Nate Silver’s Actual Problem…He’s No Longer an Adult Fighting Children.

Nob Akimoto

Nob Akimoto

Nob Akimoto is a policy analyst and part-time dungeon master. When not talking endlessly about matters of public policy, he is a dungeon master on the NWN World of Avlis

Related Post Roulette

65 Responses

  1. Avatar KatherineMW says:

    I think you’ve hit the nail on the head here. He’s gone from a topic (elections predictions) where he can compete with the mainstream media and outcompete them pretty effectively because there wasn’t any well-known systematic analysis prior to his work, to a wide range of topics in economics, the social sciences, and sciences where he’s going up against academics.

    A column doing some analysis of polls, which is competing with random speculation or simple averages of polls, is going to look good. A column doing some basic analysis of socioeconomic (or climate) indicators which is competing with people who’ve written academic papers on these subjects and have extensive in-depth knowledge of them isn’t going to look so good.Report

    • Nob Akimoto Nob Akimoto says:

      Here’s the thing, too, Sam Wang actually outperformed Silver’s predictions by using a simpler poll aggregation model without the various adjustments Silver made in his prediction model.Report

    • Seconded. He’s no longer drawing on his (considerable) strengths so much. He’ll either have to step up his game in other areas (which might not be impossible, after all), or learn that he can’t.

      That said, I could never do what he does. Never.Report

      • Avatar North says:

        Thirded, I love me some Nate (and his election forcasts have me feeling pretty glum at the moment) but I suspect he may be wading into a shark tank wearing a chum suit with some of his stuff on 538. I wish him well but I suspect it’s going to be bloody.Report

      • Avatar Kazzy says:

        In a sense, he’s going from being a big fish in a small pond to being a small fish in a big pond. Some small fish thrive. Some grow into big fish. We’ll have to see how Nate responds.

        I don’t have a problem with him trying to step up his game. In fact, I admire it. I am experiencing my own set of frustration at my current place of employment because (at the risk of sounding arrogant but, really, when do I not run that risk?) I am too often the smartest person at the table. It means I have limited opportunities to grow and that I might represent our institution’s ceiling… neither of which I think is good thing.

        Thus far, Nate hasn’t handled himself particularly well. If you consciously say you’re going to get in the ring with the big guys, you need to get ready to be punched and can’t whine when you do. It strikes me that maybe Nate already thought he was a big fish. Time will tell…Report

      • Avatar North says:

        Not to snark Kazzy but doesn’t your table consist of you and a whole slew of single digit year olds?Report

      • Avatar Kazzy says:

        Hahahaha, I thought about clarifying that for just that reason.

        Yes, in the classroom, I am generally the smartest, strongest, fastest, and tallest person, basically by definition. I’m referring to the other “tables” I sit at… during faculty meeting, academic council meetings, etc.Report

      • Avatar Glyph says:

        “I wish I could play little league now. I’d be way better than before.” – Mitch HedbergReport

    • Avatar Barry says:

      It is a good summary. Somebody coined the term (either me, or Krugman) ‘intellectual arbitrage’. There’s a massive and highly scientific polling industry, most of which exists to come up with good, reliable and predictive numbers. Then there’s the political punditry.

      BTW, here’s a blog posting taking Silver out to the woodshed for the parts of his book on financial modeling (http://mathbabe.org/2012/12/20/nate-silver-confuses-cause-and-effect-ends-up-defending-corruption/)

      “Nate Silver confuses cause and effect, ends up defending corruption”

      Somebody should tell Tod about Mathbabe; she must be the one who disseminated the liberal party line, since she was critical of him so far back 🙂Report

  2. Avatar trizzlor says:

    Healy’s response captures the problem expertly, but there are two things going on:

    1) Academics in the field criticizing FiveThirtyEight for statistical rigor and taking it’s authors to task for simplifying conclusions or ignoring important confounders. For better or for worse, this is how academia tends to work; any hot-shot outsider effectively paints a giant target on their work.

    2) Liberal bloggers criticizing FiveThirtyEight because it has yielded predictions that are not consistent with the Dem narrative. I have little doubt that the major lib-blogs wouldn’t give a second look to the models if FTE was predicting a lively, rebounding economy and climate-related disasters of horrific costs.

    So far this is pretty natural. Academics enjoy taking down the popular kid, and people in general are only skeptical of conclusions that contradict their narrative. But it will be interesting to see if we get a situation where the experts approve and the bloggers condemn, which will really be indicative of a partisan bias against FTE.Report

    • Avatar Kim says:

      1) Academics have these arguments with other academics all the freaking time. This is more of a Welcome to the Adult Club sort of thing (this is the kind that comes with plastic wall vaginas, apparently — no I will not stop making fun of that art, as it’s fucking hilarious).

      2) Kos and company would (if only because the ramifications between “we have 10 years” and “we have twenty years” is a lot of difference in terms of strategy). Kos has a slightly different view of the election than Nate does, but I’m not sure who’s got better data (I think kos gets more reports from local party activists — he’s been pretty good in the past at finding folks like Tester and Webb, so I think he’s better at handicapping the unknown challenger than Nate is.)Report

      • Avatar DavidTC says:

        Kos has a slightly different view of the election than Nate does, but I’m not sure who’s got better data (I think kos gets more reports from local party activists — he’s been pretty good in the past at finding folks like Tester and Webb, so I think he’s better at handicapping the unknown challenger than Nate is.)

        And Kos is, fundamentally, pro-Democrat. There’s a lot of times to talk about ‘objectivity’, but when people on reporting on the chances of ‘their side’ winning, I really think we need to understand that they might be deliberately hyping things a bit.

        I have a rather serious problem with treating *optimism* as ‘fact denying’. This is not an example of ‘the left denying facts’, and it’s not an example of the same thing on the right, either.

        This is how politics has always worked. Everyone claims they can win, and even manages to delude both themselves and others to some extent. Almost no one has ever walked into election day publicly saying ‘Well, I guess I’m going to lose’. (I’m sure that has happened, people who run to make a point or protest or something, but not actual *real* candidates.)

        At least among the *party and supporters*. The amazing and mockable thing last election was when it infected an entire network…but, then again, I think we’ve all know that Fox News wasn’t actually a news network anyway, but actually a branch of the Republican Party, and their nonsense about skewed polls pretty much proved it. In fact, I would suggest that a way to figure out if a ‘news’ organization is politically biased or not in their news would be to check how they respond to voter polls.(1)

        Anyway, people can talk all they want about all the other criticisms of FiveThirtyEight. Complain about how they respond to climate change stuff, or whatever.

        But trying to complain about liberals going ‘Lalala, we don’t want to hear that we won’t take the House’ is sheer nonsense. Everyone has done that, always, for the entire history of the Republic.

        1) Now, news organizations will always treat the presidential election as a horserace, no matter how clear the winner is in advance. But that’s not due to political bias, that’s due to them trying to invent news. So you’d have to figure out how to ignore their ‘newsiness’ bias.Report

      • Avatar Will Truman says:

        This may be the best comment in either related thread.Report

      • Avatar zic says:

        @davidtc

        Excellent comment.

        On your footnote:
        Now, news organizations will always treat the presidential election as a horserace, no matter how clear the winner is in advance. But that’s not due to political bias, that’s due to them trying to invent news. So you’d have to figure out how to ignore their ‘newsiness’ bias.

        I think it’s not trying to ‘invent’ news so much as use ‘fair and balanced’ reporting (two equal sides, no actual analysis of the sides) to keep market interest. After elections, political reporting, interest, and participation on sites/networks so specialized takes a nose dive; blow-out races probably have similar demographics.

        It’s about keeping the eyes on your message as much as anything; that is the essence of horse-race political reporting. They generate advertising revenue when we constantly tune in to hear who’s up and who’s down. Media benefits from political polarization.Report

      • Avatar DavidTC says:

        @will-truman
        This may be the best comment in either related thread.

        That seems very unlikely. 😉 But I was just getting a little annoyed with “Democrats don’t believe polls! Democrats are becomes as irrational as Republicans!’.

        There’s a difference between ‘really disbelieving believing polls’ and ‘saying you don’t believe polls to keep morale up’, and heck, there’s even a difference between ‘really disbelieving believing polls’ and ‘disbelieving actual factual things’. Asserting that ‘Democrats now say they don’t believe the polls’ is a indicator of looming irrationality on their part makes no sense at all. No one, absolutely no one, admits they believe negative polls about them. Ever. In the entire history of ever.

        In fact, almost all the ‘looming irrationality on part of the Democrats’ that *anyone* has posted here been completely and utterly without any sort of basis at all. It’s getting a little silly.

        Especially by using left-based all the Nate Silver punching as an example. Almost all the attacks on Nate Silver are pointing out that ‘data’ does not actually replace either political journalism or opinion pieces, and the problem with politics is *not* the lack of data. That wasn’t even the problem in the election coverage he was so hailed for in getting right!

        The problem there, and in general, was people who are persistently, completely, utterly wrong in everything they say and do being taken seriously, and a media that is completely captured by those people.

        If Nate Silver wants to provide more facts to use as weapons to attack those people, fine, but there’s already plenty of facts, and yet those idiots keep getting allowed to open their stupid face on TV and write their idiotic opinion pieces and be wrong about everything, forever, in all possibly ways. Total batshit insane people, being taken seriously and being allowed on air as if their words mattered at all.

        The attacks on Nate Silver is because is he naive and a little presumptuous about how the system works. ‘Oh, I just need to throw some FACTS in there!’. The attacks aren’t not anti-facts, they’re anti-facts-will-solve-the-problem.Report

    • I think this is a fair assessment, though it’s worth noting that for the majority of folks who are familiar at all with the hubub, that familiarity is arising almost entirely through group 2.

      Also worth noting is that to the extent group 1 is significantly contributing to the public debate over this (and I think it increasingly is) and is drawing responses from Silver’s crew, then FiveThirtyEight is providing a quite valuable service, whether or not its analyses are correct on a given issue, by serving as a conduit for the public to become directly familiar with the academic analyses.

      For instance: http://fivethirtyeight.com/datalab/fivethirtyeight-to-commission-response-to-disputed-climate-article/

      Even to the extent that they don’t directly respond to the criticism, though, at least some of the attacks from Group 2 wind up directly referencing Group 1, which allows Group 1 to draw more eyeballs.

      And of course when FiveThirtyEight puts together a piece that is actually not at all controversial within the given field (and it has and will do so) but defies conventional wisdom, Group 1 benefits even more immensely.Report

    • Avatar Barry says:

      ” I have little doubt that the major lib-blogs wouldn’t give a second look to the models if FTE was predicting a lively, rebounding economy and climate-related disasters of horrific costs.”

      Bullsh*t, unless you’ve seen liberal blogs criticizing criticizing people for predicting a dismal to mediocre economy.Report

  3. Avatar Kim says:

    http://fivethirtyeight.com/features/standard-market-forces-appear-to-apply-to-hospitals-too/
    Perfect example.
    It really, really sounds like these researchers have never heard of the phrase “Trauma Center,” or the designation implies in terms of services.

    (We’ve got plenty around this city, but I’m absolutely certain that the Ambulance drivers know about them — and do overrule family requests to send patients to the closer hospital rather than the better one).Report

    • Avatar Barry says:

      Thanks for spotting that, Kim. And the author is a Chicago B-School professor (working with professors from other universities), so it’s not just a bunch of idiots. But I’ve got to say that this phrase is worthy of Thomas Friedman (he’s explaining why patients with heart attacks are taken to hospitals better equipped to deal with them, rather than the closes hospital):

      “What explains the fact that patients were making their way to more efficient hospitals? It’s hard for us to say, unfortunately, since we don’t have data on what patients know and when they know it. They might learn from the experiences of family and friends, or perhaps doctors advise them during checkups. There’s even some indication that ambulance drivers, facing roughly the same drive time to two different hospitals, tend to choose the provider they know to be better.”Report

      • Avatar Barry says:

        My brother once received an outpatient surgery at a small hospital I’d never heard of, which specializes in certain types of outpatient surgeries. It didn’t *have* an emergency room.Report

  4. Avatar Patrick says:

    I think there’s an awful lot of criticism for the site given that it’s been up a week.

    Silver’s response to Krugman was pretty clearly intended as a joke, itself, and Healy’s response to it as if it wasn’t missed the point, so everyone calling Healy’s response a giant takedown is three levels of meta away from what’s really going on.

    I think you’re on to something, here, Nob, and your point isn’t without merit. Nate’s going to have to juggle “post something every day” with “post something of quality” and that usually doesn’t work out, so either the post frequency is going to have to go down or the work is going to have to be less deep or they need more people or they’re going to burn everybody out.Report

    • I dunno. I think they can also approach their initial posts on a given subject as the start of a dialogue, an opening statement if you will. Start with a broad and blunt overview of the writer’s semi-expert data analyses (ie, they have a lot more knowledge of the broad subject area than the average person but less knowledge of the specific subtopic than an expert in that subtopic). That overview will likely do one of three things: (1) confirm both conventional wisdom and expert consensus; (2) defy conventional wisdom but confirm expert consensus; or (3) step into a topic where there is dispute amongst experts.

      If it’s (1), then we can expect the discussion to go no further, but the conventional wisdom of 538’s readers will be enhanced by at least a cursory knowledge of the basis for expert consensus. If it’s (2), there may or may not be pushback from non-experts warranting a response from 538 – if not, then 538 has successfully shifted at least some people’s conventional wisdom in favor of knowledge of expert knowledge; if so, then with each additional step of the conversation, the expert opinion of the 538 author’s sources is more widely disseminated, and, again at each step of the conversation, conventional wisdom shifts slightly in favor of expert consensus. If it’s (3), there will certainly be pushback warranting a response from 538, and as long as it responds, both the dissenting and initial expert analyses get disseminated to the general public.Report

      • Avatar Patrick says:

        This is a pretty good approach, Mark.

        Dialogue-driven writing isn’t necessarily going to produce a traffic-heavy commercially viable web site, though.

        It’s more likely to produce a web site kinda like… ours…Report

      • I think there’s plenty of a market for (1) and – especially – (2), regardless of whether they result in an ongoing dialogue. People like stuff that confirms their prejudices, and stuff that challenges a given group’s CW but affirms elitist experts has a tendency to be link-bait, both from those whose CW is attacked and those ideologically opposed to that CW. Doing a lot of (3) doesn’t itself drive much traffic, but it does allow the maintenance of credibility with dissenters.

        Whether that model is sustainable in the long run remains to be seen, but I think it can work in theory at least.Report

    • Avatar Barry says:

      Patrick

      “I think there’s an awful lot of criticism for the site given that it’s been up a week.”

      I’ll keep saying it, so long as people like you are saying it – when somebody hires Pielke to discuss climate science, that’s a bad thing, and people should not wait to criticize it.Report

  5. Avatar Burt Likko says:

    We’ll have to see where he takes his new outlet, but I’m not convinced that there was a “narrative” in Silver’s previous publications, at least not in the sense that a Fox News Channel or an ESPN provides a “narrative.” A “narrative” is editorial, persuasive, emotional — the teasing out of a pre-determined theme from developing events. It’s a powerful thing when it’s done with accuracy, because it can separate truth from spin. Silver was never about that, though: he was about where the data leads. We can question whether his methods lead there or not (empirical results in both sports and politics suggest he’s pretty darn close, at least within those universes) and we can question whether the other writers working at the new fivethirtyeight.com follow those methods or at least the principles based on them. My objection is to the idea that Silver was selling “knowitallism” or “contrarianism.” When his results confirmed the CW, he’s said so. Turns out it’s not very interesting, but I can’t think of a time he’s used editorial discretion to shy away from confirming a prevailing assumption when the data justified that result.Report

  6. Avatar zic says:

    I’m reminded of Freakonomics.

    Their approach was pretty awesome in the early days; but as they progressed, and the data domains they used fell further outside their expertise, they got frisked harder; and sometimes it was obvious they just got it wrong.

    But I would be really surprised if 538 was never wrong, too. It’s how they handle it when they are that matters to me. Admitting failure/mistake is not a weakness; it’s a strength, a sign of maturity.Report

    • Avatar Kim says:

      http://fivethirtyeight.com/datalab/fivethirtyeight-to-commission-response-to-disputed-climate-article/

      This seems like a seriously awesome way to respond. “hey, we ran something that got serious criticism. We’ll run a rebuttal as well.” It speaks well to Nate that he’s defending Pielke’s actual credentials (which appear to be decently strong), and that he’s critiquing his own editorial staff.Report

      • Avatar Barry says:

        “Pielke’s actual credentials (which appear to be decently strong)”

        Snort.Report

      • Avatar Barry says:

        I should apologize for ‘snort’ as a reply, Kim. Go read the actual blogs by climatologists, and see *what they have to say* about what Pielke says. Their arguments are not just ‘neener, neener’.Report

      • Avatar Kim says:

        Barry,
        Pielke has peer-reviewed work in relevant journals. He’s not just some shmuck who showed up to spout nonsense who has never touched the data.

        That said, he does appear to be of trollish countenance, and to be pretty bad with statistics.Report

    • Avatar Barry says:

      “Their approach was pretty awesome in the early days; but as they progressed, and the data domains they used fell further outside their expertise, they got frisked harder; and sometimes it was obvious they just got it wrong.”

      And there was some stupidity above and beyond that; for example in their climate science piece, they went and listened to one guy, who had obvious incentives and was trying to sell something, and didn’t talk to anybody else. One of the climate scientists at Chicago fisked their chapter unmercifully (but justly), and concluded with a google map showing just how many yards it was from the econ building to the climate science building.Report

  7. Avatar Tod Kelly says:

    Nob –

    I’m unsure how to respond to this as a response to my post, since I’m unclear whether you missed the point of my post entirely or are merely trying to obfuscate it.

    The issue I had with the left on Silver wasn’t that I thought Silver is always right, it’s that there is now a push against the concept of objective data and objective journalism.  You were good enough to skim parts of Krugman that sounded reasonable, but here is a quick two-minute list of quotes thrown together from the pieces I linked to you seem to have missed:

    “the intimidation by quantification practiced by Silver and the other data mullahs must be resisted.”

    “Neutrality is an evasion of responsibility”

    “the emptiness of data journalism”

    “[People who preach objective journalism] preserved the banking system in this most recent crisis. They created the institutions that allowed the military-industrial complex to form, and produced technologies like the computer, atomic power, and the internet, as well as contributing to the McCarthy era and crackdowns on dissidents.”

    ” “objectivity” belies a latent jingoist ideology”

    [Proper journalists] wear their ideology on your sleeve”

    “What interests me is the way people talk about math as if it were divinely prophetic. They seem to subscribe to a religion that simply apes the terminology of science.”

    “To listen to many of Silver’s defenders, questioning his methodology is akin to rejecting evolution or the laws of thermodynamics, as if only his model is sanctified by the god Reason.”

    “Here’s where we find Silver’s ideological commitment, I think: contrarianism.”

    “I’ve been on the fence about so-called data journalism for some time. Two of my favorite commentators — Jim Newell and Elias Isquith — have eviscerated the idea”

     
    “Data nerds”, to use Silver’s term, aren’t even foxes — they’re obsessives who use their beloved numbers to predict outcomes.”

    Those quotes aren’t anti-Silver.  They’re anti-objective data and/or anti-objective journalism.

     

    Fun Bonus Brain Tease for the Kids!:

    One of these quotes was actually penned by Jonah Goldberg in 2012, in regards to Silver’s Romney prediction, and it was criticized by every other person quoted, save one!  Can you guess which one without looking?  Have all your friends play!  It’s fun!
    Report

    • Avatar Mike Schilling says:

      I guessed right. It had to be

      “What interests me is the way people talk about math as if it were divinely prophetic. They seem to subscribe to a religion that simply apes the terminology of science.”

      Because it takes a Goldberg to make a virtue of not being able to do math.Report

    • Avatar Kim says:

      Pielke’s getting beaten up for not mentioning that there was another side. This is standard practice for “guy with big ego” in academia (um, both sides, really).

      I think Krugman has a decent point — most of the time. In his field, it’s hard to not have some bias (that said, the 538 piece on marriages versus corporations is … amusing, and I believe the writer when she/he says she/he didn’t know which way it would go).Report

    • Avatar KatherineMW says:

      The relationship between ideology and data analysis has some commonalities to the relationship between religion and science.

      Ideology tells us why Analysis tells us how.

      Good analysis of data is valuable in informing how we should go about achieving our policy goals. Our ideology – our ethical, philosophical, religious beliefs about what we value and what kind of a world we want to see – centres around what those goals are. Technocrats become problematic when they try to replace ideology with data analysis and claim that makes them impartial.Report

      • Avatar Tod Kelly says:

        Is there no objective data? Should journalists pick and choose through the information available to make the most ideologically convincing argument?

        Is total objectivity impossible? Yes. Is it likely than anyone who attempts to look at multiple sides of a story still subconsciously tilts in a particular direction? Of course. Does that mean we should vilify those who attempt to be objective or look at multiple sides?

        What I find amazing about all of this is that in your attempt to go to bat for team Blue, you’re all unintentionally making arguments for Fox News being the way news should be done.Report

      • Avatar Kim says:

        Tod,
        it’s not that there is NO objective data. It’s that there is too much.
        One can do an entire, and complete, research on Amazonian legends. It’s not too big of a field.

        But now? Miles and miles of data. What data you choose to report on is a bias.

        Is Calculated Risk unbiased? No, but they generally do reporting on “big picture numbers”. That’s a proanalytic, anti anecdotal bias.Report

      • Avatar Tod Kelly says:

        Kim – That is not remotely the point any of those people I quoted were trying to make.Report

      • Avatar zic says:

        @tod-kelly having committed acts of journalism over many years, I have an insider’s secret for you (one I know from how I did my job, from talking to other journalists, and from reading the text books and blogs of J-school): bias is unavoidable. It’s how you handle it.

        1) push back on it, and try to research the points that contradict your assumptions;
        2) go with it, and work for an ideologically-based publisher;
        3) ignore it, because someone else will have a different bias.

        Given the great ‘liberal media’ myth out there, I have to say that my guess is #2 is much more common among conservative journalists (they often write for business-oriented sites, where their bias isn’t as obvious but is 100% there), #1 is common for most responsible news outlets, and #3 is most obvious in non-news writing and blogs.

        I haven’t gone through your list yet, but I wonder: how many of those links are actual journalism vs. op-ed pieces? That merits some consideration; bias in journalism is not good; bias in opinion pieces should be expected.

        Finally, there is, I freely admit, a lot of sloppy and biased journalism; it often responsible for something called cascades, where correlation is attributed to causation. (Food and nutrition reporting is rife with this.) But for the most part, I found reporters aware of the biases and found they worked hard to either overcome them or worked for places where they were accepted.Report

      • Avatar KatherineMW says:

        I’m not trying to “go to bat” for the Democrats. I’m not sure how you got that from anything I said.
        I’m simply describing my views on how data analysis is best used. And yes, we should try to be objective when analyzing it.Report

      • Avatar Tod Kelly says:

        @zic I agree with all of what you say here, at least up to a point.

        The difference that I see happening to day is this: Twenty years ago, pundits for either side attached on another, politicians, and political causes/programs/pitches. And then one day, half of them were attacking the very idea of objective journalism, and journalism in general – -which is a very, very different thing. Fast forward to today and — I would argue — those pundits so rewrote the DNA of their own base that they made a joke of not only their side’s “news” organizations, but their political party.

        Now, for the first time ever, I’m starting to see the other side declare war not only on the other side, but on the very concept of objective journalism and journalism in general.

        This is what makes me nervous.Report

      • Avatar zic says:

        @tod-kelly I agree with that; it makes me nervous, too.

        For one thing, because the ‘sides’ are so obvious, it’s fed the increase in just present two sides, and you’ve been fair and unbiased. Somehow, this seems to excuse the actual responsibility of analysis that journalism entails; not just presenting facts, but giving context to facts so that the reader can actually understand what they mean.

        I do think there’s growing confusion about the difference between journalism and opinion. Goldberg doesn’t do much journalism, though he does some. And I generally mistrust writers who think they can regularly do both.

        Yesterday, we listened to a bit of Mark Levin on the radio. “Why do they hate America?” he asked. And I wanted to ask him why he likes having sex with baby seals; because the implied assumptions in his ‘why’ went unchallenged.

        A good journalist will always plumb those depths. And should be applauded for doing so, too.Report

      • NobAkimoto NobAkimoto says:

        Is there no objective data? Should journalists pick and choose through the information available to make the most ideologically convincing argument?
        Is total objectivity impossible? Yes. Is it likely than anyone who attempts to look at multiple sides of a story still subconsciously tilts in a particular direction? Of course. Does that mean we should vilify those who attempt to be objective or look at multiple sides?

        What I find amazing about all of this is that in your attempt to go to bat for team Blue, you’re all unintentionally making arguments for Fox News being the way news should be done.

        Not to get too Bayesian on you, but no, there is no such thing as “objective data”. Data is always filtered by the expectations and structures presented by the one doing the interpretation. In fact, the very notion that there’s no conditional probabilities attached to a dataset is absurd. There’s always conditionals, often exogenous to the variables being described.

        The thing that bothers me about Nate Silver’s site, is that despite writing eloquently about the importance of Bayesian theory to interpreting data on previous occasions, Silver then chucks that all out in hiring writers who are wont to make highly faulty degrees of belief in certain conditionals, all the while not having a sufficiently rigorous editing staff that can weed out, or at least correct the problems with their Bayesian assumptions.

        Essentially, Nate Silver needs more Bayesian editors for the work being published. What’s essentially happening right now is that there’s not been enough checking on priors and therefore he’s getting smacked (rightly) for it.Report

      • NobAkimoto NobAkimoto says:

        I guess more broadly speaking, my point is that there’s a certain inherent bias in his thinking process. I think Cathy O’Neil actually did the best take on WHAT that bias for Silver is:
        http://mathbabe.org/2012/12/20/nate-silver-confuses-cause-and-effect-ends-up-defending-corruption/

        Now to some degree, this is fine. But I have a serious problem with how he defines himself as being somehow above biases and ideology, when in fact, he has a very powerful one, just not one he’s willing to admit to.Report

      • Avatar Patrick says:

        Nob:

        FWIW, that’s the most cogent and intelligent objection I’ve seen to the new 538 yet, anywhere.

        I’ll just reiterate: it’s been a week. Give the guy some time.Report

      • Avatar Brandon Berg says:

        Yesterday, we listened to a bit of Mark Levin on the radio.

        Why?Report

      • Avatar zic says:

        @brandon-berg that’s an excellent question. I’m not sure, except that we were flipping through stations, and heard his “Why do they hate America?” so stayed a moment for sake of listening to him answer his own loaded question.

        Plus, I do try to listen to opposing views.Report

      • Avatar Mike Schilling says:

        You know the expression “a face for radio”? Levin has a voice for mime.Report

      • Avatar Will Truman says:

        I sometimes joke that I have a voice for silent film.Report

      • Avatar zic says:

        Levin has a voice for mime.

        I thought he sounded like Lindsey Graham. Also mime potential.Report

    • NobAkimoto NobAkimoto says:

      I honestly think that assuming there’s some sort of ideologically neutral set of priors that you can use to get through a news cycle, is, in fact, futile. This is particularly true given how much depth Nate Silver himself goes into regarding Bayesian theory in Signal and Noise.

      As a broader point, I was primarily noting the reason why 538 was getting a lot of bad press, particularly from policy analysts and economists.Report

    • Avatar Barry says:

      First, ‘even the liberal New Republic’.

      Second, let’s take the Stoller comment – people can judge for themselves *why* matt was criticizing Nate:

      “Lots of people (Ryan Cooper, Paul Krugman, Tyler Cowen, Climate Progress, etc.) are now upset about Nate Silver’s pretense to objectivity, and his disdain for unnamed pundits. I don’t have particular insight on Silver’s new venture, generally speaking these things require time to work out the kinks. But I do want to point out that Silver has an ideology. This is a history blog, so let’s look at Silver’s political ideology as illustrated during the most momentous historical episode in his political life – the bailouts. Now, back then, he was a partisan, so he didn’t mind making explicit statements about advocacy, about what should be done to fix the world.

      And oh but he did have a plan to fix the world. He supported, strongly, Tim Geithner’s plans, and wanted activists to clear the road for whatever Geithner decided.

      This piece, which he wrote at the height of the financial meltdown in early 2009, when Obama had just taken office and was making decisions about the bailouts, details it. Oh wait, you can’t read it, it’s not there anymore. Fortunately some people quoted parts of it and their sites are still up. So here’s how Silver started his argument, and then ended it.

      The opening.”

      [Nate:] Nobody, absolutely nobody, has more incentive to get this right than the Obama Administration. If the economy collapses — well, more than it already has collapsed — then the Democrats get slaughtered in 2010, Obama is a one-termer, health care doesn’t happen, the poverty rate increases by a couple orders of magnitude, and the imperative to fix the environment gets put on the backburner. To suggest that Obama or Geithner are tools of Wall Street and are looking out for something other than the country’s best interest is freaking asinine. Maybe their ideas are wrong — but their hearts are in the right place.

      The ending.

      This is neither the time nor the place for mass movements — this is the time for expert opinion. Once the experts (and I’m not one of them) have reached some kind of a consensus about what the best course of action is (and they haven’t yet), then figure out who is impeding that action for political or other disingenuous reasons and tackle them — do whatever you can to remove them from the playing field. But we’re not at that stage yet.
      [End Nate]

      Matt: “There’s a long history in America of technocrats loathing mass movements while demanding that democratic structures serve the needs of experts. George Washington probably could be placed in this category – he was a surveyor and mapmaker, which was a technical profession, and he disliked political faction. Vannevar Bush, the co-founder of Raytheon and in many ways the architect of the modern military-industrial complex, was like this too. As his biographer G Pascal Gregory noted, Bush feared that totalitarianism was a more efficient and better political system, and sought military control over democratic decisions to compensate so as to win World War II and then the Cold War.

      So Silver fits into a long American tradition. He has ideas about how the world should work, and these are worth understanding.

      Cathy O’Neil wrote about Silver’s theories two years ago (her blog post was cross-posted in two places, with different and equally interesting comment threads). As a commenter on O’Neil’s site says, Silver is part of “the anti-politics machine.” Here’s David Sirota, at the time, arguing vehemently against this worldview.

      [David] Silver is effectively arguing that being a citizen in a democracy means taking orders from those in power, and not questioning those in power. We are all just supposed to follow blindly as the Very Serious People in Washington and on Wall Street tell us what to do. [End David]

      Now my guess is that Silver would disagree with this assessment, and say that he is driven by empiricism. But it’s also clear that the decision to bail out the banks was not empirically driven, because it couldn’t have been. It was a question of values. And for Silver, that meant letting a group of experts make decisions, and then engaging in democratic power projection to ensure that any obstacles to those decisions be removed. It is in fact what actually happened. This ideology is real, and it has consequences. People who believe in this ideology preserved the banking system in this most recent crisis. They created the institutions that allowed the military-industrial complex to form, and produced technologies like the computer, atomic power, and the internet, as well as contributing to the McCarthy era and crackdowns on dissidents. The anti-politics technocrat crowd now runs Google-land, and the American empire. In many ways, given that the Constitutional convention was written by technocratic lawyers in secrecy, you could argue Silver’s ideology helped form America.

      Regardless of whether you think that’s good or bad, or something in between, at least it’s an ethos.

      UPDATE: If you want to Silver’s essay on Geithner, it’s here. It is worth reading. It’s also here, on his new site, but at a different url. I don’t, obviously think anyone was hiding anything, the reason I wrote that you can’t find the old link is because it speaks to how little people actually care about track records. They just sort of want a sorcerer of the moment.”Report

      • Avatar Kim says:

        Poblano’s allowed to advocate for not getting a guns and cigarettes economy (being in NYC, he’s undoubtedly aware of exactly how bad that would be for him). I’ll state that the most blue-blooded liberals I know were shitting their fucking pants at that time, over exactly how Mr.Bush might wreck everything by not bailing out AIG’s execs, who were effectively holding the world economy hostage.Report

      • Avatar Kim says:

        um, yeah. anyone wondering? Poblano== NateSilver. That was his handle on dkos from ages back.Report

  8. Avatar notme says:

    Sorry to disagree, Silver is now finding out the hard way that liberals are as likely to shoot the messenger as Repubs. I think it is wonderful to watch.Report

  9. Avatar Jesse Ewiak says:

    Don’t know where else to drop this, since the original articles are probably ignore, but what the hell. Actual numbers on how much various seats have moved and whether there’s any equivalent sharp turn to the left by the DNC recently.

    http://www.dailykos.com/story/2014/03/30/1287898/-These-charts-show-you-where-we-ve-moved-the-needle-left-in-the-Senate-and-HouseReport

    • Avatar Will Truman says:

      That’s a really interesting article. Thanks for sharing it. On the GOP side it runs contrary to the popular narrative in some respects with newer congresscritters being more moderate in some instances and more conservative in others. Mixed bag for the Democrats, with a sharper leftward tilt in the senate though a more mixed (slightly more moderate) picture for the House.Report

  10. Avatar Barry says:

    Zic: “I haven’t gone through your list yet, but I wonder: how many of those links are actual journalism vs. op-ed pieces? That merits some consideration; bias in journalism is not good; bias in opinion pieces should be expected. ”

    Most are either op-ed pieces, or are written in The New Republic, The Week, or The National Review. The last magazine is at 100% op-ed level, on a good day. The Week seems to me to be the place where it’s hard to tell, but it’s always given me the impression that things which seem at first glance to be articles are still just op-eds.

    The first is ‘The New Republic’, which is a magazine with serious schizophrenia. In fact, IIRC one of the cited articles is by their ‘literary correspondent’.

    The post by Matt Stoller is one which I dumped up the page in a comment, to point out that it’s a rather fact-based post, making serious criticisms.

    At this point, I’m frustrated by Tod’s consistent ‘inability’ to differentiate between Jonah Goldberg and people offering up serious, fact-based criticism. this might be the basis of Tod’s initial article, that he’s incapable of understanding that, and so he’s got a pundit’s point of view, that any change in how people view Nate Silver *must* be political, because who would turn against somebody based on mere facts?Report