Princeton Election Consortium

A first draft of electoral history. Since 2004

Blown away

October 29th, 2012, 11:11pm by Sam Wang


Not by Sandy, though the winds are pretty high (check out this excellent moving representation).
I’m more blown away by the web traffic over the last few weeks. Take a look.
The above screenshot is what I see backstage. Traffic has doubled every 1-3 weeks since July; lately the doubling time has been 6 days. We’re running at about three times the traffic we got in 2008. Today, the Princeton Election Consortium received nearly 200,000 views. Note the number of referrals from Reddit, Facebook, and Twitter. That’s a new development. The list used to be heavier on politically-oriented sites such as Andrew Sullivan, Kevin Drum, and Balloon Juice. Liftoff?

Thank you all for reading and visiting. Andrew Ferguson and I will try to keep it interesting in the remaining days of the campaign.

Tags: 2012 Election · Site News

243 Comments so far ↓

  • Olav Grinde

    Sam & Andrew: I would also like to complement you on your excellent maps, where the states are proportioned according to Electoral Votes. It reassures me that the USA has not become as Red as the ubiquitous “geographical” political maps might indicate. :)

    And one more thing: Your patient and lucid posts, seen as a whole, really provide quite an education in the science of forecasting.

    As a Norwegian who followed PEC from abroad in 2004 and 2008, and am now following the US elections from this more turbulent side of the pond, your site invariably offers a calm and sober vantage point.

    For this and much more, I would like to voice my gratitude!

  • ChrisD

    Nate Silver tweet:

    CAN’T BELIEVE METOROLOGISTS USED MATH AND SCIENCE TO PREDICT THIS STORM. THEY MUST BE MAGIC WIZARDS.

    • Sam Wang

      Or perhaps they are wispy and effeminate. Or they are biased toward hurricanes.

    • MarkS

      Good to see you posting, Sam. Hope all is well on the ground in Princeton.

    • Matt McIrvin

      ****ing circulation models, how do they work?

    • wheelers cat

      Or perhaps they are wispy and effeminate. Or they are biased toward hurricanes.

      or asian.
      /smug intellectual elitist snob grin

    • wheelers cat

      or have asian sounding names.

      /sadface

    • orchidmantis

      A new meme for the spammers is in fact that climate science models are all wrong (because: feelings) and so are voting models (because: feelings). And the atmosphere is subjective.

    • Olav Grinde

      Perhaps if we spread the message that oxygen has a liberal bias, then conservatives would… Oh, never mind!

  • Suja P

    I discovered PEC through 538 this election season, and am now a confirmed addict. What attracted me is not only the math and Sam’s insightful commentary, but also the comments section. Kind of tired of the name calling and ‘This is great news… [for whoever is down]” posts elsewhere, and the commenters here really have done a great deal to restore my faith in where this country is headed.

    I do plug PEC (and 538) shamelessly elsewhere, and announced to my handful of friends when Nate and Sam appeared on SciFri. Any increased traffic is well deserved.

  • Florida Swing Voter

    Oh Romney, you were looking so competent after the first debate. Now he’s making bad press and dodging questions about FEMA.

    Floridians swooned for Romney after the first debate despite having Romney as his veep… so I doubt this FEMA situation will do much. :( I’ve lost faith in Floridian humanity.

  • Peter D

    I still believe that Ohio is >3.

    Cell inclusive polls since 10/4:
    N = 11
    mean = 3.2
    SD = 3.2

    Non-cell inclusive polls since 10/4:
    N = 18
    mean = 0.7
    SD = 3.3

    T-stat (2 sample equal variance): -3.8

    Done and done.

  • Huey

    Dr. Wang,

    Are you going to do an update on Senate races this week?

  • wheelers cat

    Look. I know no one wants to say the name of the game. The name of the game is the Sailer Strategy.
    Look at Pinkers puffy history piece on red and blue states. Not once does he mention the elephant in the room. SLAVERY. This is the last time the all white GOP has a chance at beating the demographic timer.
    Its not Dr. Wang and Nate Silver that are going “all in” or “doubling down”– its Rasmussen and Jay Cost and Sean Trende.
    They are *betting* they can eke out just one more national victory on the Sailer Strategy before the demographic timer goes off.
    And part of the Sailer Strategy is you never talk about the Sailer Strategy.

  • ChrisDC

    Desperate right-wingers screaming that Dr. Wang is too effeminate to be trusted on these matters in 3… 2… 1…

  • April Pyatt

    Hi all:

    I am a long time stalker, first time commenter from Indianapolis. I love this site. I believe I like the comment section as much as I like Dr. Wang’s analysis and commentatory. I would like to thank Froggy and ChrisD for providing a breakdown of the State changes each time the MM moves. It is much appreciated! As a complete aside, could someone tell me why Talking Points Memo still has Michigan as a “toss-up”? It is driving me crazy! Thx.

    • mediaglyphic

      April, i think TPM is skewed by a recent Foster Mccollum poll that is the most recent on their list (Oct 22/23). Interestingly RCP also has this poll (but call it Baydoun Foster, the firms name seems to be Foster McCollum White Baydoun (FMW)).

      TPM doing this should boost democratic turnout. Its to no-ones advantage (except Dr. Wang) to tell the truth. Democrats risk a lower turnout, Republicans risk less money and the press risks less viewers!!

  • DP

    I like this site and 538 because I respect solid number-crunching. It’s nice to see someone with a conceptually coherent model and the guts to put it on public display like this. Win, lose or draw, I salute you.

    I’ve noticed that most of the criticisms of such sites that started this last week or so seem depressingly dumb. They don’t even seem to understand what you’re doing. Most complaints seem to be variants of “manipulating data,” which miss the whole point of having a defined model in order to try and let the data speak for itself.

  • Mike B.

    I hope I added my little bit. I have been proclaiming this website on various sites.

  • Mike B.

    I bet Raz is waiting on a bunch of nonsense state polls to push this back down.

    But the model should take care of it.

    That is why I think it works so well.

    • ChrisD

      Here are the last 10 OH polls, working backwards: R2, O4, O3, O4, O1, O2, O4, O2, O5, tie. Ras is first and last in the list. Though they customarily wait a week between state polls, for some reason Ras felt the need to do a poll on 10/28 despite having doing one only five days earlier, on 10/23.

  • virag

    Longtime listener, first time caller here.

    It is a great thing that this site and Dr. Wang’s work are getting more exposure. Finally.

    I appreciate the desire to get it right instead of pushing an agenda, a refreshing change of pace!

  • A New Jersey Farmer

    And by the way, 60,000 of those hits on your viewership graph are mine. Refresh is a wonderful thing.

  • skmind

    All I can say is thank you Dr. Wang. I am a Gary Johnson supporter, but one who would rather see Obama win than Romney.

    No, I do not consider Romney to be particularly worse than Obama, but I do consider that Romney + House Republicans = Disaster because the Senate Democrats will simply cave in.

    As a bonus, I was able to refresh my understanding of basic statistics, so it has been a rewarding educational experience.

  • NickP-LA

    As a democrat, I am very appreciative of this site and others with similar sites. However assuring this is, it still does not make some sense on some level. If this election hinges entirely on battleground states then shouldn’t the winning percentage be determined only by the polling in these states? Someone needs to set me straight.

  • Chris

    First and foremost, thank you for the invaluable services. I have been paying close attention to both your site and the 538 blog over the past two months. Admittedly, I was very alarmed with all of the stories alluding to Romney momentum last week. It seemed like the confluence of the Intrade manipulation of last week, the media’s horse-race obsession, and the obviously biased reports amongst all the news organization from their embedded Romney reporters were culminating into creating a “perceived reality” that was nothing more than an attempt to create the perception of winning. Both you and Nate helped to dispel the spin, and, for that, I am grateful. Please keep up the great work. It is both interesting and, most importantly, necessary.

    Perhaps I was blind to this in past election cycles, but it is obvious to me that there are other forces at work that are attempting to undermine Obama’s chances of winning this year. I, of course, expected as much from the opposition campaign as well as the many commercials from outside groups. What I didn’t expect, however, was to see tell-tale signs of this with the media’s horse race obsession as well as the more subtle slights such as putting a poor image of Obama on the main page of CNN and referring to him impersonally as the “incumbent”…while underscoring Romney’s lead in certain national polls, discounting Obama’s battleground poll leads, and blanketing a politics page with Romney-centered headlines.

  • Pat

    Yay! the meta-margin just ticked to 2.18%!
    Could ChrisD or Froggy break down for us which states ratings have changed? Thanks!

  • Shawn Huckaby

    The unexpected deployment of campaign resources could be due to many factors. It could be just to bluff the other guy into spending more money for a losing cause, or even just a show of strength–we’re doing so well and have so many resources that we can fight you everywhere!

    In addition, we should be watching for hints of desperation, such as Romney’s true doubling down on the the universally acknowledged lie that Chrysler is sending all Jeep manufacturing to China. When every newspaper in Ohio is loudly calling you out on something, and you double your ad buys anyway, those are some pretty big and sweaty dice you’re rolling my friend.

    • Obama 2012

      Romney trying to “expand the map” is not a sign of strength, it’s a sign of desperation.

      McCain went into Pennsylvania late last time… lost it by 11% points anyway.

  • Justin S

    Pollster is back up

  • mediaglyphic

    I wonder if anyone has an interest in saying that Obama is a slam dunk? The press would get less viewership and democratic and republican turnout and for republicans money might be lower. Not exactly mathematical analysis, but some game theory!!

    • wheelers cat

      Game theory +theoretical population genetics.
      Im going “all in”. I’m going to say the nat’l polls are WRONG.
      we shall see if im correct in a week.

  • Andy G.

    It seems at least some of the polls’ accuracy may be affected by Sandy in the coming days, perhaps all the way to the election. (I realize most of the damage is in states that aren’t in play, but still.) Just out of curiosity, how close to the final results were your one-week-to-go estimates in 2004 and 2008?

    (Great site, by the way, I still check fivethirtyeight, but Mr. Silver’s methodology always struck me as a little overcomplicated, and his confidence a little understated. Nice to see someone with a more elegant approach.)

  • Eric

    Allow me to add my two cents: this site is amazing and I thank you for all of your careful analysis as well as the hard work you put into it. It is one of only a handful of new websites dedicated to careful and accurate analyses of the presidential campaign polling data. I think there are two reasons why you’re experiencing such a huge increase in unique visits. First, it is the first election where such websites have become known. Yes, you existed years ago, but the public weren’t aware such websites existed and would not have been interested even if they knew. Second, unlike previous presidential elections it has become PAINFULLY obvious that the mainstream media are deliberately manipulating public opinion to effect the outcome of the election, and it is angering many people. Anyone who is objective and non-partisan can see that the mainstream media have NOT been pro-Obama because they’ve been hyping Romney for weeks now, trying to create the illusion of a “tied race”, which will “come down to the wire”. This illusion encourages Romney supporters, while discourages Obama supporters. But polling and other data have consistently shown Obama has been ahead all along, and his chances of re-election stable for several months. Sadly when it comes to religion and politics people tend to react emotionally rather than intellectually or objectively so they will not be able to accept the cold hard facts. The result? When Obama wins next Tuesday tens of thousands of Romney supporters will claim that it was due to a conspiracy.

  • Matt McIrvin

    BTW, RAND is interesting now: switches between candidates are dead even, intention to vote is dead even too.

    • Matt McIrvin

      …Also, today’s demographic breakdown is great. What population of Americans actually have a majority belief that Romney will win? People over 65.

    • Obama 2012

      if “intention to vote” is really dead even, Obama is going to win this election quite easily.

      that’s a very good indicator for him.

  • Madame Hardy

    It’s a pity there’s no way to track real-world referrers. I got it from my best friend, passed it on to my husband, my son, and my parents, who are going to pass it on to most of the college where he taught…

    (Does anybody else remember the Tom Lehrer song? http://www.youtube.com/watch?v=EDeRYmB4t6Q )

    Thank you so much for your work.

  • Mike B.

    I hope everyone is OK who usually follows this excellent website.

  • Anonymous in Teaneck

    Dr. Wang,
    I’m sitting in my car, charging my phone – and checking PEC. I have no power, but am thrilled that you do! Thank you.

  • Olav Grinde

    Sam & Andrew: I would also like to complement you on your excellent maps, where the states are proportioned according to Electoral Votes. It reassures me that the USA has not become as Red/insane as “geographical” maps might indicate. :)

    And one more thing: Your patient and lucid posts, seen as a whole, really provide quite an education in the science of forecasting.

    As a Norwegian who followed PEC from abroad in 2004 and 2008, and am now following the US elections from this more turbulent side of the pond, your site invariably offers a calm and sober vantage point.

    For this and much more, I would like to voice my gratitude!

  • A New Jersey Farmer

    No power, but the mm is up to 2. Split decision.

  • Brian C.

    Shameless plug: if you are a Sam Wang addict, a Google Chrome user, and need your fix… download my PEC Google Chrome extension:

    http://goo.gl/PGlUf

    Puts an icon in your Chrome toolbar that shows the Meta Margin. Automatically updates every 2 hours. Click the icon and it shows you Obama/Romney EV estimates, electoral map, link to the latest blog post, etc.

    • Pat

      Thanks I have installed your extension a few weeks ago and like it that you can now see the MM without clicking on the icon. Just a detail however: it’s a shame that the last digit is usually not fully visible (partially cut because the display is not wide enough i suppose). For example, for a MM or 1.98%, i can only see +1.9 and the 8 is almost entirely hidden on my browser.

    • Brian C.

      Hi Pat,

      Thanks for the feedback. I noticed that too — seems to happen on Windows but not Linux. Not sure about the Mac. I checked the Chrome API and it doesn’t look like there’s much I can do to control the size of that typeface. I’m going to keep looking for a solution though.

      Please do not hesitate to let me know about other things you think could be improved — there’s a “Bug report” feature on the Web Store page (right click the icon, click Princeton Election Consortium 2012 and it will take you there).

      -B

    • Pat

      I’m on Mac

    • Jay Bryant

      Endorsement: I use Brian’s PEC extension all the time, checking it several times a day. It even has the EV map. Great stuff. Thanks, Brian.

  • Martin

    Sorry to get off topic here but i just can’t resist this! Reuters ipsos yesterday had O +1 LV but O+10 RV on their national tracker. A 9 point spread? Are you serious?!?!? CNN had Florida R+1 LV but O+6 RV a 7 point spread! I would just love to know just how difficult it must be to pass 1 of these LV screens if your not a R voter!

    • Matt McIrvin

      Regardless of the election result, it’s going to be really interesting to see if these LV screens agree with actual turnout differences.

      It’s pretty clear from the crosstabs in things like the RAND survey that Obama’s support is heavily skewed toward lower-income and younger voters, which is probably were most of the projected difference is coming from. But it’s also clear that this cycle is not going to be like the 2010 midterm.

    • Olav Grinde

      Matt: …and women voters.

    • wheelers cat

      its easy– just answer your landline.
      ;)

      Matt, the average differnce in turnout over the last five election cycles is 16.1 % more turnout in presidential election years over the midterm. Also the percent white vote has simultaneously declined over the last five cycles from ~80% to ~70%.
      http://nymag.com/daily/intel/2012/05/how-white-will-the-electorate-be.html
      That is why I think the nat’l polls are WRONG.
      “…Obama, who won 43 percent of the white vote last time, could still win with 38 percent of the white vote — as long as the white share of the electorate does not increase.”
      note, like Olav says, white women still favor Obama.

    • Matt McIrvin

      Yeah, but women vote a lot, so if anything that’s a difference in Obama’s favor.

    • DaveB

      This is the BIG story of this election. The pollsters are telling us that there is a huge enthusiasm gap. Will there really be a 4-6% gap in turnout? If Obama supporters get to the polls, he will win comfortably.

    • wheelers cat

      protip for Team Romney
      white women vote…A LOT….and we all have smartphones.
      /smug elitist intellectual snob grin

    • Obama 2012

      yeah, it’s just not believable that there’s such a huge RV/LV gap.

      and it’s these LV screened results that Sam Wang is using here …

      think about that; as good as things look for Obama here, I honestly think in reality they are even better because I think these LV screens are a GOP dream rather than anything related to what’s actually happening.

      if Obama is up by 6% among RV in Florida, he is going to win that state.

  • Chris

    Dr Wang,

    Just heard you on Colorado Public Radio. Did a great job!

  • Tim

    Sam,

    I have truly enjoyed the site since discovering it a few weeks ago. I feel that many conservative voters would feel the same way if your numbers told a different story. Nevertheless, keep up the good work and I hope your predictions hold true.

  • Matt McIrvin

    I’ve been visiting this site since 2004, and I’ve especially been a fan ever since Prof. Wang took apart his own failure to make the right call then and applied the lessons learned to get fantastic accuracy in subsequent cycles. It’s not something you see every day.

    • Bay Area Blue

      Totally agree with this. Not only objectively analyzing his own mistakes to relentlessly improve accuracy, but his willingness to be completely transparent with his model make this site the gold standard for election prediction models.

  • Richard

    Great job. Just love this site. You cut away the BS and give us honest, statistical analysis. Thanks!!

  • Martin

    Sam, your page hits are increasing almost as fast as Mitt Romney’s flip flops!!!

  • Steve in Colorado

    I put up a link to your site on my latest blog post. My blog all of 5 followers.

  • Ohioan Voting Blue

    My wife and I have contributed to the whopping increase in traffic to PEC. After I introduced her to the site, she has been telling everyone about it at the University where she teaches.

    Sam, you provide a great service that puts the election in perspective. Thanks so much for the stats and commentary. We really appreciate your efforts.

    Also, I love the wind map, thanks for including that too.

  • Justin S

    Where are the polls for the MM going to come from with pollster down? Anyone know how long they expect to be down? I don’t like getting my fix from RCP

  • Pat

    Beyond the actual damage caused, it is actually a shame that Sandy hit now, because it may cause last minute opinion changes which may not be totally measured in the final polls before election day. And therefore the vision offered by the polls through this great site may not correspond to the final results. Therefore it may be other poor sites with crappy methodologies that might be lucky to get the close to the final result. I’m afraid it will make it more difficult to evaluate and rank the predictions due to the fact that if there are any last-minute opinion changes, they may not be fully included in the estimates.

    • grandpa john

      Curious as to why you think the storm could cause opinion changes . I agree that it could cause some movement toward Obama for his handling of the event, but since it is week until election time, there will be ample time for more polling and the only poll that counts is next Tuesday. Then we will know for sure.

  • E L

    Dr. Wang:

    I’ve been following you since the 2008 election season. Thank you and thanks to Andrew too.

    When this election is over, I think you and Nate Silver should appear on SNL. I’m sure SNL would be ripe for a take down of the impressionistic momentum experts. Nate could play Joe Scarborough and you could be Bobo. The possibilities are endless, in a nonstatistical sense, of course.

  • Arbitol

    Wangmentum!

  • Jefflz

    I am new to your site via Daily Kos. I find the information you provide to reassuring in a storm of highly biased data. I also finf it reassuring that across multiple analytic sites, there is a mode of 290 or greater EV for Obama when toss-up states are assigned based on lead. During these dangerous times where the Murdochs, Koch Brothers and Karl Roves of the world have so much influence over the minds of many, your site is an oasis.

  • H Levenson

    I, for one, am glad to have a virtual place to go for solace in support of President Obama amidst a raging ocean of misinformation and propaganda.

    Thanks for your direction and steady hand on this endeavor, Dr. Wang

  • Justin

    The objectivity of this site, mixed with snark, has made it my favorite. The attacks by the right on the methodology of Nate Silver are anti-intellectual and grounded in ethereal notions. How is Nate Silver less empirical than Joe Scarborough saying the race can’t be anything but 50-50? Show me the formula behind that figure (you can’t, it’s merely anecdotal).

  • Khan

    Here is the latest attack on Nate Silver from Dylan Byers :

    http://www.politico.com/blogs/media/2012/10/nate-silver-romney-clearly-could-still-win-147618.html?hp=r13

    And this from Joe Scar.

    “On MSNBC’s “Morning Joe” today, Joe Scarborough took a more direct shot, effectively calling Silver an ideologue and “a joke.”

    “Nate Silver says this is a 73.6 percent chance that the president is going to win? Nobody in that campaign thinks they have a 73 percent chance — they think they have a 50.1 percent chance of winning. And you talk to the Romney people, it’s the same thing,” Scarborough said. “Both sides understand that it is close, and it could go either way. And anybody that thinks that this race is anything but a tossup right now is such an ideologue, they should be kept away from typewriters, computers, laptops and microphones for the next 10 days, because they’re jokes.”"

    • orchidmantis

      I picture a dark meeting of mages, like in Bartholomew and the Oobleck. What must the supplicants (I assume the people who comment at NRO, roughly) do to assure a Romney victory? They must bring him the heart of Nate Silver!!!

      In all seriousness, there is a disturbing resurgence of the neo-con conviction that you create reality by insisting really hard that it is what you want. (See Romney on Benghazi, just walking into a set-up in which he explained that the president won if he had said “act of terror” on Sep 12 or 13, which in objective reality he did.) Needing to destroy Nate Silver and his model of an Obama lead seems to be enough, in their minds, to cause the Obama lead to vanish.

      So look out Sam. They’ll be coming for you too.

    • Obama 2012

      yup – that’s what passes as “wisdom” in the world of talking heads.

  • h

    I am not an Obama fan, but I check PEC several times a day. This is not because it is telling me what I want to hear. The analytical methods are clearly explained, and reasonable. The general tone of discussion is very high, both on technical matters, and political matters (see the polite responses to Dickerson a few days ago). I like the fact that Prof. Wang and his colleagues respond to comments. But most of all, the track record is quite strong. Especially look at the betting (intrade) recommendations Prof. Wang made during previous election cycles. They were all winning recommendations.

    When the election results are in, either PEC is right, and Gallup and Rasmussen need considerable scrutiny, or PEC is wrong and the state level polls need considerable scrutiny. Notice that I don’t say the PEC analytical methods need scrutiny, or at least not much. Rather (in my opinion) a PEC failure should be attributed to garbage in from state level polling.

  • Khan

    Chris Christie has so many nice things to say about how the President has handled this so far. Are we in an alternate universe or is this the real Chris Christie?

    • Michael Bartholomew

      When disaster strikes, partisan politics is put on the back burner.

    • Westy

      I think Christie may just be hedging his bets. He’ll be a frontrunner for the GOP nomination in 2016 if he wants it, and I think he (and probably other leading Republicans) sees that the only way to win at a national level is, at minimum, appear centrist and work with the other side. Romney’s late attempt shows that as well.

    • Khan

      @Westy

      You could have it spot on. I want to assume he’s just operating outside of politics right now.

      But it could certainly be that he knows what’s coming and wants no part of it.

    • Keith

      Truth be known he may secretly wish for an Obama win. If Romney is elected this year Christie will not have a shot at the Republican nod until 2020.

    • Matt McIrvin

      This is one of those situations in which competent governance *is* good politics, and I’m not inclined to split hairs about the difference. Christie probably recognizes that.

    • orchidmantis

      I think Christie is smart and has a good sense of strategy. And “in the midst of disaster I’m thinking of what I can do to help Romney” is not a good tactic.

      Recall he was the only one, pre Debate 1, willing to say Romney would do a great job and we’d be talking about a whole different race the next day. Not saying he was prescient, just that he saw value in being willing to come across as a no-BS, highly confident in Romney surrogate.

  • Olav Grinde

    I’ve been following your site since the 2004 election. It was the best predictor then — and it just keeps getting better.

    Thank you so much for all the great work, and especially for clarifying what is real movement/probability, so that I can ignore those talking heads that are just a feature of the media echo chamber and don’t really have anything to say.

  • Jack Tenold

    I wanted to make sure I got in here to thank you for this fabulous site. Thanks to you, my blood pressure is back under control. And I love your sense of humor.