Poblano: Obama would have won Michigan

[woolldog]

http://www.fivethirtyeight.com/2008/05/michigan-what-would-have-happened.html

[BlooInBloo]

1. Interesting use of k-nn analysis... Here's background for any interested...

http://en.wikipedia.org/wiki/Nearest_neighbor_(pattern_recognition)

It would have been nice if he had discussed his use of the Euclidean metric, as opposed to other reasonable possibilities, but whatever - it's still a cool use of the tools.


EDIT: It would have also been nice to see if other classification algorithms yielded substantially the same result.

[Johnny__Motown]

2. I have been saying this for months.. but does anyone believe me?

Thanks for vindicating me

[PBS Poll-435]

3. The problem is the time-line.

Using congressional districts that voted in March to project what would have happened in January, irresponsibly distorts the data.

[MercutioATC]

5. I agree.

It's revisionist history, and it doesn't serve us at all.

[woolldog]

7. Haven't the demographic patterns been stable throughout the enitre race?

This analysis would also apply to a revote if one had been scheduled in May or June.

[BlooInBloo]

10. The criticism is irrelevant - it's a projection. That means "mathematical guess"...

There's no such thing as *proving* it except to actually carry out a vote.

[PBS Poll-435]

11. No,

In Michigan, Clinton won 30-44 year olds.
In Pennsylvania, Obama won that group.

In Michigan, Clinton won 30% of the African-American vote.
In Pennsylvania, Clinton only won about 10% of the African-American vote.

In Michigan, Clinton won 73% of non-college education voters.
In Pennsylvania, Clinton won 62% of non-college education voters.

[woolldog]

12. Clinton won 30-44 year olds, 30% of blacks, 73% non college grads?

You mean out of the people who went out of their way to vote when she was the only one on the ballot? That's meaningless.

[PBS Poll-435]

13. You asked a question, and I answered it.

Why so snippy?

[RaleighNCDUer]

14. Of course, if they hadn't moved their primary up in violation of the

rules, we would not be talking about what "would have happened in January". Which is, of course, the distortion that Hillary wanted in the first place.

[BlooInBloo]

15. It doesn't matter. The variables he used for the analysis didn't change substantially...

over the last few months - the more so when you consider their relative weighting.

[RaleighNCDUer]

18. That's what I mean. That is why Hillary wanted MI and FL up front

in the early primaries - she had a strong advantage of just plain name recognition, and people who are underinformed will vote for the familiar. They are both big markets where her early financial advantage would enable her to overwhelm her opponents with media coverage, as she did in California. As the campaign progressed, had they been later in the season, the basic demographics would balance that out, removing that advantage for her.

Had Obama campaigned in MI he would have gotten substantially more support, and she, less, whether it was early or late, but the later in the campaign the better, for him.

[PBS Poll-435]

16. Alls I know is....

Until Super Tuesday, Clinton had a 100% winning record in States that border Canada.

[Cha]

4. Good! He can win it

in the General too.

[gristy]

6. Very impressive methodology and report.

I am so impressed.

[woolldog]

9. Some background on Poblano:

http://www.nationaljournal.com/njonline/mp_20080507_8254.php
THE NATIONAL JOURNAL

May 8, 2008

Over the last week, an anonymous blogger who writes under the pseudonym Poblano did something bold on his blog, FiveThirtyEight.com. He posted predictions for the upcoming primaries based not on polling data, but on a statistical model driven mostly by demographic and past vote data. His model predicted a 17-point victory for Barack Obama in North Carolina and a 2-point edge for Hillary Rodham Clinton in Indiana.

Critics scoffed. Most of the public polls pointed to a close race in North Carolina. Looking back at Poblano's efforts in Pennsylvania, pollster Dick Bennett decried the models as "stepwise regression run amok." Slate's Mickey Kaus predicted failure for "a sophisticated model that ignores... what's been happening in the campaign. Like Rev. Wright."

But a funny thing happened. The model got it right.

Obama carried North Carolina by 14 percentage points (56 percent to 42 percent) and Clinton prevailed in Indiana by exactly the 2-point margin Poblano predicted (51 percent to 49 percent). Moreover, the predictions were more accurate than any of the pollsters' results, as indicated by the graphic below (modified from a chart created by Brian Schaffner of the Center for Congressional and Presidential Studies). The model was certainly closer to the final result than our Pollster.com trend estimates based on all of the available public polling data (we had Clinton leading by 4 points in Indiana and Obama up by 7 in North Carolina).



I also asked SurveyUSA's Jay Leve how the model would have ranked on the scorecards he created to rate pollster accuracy in Indiana and North Carolina. Had it been a poll, Poblano's model would have been the top ranking "pollster" on 13 of 16 accuracy benchmarks applied in the two states.

How could that be?

When a statistical model succeeds in this way, it is usually because the it manages to quantify reality in some important way. In this case, the fundamental insight captured by Poblano's model is the remarkable consistency of vote preference in the Obama-Clinton race among key demographic subgroups.

<snip>

[patrice]

8. VERY kewl site!! Thanks much! nt

[TahitiNut]

17. As a native Michigander (adoptive Californian), that analysis seems very reasonable.

I'm (of course) fairly familiar with the entire state, with relatives living in virtually every Congressional District. I'd say any variances from that analysis wouldn't be enough to change the resulting delegate allocation. It would, of course (again), only be true IF THEY'D CAMPAIGNED IN THIS STATE. They didn't. Thus, NO DELEGATION can be rationalized in principle and equity.

[book_worm]

19. I've always felt Michigan was an Obama state

because it does have a sizable number of college students and African-American voters and Obama would have done well enough with white voters generally to won it.

[Alexander]

20. Well, Jesse Jackson won Michigan in '88, so that's not surprising.

:sarcasm: