Has Nate Silver Ruined Politics?

Is the triumph of political forecasting good for our democracy?

Jason Zengerle worries about the impact of the triumph of numbers represented by the success of Nate Silver’s forecasting model:

What’s most problematic about Silver’s and the pollsters’ triumph last night is what it may herald for the future of campaign coverage. Without a doubt, Silver’s rigorous empiricism is much, much more preferable to the lazy, gassy, vibration-sensing punditry that has made up so much of our political journalism. And yet, the biggest complaint about campaign coverage over the last twenty years has been that it’s too focused on the horse race and doesn’t pay enough attention to the substance. Silver and his fellow polling analysts and aggregators have brought a welcome degree of precision, but they’ve only made the horse race more central to the political conversation. After all, what dominated that conversation for the past month? It wasn’t a conversation about the candidate’s dueling tax plans. Rather, it was a debate about the polls. The fact that the good guys — who put their faith in the data rather than the vibrations — won that debate may turn out to be something of a pyrrhic victory.

Zengerle may have a point here. The media has been obsessed with polling and “the horse race” to the exclusion of any actual discussions of the substance of policy for decades now. Every single poll release is treated as “Breaking News” even though it’s really just one snapshot among many that may or may not be an accurate reflection of the state of the race. We fall victim to it here in the blogosphere too, with every poll becoming fodder for yet another blog post, although I did try during this last cycle (with admittedly spotty success) to limit that and focus more on poll averages and trends than individual polls. This time around, the “horse race” coverage turned into a month’s worth of coverage of the question of who was right about the polls, resulting in a complete rank amateur like Dean Chambers getting far more media coverage and conservative accolades than he deserved, especially since he doesn’t appear to have any formal training at all in mathematics, statistics, or polling. At the end of the day, Nate Silver was proven to be right, indeed more right than he had been in 2008 when he correctly predicted the outcome in 49 out of 50 states. This year he went 50 for 50 plus, of course, the District Of Columbia. I can easily see the media lesson from this being that they need to invent their own forecasting models, so that in 2016 we end up with coverage of whose forecasting model is correct.

I’m not saying the Silver should shut his forecasting model down. Far from it, actually. He’s provided invaluable insights into how polling works, and how polls should be properly evaluated. At the same time, though, it as though all the attention that’s been paid to his model this year is going to create a perverse set of incentives for the media in the future that will mean yet more coverage of the “horse race” and less coverage of substantive issues. Already, we live in a world where that kind of coverage shapes how campaigns act. If it becomes even more prevalent, then candidates and their advisers will act accordingly.

More broadly, though, I have to wonder what this actually means for our politics. Campaigns are at least supposed to be about issues and about the direction that candidates will take our country. The more we become obsessed with numbers and what the latest forecast says, the less I fear people will pay attention to what matters. After all, why bother listening to an opposing candidate if the forecasting model says that the incumbent has a 95.46% chance of winning the election?

FILED UNDER: Campaign 2012, Public Opinion Polls, US Politics
Doug Mataconis
About Doug Mataconis
Doug holds a B.A. in Political Science from Rutgers University and J.D. from George Mason University School of Law. He joined the staff of OTB in May 2010 and contributed a staggering 16,483 posts before his retirement in January 2020.

Comments

  1. Just Me says:

    I admit I get sick of the polls. I also refuse to answer polls.

    I think there is a place for polls, but I do think the media gets caught up in the polls and often fails to actually cover the candidates plans.

    I am not really certain Silver is to blame for polls, he just uses the polls of others and feeds them into his little formula.

    If Silver quit crunching his numbers, the polls wouldn’t stop.

  2. James Joyner says:

    Arguably, Silver could have the opposite effect. That is, it’s long been clear that individual polls are meaningless. I’m hardly a statshead and have been relying on the cruder RealClearPolitics model since the 2004 cycle. Silver improves on it by eliminating the bad polls and weighing the better ones more heavily. (Although, I’d note, he got 50/50 right while RCP still got 49/50 right; only Florida, which RCP had as close, went the other way.)

    So: If people start routinely dismissing individual polls and saying Look at the aggregates or Look at the modeling, then maybe we’ll get less poll talk and more substance.

    But, since a lot of political journalism, like sports journalism, is entertainment, probably not.

  3. Rob in CT says:

    Poll data has been around for a long time. All that’s changed here is that people are getting better at reading it, and making projections.

    Sabremetrics didn’t ruin baseball (far from it!).

    Also, you know… I often agree that the “horserace” soundbites/gaffe of the day stuff is bad. But there were some pretty consequential things in this race. Obama’s speech that included the you didn’t build that remark and Romney’s talk at the fundraiser with the 47% remark spoke to real, serious underlying worldview differences. Obama’s performance in the 1st debate was important (knock knock, hello, care to make the case for liberalism Barack?), as was (I’d argue) Romney’s shameless Etch-A-Sketch. Sure, there was a bunch of other inane crapola. But Silver didn’t bring that into play, and Silver (and Wang, and others) aren’t going to make all that worse.

  4. Again, I think that part of the problem is that Silver’s success implies a truth about ourselved that many of us are uncomfortable contemplating: that indvidual identity may be largely an illusion.

    People’s behavior is far more predictable and less varied than we like to imagine to be the case.

  5. rudderpedals says:

    If you’re going to the track you’re picking up the Racing Form on the way. Same thing.

  6. David M says:

    The “horse race” coverage tends to obsess over every poll released without putting it in context, so I think the polling aggregators are a net improvement.

  7. “After all, why bother listening to an opposing candidate if the forecasting model says that the incumbent has a 95.46% chance of winning the election?

    Third party candidates have been victims of that mentality for a long time. It’s a catch-22. Can’t get in debates unless you have certain poll numbers. Can’t be taken seriously without a national soapbox, like a debate.

  8. Murray says:

    How about banning all political polling between the conventions and election day?

    No polling, no projections, … no horse race.

  9. John P. says:

    I think it’s a positive that we’ll see more focus on real scientific data, because it takes the personalities out of the loop that have been so spectacularly wrong, and indeed have poisoned the discourse in the country. It would be only at most a temporary phenomenon to focus on the horse-race aspect of the polls; smart reporters will eventually start digging into the crosstabs for interesting nuggets that reflect cause-and-effect. And that will make the work of the underlying primary pollsters more valuable, and will allow them to differentiate themselves from their competitors. Sample sizes will increase and methodologies will hone in on accuracy, all in search of coming up with more and better crosstabs. On balance, I think this will ultimately help with analysis of causes of changes in the polls, as the horse race aspect becomes a trivial commodity that anyone with a spreadsheet can determine.

  10. Rob in CT says:

    3rd party candidates also suffer from the barriers to entry put up by the major parties, no? Just getting on the ballot is tough enough.

    3rd parties are thoroughly screwed here. Poll data doesn’t really change that. I agree it doesn’t help.

  11. Rob in CT says:

    @Murray:

    Consitutional?

  12. Rob in CT says:

    @Rob in CT:

    Hah, typing fail. Constitutional?

  13. Murray says:

    @Rob in CT:

    Unconstitutional?

  14. @Rob in CT:

    Sabremetrics didn’t ruin baseball (far from it!).

    Fair point. But baseball is a game, politics is supposed to be about how we choose our political representatives.

  15. superdestroyer says:

    What Silver really has shown is how unimportant money and how unimportant campaigns really are. Unless two candidates are within a percent or two of each other, there is probably no amount of money, no campaign strategy, no performance that is going to win an election for the trailing candidate. This goes along with how few competitive House and Senate races there were.

    David Axelrod and his associates have been making the rounds on the media bragging about how they ran the greatest campaign ever but anyone who believed that President Obama was going to lose was a fool. I doubt if the Democrats could have screwed up the campaign enough to actually lose the electoral college vote to Romney.

    What Silver and the other quants are going to do is take money out of the pockets of campaign strategist, political consultants, and political advertisers. The quants will probably also limit the types of elections in the future because only a delusional candidate will want to run in a race that is not a tie two or three months before election day.

  16. john personna says:

    The real story is the Obama campaign’s use of Big Data:

    The Analytics Lesson from the Obama Campaign: Keep Your Data Organized, Secret

    Compared to that … Sliver is just the hood ornament on the whole new machine.

    FiveThirtyEight’s 2012 success marks coming of age for big data in political journalism

    The campaigns will be doing it. Secretly. The only question is whether you, as a citizen or journalist, want access to the same kind of thing.

  17. John Cole says:

    This post makes no sense. Nate Silver didn’t drive the events on the ground- he just looked at the large amounts of available data and made it make sense for people like you and me.

    The polls were merely a snapshot reflection of the actual politics on the ground, so claiming Nate Silver ruined politics is like claiming I ruined hockey because I explained the neutral zone trap to my girlfriend.

  18. john personna says:

    @John P.:

    On balance, I think this will ultimately help with analysis of causes of changes in the polls, as the horse race aspect becomes a trivial commodity that anyone with a spreadsheet can determine.

    I believe someone on Morning Joe said that the Obama people were running sixty-some thousand election simulations … each night.

  19. John,

    Read it. It’s not really about Nate. It’s about how the media will likely react to what happened this cycle, which will be with yet more inane focus on polls and forecasts and all that pointless nonsense.

  20. mantis says:

    After all, what dominated that conversation for the past month? It wasn’t a conversation about the candidate’s dueling tax plans. Rather, it was a debate about the polls.

    The only reason it dominated the conversation was Republicans were denying reality. That was the conversation. We didn’t suddenly want to talk about polls all the time. We wanted to talk about the ridiculous movement to “unskew” the polls. That was a somewhat unique phenomenon, not a tradition.

    James is correct. If we could stop paying so much attention to every single poll and what it “means,” we could spend more time talking about actual policy issues. But that would require news organizations to invest in something other than partisan pundits and other bullshit artists, and that’s expensive and boring, so it won’t happen.

  21. john personna says:

    @mantis:

    While I think the focus on Silver misses the bigger story, I have no problem with a polling dashboard all of us interested citizens can hit at our pleasure or convenience. An election has a winner and a loser. It is actually a horse race until then.

    It might be interesting to think how we’d all discuss this at OTB, knowing things like “district X in Virginia is the focus of that state race.” Or “meet the people of district X.”

    … and of course I hope you all know that Obama knew you all go to OTB. And so he took a 90% guess you were in the bag 😉

  22. John Cole says:

    @Doug Mataconis: I read it, and it still seems like you don’t actually understand what Nate does. He doesn’t create data. His analysis can’t exist without the polls, and the polls are merely snapshot reflections of what is going on on the ground. And what is going on on the ground are people reacting to a variety of things- response to the positions politicians take, debate performances, campaign advertising and literature, and so on. Or, as we call it, the politics.

    What Nate Silver has done is what moving from a 13″ black and white tv to a 42″ HDtv did for sports. It doesn’t interfere with or replace the game, it just gives us a better picture.

  23. Janis Gore says:

    Well, I think the great danger is that this effeminate little slip of a man is imposing his “Homosexual Agenda” on us all.

  24. swbarnes2 says:

    @mantis:

    If we could stop paying so much attention to every single poll and what it “means,” we could spend more time talking about actual policy issues.

    If the front pagers here don’t want to do that, (they’d much rather think about what the real Romney is like than examine the consequences of Romney’s stated policy proposals, like say, letting the auto industry die, or eliminating ObamaCare, or supporting naked racism) then the media won’t be interested either.

  25. Jen says:

    One of the main reasons the polling data became a story was the volume of people who said it was wrong, and politically motivated/bias slant, etc.

    In my opinion, what the triumph of Nate Silver’s predictions *should* mean is that next time we can skip over this silliness about unskewing polls and instead focus on the real issues. We’ll have to see how well the right adjusts to the new reality of math.

  26. john personna says:

    @John Cole:

    I can’t tell from the conversation if you and Doug have read and digested the big data pieces yet, but it is a little bit more than tv to big screen. It is a shift from reduction (averaging polls) to expansion (massive simulation). To use an analogy, the polls hand Silver the dice, and then he rolls them a few thousand times, carefully recording outcomes.

    The Obama campaign took that one step further. They found the sorts of people they were winning with, and then sought more people like them.

    I thought that “dinner with Sarah Jessica Parker” was the lamest thing I’d ever heard. Of course, I now know that they new it wasn’t for me. They didn’t care about me. They were hitting a specific group with a specific message, even when that message played nationally.

    I never got a single piece of Obama mail here in California. That might have been because they knew they had me, or it might have been because they knew my area was a hard sell, but either way it’s a new kind of campaign.

    Is focused effort good for democracy? That’s a better question.

  27. gVOR08 says:

    It is difficult to see how Nate Silver and polling generally could possibly make the media’s obsession with the horse race, at the expense of policy, any worse than it was. Maybe Nate himself could tell us that his effect has been to increase the percentage of stories on the horse race from 98.0% to 98.2. And it’s hardly Nate’s fault that Dean Chambers and the whole conservative echo chamber launched off into the ozone. They were faulting polling generally, not Nate. It’s PPP and Rasmussen and Gallup and all the rest that do or don’t skew on party ID, not Nate.

    It’s the punditry that really went after Nate. He undercuts their whole generally non-falsifiable game of insider information, connections, gut feel, intuition, imagination, and BS.

  28. Rob in CT says:

    @Doug Mataconis:

    Heh, I almost included that very caveat in my post. Politics, while it certainly shares significant similarities to team sports, is not in the end just a game.

    I think my other points stand.

  29. PJ says:

    Did meteorology ruin the weather?

    Should we perhaps get back to Almenacs and people forecasting the weather by dissecting animals?

  30. “Did meteorology ruin the weather?”

    Have you watched the Weather Channel lately? 😉

    Seriously, though, I don’t see Silver’s methods making an already obnoxious freak show any worse. As others have said, they’ll actually help a little by taking a lot of the BS out of polling. Who cares of Rasmussen leans right when that can be compensated for in the model? Unfortunately, these methods won’t do a damn thing to pull talking heads out of asses. Media coverage of elections will still be sensational crap, even if that crap has a smaller poll component.

  31. michael reynolds says:

    @john personna:

    It’s fascinating, because what they’ve done is change the game in fundamental ways that pundits and bloggers just aren’t getting yet. (And may the Republicans remain clueless.) These guys are using smart bombs while the other side is tossing grenades out of a Sopwith Camel.

  32. legion says:

    No, no, no, no, and no.

    Nate didn’t “ruin” politics. What he ruined was political punditry. One can understand, if not excuse, the media’s inability to tell the difference between the two.

    They make money only if there’s interest in the race. There’s only interest in the race if there’s uncertainty. Having good numbers limits that uncertainty. Ergo, political pundits hate people like Silver with the white-hot heat of a Republican who’s just had his tax shelter taken away.

    But this has _nothing_ to do with “politics”.

  33. john personna says:

    @michael reynolds:

    It’s tremendously interesting, isn’t it?

    Who will get access to Obama’s data horde?

    Some 2016 candidate will be able to pre-run a few million elections at his leisure, as he drafts his platform before entry into the 2016 primaries.

    That beats “gut” any day.

  34. PJ says:

    @Eric Williams:

    Have you watched the Weather Channel lately? 😉

    Well, the Weather Channel is to weather as 24/7 news is to politics. 🙂

  35. mattb says:

    @John Cole:

    it still seems like you don’t actually understand what Nate does. He doesn’t create data. His analysis can’t exist without the polls, and the polls are merely snapshot reflections of what is going on on the ground.

    Silver may not create data, but to JP’s point he is creating meta-data (or rather meta-polling-data). The thing is that meta-polling-data also immediate becomes additional data that is fed back into the system (if you will) in a way that traditional punditry and analysis is not.

    Back off man, he’s a statistician and he’s doing science.

    And, like it or not, Silver’s win this year means that in the next Presidential Cycle a lot of people are not only looking to him for a barometer of where things are currently, but for the predictive aspects of his model. And, to get a bit quantum, the act of measuring something invariable changes the thing being measured.

  36. As I thought about all the buzz around Nate Silver (and all the sports metaphors in electoral politics), I was reminded of something. Anybody else remember when Super Bowl outcomes were predicted by simulating hundreds of games using Madden NFL? Is that still done?

  37. superdestroyer says:

    @Eric Williams:

    Do you really think the media will spend time covering any election that Silver and the other quants say is a 90%+ lock for a candidate? What Silver does is take the unknown out of the election before voting day. Thus, there is little reason for the media to be interested in elections in the future when so few of them will be competitive.

    Image what happens in 2016. If the same Democratic candidate wins the Iowa caucus and the New Hampshire primary, everyone will know who the next president will be almost a year in advance. It will lead to the longest transition in history. Of course, the long term issue that Silver faces is that fewer competitive elections in the future, there will be less for him to model and write about.

  38. michael reynolds says:

    @superdestroyer:

    Obviously you didn’t follow Silver’s blog. It went from a high of 90% to a low of about 60%. 60% is nowhere near being a sure thing. And the fact that the number changed proves that events can alter outcomes.

    Duh.

    So whatever the initial number, a candidate would be well-advised to create some event that alters the outcome. Some of us like to call that campaigning. Which will go on, regardless.

    Life is not finally predictable. You know Frederick the Great? He wasn’t always so great. At one point he was essentially guaranteed to lose all of Prussia to a combo of enemies including Russia. They had him cold. He admitted he was beat. He was talking suicide. Then the Czarina with a big hard-on against him up and died and was replaced by the Czarevich (then Czar) who loved Prussians like a teeny-bopper loves bad pop songs. Russia made peace, Freddy lived.

    Nate would have called it 99% chance Freddie goes down. Events, dude. Sh!t happens.

    As for your notion that the media will stop covering elections, you are evidently under the misapprehension that they do this as a charity. They make money on elections. They’ll keep doing what makes money.

  39. “Do you really think the media will spend time covering any election that Silver and the other quants say is a 90%+ lock for a candidate? What Silver does is take the unknown out of the election before voting day.”

    Hogwash. Polls still only reflect the current state of the “game”, much like a photo of a Risk board would. What does it mean to say a candidate has a 90% chance of winning 6 months out from the election? It mostly says they other guy has his a lot of work to do. It’s just a conditional probability based on current data. It’s not the marginal probability that X will win, and not Y. It’s P(winner = X | current data). In Bayesian statistics, as data are accumulated, posterior probabilities are updated. If candidates say or do things that change minds, polls will eventually reflect that, the model will incorporate those data, and new predictions will be made.

    Too few people understand what a Bayesian means when he says an event has a 90% probability.

  40. grumpy realist says:

    @mattb: Does anyone know if Nate is actually running Monte Carlo simulations for his predictions? That would be cool.

    The only reason people are squawking is because Nate has taken away their chance to gas off before a TV camera and pretend they’re sounding intelligent.

    And if the data in ain’t correct, you’re going to not be correct in your predictions. That was where the whole CDO bullcrap went wrong: people forgot that they were using historical data when predicting default rates and well, if you make loans riskier, you’re going to get more defaults.

    It’s only when dealing with statistical mechanics that you can really assume the CLT actually holds. In all other cases you’ve got to remember that people act like, well, people.

  41. john personna says:

    @superdestroyer:

    Do you really think the media will spend time covering any election that Silver and the other quants say is a 90%+ lock for a candidate?

    Every candidate should model in preparation, and no candidate should come into a race with a 10% chance of winning. He’s just wasting our time.

  42. sam says:

    @michael reynolds:

    These guys are using smart bombs while the other side is tossing grenades out of a Sopwith Camel.

    Kevin Drum put it this way:

    In the end, it turned out that one side ran its campaign like a business, while the other side ran its like a local PTA. Ironically, it was the ex-community organizer who did the former and the ex-CEO of Bain Capital who did the latter.

    Quote of the Day: America’s Billionaires are Pissed Off at Karl Rove

    So much for all the vaunted bidness-management expertise horseshit.

  43. michael reynolds says:

    @superdestroyer:

    Let me try to explain something fundamental to you. You’re a guy in desperate search of a Single Unifying Theory. In your case you think you’ve found the answer and it’s race.

    It’s not. There is no Single Unifying Theory. It’s a delusion you have. It’s an obsession. It’s not real. And everyone reading you knows it’s not real. Do you understand that at least? Do you understand that everyone else sees that you’re obsessed and that you’re unbalanced?

    When everyone knows you’re wrong you should at least consider the slight possibility that you are in fact wrong. You are not Galileo, you’re just some cracker with an obsession. You do not have unique insight. You have a mental problem.

    And honestly since we don’t know you and therefore don’t really feel for you, your mental problem is just a huge crashing bore to everyone else. We all know you’re nuts, like some street person ranting about the CIA. It’s funny for a while, then it’s just boring.

  44. Geek, Esq. says:

    On a related note, Romney’s campaign essentially took the Dean Chambers approach:

    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.

    Those assumptions drove their campaign strategy: their internal polling showed them leading in key states, so they decided to make a play for a broad victory: go to places like Pennsylvania while also playing it safe in the last two weeks.

    Stunning, absolutely stunning that a numbers guy like Romney would do this to himself. They chose to lie to themselves rather than look at math that made them uncomfortable.

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

  45. Argon says:

    As long as it eliminates one more source of non-empirical BS I’m for it. Let them spin real numbers at least.

  46. superdestroyer says:

    @grumpy realist:

    I have read that Silver is just using Excel. One can do some Monte Carlo inside of Excel but it is hard to do and crude unless he purchases an add-in like Crystal Ball.

    However, there should be other quants that are using Monte Carlo modeling. It is not that hard and the uncertainties are established by the ranges of the polling data with appropriate weights.

  47. swbarnes2 says:

    @john personna:

    Some 2016 candidate will be able to pre-run a few million elections at his leisure, as he drafts his platform before entry into the 2016 primaries.

    That beats “gut” any day.

    What Nate does it take a hell of a lot of different polls, and figure out how to weigh them to get a results more statistically powerful than any one poll could be.

    He can’t magically tell the future. He can’t tell you how one policy will move the electorate. If Elizabeth Warren runs a couple of surveys prior to a Presidential run, Nate Silver can’t magically foresee what would happen in a real vote a year later.

  48. superdestroyer says:

    @michael reynolds:

    But there were many other elections that Silver called much eariler. Also, Silver but broader uncertainties on his predictions in the early months of an election. My guess is that Silver will narrow with uncertainties in the future and sharpned his predictions. Once again, when it was shown that the Republicans would have to but resources into North Carolina, Indiana, and Virginia, it should have been obvious that Romney was not going to win. The only uncertainty was how badly Romney would lose.

  49. superdestroyer says:

    @michael reynolds:

    MR,

    People find the idea that demographics trumps campaigns, issues, the media, because they do not like where the demographic trends are taking them. Progressives absolutely refuse to think about demographics. Wonks and Pundits have only started thinking about demographics in the last few days.

    What the Republicans have to understand is that the white vote in the U.S. can vary between 60% for Republicans to around 50% for Republicans. Campaigns and performance can maybe change the percentage by a single point or two.
    However, as the exit polls have shown, Blacks voted over 95% for Democrats, Hispanics and Asians voted over 75 % for Democrats. That means that the Republicans cannot swing the white vote enough to make up for the massive loses they have with non-whites.

    Given that the U.S is becoming less white, there is no way for a conservative party to exist. Even slate.com has been running stories about how California is a one party state and faces many problems even with total control http://www.slate.com/blogs/moneybox/2012/11/07/california_everywhere_the_risk_of_democratic_dysfunction.html

    The media is full of stories concerning the failure of the Republican Party as a national party and the demographic trends that are all against the Republicans. The real discussion is whether the Republicans try to maintain the status quo and let demographic bury them or commit suicide by becoming the second, big government party.

    My guess is that the Republicans will commit suicide and that no one will be interested in Silver’s predictions in the future because it will be too hard to model the New Hampshire primary or the Iowa caucuses.

  50. @superdestroyer:

    “Silver but broader uncertainties on his predictions in the early months of an election.”

    That makes perfect sense. As the opportunities to change minds dwindle, certainty increases. In early polls, prior data are weighted very highly, making the estimator strongly biased (in the statistical sense). The small number of early polls also give the estimator high variance. As the number of polls increase, pre-polling data are given less weight, polls begin to replace the priors, and the larger, more diverse nature of later polls bring down variance. Several consecutive polls with strong agreement also reduce variance. By the time election day arrives, it would take a major event or really awful polling for the prediction to be very wrong.

  51. Woody says:

    Here are my predictions:

    The insipid tv political horserace punditry will become infused with a product called Quants . . . Now With More Data Than Brand X! Every outlet will have their own Math Team that will claim they have the secret sauce inside data.

    Nate Silver will post a 65% prediction that will go the other way. This will be seized upon by the savvy pundit set as unbreakable proof their way is superior (great post about Frederick, Mr. Reynolds).

    The timing of calculated campaign folderal (leaks, accusations, etc) will be increasingly better calibrated to address niche voter groups.

    Political junkies like ourselves will continue reading and arguing because we just can’t help ourselves – and because I think most here are very aware indeed that it is indeed not a game (be nice if the elites would recognize this fact as well).

  52. john personna says:
  53. john personna says:

    @swbarnes2:

    I didn’t say Nate, did I?

    I believe the campaigns have tremendous volumes of position polls tied to congressional districts. Thus if I say “raise social security age to X” I can mine that data and know, at least at the time of the poll, my initial capture of voter opinion.

    We know the old large grain version of this. Y% of Americans agree with my plan. The problem is, Y% of Americans does not give me projected electoral victories.

  54. michael reynolds says:

    @superdestroyer:

    Dude. Is gender not demographics? How about age? Women and youth both went for Obama. But you see only race because that’s your obsession. If you shifted the women’s vote a few points it would trump the minority vote. Conveniently the GOP is focusing on race because it makes their racist voters happier than admitting the problem is women. Harder to demonize women.

  55. superdestroyer says:

    @john personna:

    Yes, the Republicans are discussing whether to let the chronic disease of demographic changes destroyer them while trying to remain conservative in some form, or commit suicide by adding millions of third world immigrants to the voter rolls in a country where the government gives set asides and quotas to Hispanics.

    Of course, the idea of adopting positions that help middle class American citizens has not been discussed in the past and will not be discussed in the future. All the article shows is that the cheap labor Republicans will probably win the day and the Republicans will choose suicide instead of a slow death.

  56. James in LA says:

    It appears we now have sufficient sensors down to the precinct level to accurately detect swings in the electorate. Silver and most of the other predictive models were saying the same thing for all year. Pick any 6th day of the month in 2012, run the election, get the same results. And so, a fairly accurate rough outline of the future has always been known.

    Sounds like Azimovian physchohistory to me. But it only works if people vote. No better way to show the need to vote than accurate predictions of what happens.

    It deals a deadly blow to the era of Hunch Politics for sure. Can Rove really rise again with a 1% ROI?

  57. James in LA says:

    @superdestroyer: Given that the U.S is becoming less white, there is no way for a conservative party to exist.

    If this is your metric, good riddance.

  58. superdestroyer says:

    @michael reynolds:

    Actually married women voted for Romney. It is single women that vote overwhelmingly for the big government party. And yes, youth went for the big government party. When you add up blacks, Hispanics, single women, and youth, it is four demographic groups that pay little in taxes, have little issues with government regulations but benefit from entitlement programs.

    As the U.S. heads toward a period of the Democrats being the dominant political party, the real question is what policies will be adopt. And what will the quants do in the future when the Democratic Party is the one dominant party?

  59. superdestroyer says:

    @James in LA:

    But then the question is what happens to politics, governance, and policy with one dominate political party. What will the Democrats push for when they no longer have to worry about the Republican Party? What policies will be adopted when the Democrats do not have to fear about being voted out of office? However, I guess worrying about polling data in noncompetitive races is more interesting than thinking about what the government will look like in the future.

  60. Tsar Nicholas says:

    Exactly how many layers of irony are there in a post about political forecasting that somehow ignores the giant flaming neon mathematical elephant in the room of the election that just happened a couple of days ago?

    Silver’s final call was a 91% probability of an Obama win, wasn’t it? IOW according to Silver if we had conducted this election 100 times Obama would have won 91 times.

    Well, hello, Obama’s winning margin in the Electoral College was approximately 405,000 votes (~120k in VA, ~55k in Fla., ~110k in OH, ~120k in CO). That’s out of ~119 million total ballots. That’s one third of one percent (0.34%). If a little over 200,000 people out of the roughly 119 million who voted had changed their minds on their ways to their respective polling places right now we’d be discussing whom Romney should choose for his cabinet, Nate Silver would be lining up a job carrying Chris Matthews’ jockstrap, and the political left in this country would be swimming in vats of Zoloft, Paxil and Xanax.

    Come on, let’s not smoke ourselves blind. Silver did make the correct call. Obama won. He got around 50% of the total vote. He narrowly won the four key states that gave him the necessary electoral votes. The Fair model by Ray Fair in prior cycles was predicting electoral outcomes even with greater certainty. There are no magic oracles.

  61. Ben Wolf says:

    @James in LA: I wouldn’t bother responding to superdestroyer. In all honesty I suspect he’s a functional schizophrenic and simply can’t help the obsessive repetitions on race and one-party states. Best to ignore it.

  62. superdestroyer says:

    @Ben Wolf:

    I only point it out when people write about politics, elections, or governance as if everyone is an upper middle class white family.

    You may want to notice that many media outlets are beginning to notice demographics, trends, and the state of the Republican Party that I have been noticing for years.

    What are so many people interest in politics trying to be like the Dean Chambers and refusing to look at data that does not support their POV.

  63. michael reynolds says:

    @superdestroyer:
    Which demographic group is the largest consumer of government services? Old people. By orders of magnitude.

    They voted Republican.

    Explain.

  64. Jib says:

    @Tsar Nicholas: Oh please, come on. You have to understand that the size of the victory does not indicate the probability of victory. 3 points is nothing in the NFL but an NFL team with a 3 point lead with 3 minutes left in the 4th qtr has a 75% chance of winning.

    If you run good polls all season and in each poll, they show the same narrow outcome, and there is very little change between the polls in the outcome, the chance of the actual voting poll showing a major shift is very small. So the size of the victory may be thin but the chance of that victory is still high.

  65. superdestroyer says:

    @michael reynolds:

    The elderly see that they are getting their money back after they have paid in for decades. Tim Kaine in Virginia ran advertisements making this exact claim. Of course, the Repulbicans, from time to time, have tried to talk about the massive amounts of entitlement spending on the elderly. However, the Democrats have demagogued and the idiot Republicans decided to double down to pandering to the elderly.

    In the long run, the government is not going to be able to deliver on all of the promises that is has made or Americans are going to be asked to give up a lot to fund all of the entitlement spending. That is one of the reasons that I believe the Democrats will be dominant and the Republicans will be irrelevant. There is no reason for two political parties in entitlement spending, entitlement growth, and entitlement funding is off the table. The only real discussion is what are Americans willing to give up to fund all of the entitlement spending in the future.

  66. MM says:

    @Murray: There will always be the horse race. “Is Sandy good for Obama?” “Did Stacey dash make it OK for blacks to vote republican?”

    Political pundits and the consuming public are not, by and large, bright people. Couple that with the fact that fact-checking is largely considered taboo, and you have nothing left but hirse race journalism.

  67. KariQ says:

    @john personna:

    I never got a single piece of Obama mail here in California. That might have been because they knew they had me, or it might have been because they knew my area was a hard sell, but either way it’s a new kind of campaign.

    It’s interesting. I didn’t get a single piece of mail, either, but I did get a phone call from the Obama campaign, asking if I still supported the president. I wonder if their data horde is fine tuned enough to know that I was a frequent third party voter until W. made a partisan Democrat out of me. Perhaps they were worried I was slipping back to independent/third party status?

  68. KariQ says:

    @superdestroyer:

    Even slate.com has been running stories about how California is a one party state and faces many problems even with total control

    As usual for a national media piece, Yglesias completely overlooks the fact that we had divided until 2011. When the Democrats took over complete control in 2011, they were given a state that had massive budgetary problems caused in part by the recession and in part by overly optimistic revenue assumption made by the previous Republican governor. You don’t fix those kinds of a problems in one or two years.

  69. superdestroyer says:

    @KariQ:

    Once again, progressives manage to blame Republicans for everything and take no responsibility themsleves. The state budgets have been pass by the Democratic control legislature every year for a long time. Trying to blame Schwarzenegger for everything that when wrong when the Democrats controlled every other state wide office and both houses of the state legislature is laughable. Who will the Democrats blame when there are no more Republicans in California?

    I guess you will be one of the progressives who will be lamenting in january 2017 that the Obama Administration was hurt by the Bush economy for the entire eight years that President Obama was in office.

  70. swbarnes2 says:

    @KariQ:

    SuperDestroyer is a troll, do not feed him.

    Even though Dems had a majority in the state legislature, they lacked the super-majority required for a lot of things, like raising taxes. I believe this election has just changed that.

    So no, California has never been a one-party state. Republicans in the state legislature had enough numbers to block what they wanted to block.

    And really, California is majority Dem, but there is a lot of red territory out there. Its just sparsely populated, and LA and the Bay Area greatly outnumber the rural areas.

  71. KariQ says:

    @superdestroyer:

    Right, I forgot. Nothing is ever the Republicans fault. W. left office in 2007 when the economy was booming, Obama caused the recession that started a full year before he was inaugurated. Schwarzenegger had no impact on California’s budget while governor, even though it couldn’t become law unless he signed it and he had a line item veto; and the Republicans in the legislature had nothing to do with the budget in spite of the fact it couldn’t pass with Republican votes due to the 2/3 supermajority requirement that was in place until 2011.

    How silly of me.

  72. It’s just occurred to me that it’d be helpful if bloggers published lists of prominent trolls above comment entry forms with a warning to not feed them. Folks who aren’t regular readers don’t know who to plonk.

  73. superdestroyer says:

    @swbarnes2:

    Once again, if there is one Repbulican in a legislature, on a committee, or involved in politics, they will be blamed for everything that does wrong. It is like the Democrats in Maryland or Mass., trying to blame the few Republicans that exist for their own budget issues.

    Considering that most deep blue states had massive budget problems, it is hard to say that it was the Republicans fault because they would not let the Democrats raise taxes through the roof. I guess the idea of actually cutting the budget never occurs to Democrats when they are in the majority.

  74. KariQ says:

    @swbarnes2:

    I know about SD. I was more annoyed that someone like Yglesias, who really ought to know better, was giving the “one party” line which is not just false but part of an alternate reality. I’ll put California’s future as a all Democratic state up against that of any all Republican state, without hesitation and at whatever odds anyone wants to give me. We’ll be fine.

  75. superdestroyer says:

    @KariQ:

    You should look up the test scores of the public schools in California. You should look up the unemployment rate for California. The unemployment in California has been higher than the unemployment rate in Texas for over a decade. Considering that the population in California has started to grow slower than the country as a whole, I would not be so optimistic about the future. Do you really believe that California will continue to import enough immigrants to off set all of its economic problems.

    What part of high taxes, bigger public sector, and smaller private sector is going to make you see a bright future in California?

  76. Franklin says:

    @Tsar Nicholas:

    If a little over 200,000 people out of the roughly 119 million who voted had changed their minds on their ways to their respective polling places right now we’d be discussing whom Romney should choose for his cabinet, Nate Silver would be lining up a job carrying Chris Matthews’ jockstrap

    You still seem to understand a little but take it to a bizarre conclusion. This isn’t the only election he’s done. In fact he’s pretty accurate at other races, as well, and has been for a few years now. At some point, yes, he’ll have some result at a low probably (let’s say at less than 10%) but it will happen. This will only imply that his variance model is pretty accurate, too, assuming these “errors” happen less than 10% of time. Only idiots will relegate his model to the dustbin as you suggest above.

  77. James Joyner says:

    @superdestroyer: This is your last warning on the one-party state meme. We’re all tired of it. Find some new material or take it elsewhere.

  78. michael reynolds says:

    @superdestroyer:

    Nice evasion. Let’s try this again.

    Your theory is that the brown people are by nature takers. That there can never again be a conservative government because brown people are now more numerous and they will always vote Democratic.

    But the brown people are not the “takers.” Old people are the takers, and they vote Republican.

    In addition, professionals tend to vote Democratic, and they are clearly not the “takers.” Right?

    And young people are not takers but they vote Democratic.

    In other words, you can not match up “Democrat” and “taker.” Nor can you match up “brown people” and “takers.” Which leaves you with nada.

  79. michael reynolds says:

    Oops, did not see James’ comment above. Now breaking off contact with SD.

  80. An Interested Party says:

    Well, hello, Obama’s winning margin in the Electoral College was approximately 405,000 votes (~120k in VA, ~55k in Fla., ~110k in OH, ~120k in CO). That’s out of ~119 million total ballots. That’s one third of one percent (0.34%). If a little over 200,000 people out of the roughly 119 million who voted had changed their minds on their ways to their respective polling places right now we’d be discussing whom Romney should choose for his cabinet, Nate Silver would be lining up a job carrying Chris Matthews’ jockstrap, and the political left in this country would be swimming in vats of Zoloft, Paxil and Xanax.

    This is similar to the whining that some Democrats did in 2004, particularly talking about Ohio…despite the pathetic display above, a win is a win is a win, as a loss is a loss is a loss…some people need to just get over it…

    I’ll put California’s future as a all Democratic state up against that of any all Republican state, without hesitation and at whatever odds anyone wants to give me. We’ll be fine.

    Exactly right…anyone who wants to compare California to Mississippi or Alabama is certainly free to do so…

  81. al-Ameda says:

    Honestly, I fail to see how Silver, a statistician who has accurately forecast the results of the 2008, 2010, and 2012 national elections, has ruined politics for all of us.

    We are who we are and Silver knows where we are and how we vote – it doesn’t change a thing.

  82. Eric the OTB Lurker says:

    Hey, folks, just thought I’d pop in and ask… Has anyone seen Jan or Smooth Jazz? I’ve been eagerly awaiting their insight and considered opinion. They were for Obama, right?

  83. Eric the OTB Lurker says:

    Oops, found Jan in another thread. Smooth Jazz? Yoo-hoo! Where are you?

  84. superdestroyer says:

    @James Joyner:

    James, OK. However, it is really hard to discuss politics these days without mentioning the biggest issue around. How does one discuss the future of politics or governnance without discussing the current paths that the two political parties are on.

    I guess I will be forced to find the politically correct terms that most everyone is forced to use to avoid discussing the real issues these days.

  85. Ken says:

    @Tsar Nicholas: Come on, let’s not smoke ourselves blind. Silver did make the correct call.

    No, Silver didn’t “make the correct call.” he made fifty correct calls. Out of fifty contests. Combined with his 49 correct calls in the last presidential election, it’s pretty solid evidence that his methods are sound.