Obama up 6 in VA. Didn't it used to be more? Did that one tighten up?
Posted by CG at October 13, 2008 04:32 PMJust adding that for some reason, I really want Obama to win VA. Maybe because I live next door to it, but I'd love to see it go blue. These polls are certainly heartening. So many ways to get to 270. I feel like I'm holding my breath and will be for the next 3 weeks.
Posted by CG at October 13, 2008 04:36 PMObama is off a point from his national tracking poll average peak, so that accounts for part of the slide in Virginia (and elsewhere). Rasmussen's last two polls in Virginia gave Obama leads of just 2 points last week and 3 points this week.
Be prepared for a little tightening in the polls over the next 3 weeks although I think Obama will outperform the final polls.
Posted by CA Pol Junkie at October 13, 2008 05:33 PMoh yeah, something's gonna tighten in the last few weeks.
Posted by T2 at October 13, 2008 06:01 PMIs there any polling that directly addresses the "Obama palls around with terrorists" campaign?
Posted by James E. Powell at October 13, 2008 07:16 PMTHE BRADLEY EFFECT...
Don't count your chickens. We want to see at least 6 pts in the important battleground states to breath easy. Any less and its a statistical dead heat, polling isn't going to account for this effect.
Posted by Mike Montfort at October 13, 2008 07:22 PMWith high black voter turnout this year, Obama may not need 30% of the vote to win in Georgia. I'm beginning to think that the fundie vote will be somewhat less this year than usual, which hurts "Keating Five" McCain as well.
Posted by libhomo at October 13, 2008 07:55 PMMike Montfort,
Check out 538, for some articles regarding the Bradley effect.
http://www.fivethirtyeight.com/2008/08/persistent-myth-of-bradley-effect.html
Posted by Will at October 13, 2008 08:03 PMI worked in a predominately black area to register voters. I can tell you they can't fill out the paperwork fast enough to register. Interesting, they are well aware of how much their vote counts, and it isn't because Obama is black that they are registering to vote. It is to see that John McCain isn't elected.
Posted by Judith at October 13, 2008 08:07 PMIt's not going to be close.
Forget about day-to-day movement ... this is mostly noise. Even Rasmussen, where McCain has his best showing in days at 50-45, writes that for the past week or so Obama has been steady at 50-52 and McCain steady at 44-45, so they don't consider today's reading to be anything but noise. The lead is solid and not varying much at this point. Depending on your sampling and weighting methodology the Obama lead is from 4 to 11 points nationally, and *slightly higher* in the battleground states.
Furthermore, all the various studies and date regarding the so-called "undecided voter" are showing that this is an overrated phenomenon. Only in a very close race can they make a last minute difference, as in Ford-Carter or the two Bush elections.
The two Republicans I respect the most, because they speak truth no matter what the situation, are Ed Rollins and David Gergen. Rollins is pro-McCain, Gergen hasn't committed but it's pretty clear he respects Obama's ability. Both are predicting landslide for Obama, not out of frustration, but just based on their long experience. They both see this year as 1980 in reverse.
I know, none of us will sleep soundly until we've seen the acceptance speech the night of November 4. But as I've been saying here since early August -- it's going to be a clear victory for Obama, with a big Senate majority to go with it.
Posted by Anonny at October 13, 2008 08:46 PMHow do you reach the conclusion from the North Dakota poll that Obama is "ahead" there?
Posted by DeanOR at October 13, 2008 09:45 PMForgetting the day-today movements and stepping back and looking at, say, the Pollster.com national polling trends, the aggregate polling for the past two weeks has been remarkably steady at about 49-50 for Obama and 42-43 for McCain.
During that time we've seen 1 and two point movements in both directions in individual polls, but that is expected as normal statistical noise. There's been a lot of activity during the past two weeks: two debates, stock market crash and partial rebound, and McCain's shit-slinging. Yet the polls have been constant.
The state polls, which tend to be lagging indicators in large part because there are fewer of them per state, have also been, on aggregate, steady for the past week or so. Any sudden jumps, like North Dakota, are explained by the fact that there are few polls in that state.
This sort of two-week polling stability is historically unusual at this point in an election. In the past there is typically movement one way or the other, and if one candidate has a substatial lead then the race tends to tighten during October.
As to the cause of the polling stability these last two weeks we can only speculate. It may be because a lot of swing voters came to their decisions earlier than normal, possibly due to the extensive early voting this year or due to the media focus on the econmic crisis, which made people think about the Presidential choice sooner than normal.
Or, alternatively, it may be a case of bad data in the past. We have a cornucpia of daily tracking polls in 2008, all with advanced sampling and filtering methodologies. Historically there have been far fewer polls, and they were much more susecptible to statistical noise (Gallup in 2000, for example, was notoriously bad). It may be that having a stable electoral preference throughout October isn't that uncommon, but in the past poor polling methodologies gave the illusion of variation in voter preference when there wasn't any.
Posted by Anonny at October 14, 2008 05:11 AM