Sunday, November 11, 2012

Un coup de dès jamais n'abolira le hasard

About a week ago I wrote a post adding to my ongoing series on puerility, observing how the cultural phenomenon of the FiveThirtyEight blog and the conflicts surrounding it exemplified a discourse in which discrete, mutually exclusive outcomes are the only imaginable ones. Then, while I drove to Maryland for a workshop, stopping in Philadelphia on the way back, about eighty people commented on my post to let me know that they were persuaded (in error) that I was somehow defaming Nate Silver personally and statistics as a field, and that it was up to them to defend both.

This seems to me to suggest two things.

First, that the same logic that gives us "Obama or Romney?" as the paramount question one can ask about an election also gives us "for or against?" as the paramount question one could ask about the FiveThirtyEight blog. This is in no way an interesting question to me, but for many people it was the only question, and therefore my post could only be read as answering it. This reduction to discrete, mutually exclusive, and usually binary outcomes legible in the terms of a game is of course what I was identifying as a form of puerility in the first place. Obama or Romney? Statistics or "gut"? Nate Silver or Politico? Quants or scouts? These questions clearly generate a great deal of pleasure, as evidenced by the enthusiasm with which they are debated, but there are other questions, involving words like "why" and "how," that are worth discussing.

Second, that childhood is so overwhelmingly treated as a debased category that to invoke it is considered an insult.* In addition to its literal meaning of "boyish" (Latin puer), "puerile," in common usage, carries a pejorative connotation, of course, but that's its least descriptive and least useful aspect, which is why I set it aside. Puerility, in the sense of the performance of child masculinity, is one of the most powerful political forces in the present moment; that's why there is a "Nate Silver phenomenon" in the first place. It should go without saying that anyone can engage in this performance, but it is also worth noting (so I noted it) that Silver's public persona (white, male, youthful, virtuosic) makes him a particularly good candidate, out of the many people and organizations aggregating polls, to emerge as the celebrity of popular political statistics.

On election night it was interesting (though not surprising) to observe how, once the presidential race was called, Silver began to be celebrated on Twitter (elsewhere too, I'm sure, but Twitter is time-stamped) as if he were the magical wizard that, prior to the election, Silver himself so patiently tried to explain that he was not. A lot of the tweets were really funny (funniest), but many of them oddly called the Obama victory "a win for statistics" or even "a win for reality," as if to suggest that the validity of either were contingent on who won the election.** Some even explicitly trod into "Nate Silver IS a wizard!" territory:

Such celebrations seemed to concede (erroneously) that Scarborough et al. had a point in the first place— that a Romney win would have falsified Silver's model, and that Silver's model were based on occult wizardry rather than weighted averages of widely available polling data. As Siva Vaidhyanathan put it:

And surely many of the people declaring Silver the real winner of the election knew this, and had even, prior to the election, said it. This put no damper on the explosion of Silver jokes, however; the pleasure of play trumped the basic premises of the very thing being celebrated. The cultural meaning of statistics was precisely puerile at this moment, openly signifying "winning team" more than it signified the actual principles of statistics.

Another form of data analysis was also declared a winner after the election, it is worth noting—the data-mining that enabled the Obama camp's microtargeted get-out-the-vote effort. This was swept into the same category as Silver's poll averages and made a cause for celebration. But as Zeynep Tufekçi, who had earlier argued that work like Silver's had the potential to limit the puerile logic of the horse race, observed, data-mining is ethically neutral at best, and is as eagerly pursued by Target as by the Democratic Party:

Or as Alexis Madrigal put it, "Data Doesn't Belong to the Democrats". "The left's celebrating the analytical method right now as if it belonged to them," Madrigal writes. "But it doesn't. [...T]his election was not a triumph of data over no data, of rigor over hunch. The 2012 election was a triumph of Democratic data over Republican data." What Madrigal predicts next is a data analysis war, as Republicans struggle to catch up with and exceed the facility already achieved by Democrats.

This is indeed probably what will happen in 2016, and it is about winning. In such a discursive environment, we can easily have another election in which drone strikes are not up for debate at all. But who wins—the question that FiveThirtyEight and the political parties' data-mining efforts each, in their different ways, attempt to answer***—only ultimately matters in the context of policy questions. Are we able to ask them?


*This is complicated, to say the least.

**This is in contrast with the predictions in individual states, which, taken together, are rather more meaningful for evaluating the model.

***I.e., Silver tries to answer by prediction based on polling data, while the data-miners try to answer by trying to secure a particular outcome.

Friday, November 2, 2012

The Passion of Nate Silver (Sort Of)

Needless to say, I am fascinated by the recent Silver Wars, which erupted, oddly, in the midst of a gigantic hurricane bearing down on the east coast. (No major damage up in my neck of the woods, although an uprooted tree on the green did reveal a skeleton. h/t Miriam.)

The "Nate Silver phenomenon" is a perfect example of Second Gilded Age puerility, a form of political commentary that is concerned not with meaning or ethics but rather with phenomenality, especially as translated into abstract forms, chief among them numbers. When I use "puerility" in this way, I don't mean it pejoratively but literally: this is a form of boyishness, as boyishness has been constructed in U.S. history. It's concerned first and foremost with abstract play—even a certain virtuosity with play—and it is entirely bound up its own game. And it is a game that may be a little ruthless, a game that implicitly must be played by a white, boyish figure, a Tom Sawyer who insists on playing even when a slave's freedom is at stake.* Silver's Wunderkind image creates kind of persona from whom we are prepared to receive statistical models; it is entirely appropriate that his statistical forecasting began not in politics but in sports.

Nate Silver's models can tell us how likely it is that Obama will "win" (the game). They can't, and absolutely do not aim to, explain, say, the role of race in the election. And they cannot give definitive predictions either, only probabilities: that's the point. Statistics always pulls back from the claims it makes; if it did not do so, it would not be statistics. Statistics is an inherently puerile discipline, not because it is dominated by men but because its principles concord so strongly with the way we have constructed boyhood—an unrelenting commitment to the play of abstract forms above all else: above wishes, above belief, above ethics, its only ethics being a commitment to the rules of the game. It presumes being unable to really know "the answer," except as defined and bounded by the game.

The Silver backlash has a huge problem with this. In the Politico piece that seems to crystallize the backlash, Dylan Byers quotes the NYT columnist David Brooks:
"I should treat polls as a fuzzy snapshot of a moment in time. I should not read them, and think I understand the future," he wrote. "If there’s one thing we know, it’s that even experts with fancy computer models are terrible at predicting human behavior."**
The key here is the word "understand." Brooks thinks that Silver thinks he "understands" the future. But understanding has nothing to do with it; there are simulations, and they indicate the probability of potential outcomes. It's not understanding; it's pointing.

The backlash to the Silver backlash is even more interesting. Silver's most ardent defenders are wholly immersed in the logic of puerility: new-media, moderate-left, youthful, exclusively male***—most notably Silver's fellow statistical Wunderkind, Ezra Klein. Their chosen tactics, moreover, have at times taken forms that we might associate with the other sense of "puerile." An article in Deadspin, Gawker Media's sports site, for instance, delights in the flamboyant immaturity of name-calling: "Nate Silver’s Braying Idiot Detractors Show That Being Ignorant About Politics Is Like Being Ignorant About Sports."

The Defenders accuse the Backlashers of two things—ignorance of statistics and a reflexive personal hatred of Silver founded on defensiveness—and suggest that the two are, in essence, identical. The latter, it is interesting to note, is a highly psychologized accounting. In TechCrunch, Gregory Ferenstein writes,
Why does Silver, who is really just an apartisan puzzle-solver, inspire so much loathing? Because his results reveal a psychologically disturbing fact: we live in an uncontrollable, unpredictable world.
As far as Ferenstein is concerned, the reason Silver is being criticized is that he reveals a truth that some people can't handle. That truth is, essentially, that statistics is a valid method of producing knowledge about reality; in other words, decrying Nate Silver reveals an ignorance that is the same as a psychological weakness. Klein's take is less cosmic but equally psychological:
Klein packs a slight dig at the Backlashers' masculinity in that phrase, "makes them feel innumerate." "Innumerate" is code for "inadequate," but a particular kind of inadequate; it's a castration complex built on an ignorance of statistics. Silver, as a methodologist and as a person, is "threatening" to traditional journalists (Ferenstein). The Defenders impute wrong feelings to the Backlashers, wrong feelings that are indistinguishable from wrong knowledge.

What is so interesting to me about the Defense—which is, if anything, more impassioned than the Backlash—is that it finds in the Backlashers a profound moral failing. Ferenstein's TechCrunch piece literally includes a picture of Galileo, calling up a larger-than-life myth of the forces of dogma unfairly pursuing a scientific crusader whose, as the Indigo Girls would have it, "crime was looking up the truth." Yup; Nate Silver, Galileo; I totally see it.

But the real failure of the Backlashers is a little more complex: not a moral failure, but rather a failure to be okay with the moral absence at the heart of statistical methods. The Silver backlash wants an answer, a position; it wants Silver to stop playing around. In other words, it reads statistics itself as waffling and double-tonguery. It's not wrong in that sense. It just fails to appreciate that that is more or less the entire point of statistics: to measure what is irreducibly uncertain.

There's certainly a basic failure to grok statistics that underwrites the comments by Joe Scarborough, David Brooks, et al. (the Backlash). And undoubtedly they are craven, miserable, petty people, as Ferenstein, Klein, and others suggest, although I have my doubts about proving the latter by way of the former.

But it's also important to break out of the puerility sandbox for a minute. We do not have to suppose, as Robert Schlesinger does in U.S. News, that the only alternative to "quants" is "gut feeling traditionalists" and "conventional wisdom"—that is, non-knowledge. There are good reasons to be wary of the statistical mode, if not reasons dreamed of by David Brooks, and they do not necessarily involve siding with the Ancients in a battle against the Moderns. Klein writes that
If you had to distill the work of a political pundit down to a single question, you’d have to pick the perennial “who will win the election?” [...] Now Silver—and Silver’s imitators and political scientists—are taking that question away from us. It would be shocking if the profession didn’t try and defend itself.
Perhaps so. But what if that weren't, after all, the question?

A Nieman Lab defense of Silver by Jonathan Stray celebrates that "FiveThirtyEight has set a new standard for horse race coverage" of elections. That this can be represented as an unqualified good speaks to the power of puerility in the present epistemological culture. But we oughtn't consider better horse race coverage the ultimate aim of knowledge; somehow we have inadvertently landed ourselves back in the world of sports. An election is not, in the end, a game. Coverage should not be reducible to who will win? Here are some other questions: How will the next administration govern? How will the election affect my reproductive health? When will women see equal representation in Congress? How will the U.S. extricate itself from permanent war, or will it even try? These are questions with real ethical resonance. FiveThirtyEight knows better than to try to answer with statistics.**** But we should still ask them, and try to answer them too.*****


The first woman I have seen to comment on the Silver Wars is Margaret Sullivan, the New York Times's public editor, and it was to reprimand Silver for playing around—literally. Silver offered to make a bet with Backlasher Joe Scarborough about the outcome of the election. Aunt Polly Sullivan writes that "[i]n a phone conversation, Mr. Silver described the wager offer as 'half playful and half serious.'" Which is, of course, the essence of the appeal of FiveThirtyEight.


*I am of course alluding to the ending of The Adventures of Huckleberry Finn.
**It's bad enough to have to link a Politico piece, but I'm not linking Brooks.
***I have not yet seen a Silver defense by a woman, although I suppose they must exist. I have seen many, prominently placed, by men, however.
****Silver's NYT colleagues, Dubner and Levitt, do not know better, unfortunately.
*****I am all too aware that a perfectly plausible gloss of this post would be: "don't hate the player; hate the game."