Princeton Election Consortium

A first draft of electoral history. Since 2004

PEC on the BBC (and other venues)

November 7th, 2012, 11:17am by Sam Wang


At noon (in about 45 minutes), I’ll be on BBC Radio 4. At some point soon I’ll be on ‘More Or Less,’ a programme on the use of numbers in the news.

Adam Gopnik in the New Yorker. Also Mashable, Gawker, the Chronicle of Higher Education, Simply Statistics, Business Insider, NYTimes, Forbes. It’s fun to see the regular press get so interested in good statistical analysis.

Photo: Ian Holmes. Here’s a cake video. Happy birthday, Jay!

Tags: 2012 Election

266 Comments so far ↓

  • Stephen Coltrane

    CBS reports that Romney was “shellshocked” by the defeat:

    http://www.cbsnews.com/8301-250_162-57547239/adviser-romney-shellshocked-by-loss/?tag=socsh

    Significant paragraphs: “They thought intensity and enthusiasm were on their side this time – poll after poll showed Republicans were more motivated to vote than Democrats – and that would translate into votes for Romney.

    As a result, they believed the public/media polls were skewed – they thought those polls oversampled Democrats and didn’t reflect Republican enthusiasm. They based their own internal polls on turnout levels more favorable to Romney. That was a grave miscalculation, as they would see on election night.”

    Nothing we didn’t know, of course. ;-)

    • Michael

      ” poll after poll showed Republicans were more motivated to vote than Democrats”

      Of course, poll after poll also showed that more Democrats were planning to vote than Republicans. More voters who are motivated beats voters who are more motivated. But there we go with the math again.

  • Fabulist

    http://gawker.com/5958847/lets-play-drunk-nate-silver-the-hilarious-new-twitter-game

    See, Dr. Wang? Sometimes it’s better to let someone else take the media heat.

    • RandyH

      Ridiculous how much crap they gave Nate,Nice guy just like Sam who just did his math!

      Repubs don’t believe in that remember.

  • Joshua

    Sam -

    Don’t know if you’ll get this far down in the comments – but if you (or anyone else) can tell me how to find the stats on the final tally for the House of Reps popular vote, it would be appreciated.

  • Iseeurfuture

    MR has conceded Florida. Add 29 more votes to the tally.

    • 538 Refugee

      Conceding in this context just means he won’t dispute the results. The vote will still play out. If the outstanding votes give Romney the edge, that is how it goes. Sam is still hoping…. ;)

  • Reason

    Sam, may as well move FL to blue. Romney conceded FL as he admitted the outstanding absentee ballots were in all democratic counties. WOO HOO!

  • Steve16748

    Reason noticed byUSA Today, today.

    Pundits should give more weight to polls even when they don’t like the results, Wang says — comparing it to ignoring hurricane forecasts because you don’t like the idea of your house blowing down. “Their gut instincts were wrong, and it would be a good idea for them to ground their judgments in quantitative data.”

    Well said, as usual, Professor.

  • Larry Gonzales

    Thanks so much, I made so many bets with Republicans at the Rose Bowl. They kept saying Romney would win. I showed them your prediction and one after the other exclaimed, no way! I bet each 5.00 that Obama would have around 334. Let’s just say my next game will be on them. I even got called crazy for taking this bet.

  • 538 Refugee

    Sam and Andrew are silent. You think they are hoping we will go away now? ;)

    • Steve16748

      Maybe they are 538ref. I think I feel like you, I would pay to hang out and comment here, but not at 538.

  • Olav Grinde

    Breaking news: Senate Minority Leader Mitch McConnell says he’ll double his efforts now to ensure that Obama is a two-term president!

  • Steve in Colorado

    Looking at the EV numbers and the margin of victory, it appears as though Colorado was the tipping point state again this year, like in 2008.
    Vote was closer in OH and VA, so 303-31= 272, still an Obama victory. CO was the next closest, so 272-9= 263, not enough, so it was CO who tipped the country over to Obama.

    • E L

      I saw adds for Rehberg 2 hours after polls closed in Montana from a SuperPac. I think they bought blocks of time.

    • Reason

      Steve, uhh, what? He won VA by 3% of the vote and Ohio by almost the same. Where do you get. Also, FL was just called for Obama. So, where are you getting this?

    • Michael

      I think you’re misunderstanding Steve’s point. Rank the states by margin of victory and Colorado is the one that puts the President over the top. He could have lost all the states that he carried by smaller margins (VA, OH, FL) and still been re-elected.

  • Hayford Peirce

    Now that the election is over, whatever happened to the deluded Colorado professors who predicted that, because of economic factors, Romney would easily win the election? Kenneth Bickers and Michael Berry are their names. By cherry-picking their data (with great skill, I imagine) they retro-projected all the election winners going back to, what?, 1940. Flawlessly! The right-wing tinfoil-hat brigade was fond of bringing them up.

    Has anyone heard of any comments from these gentlemen subsequent to the events of November 6th?

    Which reminds me: there was *another* professor who, six months or so ago, said that for various factors Obama would clearly be elected and that no one even had to bother to look at the polls. He claimed that he had been calling elections *in advance* correctly back to, oh, 1980 or so. A contemporary Hari Seldon….

    I’m pretty sure that Nate Silver wrote an entire column about this guy, basically accusing him of hindsight and fudging on his claims….

  • dava

    Hi all,

    I’ve enjoyed this site immensely! Thanks Sam and Andrew..
    I’ve given some thought to the decline of the MM after the 1st debate.. Remember, this is when pollster’s were making the switch from RV’s to LV’s. I’m thinking this in itself was a substantial contributor to this decline.

    Dava

  • Craig

    Latest vote count has Mitt perilously close to 47% (point something).

    Is anyone else still seeing Republican ads on TV? They can’t possibly have been this incompetent, could they?

  • securecare

    About the House vote.

    “Although a small number of ballots remain to be counted, as of this writing, votes for a Democratic candidate for the House of Representatives outweigh votes for Republican candidates… 53,952,240 votes were cast for a Democratic candidate for the House and only 53,402,643 were cast for a Republican — meaning that Democratic votes exceed Republican votes by more than half a million.”

    http://thinkprogress.org/justice/2012/11/07/1159631/americans-voted-for-a-democratic-house-gerrymandering-the-supreme-court-gave-them-speaker-boehner

    • wheelers cat

      that is Dr. Wangs redistricting advantage. he predicted this as well.
      its anti-democratic but so is voter suppression.
      will we see SCOTUS act on either?

    • 4BOinKY

      Thank you for posting this.

    • Michael

      There’s also some self-disenfranchisement revealed by those numbers. Ten million people who voted for President didn’t vote in their house races.

    • 538 Refugee

      Or their votes weren’t counted?

    • securecare

      “…Ten million people who voted for President didn’t vote in their house races….”

      Undervote across the House races was significant, interesting.

      There is also this which I think is worth noting. There should be numbers available that reflect the possibility of this being correct.

      “…the punditry is missing one of the available analytic “angles” in thinking about this election. IMO a lot of rural white people may have stayed home rather than vote for either BHO or Romney…If you seek a net effect from my hypotheses, it seems likely to me that the conclusion reached by quite a few citizens was that they preferred not to vote at all rather than vote for either of two unacceptable candidates.

      That effect may have played a significant role in the result…The potential white vote out there in “flyover” country may be larger than what was showing in this election….”

      http://turcopolier.typepad.com/sic_semper_tyrannis/2012/11/httpenwikipediaorgwikiunited_states_presidential_election_2012.html

  • wheelers cat

    Since Dr. Wang is apparently very busy with his new and well deserved fame, allow me to flash my nerd chops.
    The GOP is about to go to war, within itself.
    http://thehill.com/blogs/ballot-box/presidential-races/266649-conservatives-lambast-romney-vow-to-take-over-gop
    Pascal Boyer, cognitive anthropologist and evolutionary biologist on fundamentalists (e.g. the Tea Party)
    “We can get a better sense of fundamentalist reaction if we describe more precisely what is so scandalous about modern influence in a religious milieu and if we take into account that the reaction is a matter of coalitional processes. The message from the modern world is not just that other ways of living are possible, that some people may not believe, or believe differently, or feel unconstrained by religious morality, or (in the case of women) make their own decisions without male supervision. The message is also that people can do that without paying a heavy price. ”
    Tea Party “conservatives” view moderation as defection. So expect Romney to pilloried as insufficiently conservative by the Tea Party segment of the GOP.

    • 538 Refugee

      I mentioned this somewhere in this jumbled heap. I look for a third party effort out of them next cycle because the Republican party isn’t ‘pure enough’.

    • wheelers cat

      there will be war first.
      In 2010 the GOP elites used the tea party to GOTV and lather up enthusiasm. Rove didnt stab their darling (Palin) in the back until two weeks before the election. The elites may run the same strategy for 2014. Alternatively the elites may try to force the base towards to the center with an eye to 2016.
      The only significant bounce Romney got was his 180 to the center at Debate1.
      who can say?

      “War is like any other bad relationship. Of course you want out, but at what price? And perhaps more importantly, once you get out, will you be any better off?”
      quellcrist falconer, Campaign Diaries

    • wheelers cat

      A declaration of war from Mark Levin.
      http://www.realclearpolitics.com/video/2012/11/07/mark_levin_gives_unvarnished_truth_on_romney_loss.html

      he just cried havoc and let slip the dogs of war.
      538 refugee may be right, and this will result in a third party, but there will be war first.

  • wheelers cat

    The Nerdiest Election Ever
    http://andrewsullivan.thedailybeast.com/2012/11/nerdiest-election.html
    This really is the Triumph of the Nerds– and not just the Poll Jedi like Dr. Wang.

    I personally know its not all about micro-targetting because I worked on GOTV here in Colorado, and the email and phone trees are purely SNT (social network theory).

    • counsellorben

      wheeler,

      I saw the SNT in the emails I got from the campaign while poll watching on election day, but the GOTV canvassing targeting was extraordinary, going after less likely voters, and the scripting was excellent.

      I did not make many contacts, but those I did make were well-leveraged, and each contact had a significant probability of moving one or two more less likely voters to show up and vote.

    • wheelers cat

      in the end it comes down to red/blue genetics. common sense and authoritarianism (top down dissemination of power and info) worked well in the past for the GOP.
      But deeply ingrained belief in privileging common sense and obedience to authority utterly fooled the GOP about Team Obama’s grassroots advantage and the reality of the math.
      this is genius here.
      “Obama campaign officials noted Wednesday that they had years to build up a field operation that was often not visible to the other side. The director of Obama outreach to African-Americans in Ohio oversaw a barber shop and beauty salon program that helped register new voters and distribute literature. A Congregations Captains Program helped the campaign arm supporters in traditionally African-American congregations with what they needed to mobilize other parishioners.”
      it was not only invisible to humans with organic conservative tendency, it was INCONCEIVABLE.

    • counsellorben

      A great analysis.

      So now it is time for Obama for America to become Organizing for America, and to use this infrastructure to support state and local Democratic candidates (especially with GOTV in off-year elections) starting in 2013.

    • Shawn Huckaby

      I’d like to see it go beyond just elections and extend into creating networks that can be mobilized to push for specific legislative and/or societal goals. We need to keep our eye on the ball and continue to hold our representatives accountable. A rapid response type network will be invaluable going forward.

      For a glimpse of the enormity of the task ahead you need look no further than who the GOP is planning to put on the House Science and Technology subcommittee.

    • E L

      @Cat Republican internal war? Let us hope it lasts at least from 1618 to 1648.

  • 4BOinKY

    Does anybody know where to find raw totals for all House races? I want to tabulate the seat distortion created by calculating the difference between the projected House seat allocation and what that would look like if you allocated them proportionally.

  • counsellorben

    Sam,

    Congratulations!

    Going forward, I would want to hear your opinion re: the paper from David Rothschild and Justin Wolfers (http://www.brookings.edu/~/media/research/files/papers/2012/11/01%20voter%20expectations%20wolfers/01%20voter%20expectations%20wolfers.pdf) regarding methods to incorporate expectations questions into opinion polling.

    This seems as if it is right in your wheelhouse, and could represent an important methodology for opinion polling and poll aggregation in the future.

    My only caveat: I don’t know if you can continue to be agnostic about your polling data sources, as I expect Florida would have been on the other side of the razor’s edge in your model without the one-sided house effects of certain pollsters. However, trying to cure this problem may be worse than living with it.

    Looking forward to your next endeavors!

  • wheelers cat

    Romneyworld: The Reckoning
    http://dyn.politico.com/printstory.cfm?uuid=5DFEA4D9-B969-4382-B8D9-3994D048F61D
    “On a rainy and dreary post-Election Day here, as Romney staffers *turned in their Blackberries* at campaign headquarters and huddled for a late-afternoon staff meeting, most were more focused on housekeeping than re-litigating strategic decisions. They were disappointed but circumspect, although a more exhaustive review is planned in the next few days.”

    AHA! that explains it.
    Team Obama had iphones.
    /kidding, but really not

    I have long thought that the enemy of conservatism isnt liberalism– its evolution. The GOP lost in almost every dimension of 21st century culture– they lost on twitter, they lost on facebook, they lost on technology, they lost on pop music (even Christie fell to the Boss), they lost on hollywood, they lost on academe, they lost on science, and in the end, they lost on the math.
    The only place they can win anymore is in church.
    Its what I said about the farmer method.
    The subliminals were against them.

    And now the GOP is in for a long hard winter with no provisions.

    tant pis

    • wheelers cat

      not to mention how badly the pollsters whiffed on the hispanic vote.
      http://2012.talkingpointsmemo.com/2012/11/poll-latino-vote-devastated-gop-even-worse-than-exits-showed.php?ref=fpa
      87-10!!!!
      holy kurtosis.

    • wheelers cat

      And i’d like to point out– Scott Rasmussen KNEW he this weakness in his LV model in 2010 in CO and NEV.
      the ninth circle of hell is too good for him.
      but maybe the Koch bros and Sheldon Adelson will take him out.
      ahhh, I love the smell of the free market in the morning.

    • E L

      The Republican solution:
      After the uprising of the 17th of June
      The Secretary of the Writers Union
      Had leaflets distributed in the Stalinallee
      Stating that the people
      Had forfeited the confidence of the government
      And could win it back only
      By redoubled efforts. Would it not be easier
      In that case for the government
      To dissolve the people
      And elect another?
      —Bertoit Brecht

    • grandpa john

      maybe the wingers should rethink their stand on evolution, after all one of the constants that we live with is that things change and that includes politics . where evolution comes in is that in order to survive changes there has to be adaptation else we are left with the old addage “adapt or die
      personally I hope that in the case of the Republican party that they are to pig headed stubborn to accept any blame and refuse to adapt choosing instead to die.

  • E L

    It’s 8:15 am EST and the snapshot numbers haven’t changed. What’s wrong? Oh, never mind….

  • Rudy

    The Consortium: your approach has been validated again and the nerds won. The pundits and talking heads can now marinate in “momentum”, “transition office”. I detected that once PA was announced for Obama, the campaign team had all- knowing smiles that seemed to signal that the battleground States was going to the killing field.

    Sam, food for thought the EV model could benefit from considering externalities, much as one can factor “uncorrelated” noise in estimation theory. Is there a way the model could model in the superb ground game of one campaign compared to the other. To me the main message was the superb market segmentation strategy executed by BO team. Is there a way of factoring this in initial State poll data in your analysis?

  • Princeton Voter

    Congratulations on a very accurate predication. (as expected!)

    One minor note– Is there a typo in Forbes article? Should the median be +2?

    “My use of median-based statistics is a way to get rid of outliers: if the numbers say Obama +1, Obama +2, and Romney +9, the right answer is probably Obama +1, the middle value (median), not the average.”

    Thanks!

    • Duhawk

      No typo, just confusing in the way they wrote it. O+2 is better than O+1 is better than R+9 so the middle is O+1

  • Fabulist

    Just wanted to lend my voice to the chorus. I’ve been a lurker for years, Dr. Wang; your site, and a handful of others, were rare islands in a tossing sea of bulls&%t this election season. Thank you. Checking in with your numbers was an enjoyable part of the day. Err, parts. Like five or six parts. Maybe ten checks, max. Really, I had a problem.

    I also have a theory that wheelers cat is an emergent, grrl-powered, third-culture rogue AI with plans to Occupy the internet, born of social-justice Anonymous actions and melting financial sector computer networks, but this is hardly the place to go on about that, is it?

  • Special Agent Dale Cooper

    Mr. Wang,
    Just wanted to say thanks for all of your hard work…one of the many sites that provided daily, well-thought, and articulate reassurances for me. Here’s hoping that you get some of the same credit and acclaim that Silver is (rightfully) getting as well…be well!

    • Olav Grinde

      Dear Ms Sheckley,
      I for one will give you my gratitude. And, by the way, your posts always provide amusement and often substance, too. I look forward to checking out those links during my next break. Meanwhile, there’s work to do. Cheers!

    • wheelers cat

      Ms. Jay, I have enjoyed you so much.
      I will miss you.
      Im packing up my election camp gear even as we speak…but I think we can look forward to a campers reunion in 2014.
      Meanwhile, the Meta-Margin on my Colbert prediction is shooting up.
      darn, I could have made a killing.
      ;)

  • Joe from Florida

    I came across this article,
    http://www.chicagotribune.com/news/politics/sns-rt-us-usa-campaign-rovebre8a7011-20121107,0,2643102.story

    They claim that American Crossroads was1% effective, whereas Crossroads GPS achieved 13%. Neither did that well, but GPS did better. I wonder how much of that might have been because informed donors were steered that way from here.

    Also, did Romney and his top campaigners really swallow their own koolaid or were they just cynically creating the momentum perception all along?

    • Matt McIrvin

      I’ve heard that the mere experience of working on a political campaign can often convince people inside the organization that they’re going to win, regardless of the facts on the ground. You get energized by being on the team.

    • Olav Grinde

      Koolaid? Nah, Romney & Rove & Co drink Chivas, a foul drink that is heavy on marketing and comes woefully short on quality. Just like the Romney/Ryan campaign.

    • Olav Grinde

      Ah, forgive me. Mitt might raise a glass of Martinelli’s, foregoing the Chivas. Come to think of it, that’s the only principled choice I’m aware of, but a policy I’d support.

  • AySz88

    Uh-oh… That brag being picked up in the media about perfect popular vote might have been premature… The total from the AP now via C-SPAN says 51.2% (51.199 to be more precise) – that isn’t quite braggable (538 was guessing 51.2(6)ish). Not that any of this actually means anything in any statistical sense – more of an entertaining _fill-in-the-blank_ measuring contest.

    • AySz88

      I should have said that those are two-party share calculated from the vote totals: 60,652,238 vs. 57,810,407 at this moment.

    • 538 Refugee

      Yeah, in the end the general outcome was way closer than the talking heads. I’m waiting to see the final analysis. That is probably weeks away. Ohio doesn’t even count their provisionals for a while. Then we see how the polls did compared to the final results. In the end, it is the polls being tested here. The aggregation models only try and make sense out of the competing polls. In general the aggregators have proven accurate enough. My general sense is that on the whole Romney may have outperformed his polls. Maybe that is because I live in Ohio? Seems there were a few states a little closer than I thought they would be though.

    • Ms. Jay Sheckley

      Yeesh, four days ago they were all “Romney has momentum.

    • Joel

      @538 refugee

      Romney outperformed in Ohio , but of the swing states that’s the only one I can think of. He obviously underperformed in Florida (I had him winning there), but also came below the PEC margins in Iowa, Nevada, Pennsylvania, Virginia, and did much poorer than expected in New Hampshire.

    • Matt McIrvin

      Absentee ballots, which tend to get counted later, might boost Romney’s totals a little; I think a Republican edge in absentees has been the pattern in previous cycles, probably because of elderly voters and military households. Though you never know.

    • Matt McIrvin

      Ohio was also very unusual in that the drop in Obama’s margin predicted by the polls from 2008 to 2012 was abnormally small. So it may just be a sort of regression to the mean.

    • Matt McIrvin

      …on Tuesday night it seemed like Romney might be outperforming his late polls in Virginia, but I think that was largely the effect of election-night numbers from Fairfax County and Norfolk coming in very late. The same thing happened in 2008.

    • MAT

      @Matt

      When absentees are counted depends on the state. Here in NC, we count all of the ones we have received as of 2 pm Election Day on Election Day – I spent a very long and trying several hours cross referencing and feeding a bunch of folded absentee ballots into the voting machines Tuesday afternoon, so that we could retrieve the results when the polls closed later that evening. Provisionals and any remaining absentees we get that were postmarked in time get totalled up by next Friday when we certify the official results). So in our case, the bulk are already baked in. Other states have different methods.

      Also, for @Olav, who I know is concerned about such things, I’m just about to leave for what we call a hand to eye recount – after each election, the state Board of Elections sends us a randomly selected statewide race, where we hand count the results in one or more precincts (also randomly selected) to make sure the scanned voting machine totals match what the ballots say. A sanity check on the voting equipment,as it were.

  • E L

    I’m 100% Irish. Grandparents were off the boat. So here’s my final salute to George Wills, Dick Morris, and, in particular, the MSNBC’s Joe Scarborough: http://www.youtube.com/watch?v=jbrzZWLu6Qw&list=FLoFPKOpwl8m3oxxucTeMXnA&index=40&feature=plpp_video

    May Fortuna smile on you, Sam and Andrew, as kindly as she has on our President and all who come to PEC.

  • Alan Cobo-Lewis

    The New Yorker comment about unpredictability of lots and lots of actors was funny. Law of Large Numbers, anyone?

  • Ms. Jay Sheckley

    NICE PICTURE :D

    Say, this thread is for_ all_ venues mentioning PEC, right? Would you believe PEC is mentioned in a new youtube video?

    John Sawyer and I just couldn’t cut those pretty charts. So, in one addle-pated take, we made our error-ridden YouTube debut, extolling Sam Wang’s historic Princeton Election Consortium and its anxiolytic charts.

    As the Talking Heads said,”My god! What have I done?”
    This: http://www.youtube.com/watch?v=TpLi4__okCs

    Can you spot the data-inversion?

  • Dean

    That’s a great cake! It would be really cool to have a t shirt with PEC and the Median EV Estimator with the phrase, “It’s the Math, Stupid,” or some other catchy and true phrase.

  • Deb

    Thanks Sam, first of all for making my yesterday far less anxious than it would have been, and secondly for your spectacular data crunching. (never heard of you until I read the Chronicle article yesterday)

    • pechmerle

      Got a good laugh from the one and only commenter at Forbes on the aggregator story: He “knew” that Ron Paul was the only Republican who could beat Obama!

      It’s thinking like that that has made the Republican party what it is today.

    • Ms. Jay Sheckley

      An enjoyable read. Thank you.
      Maybe we’ll be getting more Sam, not less.

  • Aravind

    Wish we elected a president once in two years instead of four just so we can come read Sam’s posts and the status of the MM.
    Habitually I keep checking here every often.

  • jharp

    Many many thanks to you Sam.

    You are a hero.

    Peace and good health to you and your loved ones.

  • Maurice Pinzon

    Prof. Wang,

    Adam Gopnik in that article you linked to in The New Yorker called your approach “an entirely different…epistemology,” and a challenge to the punditry class. Now that’s impressive and it reminds me of that old song, “Video killed the radio star.”

    As someone previously mentioned in another post, I also hope you work the numbers on other interesting projects. I’m eager to continue to read. As it stands, I’m having withdrawal systems already.

    -Maurice

    • Marty

      Yes! I’m trying to deny this, but it is true–withdrawal symptom–I’m obviously addicted to this site. Not looking for a cure, though.

      Marty

  • Muhahahahaz

    I extrapolated the current numbers from each state, and I came up with the following.

    Obama: 67,580,339 (51.553%)
    Romney: 63,508,751 (48.447%)

    For instance, in California the current count is 5,573,450 to 3,635,571, but that’s only estimated to be 69% of the vote.

    (Yes, some sites mention that “100%” of CA is reporting, but that just means every precinct has returned some portion of their votes. They have until December 7th to finish counting all of the votes, in CA.)

    If we assume that the current percentages will stay about the same, e.g. 59% to 39% in CA, then we can estimate how many of the remaining votes will go to each candidate. For instance, in CA this would give a final total of 8,077,464 to 5,268,983.

    Doing this for each state, we find that there are about 13 million votes remaining, with about 7 million going to Obama and about 6 million going to Romney. Of course, these assumptions may be wrong (and the “percent counted” estimates could be off for some states), but this leaves us with the totals above. :D

    Anyway… does this seem reasonable? Discuss…

    • TAW

      I’m not sure about the extrapolation to get a popular vote total, but am sure that this is a topic that needs more exposure in the main stream media.

      After the MSM gave up on the possibility of a Romney win in the EV, the reported virtual tie in the popular vote became a huge talking point.

      And now the ‘razor thin’ margin is porking up.

      I did notice that some pundits were happy to recast this as a beat down of the GOP, which sets up discussion of what they can/have to change.

      Seems important to me.

    • Matt McIrvin

      That just illustrates further the innumeracy of pundits: they were emphasizing that Romney was leading in the popular vote in the hours before the West Coast had been counted, even though it was obvious that California alone was going to put Obama’s PV total over.

    • Duhawk

      At the very least, your percentages are wrong. They shouldn’t add up to 100%, they should add to somewhere between 98 and 99%. Other than that I think this is about as good an estimate as you can get. The only other way would be to find where the votes are still out and use 2008 numbers scaled using this years current percentages, but that would take a lot of work (if it is even possible) and in all likelihood would end up at your projection.

  • Huey

    Thanks Dr. Wang & Andrew. Been following you since 2004. Since then I never listen to or follow the MSM during an election year.

    You guys, Nate Silver and Drew Linzer keep me composed during the elections as I always trust the math!

  • Paul C

    Congratulations.

    May I humbly suggest you stick clearly to the limits of what poll aggregation is good at, and that you differentiate, as you have on the blog, between “pure” poll aggregation and hybrid aggregation/modeling, and that you be particularly careful about discussing when poll aggregation finally converges upon and becomes good at predicting actual outcomes.

    Some poll aggregators/modelers have a tendency to overemphasize the predictive possibilities of poll aggregation. This is actually not its strength. Don’t get sucked in.

    On topics there is “what’s so” and there is “so what?” For too long, pundits and reporters have been misreporting what’s-so regarding what polls say. What polls say is best determined by poll aggregation (without modeling). Reporting that belongs to journalists, how it gets reported, belongs to the standards of the news organization.

    The question, “So what?” still belongs to pundits.

    Good luck, and, again, congratulations.

  • poll_wonk

    Thanks Dr Wang!
    You got many of us thru some uncertain hours.

  • Hedgehog

    Congratulations to the winners, this has been very educational. Now if I can just figure out how to use it to change results rather than just predict them.

    • wheelers cat

      hedgehog, you were WRONG.
      Pure-D unadulterated WRONG.
      Your wrongness pegged the wrong-o-meter at the apogee of the wrong curve.
      now go away.

    • wheelers cat

      and besides, you were rude.
      go bet on Intrade or sumthin’

    • Hedgehog

      @wheeler – still classy even in victory I see. As usual your factually incorrect since I long since acknowledged that these models looked correct even though I hoped they weren’t.

      However, it’s not your board, I’ll stay or go until those actually do own it disinvite me.

    • wheelers cat

      hedgehog you were the one bitching about me using french aphorisms and big words you didnt understand.
      I’m an anti-capitalist and an Occupier. I cooked pinto beans and gifted venison for Occupy Denver and slept under a tarp in the snow.
      I can smell your glibertarian free market Intrade stench a mile away.

    • wheelers cat

      tete du merde.

    • Sam Wang

      Well, that got through moderation.

    • Hedgehog

      Actually, it was the other guy at Votamatic who complained about your French (and how bad it was). I was the one who pointed out how hypocritical it was of you to complain about the rudeness of others. You seem to insist on proving me right. Note my lack of swearing at you in any language foreign or domestic.

      Congrats on being such a hard core Occupier – we’re all impressed.

    • Ms. Jay Sheckley

      Hedgehog wrote,” we’re all impressed.”
      I see you’re speaking for those who aren’t impressed with wheelers cat.
      Sorry to hear you have tapeworms.

    • wheelers cat

      “Well, that got through moderation.”

      yeah, but Klytemnestra didn’t. :(

  • Anita

    Awww! The cake!!

  • garryb

    Many thanks not only to Sam for the analysis but to all the contributors to the site for the civil discourse.
    I’m hoping that the site will continue to discuss issues other than the elections. I dont want to wait two years to discuss numbers again.

    • Marty

      I agree. And while I guess we can’t ask Sam to give up his day job to moderate such a discussion, I don’t see why there couldn’t be a continuation with some round-robin moderating? There are so many things I would like to get a better understanding of, from a quantitative point of view, without a political approach . For example, the ‘cap and trade’ approach to carbon emission controls.

    • Simone

      I am actually replying to Marty’s reply to this comment to say that I am intrigued by the idea of a quantitative discussion of ‘cap and trade’ (or other) approaches to carbon emission controls. Should Sam be in search of another topic that needs a boost from a fresh-thinker over the next few years…

      Also, kudos Sam and Andrew. Your site has kept me sane these past few months! Thank you for that.

  • 5k0rp

    Nerds: 1
    Pundits: 0

    This was a trailblazing effort. We will see the mainstream media pay a lot more attention to this sort of analysis in future elections.

    • skmind

      Wanna bet?

      The media is reliant on selling their punditry.

      There is NO way they let boring math replace entertaining pundits.

      http://bit.ly/RJIATA

    • jayackroyd

      Oh it’s worse than that. I don’t watch much at all anymore, but I had CNN on a month or two ago, and realized that the content (talking points delivered by competing operatives) was the same as the advertising (the same talking points delivered by marketers).

      And now we’re going to see a jump shift (saw it on election night, actually, with Brokaw among others) to the need to dismantle the New Deal.

  • Bruce Wayne

    Any thoughts on the huge drop in voter turnout?4 years ago, Obama got 69M to McCain’s 59M. This year it’s looking like 60M to 57M for Obama. The total is slightly less than in 2004.

    Did some 11 million voters fall off the earth over the past 4 years?

    • Hedgehog

      I’m thinking it’s cause all the action is in 10 states or so. Where I’m living, there’s no doubt who was going to win every race – people here only voted out of a sense of duty. Maybe the people who want to ditch the EC are right. I’m tired of hearing about Ohio and Florida.

    • Michael

      Turnout is not down. There are millions of votes that haven’t been counted yet. Probably more than 3 million left to count on the west coast along. California won’t certify a final count for a month.

    • Joel

      1. Hurricane Sandy suppressed turnout in the (generally uncontested) northeast.
      2. All the votes haven’t been tallied yet. Wait a month or so.

    • Matt McIrvin

      I’ve been hearing that minority turnout was actually UP from 2008. Which, if the total is equal or lesser, would imply that white turnout was down.

    • David

      Young voter turnout was up to 19% from 18%! So much for the “young voters won’t come out again” argument.

  • wheelers cat

    im done with politesse.
    drink up me hearties.
    its even better than tequila, and lower in carbs.
    http://peterjsullivan.files.wordpress.com/2008/11/gop_tears.jpg?w=200&h=350

  • Richard Vance

    So sad we didn’t retake the House. I recall Sam guessing that the disaster 1st debate actually did our down ticket races a favor as it rejuvenated the Romney campaign and kept the focus upticket thus denying the Republicans all that down ticket money. It sounded good at the time. But it looks that what really happened was the 6%MM prior to the debate had the Democrats all across the icket energized and our House hopes were up over 50%. The president’s horrible (he was an empty chair) performance in debate 1 and the precipitous decline in MM worked across the board… Now with the gerrymander of Republican 2.5% advantage the final MM was insufficient for the President to have coattails. So although the President recovered and won he is responsible for losing the house.

    Harsh? yes. True? yes.

    • Muhahahahaz

      Seems logical, captain.

    • wheelers cat

      no Richard Vance.
      it was the redistricting advantage.

    • Joel

      Dems need to take control of State Legislatures. And then get algorithmic redistricting passed.

    • David

      I think we have possibly entered a phase where the Republican Party dominates at the state and congressional levels while the Democratic Party dominates the Senate and Presidency. This will generate a lot of gridlock, though I’d rather be the Democratic Party in that setup, obviously.

      I think Obama’s organization needs to ramp up on off years. Have to convey to low-turnout voters how important those elections are.

      May have to put money and organization around Blue Dog Democrats against Republicans for congress considering redistricting. Blue Dogs may not be optimal, but they can be worked with.

    • bob McConnaughey

      Not true in NC. We now have 3 Bantustans – err – congressional districts with ~ 75% Dem majority; 8 w/ a very safe 55-60% Rep. majority and 2 that are vaguely competitive. Our house has been “moved” out of a tea-partyette’s district since Pittsboro, a small town 20 miles south of Chapel Hill, is incongruously Dem. :-) and the statehouse wanted a safe district for Ellmer – who’d defeated a traditional moderate/ conservative Southern Dem in 2010. No debate outcome was going to change that – through possibly the presidential outcome might have been reversed.

    • Howie Weiner

      I don’t think the President’s performance had much to do with the House. A 6% MM was not sustainable. The electorate was prepared to give him a MM of about 2-3% as Sam pointed out way back in the summer. The debate performance was the precipitating event leading to its swift decline but the decline was happening prior to debate. Just declined quicker. House was gerrymandering and explains why the number of seats picked up by Dems underperforms the popular vote margin.

    • grandpa john

      David; the reason that reps are dominating at the state level is because the Dem leadership abandoned the 50 states strategy, giving the Reps carte blanche to do whatever they please within state gov.controls Then the dems bitch and moan when the reps do this extreme redistricting and pass voter suppression acts.
      The way to stop this crap is to develop strong dem
      parties within the state and take back control of governors office and legislatures,

  • Olav Grinde

    Memorable quote:

    “You know your party is in trouble when people ask, ‘Did the rape guy win’, and you have to ask, ‘Which one?’
    . — Alec Baldwin