Why political journalists can’t stand Nate Silver: The limits of journalistic knowledge

The more I think about the rift between political journalism and Nate Silver, the more it seems that it’s one that’s fundamentally an issue of epistemology — how journalists know what they know. Here’s why I think that’s the case.

When we talk about the epistemology of journalism, it all eventually ties into objectivity. The journalistic norm of objectivity is more than just a careful neutrality or attempt to appear unbiased; for journalists, it’s the grounds on which they claim the authority to describe reality to us. And the authority of objectivity is rooted in a particular process.

That process is very roughly this: Journalists get access to privileged information from official sources, then evaluate, filter, and order it through the rather ineffable quality alternatively known as “news judgment,” “news sense,” or “savvy.” This norm of objectivity is how political journalists say to the public (and to themselves), “This is why you can trust what we say we know — because we found it out through this process.” (This is far from a new observation – there are decades of sociological research on this.)

Silver’s process — his epistemology — is almost exactly the opposite of this:

Where political journalists’ information is privileged, his is public, coming from poll results that all the rest of us see, too.

Where political journalists’ information is evaluated through a subjective and nebulous professional/cultural sense of judgment, his evaluation is systematic and scientifically based. It involves judgment, too, but because it’s based in a scientific process, we can trace how he applied that judgment to reach his conclusions.

Both of those different ways of knowing inevitably result in different types of conclusions. Silver’s conclusions are at once much more specific and much less certain than those of the political punditry. The process of journalistic objectivity can’t possibly produce that kind of specificity; that’s outside of its epistemological capabilities.

This is what David Brooks is saying when he tells Politico’s Dylan Byers, “If you tell me you think you can quantify an event that is about to happen that you don`t expect, like the 47 percent comment or a debate performance, I think you think you are a wizard. That`s not possible.” There’s no place in Brooks’ journalistic way of knowing for quantifying the possibility of future events, so he can’t see how anyone else can know that, either. It’s simply impossible.

Joe Scarborough gets us even closer to the clash between processes of knowing when he tells Byers, “Nate Silver says this is a 73.6 percent chance that the president is going to win? Nobody in that campaign thinks they have a 73 percent chance — they think they have a 50.1 percent chance of winning. And you talk to the Romney people, it’s the same thing.” How does Scarborough know that Silver’s estimate is incorrect? He talked to sources in both campaigns. In Scarborough’s journalistic epistemology, this is the trump card: Silver’s methods cannot possibly produce more reliable information than the official sources themselves. These are the savviest, highest inside sources. They are the strongest form of epistemological proof — a “case closed” in an argument against calculations and numbers.

The other objection political journalists/pundits have to Silver’s process is evident here, too. They don’t just have a problem with how he knows what he knows, but with how he states it, too. Essentially, they are mistaking specificity for certainty. To them, the specificity of Silver’s projections smack of arrogance because, again, their ways of knowing are incapable of producing that kind of specificity. It has to be an overstatement.

In actuality, of course, Silver’s specificity isn’t arrogance at all — it’s the natural product of a scientific, statistical way of producing knowledge. Statistical analyses produce specific numbers by their very nature. That doesn’t mean they’re certain: In fact, the epistemology of social science has long been far more tentative in reaching conclusions than the epistemology of journalism. As many people have noted over the past few days, a probability is not a prediction. Silver himself has repeatedly called for less certainty in political analysis, not more. But that split between specificity and certainty is a foreign concept to the journalistic epistemology.

TL;DR: Political journalists are skeptical of Nate Silver because they don’t understand and don’t trust the means by which he knows what he knows. And they don’t understand it because it’s completely different from how journalists have always known things, and how they’ve claimed authority to declare those things to the public.

No, even that’s TL;DR: When journalistic objectivity is confronted with scientific objectivity, its circuits are fried.

Share

31. October 2012 by Mark
Categories: political journalism | Tags: , , , , , , , | 148 comments

  • dsjoerg

    “It is difficult to get a man to understand something, when his salary depends upon his not understanding it!”

    • pauldwaite

      True. When Brooks says “If you tell me you think you can quantify an event that is about to happen that you don`t expect, like the 47 percent comment or a debate performance, I think you think you are a wizard.”, he’s misunderstanding what Silver is doing. He’s not saying he can predict that stuff. He’s saying that stuff doesn’t actually matter very much.

      • http://twitter.com/pinkbunny70 PB for Obama 2012

        Actually these events can have an impact, but over time, they get smoothed, and the impact lessens, as other new pieces of information gets absorbed.

        The DNC and the 47% clearly provided momentum and trajectory for President Obama, over time, it was impacted with Debate Performance 1, then even that was smoothed out,and President Obama started again his momentum.

        The impact of the event (coefficients) depends on relative importance of key factors affecting voters (variables)

      • kahner

        I think its not that those factors don’t matter, its that their effect (or lack of effect) is evident in freely available polling data.  You don’t have to try to read the tea leaves or guess how it will play with white suburban soccer moms.  You just have take poll data and do the math.  If white suburban soccer moms hate Romney’s position on abortion it shows up in the poll numbers.  All the punditry analysis is at best pointless or at worst misleading.

  • Just the facts, Ma’am

    Moreover, media love the horse race:  Mitt was down, now he’s up and  had momentum.  Obama was down, then up but no momentum, then Sandy hit now Mitt is stalled…blah…blah..blah.  That keeps us all glued to them to tell us the next wrinkle. But what if that narrative did not truly prove out in voter polling data? Now  comes along some guy to blow up the media horse race narrative with cold hard numbers that expose some of the false drama and makes pretty clear who will likely win…and the media reaction:  well that simply can’t be so because we didn’t say it was so!

    • pdxuser

      Yes, having a tied race is good for ratings, clicks and circulation. Being able to create narratives about a tied race is a major financial asset, because a tied race commands more attention and generates more interest. So now that someone is threatening that economic model by denying those narratives credibility, some in the business are trying to defend their moneymaker by attacking the threat.

  • http://www.facebook.com/people/Brett-Turner/1684817908 Brett Turner

     This is very similar to the battle between baseball scouts and baseball sabermetricians, to which Nate devotes a chapter in his book.  After 10-20 years of debate, most ball clubs now agree that there is value in both approaches.  In fact, because the two approaches are so fundamentally different, an approach which relies upon both sources is actually more accurate than an approach which relies upon either one alone.  That is where the Joe Scarborough vs. Nate Silver debate will eventually end up—but we may need a few decades of fighting to get there.

    • moomistercowman

      asdf

  • http://twitter.com/claydelk Clay Delk

    It reminds me of the NPR story yesterday about the first computers used in election night coverage in the 1950′s.
    http://m.npr.org/news/U.S./163951263

    The most interesting part is the story of Lou Harris, a noted pollster at the time, who beat the computer in 1952. But then we went to work for IBM and helped develop programs that crushed the competition (human and computer) in the next election.

    Nate Silver wasn’t a political analyst when he started doing this in 2008. Maybe he’s more politically savvy now, but at the core he’s a data guy. 

    What really scares them is that his predictions were almost perfect last time without the type of political experience most of them have. Now–to steal from Romney’s messaging–he’s four years closer to being a political expert too.

  • http://twitter.com/vr1000 Rich Hastings

    Great quote that relates to these ‘medidiots’ in Moneyball from the Red Sox owner, John Henry:

    “I know you are taking it in the teeth, but the first guy through the wall… he always gets bloody… always. This is threatening not just a way of doing business… but in their minds, it’s threatening the game. Really what it’s threatening is their livelihood, their jobs. It’s threatening the way they do things… and every time that happens, whether it’s the government, a way of doing business, whatever, the people who are holding the reins – they have their hands on the switch – they go batshit crazy.”

  • Walt French

    I’m not sure that Morning Joe and Fox are really “journalistic” enterprises. They know what their viewership wants to hear; drily objective facts that don’t favor the Right are NOT what they want. In fact, it’s more fun to continue the bashing of anybody who isn’t part of the club.

    They’re tribal, not journalistic outfits.

    • http://economistsview.typepad.com/ Guest

       Coddington used two Republican entertainers (Brooks and  Scarborough) to build his case causing him to miss the point.

      The reason the conservative operatives are trying to destroy Silver is because he has the tools and (with the NY Times backstopping) the resources to publicly spot and call out electoral fraud – this scares the crud out of the right more than anything. It is definitely going to be harder to steal this election.

      • NeonVincent

         He was able to do that with the Iranian election.  The numbers looked both too good and too fishy at the same time,.

  • Dan Vergano

    I disagree. It is a mistake to confuse the pathology of our political reporting/punditry caste, which is a serious problem in journalism, with a failure of the entire journalistic method. The ideas known as ‘precision journalism’ popularized more than a decade ago roughly aimed to combine both the primary sources of reporting and data-driven reporting that Silver and other database reporters pursues. Sadly, economics have conspired to destroy a lot of room for precision journalism at news outlets today, leaving us with our modern era of opinion writing for page views.

    Since 1921, when Walter Lippmann wrote Public Opinion, all serious thinkers about what we do have known that ‘objectivity’ in journalism is a shibboleth; only academics have continued to wield this strawman as a club to beat up on the profession in a bid to assert their own agendas. I’d urge you to reconsider the relevance of the sociological literature to the real problems of the profession. There are real problems and shortcomings inherent in the journalistic method, but to hook onto the axe-grinder’s idea of objectivity as the root cause of the problems is a mistake.

  • Marvin Brown

    VERY astute analysis. Thank you.

    • Marvin Brown

      Also, Hook ‘em.

  • jetjocko

    I’m not sure the “rift,” if there is one, is between political journalists and Silver. Between pundits and Silver? Sure. Because they make their living based on gut intuition (and as a result do about as well as chance would predict). But the real reporters aren’t attempting to predict; they have their hands full just reporting on what’s going on in the present.

    • markcoddington

      I think you have a point. I wrestled with referring to “political pundits” throughout the post, as opposed to “political journalists.” I ultimately decided against it because when it comes to how they know what they know (which is what I’m most interested in here), there’s no real difference between the two. 

      As you noted, they might do different things with that knowledge – report on the present vs. predict the future – but the process by which they come by that knowledge is the same, and it’s fundamentally journalistic in origin.

      • http://twitter.com/aberg7 Andrew Berg

        And if one was strictly a “reporter,” there would be scarcely little to report WRT Silver. The pundits are the reporting elite who get to pepper in their feelings, so they necessarily stand in as mouthpieces for the fourth estate.

  • Gina Mosko

    You’re attributing scientific objectivity to opinion polls?  That’s rather bold.

    • markcoddington

      No, I wouldn’t attribute scientific objectivity to opinion polls. That would be, yes, foolhardy. The “scientific objectivity” is a reference to Silver’s methods of analyzing the poll data – not that his method is really fully scientifically objective, but that it’s patterned after that model.

      Political journalists don’t seem to have a problem with poll data (on the contrary, they seem to love it) because it’s quite close to what they do, epistemologically – talk to lots of people, then extrapolate that to a description of reality. Their problem here is with what Silver does with that poll data – building models and attempting to describe reality through mathematical and statistical principles. That’s where the conflict lies, not with the poll data itself.

  • Pingback: Old story: The insiders go after the stat guys. New twist: in politics | HardballTalk

  • Dgarant

    Thanks for the post & comments, but I see 2 important contributing factors that aren’t being addressed: the electoral college, which is difficult to understand, explain, & defend, especially in the context of news media; and the ad-driven business model of most media, which has built-in incentives to portray elections as nail-biters & cliff-hangers.

    Ohio is mentioned as a key swing state but with little discussion about why; the electoral map means that for Romney to reach 270 he must either win Ohio, or he must win nearly every other swing state, and this is a situation Nate’s models have consistently predicted even before the primary campaigns began.

    For the most part, Nate’s critics either have a vested interest in him being wrong, or they make their livings ginning up controversy but are never held accountable for bad predictions, indeed a study showed that mainstream pundits are right only about one-third the time. Nate’s considered predictions — based on actual data — are, for these media entertainers, like the old joke about the jerk who tells the reader of a murder mystery “the butler did it”.

  • http://twitter.com/leicaman leicaman

    I think this is one big heap of sophistry. Being a journalist myself, I can clearly state that the claim here of what is the epistemology of journalism is a straw man. No legitimate journalist believes that, and no legitimate journalist makes a claim for objectivity. A lot of us talk about the ideal state of objectivity, which is impossible to arrive at by anyone. (That’s the certainty part.) Statistics are not about objective facts, but analysis with less than absolute certainty.

    The idea that journalists obtain secret knowledge from official sources that they then use “news judgement” to determine if it’s true or not and then pass on to their readers is simply laughable. I don’t know any real journalists who would give “official sources” such credibility.

    Sure, there are a lot of hacks in the journalism world, who bandy about the word “objectivity” like a snake oil salesman telling you he’s going to cure your gout. But that hardly makes this specious but fallacious argument meaningful or factual.

    • kringlebertfistyebuns

      I think if you substitute the word “pundits” for “political journalists,” you’ll find the argument made holds up far better.

    • pdxuser

      “I don’t know any real journalists who would give “official sources” such credibility.”

      So then why do they quote ”senior administration officials” so much?

    • http://twitter.com/jonathanstray jonathanstray

      When a political reporter writes a politics story, where do they get their information? A large fraction of the sources are campaign officials or politicians, with polling organizations, industry analysts, and the occasional academic expert thrown in. If those aren’t “official sources,” what are?

  • ceolaf

    While I think that Nate Silver is doing the best election work out there, we should not foget that he does not reveal his model. Thus, it is impossible for others to evaluate, comment upon or improve it. So, though I approve with his approach, I — like everyone else — cannot actually comment on whther or not he is making mistakes in his work that he should catch. 

    • learnedhand64

      If Silver is doing the best election work (I agree), should we care about his model?  Whatever it is, it has to be more accurate than “a reliable source”.   

  • yetanothersteve

    First of all, most people don’t understand math and numbers. 73% chance of winning is a close race. 3 to 1 is NOT “called.” It’s a lot of uncertainty. But I think it sounds like 73%-27% in the polls where the probability would round to 100%.

    And if you’re in one of the campaigns 73% or 27% feels like 50%. 27% events happen A LOT. It’s a lot of uncertainty.

    • RandCraw

      A ‘prediction’ (description, really) is even less convincing when it’s based only on a likelihood value (e.g. 73%) which lacks context for that value, that is, a confidence interval (e.g. < 5%) and an assurance that the sample is unbiased and distributed consistently. 

      Silver's credibility hinges on our faith that 1) his methods are selective enough to reliably discriminate between signal and noise, 2) his methods are fair and unbiased, and 3) he honestly reports all of his results and doesn't cherrypick for entertainment's sake, 4) his methods have worked in the past.

      Lacking the political street creds of political talking head, Silver's value depends entirely on his personal honesty and technical ability.  And that's a racehorse of an entirely different color.

      • http://twitter.com/balehman Ben Lehman

         538 gives confidence intervals. They’re slightly hidden, but if you mouse over the graphs you can see the confidence interval.

  • http://www.facebook.com/profile.php?id=10133982 Mark Poepsel

    It is too convenient to call Mark’s discussion of objectivity in political reporting a straw man.

    Perhaps “legitimate” (whatever that means) political reporters don’t claim to report with pure objectivity. Perhaps scholars rely on “objectivity” as vaguely defined in textbooks and more clearly defined by Kovach and Rosenstiel too often as a means of describing the ostensible truth strategy of most political reporters and pundits.

    Even if that is the case, if Mark is misrepresenting the truth strategies of journalists, what is a better term for political journalists’ dominant truth strategy?

    Are you saying C-A-R supplemented by qualitative information is best?

    By calling Mark C. on a potential logical fallacy and suggesting he’s misrepresenting journalists’ truth strategies without offering a better term and definition, it seems you’re (Dan Vergano and leicaman) only offering half an argument. If Mark’s interpretation of the dominant epistemology of political journalists’ is wrong, what’s right?

    Isn’t cutting funding for investigative journalist handing a great deal of control over to those who work within the most vague paradigm of objectivity (often broadcast journalists)?

    Or do you just want Mark C. to specify that, although journalistic “close as we can get to it” objectivity is the dominant epistemology for political journalists and pundits, there are those who do something else, who are more honest about the impossibility of scientific objectivity in journalism and less reliant on a vague kind of objectivity based primarily on source credibility and anecdotal information?

    If all you want is a shout out for those who do it “better” – why not skip the attacks and just tell us all how political journalism should be done so that the old dog scouts and the Sabermatricians can do what we all really do best – drink?

  • Pingback: Aktuelles 2. November 2012 — neunetz.com

  • http://profile.yahoo.com/GMLZOIQ75GREUYG42TN5Y2ZJHY KevinK

    As someone who has “predicted probabilities” on Wall Street, I can assure you Nate Silver is full of ****.

    He has Obama at 80% to win?  Okay, ask him if he’ll bet you 4 dollars to your 1 dollar on Romney?  SPOILER ALERT: He wont make that bet.

    Better yet, give him fantastic odds (by his model) of merely 3 to 1 that Obama will win. SPOILER ALERT: He wont make that bet either.

    Translation?  Nate Silver knows he’s full of crap.

    • Mrbenzduck

      KevinK, what makes you so certain Silver wouldn’t take that bet?

      Offshore betting sites like Ladbrokes in the UK currently have Obama at 1/4. These sites post odds to ensure equal money is bet on each side, so one can assume a sizable number of bettors are taking the same bet you mock Silver for confirming as a legitimate line — as many bettors are taking Obama at 1/4 as Romney at 4/1.

      Look at it this way. If a pro football team has a 3 point lead with two minutes remaining, statistical odds of winning the game are about 80%. You’d like to be the team in the lead, but you’re nervous as hell.

    • Claus Portner

      Actually, his predictions are very close to what you get from the actual betting services and they are purely in the game to make money. They would not be offering essentially the same odds if they believed they were going to lose money on the bet. Ladbrokes in Britain has Obama with odds 1/4, while Romney is at 11/4. Hence, if you put in $1 on Obama and he wins you get $0.25 in profit, whereas if you put $1 on Romney and he wins you get $2.75 in profit. Nobody seems to be claiming that Ladbrokes is a raving Obama supporter who is just out to confuse and annoy people, but then again you never know with those brits. Oh, and Silver did actually put up a bet offer to Joe Scarborough (and got in hot water with the NYT’s public editor - http://publiceditor.blogs.nytimes.com/2012/11/01/under-attack-nate-silver-picks-the-wrong-defense/ ).

    • Alan83059

       Not what his track record shows…guess if he had Romney as the favorite you would have no issues…by the way I’ll take either of your bets…I’ll take intrades odds as well…..

    • ICDogg

      His odds are very close to what the actual books are offering. If you want to make that bet on Romney, go to BetFair. There’s money to be made…if you’re right.

    • Neil Kandalgaonkar

      He already challenged Scarborough to a bet.

      https://twitter.com/fivethirtyeight/status/263948379497840641

  • RMR

    Yep.  Quantitative vs. Qualitative.  It goes way back.  Reasonably smart people have appreciated the benefits and drawbacks to both approaches, Silver included.  But because the quantitative stuff is less intuitive, that side of the “aisle” comes around much more slowly, if at all.

  • Albert Monroe

    I think that (most) journalists are bad at math, and they denigrate what they don’t understand.

  • Toklatak

    Saint Thomas Aquinas held in his philosophy that the argument “from authority” (something is true because an authority said it was true) is the weakest argument in a proof. It is in fact a tautology. Maybe the problem is these people are not conservative enough and we need to go back to scholastic philosophy.

  • http://twitter.com/BrashEQLibrium Brash Equilibrium

    It isn’t journalists that hate Silver. It’s conservatives, and it’s simply because Silver’s model happens to predict that Obama will win. End of story.

    • Dave N.

      Exactly. Nate got the 2010 midterms almost to the number, and that was a heavy Republican shift. No one seemed to have a problem with it then. His cumulative score from 2008 and 2010 is 97.5% correct. Small sample size, yes, but I’ll accept his models.

  • Alan83059

    I only wonder how much we would be hearing from the right if Silver had their guy on top with an 80% chance of winning.  Interesting we heard nothing during the 2010 elections when Nate was spot on with his predictions, blowing away most of the journalists opinions.  At the end of the day this is a discussion of agenda, if Nate’s model works to the benefit of your guy, he’s a god.  If not he’s a dog, unfortunately, this dog can hunt and has a track record to prove it!!

  • Pingback: Today’s War on Nate SIlver: Quiet Flows the Don Edition | FavStocks

  • crowdog

    thanks for this.

  • Timfsull

    Is Joe Scarborough, in any traditional sense, a journalist? Is Dylan Byers? Is mark halperin? They type yes, but they dont seem to me to be the sorts who strain after some ideal notion of objectivity, even if the model for journalism you posit is accurate. I think they are basically merchants of gossip and/or bluster and they are deeply threatened by the fact that silver and the other election quants are generating commentary that is, however uncertain and scientifically humble, much more predictive and persuasive than the off the cuff reasoning and gossip dropping they engage in.

  • simsalabim

    Very good post. I’d like to add one thing: Irrespective of political leaning there seems to be a misunderstanding of how to evaluate his model (, or in terms of your post: how to assess the specificity Silvers model provides in assigning probabilities to uncertain events).

    People attacking him seem to be of the opinion that he and his model will have failed if Obama doesn’t win. Or that his model has already failed because it “predicted” the wrong winner in the Nevada Senate race in 2010.

    People defending him always point to his 2008 results with statements like “He predicted 49 out of 50 states” as if the value of the model always increases with the number of states it “predicts” correctly.

    Both sides are wrong. The only reasonable way to assess the quality of his model is to look at whether it behaves as expected.
    Assume, for example, that the model assigns every state to either Romney or Obama with a probability of 80%. Then the expected number of correct “predictions” would be 40 states. If his model was right in all 50 instances, the probability that the model has an incorrect way of assigning probabilities is quite high and I would think it is hedging too much. I would, on the other hand, see failure in predictions (at least in the range of 8-12) as a crude validation of the model.

    I don’t know whether Silver or anybody else has actually compared actual performance to expected performance, but it would be nice to see.

  • Joe

    Well done. Thank you.

  • rmanvil

    Mainstream pundits, particularly the ones who appear on television, are intimidated by Fox News and the right wing. If Nate Silver calculates that Obama has a 75% chance of winning, Silver’s analysis cannot have have any validity, because it isn’t “fair and balanced.” Similarly, if Mitt Romney tells twenty lies and Barack Obama tells five lies, then only five of Mitt Romney’s lies may be discussed in polite company. To do otherwise would be to exhibit bias and incur the wrath of the fairness police.

  • blwood

    I think this is exactly right. The same controversy has long existed in clinical psychology, though it should have been resolved by Paul Meehl’s classic “Clinical vs. Statistical Prediction” published in 1954″: http://www.psych.umn.edu/faculty/grove/114meehlscontributiontoclinical.pdf

  • Pingback: News › What Does Nate Silver’s ’80.9% Chance of Winning’ Mean, Anyway?

  • http://www.facebook.com/roy.basch.9 Roy Basch

    The conversation about Journalistic epistemology is long overdue — perhaps in general, but at least for folks like Joe Scarborough (as an archetype rather than as an individual).

    Lawyers have known for years that eyewitness evidence is often the most susceptible to being wrong but that “circumstantial evidence,” while it gets bad PR, is often orders of magnitude more reliable in getting to the capital T truth.

  • VirtualGuest

    The “I’ve got a gut feeling that the race is a coin-flip because Mitt’s got momentum!” school of punditry is quickly becoming deader than 70′s era disco pants and men’s platform shoes.

    Expecting them to like it is like expecting the Dinosaurs to “embrace the comet!”

  • Pingback: Nate Silver, the New York Times, and the Challenge of Being Your Own Disruptor « Groundswell

  • Tall Guy

    Honestly, I think most journalists (and most people unfortunately) don’t understand any probability other than 50% and 100%.   

    If Nate Silver predicted 100 things at 80% each and 80 of them came true, they would say he was right about 80 and wrong about 20.  They just don’t get it.

  • tga999

    I think they’re just reinventing the kind of critiques and caveats of these methods that you’d hear about in an introductory statistics  course.  Yes, there is a risk of model error.  No, that doesn’t mean that the models’ predictions are useless.

  • http://twitter.com/shinobi42 Shinobi

    As a statistician I have long believed that journalism is the natural antithesis to statistics.  Journalism relies on anecdotes and sensational stories, where statistics relies on consistent measurement and treating information methodically. 

  • Pingback: This Week in Review – Nieman Journalism Lab | The Journalism Site

  • skmind

    “Nobody in that campaign thinks they have a 73 percent chance — they think they have a 50.1 percent chance of winning”

    And Joe, how did you compute that number?

  • http://twitter.com/jonathanstray jonathanstray

    All true, I think. A few thoughts:

    - Journalistic objectivity was originally modelled on scientific objectivity. Funny to see everything come around to discover that they’re different things.

    - Part of the attacks on Silver are coming from outraged Republicans. In such cases, it might have less to do with epistemology and more to do with partisanship.

    - The epistemology of journalism is very rich indeed.  But while I’m a strong advocate of quantitative methods where applicable (as here) ultimately they are not strong enough or universal enough to form the basis of journalistic truth.

    - Two books I recommend highly in this regard: Kahneman’s “Thinking Fast and Slow” and Hueur’s “Psychology of Intellgence Analysis” (available free online.)

    I’ll be digging into this a lot more, and making some suggestions in the future… I think it’s a travesty that journalists are not trained in the state-of-the-art in truth finding, and I think we’re going to need some j-school curriculum changes.

    • markcoddington

      Thanks for the comments and suggestions, Jonathan. I’m curious to see what you come up with as you think about this further.

      As to your second point, I think you’re right that there are two distinct sources of critique, and that they arise from completely different places theoretically. The origins of that conservative critique don’t interest me nearly as much as the journalistic one, mostly because I’m not nearly as interested in political science and psychology as I am with sociology. But there is something there, I think, that’s deeper than just “the right hates him because he makes Romney looks good.”

  • http://twitter.com/Augustine25 Augustine 25

    I’m trained as a political scientist. I think you are missing a very simple solution. The polls Nate Silver is relying on add up to an extraordinarily high level of probablity of an Obama win only because they are using a voter turn out model based on the 2008 results. All of this is disclosed in the polls themselves. Nate also adds in an economic adjustment that is not at all scientific.

    All in all, you can get a more accurate forecast simply relying on Rasmussen or Gallup. Sadly, I think you are overestimating Nate’s skill in science or forecasting. Last time I checked, Nate only has a B.A. degree. He really isn’t trained to handle these tasks with the sort of authority the left gives to him.

    • http://www.facebook.com/klgallaher Kenneth Gallaher

      that meme has gone round and round and is FALSE.

  • Pingback: What Does Nate Silver’s ’80.9% Chance of Winning’ Mean, Anyway? | iTAX – tax news

  • Pingback: Epistemology of Journalism | Physicalism

  • Pingback: Abbreviated Pundit Round-up: Tales of the election | Hotspyer – Breaking News from around the web

  • http://www.facebook.com/klgallaher Kenneth Gallaher

    Being a bit of a stats nerd myself – having made my living at it for many years I KNOW that it is possible to extract GOLD from noisy data.  That is what Nate does.  To those on the outside it looks like alchemy – but it’s not.  What the righties say about Nate’s results is a combination of ignorance and delusional thinking and pixie dust – the same place Mitt’s budget comes from.

  • Thecosmiccowboyranch

    In politics today there are very few who have the depth of demographic and political knowledge required to even argue with Nate Silver…

    And you don’t hear Chuck Todd arguing with him…

    Nuf said…

  • superseiyan

    Part of this is traditional journalists fear of agenda setting.  Look at Sullivan’s (NYT editor) condescending comments to Nate Silver, talking about he was elevated by the institution. But isn’t it vice-versa? 

    Also look at some of the malpractice from mainstream journalists.  You have Wolf Blitzer calling 4 point Obama leads a “tie” (yet other closer races with Romney leading aren’t reported as such).

    Or Chris Cilizza deciding that Ohio, which Obama has greater leads than Romney does in North Carolina, is a toss-up because of the “necessity of Romney to win there”.  

    Traditional journalists are no longer the sole agenda setters now, there can be dueling narratives.  

  • Pingback: What Does Nate Silver’s ’80.9% Chance of Winning’ Mean, Anyway? | Web Development, Search Engine Optimization, Social Media Marketing Guru

  • http://www.facebook.com/steve.parsons.9277 Steve Parsons

    I don’t think it is so much *objectivity* )or technical know how)as it is that Nate makes topline statements. Every journalist at one time or another interviews “experts” but they (the journalist) asks the questions and synthesize the information into the topline summary.  Nate poses his own questions, answers them and reaches conclusions.  What’s more, the Times has given him a megaphone.

  • Pingback: The Privileged Epistemology of Journalism | Dustin Boyer