2016 Postmortem
Related: About this forumObama's Data Team Totally Schooled Gallup
In the new issue of Bloomberg Businessweek, I have a short feature about how Googles (GOOG) executive chairman, Eric Schmidt, is investing millions in a new company with Dan Wagner and two dozen other veterans of the Obama campaigns data analytics team. Schmidt was, not surprisingly, full of praise for the engineers, statisticians, and scientists whose work informed the strategy of a campaign that won by 5 million votes. But given the collaborative effort and many moving parts of a modern presidential campaign, it can be difficult to isolate the data teams contribution. One great example, though, is the intricate mathematical models of swing states that Wagner and colleagues built that were meant to offer an alternative glimpse of the state of the race to public polls and even the campaigns internal polls.
As David Plouffe, then a senior White House adviser, explained in my story, the data teams models proved to be much steadier and more accurate than even the traditional tracking polls the campaign was also conducting. A number of Obama vets repeated this claim to me, so I asked them to provide some evidence to back it up, and they did. Here, for the first time, is a chart based on internal data that shows how the Obama campaigns swing state model performed against the much maligned Gallup poll over the last several months of the race. This was the campaigns daily horserace projection of the outcome, based on a nightly survey of 10,000 people.
To me, a few thing jump out: Gallup indicates that the selection of Paul Ryan as running mate hurt Mitt Romney, but Obamas model really doesnt; Gallup suggests, incredibly, that the 47 Percent flap hurt Obama and moved the race back in Romneys direction; and, biggest of all, Gallup shows a huge drop for Obamareally, an outright collapseafter the debacle of the first debate. At the time, Obamas staffers were claiming to the press that, yes, their internal numbers showed the presidents weak showing had hurt his support, but that the fall was brief and quickly stabilized right about where his level of support had been all along. As a reporter, you never know if youre just being spun when campaigns tell you this, because even if they really were collapsing the way Gallup suggests, theyd probably lie about it and say everything was fine, so as not to feed the panic. Based on this data, though, the Obama campaign looks to have been telling the truth.
One last bit of color I wanted to include in the original piece but couldnt fit. David Axelrod, for one, had enough faith in his data guys that he was willing to bet his mustache on it. A week before the election, Axelrod wandered back to The Cave, where the analytics team worked, to ask point-blank if Obama would win Michigan, Minnesota, and Pennsylvania. Newspapers were reporting surging crowds at Romney rallies and waning enthusiasm for Obama. Wagner said he thought he would. The next day, on MSNBCs Morning Joe, Axelrod announced he would shave his mustache if Obama were to lose even one of those states. He won all three.
Source: http://www.businessweek.com/articles/2013-05-31/obamas-data-team-totally-schooled-gallup#r=rss
illegaloperation
(260 posts)Without a doubt that Obama would have had North Carolina in his bag.
DCKit
(18,541 posts)They own that state, as bizarre as that is.
Lordquinton
(7,886 posts)that the first debate thing focuses exclusively on Obama, and ignores the fact that Romney did a complete 180 for it, then dropped the act after. No one talks about that.
TexasTowelie
(112,065 posts)However, take solace that we get to gloat over President Obama's victory in the election. That burns the souls of Teabeggers forever and it can't be taken away!
Now we need to kick butt in 2014 and elect another Democratic president in 2016!
Lordquinton
(7,886 posts)four years from now we'll have a heck of a fight.
emulatorloo
(44,100 posts)Flat out denied things he had been saying. Very difficult to debate someone like that.
Romney was caught in more lies in that debate, and yet people considered him the winner?
dsc
(52,155 posts)The fact is Obama was on a high before the first debate (he was at about 53.5 and won with about 51.7). That is a difference of 1.8 which is less than the margin in NC (2.1%). Now, of course, it is possible he lost more in NC or would have fought harder for NC if he hadn't lost the support post first debate. But quite likely he still would have lost here and there was no down ticket damage done either since we stunk up the joint all by our lonesome selves in that regard.
backscatter712
(26,355 posts)Obama was right on the facts in the first debate. The only thing Romney succeeded in was in delivering a blizzard of bullshit so thick Obama couldn't figure out which of the hundreds of lies to debunk in the time allotted.
Biden, then Obama himself showed how to cut off the Gish Gallop at the knees in the following debates.
Hestia
(3,818 posts)Cha
(297,032 posts)snip from your link..
"To me, a few thing(sic) jump out: Gallup indicates that the selection of Paul Ryan as running mate hurt Mitt Romney, but Obamas model really doesnt; Gallup suggests, incredibly, that the 47 Percent flap hurt Obama and moved the race back in Romneys direction; and, biggest of all, Gallup shows a huge drop for Obamareally, an outright collapseafter the debacle of the first debate. At the time, Obamas staffers were claiming to the press that, yes, their internal numbers showed the presidents weak showing had hurt his support, but that the fall was brief and quickly stabilized right about where his level of support had been all along.
That's what I remember, too.
***snip
"...biggest of all, Gallup shows a huge drop for Obamareally, an outright collapseafter the debacle of the first debate. At the time, Obamas staffers were claiming to the press that, yes, their internal numbers showed the presidents weak showing had hurt his support, but that the fall was brief and quickly stabilized right about where his level of support had been all along.
snip***
Yeah, PBO "lost North Carolina because of the 1st debate"..
TexasTowelie
(112,065 posts)The polling methodology for Gallup and Rasmussen were suspicious throughout the last two months preceding the election.
I recall reading a different article a few months ago. IIRC, the Rasmussen poll calls a phone number and if nobody answers, then they immediately move to the next randomly chosen phone number and don't return back to that original number in their survey. They also conduct their polls within a tight four hour time frame. With those restrictions, their demographics become skewed and a larger percentage of older people are sampled than are in the actual voting population.
There was a similar flaw with the Gallup poll that is related to not allowing for adequate sampling of cell phones. With the high percentage of land lines contacted it also skews the demographics.
Aloha and thanks for your reply. We are creeping up to 4:20 in the Central time zone so it's about time for a catnip break.
davidpdx
(22,000 posts)I'm going to read it after I get my studying done. It sure will be more interesting then my studying.
Ps-I was the 10th Rec
TexasTowelie
(112,065 posts)At this point there are 119 comments on that link if you want to scour the comments. Thanks for the rec.
Good luck with the studying. I'm thinking about taking some data management courses that will be offered online next spring at the University of Texas so I may have to get back into that mindset again. I got my BS in 1987 and a professional certification in insurance data management back in 2004, but it's been a while since I cracked a the text books with the expectation of taking an exam afterwards.
I'm glad the bluster over the North Korea situation has died down for awhile--it looks you called the situation correctly a few weeks ago.
davidpdx
(22,000 posts)I started up again in 2010, seven years after I finished my master's and found the first few months daunting. One possible thing you could do is look at some courses on Coursea and take them as a "warm up". They are free and there are all different kinds of classes. I'm taking some law classes just for the hell of it because I like law and there were only two law classes in my master's work and one for doctoral. The two professors I've had gave great lectures. My international criminal law class is interesting as it has reading, a lecture, and then simulations where you have to argue for or against a specific case in the reading. This is of course watered down from a real law class, but given it's free what can you expect. The classes last anywhere form 5-8 weeks usually.
The online class format has become very popular. My master's work was a combination of online, class once a week, and compressed weekend (Friday, Saturday, and Sunday). My entire doctoral program is online and I contact my professors via the phone or sometimes Skype video conferencing.
Laelth
(32,017 posts)-Laelth
TexasTowelie
(112,065 posts)Gallup seems to have an interest in portraying the race as being winnable by Romney. They probably thought that they could be hired to conduct more polls in individual battleground states if the race was close.
Grateful for Hope
(39,320 posts)Suggests that the only poll we should have watching was the internal tracking.