The other one is a bit less interesting. MS has moved from Weak McCain to Strong McCain. A few people (myself included, early on) had hoped that Mississippi's large black population would help move it towards Obama, but that's not looking good at present. Of course it all comes down to turnout, so who knows. I wouldn't be betting on MS personally, though.
UPDATE: I guess I was in wishful thinking mode when I wrote this yesterday, because I said ND moved from Weak McCain to Barely McCain. It actually moved from Weak McCain to Strong McCain.
| If the election were held today and... | Obama | McCain |
| ...all McCain <5% went to Obama | 383 | 155 |
| ...all Obama <5% went to McCain | 286 | 252 |
| Shift since yesterday: | +0 | +0 |
| Shift this week: | +3 | -4 |
(See this post for an explanation of these numbers.)
For the raw poll data and more details, see this day's page on Electoral-Vote.com:
8 comments:
I sense a sine curve direction change....
Not really... If you read my post from the 13th, you'll see that I noted that ND had swung more than 10 points towards Obama over a period of about 3 weeks. That suggests that the Oct 13 poll was quite possibly an outlier. The fact that the new poll shows that it went from +2 Obama to tied isn't enough to suggest that things are headed the other way. It's also important to note that even if ND is currently tied, 1 month ago, McCain had a 10+ point lead there. Even if Obama ends up losing ND by a point or two, that's a fantastic result. ND is not a state that McCain should have even had to worry about. The fact that it's even close means that McCain has to spend time and money and resources there that he would much rather spend in states like VA and NC.
North Carolina is the other state that accounts for the apparent downward trend for Obama, but it has generally been trending in Obama's favor. Andy doesn't seem to list all the polls for NC, so if you factor in the others, it would actually move into to Barely Obama category. In particular, he doesn't seem to list the polls from PPP, which have showed an Obama lead recently.
What's much more likely than a downward trend for Obama is that he's reached the plateau of his support. There is a theoretical limit to the support that any candidate can have, and once they reach it you wouldn't expect it to continue to go up. It's probable that Obama's numbers will stabilize right around where they are, and probably even tighten a bit in the national polls. One or two of the very close states could switch back towards McCain, but the reverse is probably equally true.
Short of a game changing event, I don't see any reason to expect that there will be any mass movement one way or the other at this point. And remember that a game changer could go either way.
My hopes for the election:
* Good: 270+ (wins)
** Better: 371+ (beats Clinton '92)
*** Best: 380+ (beats Clinton '96)
I'm abandoning any hope of an electoral landslide over 400. I think the country is still too evenly divided for the kind of landslides of 1964 to 1988.
1964: 486-52
1972: 520-17
1980: 489-49
1984: 525-13
1988: 426-111
Johnson, Nixon, Reagan -- the bigger they come, the harder they fall. Just thinkin' you got "political capital to spend" seems to be the problem.
Support for the sine wave cresting theory (nod to diogenes): http://www.pollster.com/08USPresGEMvO.png (that, and the supertracker on 538)
A decline to the point where McCain could win is extremely unlikely. Obama's previous sharpest decline was probably Sep08 (O 301 vs M 204) to Sep19 (O 243 vs M 274). It caused a move of about 60 (predicted) EV down for Obama, 70 EV up for McCain, in a mere 11 days . If that were to occur now, the 347-367 EV predicted for Obama would drop to around 287-307. He's probably built up too much of a lead to be in much danger now (altho the McC win %age at 538 has opened to 6.4%).
Counter-evidence of any downturn: Sam Wang has Obama bigger than ever. Check the Median EV Estimator. (367)
Another site I like very much (by astrophysicists), http://www.colleyrankings.com/election2008/
has been very conservative, as they have a very simple methodology using the MEDIAN poll from the past month. That automatically tosses out outlier polls without extensive computational processing. They were stuck at Obama 273, McCain 265 for 3 weeks. They have Obama up at 367 now, too. If that is close to reality, add MT to get 370, and add IN or GA to get over 380. You could argue that their method will give a lagging indicator if the support veers either way at the last minute, though.
3bluedudes has 80 presidential projection sites. All 80 show Obama ahead. Of those updated 10/18 with no ties allowed, 349 or 350 is about the current median projected outcome/snapshot.
If you read Nate's comments, he argues that the polls aren't going down, merely plateauing. He addresses the national polls in that post, but most importantly he talks about the state polls:
"McCain's other problem is that the polls in battleground states have not really tightened at all. Obama gets good numbers today, for instance, in North Carolina, Wisconsin, and Florida. Obama presently has something like a 3:1 advantage in advertising, and most of that advertising is concentrated in battleground states. As such, this may serve as a hedge against any improvements that McCain is able to make elsewhere in the country."
Here's another interesting tidbit from ARG:
"Because of McCain's electoral inefficiency, even if McCain were to proportionately increase his national ballot share to 49% for a tie, Obama would still lead with 316 electoral votes. Obama is running a race in the Electoral College while McCain is running a national race."
Both of these points are why I'm still not at all concerned about the state of the election.
Let me make one thing clear, though: It is quite probable that the race WILL tighten up between now and election day. You're not seeing that in the polls now, but it's quite possible that you will in the coming days. Nate has been predicting such a tightening all along, and such a tightening is normal before most elections. But I stand by my assertion that unless there is a game changing event between now and election day, Obama will win and he will do it decisively, both in the popular vote and the electoral college.
538: "Most likely Obama EV totals: 369, 364, 380, 349, 338."
I'm rooting for the third spike.
Nate defines a "landslide" as over 375, still a 32.75% probability. It's roughly the same as the chances of a filibuster-proof Senate, and if we can get one we'll probably get the other, too.
UPDATE: I guess I was in wishful thinking mode when I wrote this yesterday, because I said McCain moved from Weak McCain to Barely McCain. He actually moved from Weak McCain to Strong McCain.
Huh???
Oops... Corrected again.
Here's what it should have said (and does now):
I guess I was in wishful thinking mode when I wrote this yesterday, because I said ND moved from Weak McCain to Barely McCain. It actually moved from Weak McCain to Strong McCain.
That still cracks me up. Paranormal researchers measure McCain's electromagnetic fields as internal pollsters attempt to cross-reference his answers to determine how McCain he is today. Who is he today? There is a bit of truth in that slip. Yes, McCain has moved from Weak McCain to Barely McCain again.
Be sure to get enough sleep, Poll Junkie, or you may get incoherent or irritable or even McCain.
I am looking at the new Oct20 e-v.c right now and thinking ND and MT are not freak waves but a trend. McCain is losing the Lutherans and corn-growing ethanol supporters and wind farmers, etc. That could mean he is losing SD, NE, and others, it's just not detectable.
Some of these red states haven't been polled since late Sept, McCain's hi-water mark, and they were barely 10% +McCain then. The polls have moved 10 points away from McCain since then.
People may say "McCain" when they answer Qs on the phone, but will they get out and drive to a polling place? Would they even go to the mall to meet or pick up McCain? The answer is no, they don't even bother to attend his rallies.
I am starting to think it may be a "landslide". A Blue-State drive-path coast-to-coast. I am turning into DIck Morris.
Post a Comment