Official South Dakota and Montana Democratic Primary Results Discussion Thread

[Alcon]

Might as well, honestly.

A note about South Dakota:  they have an unusual counting method.  Ballots are not tabulated at the polling places in any way.  They are driven to the county seat where they are counted up as optical scan ballots.  The lack of a verification process at the polls speeds things up slightly, I'd think, but this year the news outlets are implying this method slows stuff down.

Also, Shannon and Todd Counties do not have county seats and must drive their ballots to Hot Springs (90 miles from Batesland) and Winner (65 miles from Parmelee), respectively.  In other words, their results are going to be really late.  If anyone remembers Tim Johnson's win in 2002(?), that was what happened.

[Alcon]

Let's not forget that, faced with the opportunity to stir up bad news for Clinton and good news for Obama, Drudge going to jump at it.  He's clearly not un-fond of Obama, and we all know what he thinks of Clinton.

[Alcon]

I would like to be the first to give a pre-lol to ARG.

[Alcon]

South Dakota:

Should Obama pick Clinton as VP?
Yes 55%
No 41%

Among Obama supporters:
Yes 40%
No 56%

"fascinating"

[Alcon]

Don't those numbers indicated that Obama's in trouble in South Dakota? For there to be a 15 point positive spread when the two groups are combined as opposed to a 15 point negative spread with Obama on their own?

Exactly what I was calculating...for a tie, over 70% of Clinton supporters would need to say "yes" to that question.

Not particularly great news for Obama.  Keep a look out for the Clinton supporter numbers.

[Alcon]

Don't those numbers indicated that Obama's in trouble in South Dakota? For there to be a 15 point positive spread when the two groups are combined as opposed to a 15 point negative spread with Obama on their own?

Exactly what I was calculating...for a tie, over 70% of Clinton supporters would need to say "yes" to that question.

Not particularly great news for Obama.  Keep a look out for the Clinton supporter numbers.

Actually though, don't you think that would be probable, that 70% of Clinton voters would support that?

That's a good question, really.  It depends on how high the delusion vote is, and how the question is asked.

[Alcon]

The South Dakota numbers are weird.  Nearly 70% consider Clinton honest and trustworthy?  That's a big WTF -- she hasn't seen those numbers since West Virginia and Kentucky.  Then again, in WV & KY, Obama was in the negatives in honest/trustworthy.

As for Montana, those numbers nearly exactly match Indiana, maybe each being a little less approved-of.

Could ARG be right???

[Alcon]

Honestly, judging only by the exit polls, I'd give up on South Dakota now and worry about Montana at this juncture.

Of course, we could see a different dynamic here than in previous races...

[Alcon]

I'm not sure what you mean, Verily.  Both say that in South Dakota, 55% say Clinton should be the VP.

[Alcon]

CNN:

South Dakota:

18-29 Years Old

Obama 65%
Clinton 35%

65+ Years old

Clinton 65%
Obama 35%

That's probably the biggest age split we've ever seen...which, in South Dakota, is not particularly great news for Obama.

But, yes, that isn't Clinton +26.  You don't win 18-29-year-olds 2-to-1 while losing by 26 points.

[Alcon]

Poblano has Obama winning SD by 4%-5%. He has been incredibly accurate during this entire primary season:

www.fivethirtyeight.com

He projects primary results based on demographics and statistics, very in-depth stuff.

I'm the first to trumpet Poblano, but demographics can only say so much.

Plus, one word: Kentucky

[Alcon]

South Dakota

Economy voters:

Clinton 58%
Obama 42%

Iraq:

Obama 61%
Clinton 39%

Looks like a very probable Clinton win, but still a definite "lol ARG"

[Alcon]

He'd lose no matter what, actually, in all likelihood.

The "other" issue is mostly Healthcare, which is usually just slightly worse for Clinton than Economy.  I'm going to give her that issue 55-45.

Past issue allocations by state, extrapolated to South Dakota:
Pennsylvania: Clinton +4.3
Oregon: Clinton +2.4
Kentucky: Clinton +7.6
Indiana: Clinton +8.3
North Carolina: Clinton +6.4

The range here goes from a tie (Vermont) to the upper single digits.

[Alcon]

Every year we get some new idiot that everyone trumpets as knowing everything who knows very little actually.  Kind of annoying.

err, where did that come from?

Although it is impossible to actually guess real numbers, a Clinton win of 5% or so would mean that ND and SD aren't really as different of states as I've thought - or at least their Dems aren't.  SD is a closed primary, right?

Yes.

[Alcon]

Well, then Sam isn't being fair to Poblano.  Poblano has never advanced his model as anything but a statistical model, and that only goes so far.  It's people who take his (pretty excellent) record and assume his model is sacrosanct, who really deserve the ire.

And it really seems to be the best predictive model I've ever seen, even if it isn't near perfect.

[Alcon]

I was basing those % within the issues off the other states, too. So, when Obama won health care in Oregon, he won by closer to his Iraq margin than to his economy margin (so here it would be around 55-45 Obama). But that's obviously not going to happen anyway.

True, which is annoying.  Assuming the exit polls haven't reverted to being crappy, this is looking like Clinton +2-7 to me.

New unnecessary information!!!

South Dakota

White, working-class voters (AKA no college degree)
Clinton 60%
Obama 40%

College graduates
Obama 53%
Clinton 47%

[Alcon]

This is...interesting.  The college education rates, unless South Dakotans are lying about their education rates more than Indianans, point to a double-digit victory.  If they lie at Indiana rates, it would be about 62-38 non-college.

(0.62*0.4)+(0.38*0.53) = 44.9% for Obama
(0.62*0.6)+(0.38*0.47) = 55.1% for Clinton

What's going on?  Is my math wrong?

[Alcon]

This is...interesting.  The college education rates, unless South Dakotans are lying about their education rates more than Indianans, point to a double-digit victory.

What's going on?  Is my math wrong?

Your math is right, but it doesn't square with the other data from the exit poll. So they may have lied more than usual, the weightings may end up removing that inconsistency, or something else is going on. We've have internally inconsistent leaks of exit polls before.

For example, Utah supposedly had a majority of voters college-educated. So high rates of lying are not unlikely.

Yeah, but Utah's bachelor degree rate is 26.1%, and their Democrats tend to be bohemian, so that about fits in with the Indiana number.

But if the lie rate in South Dakota doubled, Obama would still be down by very high single digits.

[Alcon]

Eh, we'll see I suppose.  I still maintain that Democrats in Utah are probably unusually educated relative to Republicans, but we shall see.  Not that it really matters.  A loss is a loss for Obama, and this is a loss; at this point it really just matters for how humiliated ARG will be.

Seriously, posting would get boring if people knew what to expect all the time.

Content?

[Alcon]

Candidate who brings about change:

Montana: 55%

South Dakota 49%

Sorry, what's the question on that?

[Alcon]

Montana Clinton voters:

Obama 60%
McCain 25%

Should Obama pick Clinton as VP? (all respondents)

Yes 49%
No 45%

[Alcon]

Is Lou Dobbs drunk?

[Alcon]

Are you...sure?  They're usually more subtle about that.

That would point to Obama +10-12.

[Alcon]

Are you...sure?  They're usually more subtle about that.

That would point to Obama +10-12.

Turn to Fox News if you have it.  I'm still believing CNN's current count of 2,108

Sorry, I'm doing unclear quoting - I was talking about the Obama exit poll.

[Alcon]

Economic voters in MT:

Obama 49%
Clinton 45%

Where is the rest of the vote?

I think it's time we start ignoring exit polls again