most of them are over-weighting the Democratic vote by mistakenly applying data from the 2008 rather than the 2010 races, an error that the "Unskewed" site claims to be avoiding. We'll see soon enough.
"My husband suggests that, if this electoral map from Unskewed Polls is even close, David Axelrod will be obligated to get a full body wax. "
ReplyDelete//image of Gollum, only uglier manifests somewhere under pointy part of head//
MEDIC!
This is, at least, what the vote would look like if journalistic watchdogs made both sides play by the same rules, instead of burying some stories and highlighting others.
ReplyDeleteThere are significant methodological problems with "unskewing" polls and then comparing them one to another as if your modification made them apples-to-apples. So I wouldn't expect anything like this model to play out next week.
ReplyDeleteI still think the polls are presenting us with a highly unlikely vision of the future, however. It doesn't make sense that we would endorse four (or even two) more years of the same power division that we have now, given how much everyone hates the government and is furious about the continued recession. I'd think this would be an anti-incumbent wave year; but even accounting for gerrymandering, I'd expect more movement in statewide and national races than the polls suggest.
Powerline had a post about that this morning.
ReplyDeleteI thought Gollum was the Sr Sen from Nevada....? Either way, the mental image is enough to need the VC Especiale brainwash -- at least a case worth.
0>;~}
As I understand Unskewed's unusual approach, they make no attempt to mold their poll responses to any model that predicts a D/R/I distribution. They just publish the responses as they get them. I don't know whether that tends to give a more accurate picture or not, but it does avoid the problem of guessing which D/R/I distribution is most realistic. The other pollsters try to guess whether the D/R/I distribution will look most like 2008 or most like 2010 or some other picture that's arrived at by some abstract means.
ReplyDeleteBut all polling methods are getting more difficult now that the response rate has dropped to about 1 in 9. I've been reading, also, that there's a big disparity among polling companies in how hard they try to screen out "unlikely" voters. Some simply ask if the person is likely to vote, but statistically that produces answers that would suggest turnout rates far in excess of what we ever get. Other polling companies ask questions about whether the person voted in the last two elections and, even better, if they can clearly identify the location of the polling place. The latter approach is much more expensive, as it requires calling more people and screening out more responses.
T99, do you still have my email addy? I have...uhh...ummm...misplaced yours...so to speak.
ReplyDeleteActually, I have no idea WTF happened to it during the move, but it's no longer in my contacts.
I've emailed you separately -- found your address -- but mine's easy to remember: texanninetynine (spelled out) at earthlink dot net.
ReplyDeleteMy email to you bounced back, so I'm relying on you to email me now.
ReplyDeleteYou have mail....hopefully.
ReplyDelete0>:~}