Saturday, September 10, 2016

Power to the Numbers

Power to the People Numbers
            The spotlights flare to life, and applause swells - egging the elusive celebrity to the stage. Neither wild rock star nor raucous stand-up comic, instead a small bookish man enters from the left in small strides, fixing his glasses as he goes. Not the headliner a sold out show would suggest, but an individual beloved by his audience all the same. Against all odds (which he surely could calculate) Nate Silver has made statistics somehow sexy.
            Of all the eminent statisticians of the past hundred years - Yates, Cox, Tukey, Pearson, Silver, among many more - only one numbers man is a household name (outside of a few geeky bloggers that is)2. And in comparison to the esteemed company (predominantly in of the academic community), Nate Silver’s inclusion in the ‘famed statistician’ category is a dubious one3. After all his notable contribution to the field of statistics hasn’t been a novel approach to multivariate correlations, but in fact a simple application of the very algorithms pioneered by the four aforementioned statisticians in his prediction of the 2008, 2012, and 2016 presidential elections. To the purists of the statistics world, Silver is nothing more than a glorified poker player betting on the largest national game4. And to the average American, Silver is just “the numbers guy.” He’s the man who calculates if their football team will win on Sunday, and once every four years reads the national polls for you. Barely intellectual and barely public? Has Silver’s own distance from the “research-and-publish” statistics circles and his numerical entertainment to the slightly-geeky-American cost him his seat as a distinguished public intellectual? As a pure statistician, probably yes. However, Silver still is undoubtedly a public intellectual, just not the one that his statistics pedigree might indicate. His fundamental contribution to the sphere of public intellectuals comes as a data journalist.
            Before breaking down data journalism, we need to first jump back to 1970 (not the year that Nate Silver was born, neither when he bought his first calculator). As author Jonah Keri highlights in his novel Baseball between the numbers, 1970 marked the introduction of mass statistical compilation in major league sports5. For the first time statistics – at a scope as expansive as the US census – became available and of interest to the public5. Overnight, numbers became part of our vocabulary as a nation. How many yards did Walter Payton rush for after contact? What’s Nolan Ryan’s strikeouts per appearance? The intertwining story of stats and sports is more than just fascinating subject for a novel or an Oscar-nominated feature like Moneyball. It is a microcosm emblematic of a much greater macroscopic shift towards a statistics driven media over the past 50 years. And just as sports coverage has continued with its jargon heavy coverage, our national news has all the same become irrevocably tied to “the numbers.”
Fatalities. Exit polls. Tax Dollars. Cancer Rate. Murder rate. Credit Default Swaps.
Before moving on from sports completely, it’s of interest to note that in more than one way Nate Silver has sports to thank for his career. Not only have athletics been fundamental in developing the stat-driven-media-landscape that Silver has become a central commentator upon, but in fact prior to his journalistic pursuits he was employed as a sports statistician6. In his early work, Silver developed several important baseball statistic systems used today across MLB for scouting, trading and roster building6. After six years of work within baseball, Silver found a new calling, but one not so different from the sports statistics he was familiar with:

“My fulltime occupation has been as a writer and analyst for a sports media company called Baseball Prospectus. In baseball, statistics are meaningless without context; hitting 30 home runs in the 1930s is a lot different than hitting 30 today. There is a whole industry in baseball dedicated to the proper understanding and interpretation of statistics.
In polling and politics, there is nearly as much data as there is for first basemen. In this year’s Democratic primaries, there were statistics for every gender, race, age, occupation and geography – reasons why Clinton won older women, or Obama took college students.
But the understanding has lagged behind. Polls are cherry-picked based on their brand name or shock value rather than their track record of accuracy. Demographic variables are misrepresented or misunderstood.7

In this Op-Ed published in the New York Post in June of 2008, Silver outlined the dogma of his new blog ‘FiveThirtyEight’, and with it began to popularize a new brand of political commentary, “data journalism.” Why were the news anchors and journalists getting the coverage so damn wrong? Transferring his zeal for sports statistics, Silver began working to contextualize the numbers seemingly every news organization was botching. Take for example the media accusation “Barack Obama has problems with working class voters.”7 The claim circulated during spring of 2008 that the young Illinois Senator was, “out of touch,” and prognosticated it would cost him primary voters8. In truth the connection was geographically related to Obama’s polling Kentuckian and West Virginian voters, not to his reputation with blue collar workers7. Careful studies of Obama’s polls among industrial workers in the northeast would have easily disproved the specious connection7. Unsurprising to Silver in the subsequent weeks, Obama performed exceptionally well in this demographic in Oregon, as well as the swing state of Wisconsin8. The next day the New York Times front page read “After Big Loss, Obama Woos BlueCollar Voters.”8 What was viewed by the media as an “upset to Clinton” was in fact erroneous statistics predicting a false outcome. As Silver alludes in his Op-Ed, a number is worthless without a context. In direct response to wanton statistical misinterpretations plaguing the 21st-century-media-landscape Silver developed his blog FiveThirtyEight. The blog’s doctrine posits if the media is founded upon statistics, the most correct media is directly correlated to the most correct interpretation of statistics. And thusly Silver’s data journalism was born.
The numbers-first-format is what distinguishes Silver’s brand of journalism from all other media outlets, Fox News and New York Times alike. In its purest form, data journalism is strictly the complete aggregation and thereby complete analysis of numbers, with as little outside narrative, spin or interpretation: A chart, and a title, “8:15 PM Update.9 FiveThirtyEight’s coverage of the 2008 presidential election - of the three elections detailed by Silver – best encapsulates this incessantly empirical nature of this commentary. Day to day and week to week, the narrative wasn’t a catchy headline, but rather a percentage point. Percent Obama. Percent McCain. For Silver, good numbers with good analysis told the story themselves.
 In contrast, whether it’s Paul Krugman writing on the benefits of Obama’s stimulus package or Bill O’Reilly decrying it, each author started with a political narrative. And by political, I’m not referring to the two-party-system (although party ties certainly does rear its ugly influence in our news media more than it should), instead I’m referring to it in its broadest definition, “adhering to a certain group of ideas or policies.” As journalists develop their stories they are assembling a narrative, one they hold to be true. And in that process of writing they working to select facts and figures that support: “adhering to the ideas or policies.” Facts that don’t support adhere are put into a pile to rebutted or justify. And that is the flaw that Silver’s process exposes. By dividing statistics into those ‘for’ and ‘against’ ones argument, the interpretation becomes “why the for are correct” and why the “against are incorrect.” In comparison Silver’s method looks to analyze the gamut of statistics in totality – in as few words as needed.
This isn’t to explicitly accuse Krugman or O’Reilly of cherry picking or sloppy journalism, but to highlight the inherent bias our media outlets have in this process of narrative development. As Champkin synopsizes, “the sources cited in our national media are not factually incorrect, but rather the improper context and the limited data sets undermine the correctness and interpretation of their narratives.”6
As Silver’s work has taken a more typical journalistic format in recent years, he has actively worked to reverse engineer the typical journalistic method and stay true to the numbers. In a rather self-aware retrospective after writing at FiveThirtyEight for six years, Silver wrote “data journalism, as I see it, is to apply the scientific method to the news.”10,11 In recent years, this methodology has become more and more important as FiveThirtyEight’s content has expanded beyond strictly charts and figures, and into article formatted content. Notwithstanding the process is still the same; before the math there is no narrative, only numbers. Nevertheless, Silver acknowledges that as FiveThirtyEight has grown, there too has been a growing to adapt data journalism beyond a dogmatically numbers only media, but as a standard for data driven journalism nationwide. He explains in perfect FiveThirtyEight fashion, graph and all:

“The point is that data journalism isn’t just about using numbers as opposed to words. To be clear, our approach at FiveThirtyEight will be quantitative — there will be plenty of numbers at this site. But using numbers is neither necessary nor sufficient to produce good works of journalism. Still, I would never have launched FiveThirtyEight in 2008, and I would not have chosen to broaden its coverage so extensively now, unless I thought there were some need for it in the marketplace. Conventional news organizations on the whole are
lacking in data journalism skills, in my view.”10


It is here that Silver’s journalistic integrity and intellectualism in fact supersedes his statistical prowess, and by virtue of his growing influence within the American media at large where Silver has his public footing. Like his mild demeanor, Silver’s media presence is one of equal subtly. Though Silver’s data based journalism contradicts conventional journalists, the purpose of the piece is never to lambasting like Dawkins or Hitchens might. But all the same, Silver is a public intellectual. As Mack notes, “Those concerned with public intellectuals as a class will inevitably fret about the health of that class.”12 Within this concept of the intellectual, Silver operates as the obedient watchdog of our national political discourse. Though his bark is quiet, Silver by the numbers highlights the statistical transgressions of his fellow journalists. Like a compass gives bearings to a lost sailor, Silver’s data journalism acts as an objective center to orient the often sprawling and abstruse media maelstrom. And that while golden statistical standard is a subtle one, Silver’s FiveThirtyEight machine, his television appearances, sold out shows, bestselling novels, and reputation do have an unmistakable effect on shaping our media that establishes Silver among the ranks of top intellectual journalists operating today.
Beyond the simple implications within the political intelligentsia, Nate Silver offer’s evidence to the fundamental role of intellectuals in maintaining democracy. Donatich posits the notion that consolidated experts are inherently corrosive to society, history bound to repeat12. What he fails to acknowledge is the potential harmony of intellectual community held to checks and balances. In an appropriately democratic metaphor, Nate Silver operates within the media community as the Supreme Court does to the White House and Congress. And within a democratic system, public intellectuals can in fact be healthy influences, indispensably educating the public on issues beyond their scope of knowledge.
Nate Silver is best known as a statistician, and an association doubtful to ever change. But unlike statisticians, his work is not focused on the research and development of statistical methods as a university doctorate would. Silver’s work is nothing new, but simple application of thoroughly designed statistical systems and models. It is rather the underlying journalistic application of Silver’s work, not the statistical quality that defines it. By elevating an empirical, data driven conceit to the forefront of our media circles, and challenging the status quo of numerical malapropisms Silver has established himself in his own rarified sphere of the public intellectual body. His influence is a subtle, but steady force in accurately shaping the numerical discourse that channels through our twenty-four-hour media circuit. And certainly his football predictions each Sunday doesn’t hurt his public popularity either. Perhaps Silver hasn’t made statistics sexy, but he has made them palatable, understandable, and all the more powerful when properly used in our modern media.

Endnotes
1Conversations with Tyler: A Conversation with Nate Silver. (2016). George Mason University.    Retrieved from http://mercatus.org/events/conversations-tyler-conversation-nate-silver
210 Famous Statisticians of the 20th Century. (2013). Retrieved from             http://famestatisticians.blogspot.com/
3Wang, S. (2012). Nerds Under Attack. Princeton University. Retrieved from
            http://election.princeton.edu/2012/10/29/nerds-under-attack/
4Trende, S. (2016, May 12). The Value of Data Journalism. Real Clear Politics. Retrieved from
            http://www.realclearpolitics.com/articles/2016/05/12/the_value_of_data_journalism_130   534.html
5Keri, J. (2006). Baseball Between the Numbers: Why Everything You Know About the Game Is    Wrong. New York: Basic Books. pp. 4-20.
6Champkin, J. (2013). Nate Silver. Reference: Statistics Journal. 10 (6), pp. 36-39.
7Silver, N. (2008, June 1). Margins of Error. New York Post. Retrieved from
            http://nypost.com/2008/06/01/margins-of-error/
8Nacourney (2008, May 13). After Big Loss, Obama Woos Blue-Collar Voters. The New York      Times. Retrieved from http://www.nytimes.com/2008/05/13/us/politics/14cnd-       campaign.html?_r=0
9Silver, N. (2008). 8:15 PM Update. FiveThirtyEight. Retrieved from
            http://fivethirtyeight.com/features/815-pm-update/
10 Silver, N. (2014). What the Fox Knows. FiveThirtyEight. Retrieved from
            http://fivethirtyeight.com/features/what-the-fox-knows/
11 Scientific Method (1999). University of Rochester. Retrieved from
            http://teacher.nsrl.rochester.edu/phy_labs/appendixe/appendixe.html

12Mack, S. (2016). The Supposed Decline of the Public Intellectual. The New Democratic Review. Retrieved from    http://www.stephenmack.com/blog/archives/2016/08/the_supposed_de_1.html#more

No comments:

Post a Comment