“Sir, the possibility of successfully navigating an asteroid field is approximately 3,720 to 1.”
“Never tell me the odds!” ―C-3PO and Han Solo
Yogi Berra famously threw the fat lady off her stage in 1973 when he said, “It ain’t over till it’s over.” With the rise of and reliance upon data-driven modeling of elections and sports we might just as well rephrase it as, “It’s over before it begins.” But we’d be wrong to do so.
Like most oddsmakers going into Super Bowl LI, Nate Silver’s FiveThirtyEight.com, owned by ESPN, predicted the New England Patriots to win. Going into the half-time as the Falcons were up 28-3, the site gave the Patriots a less than 1 percent chance of winning. FiveThirtyEight tweeted: “That Patriots drive took another 5:07 off the clock and actually dropped their win probability from 1.1% to 0.5%.”
Of course we all know what happened next. In yet another brilliant statistical upset in a year of upsets, the Patriots defied all probability after the half. They scored 25 unanswered points, taking the Super Bowl into an historically uncharted overtime which they then proceeded to win—giving America, and the world at large, a clinic in determination, momentum, and the ability of human beings to surmount even the greatest of statistical odds.
It was a lesson in the value of risk taking and accomplishment; values that were once core elements in the American mythos but that increasingly have fallen out of favor in exchange for the perceived infallibility of data-driven models and analyses.
Since the mainstreaming of data punditry, exemplified by Nate Silver’s meteoric rise and FiveThirtyEight’s hallowed place in the culture, we’ve seen a cultural shift with regard to the use of statistics and data. Big Data, polling, and more specifically, Silver’s predictions, have become the equivalent of a mic-drop in any conversation about sports or politics. Throughout the election cycle, on TV shows and social media feeds across the country, his pronouncements were treated as sacrosanct papal bulls. His data-driven analysis, whether accurate or not, provided gravitas for those seeking a more commanding way to eviscerate opponents in debate. “Silver gives Hilary a x percent chance to win the election” became the trump card in any conversation.
We’d moved to a point where we seemingly were willing to assign data modeling more value than the possible variances, irrationality and risk-taking inherent in human decision-making. This happened during the Super Bowl just as it happened during the election. In both cases, statistical models were held up as unassailable predictors.
And in both instances, they were wrong.
For his part, no matter how certain Silver was of his model, he’d often hedge. In October 2016, under the headline “Clinton Probably Finished Off Trump Last Night,” Silver wrote: “I’m not sure I need to tell you this, but Hillary Clinton is probably going to be the next president. It’s just a question of what ‘probably’ means.” (emphasis added) He then spent the bulk of the article convincing us that Clinton would win, but at the end noted the possibility he could be mistaken. When results of the Republican primary, the Michigan Democratic primary, and the general election proved him very wrong, Silver’s postmortem explanations moved the goalposts, claiming event X or event Y was unprecedented, thus skewing the initial models. Even after the Super Bowl, in an attempt to make light of the situation, he tweeted: “At least the Falcons won the popular vote.” To which a user responded, “Nate, you don’t get to make election jokes.”
Silver also acknowledged in a lengthy post-election analysis that subjective best guesses and metrics are often baked into the stats when unprecedented things happen. By saying this Silver, admits that pure stats—facts, figures, polls, and data—might work for averages and as descriptors, but they cannot accurately adjust for extraordinary events and people. This was best summed up by David Morris, when he wrote about Silver’s failure to predict Trump’s victory in the Republican primary: “Unlikely events like the Trump nomination are, by their very nature, impossible to predict.”
The models, thus, don’t ever really predict the future. They are informed best guesses that describe how current events would likely play out if those events and the responses to them conformed to the past. The trouble with trusting the Oracle, however, is that when history occurs, it is often a break with the past.
Silver’s accuracy is not the issue here. Everyone get things wrong from time to time. It’s just that despite being fabulously wrong over and over again, and despite his admissions of fallibility, people still cling to his pronouncements as the ultimate argument from authority. This signals a more profound structural problem with the culture—one too eager to find quantifiable solutions to complex and often unquantifiable situations, especially when those quantifiable solutions comport to their views of the world as it should be.
It’s not Silver. He’s just the fetish for the phenomenon.
Jason Rhode, in Paste Magazine, opens his withering critique of Silver with a quotation from Federalist 55: “Nothing can be more fallacious than to found our political calculations on arithmetical principles.” And yet today many seem to believe that Silver is arithmetic made flesh, as such he’s an avatar of a cultural desire for statistical certainty in light of a constantly changing and often unpredictable world of human interaction and politics. He is Hermes bringing us the word of the gods. But sadly, we miss the point of hermeneutics, that discipline of critically assessing the nature of Hermes’ message.
Instead, invoking FiveThirtyEight seems to bestow upon the speaker of the Silverian incantations an air of both intellectual superiority and mathematical indifference. “Nate Silver predicts…” is akin to saying “Shut up idiot, what do you know? The numbers don’t lie, don’t doubt your betters!” But that appeal to Silver is really an appeal to the illusion of a fully predictable future.
Ultimately, an overreliance on Silver—and Big Data in general—is a quasi-religious attempt to bring order out of chaos, an almost fundamentalist approach that borders on number zealotry. It’s an attempt to overlook how little we know about what we imagine we can design.
The American zeitgeist until quite recently has been opposed to this view of human nature and events. From our movies which stress against-the-odds comebacks to our national mythos as the set of upstart colonies that managed to defeat the strongest empire on earth at the time, we have reveled in being exceptions to the rule. This thinking, in turn, has lead to a national character that stressed self-reliance and risk taking.
But now, with a large segment of the population and an even larger segment of our leaders all too happy to reduce human interaction to data points, we run the risk of becoming an increasingly risk averse and technocratic society where people value comfort over vision, ease over innovation and utility over passion.
Statistics are an integral component in decision making, but as the caveat during every infomercial tells us, “Past performance is not a guarantee of future results.” Ultimately, the Super Bowl, the election of 2016, and so much of history show us the problem with technocrats and those that would use the pronouncements of statisticians as some guaranteed proof of outcome. They can’t take into account human ingenuity, grit, and the ability to create hope and momentum in the face of time decay and defeat. Silver’s failures of late are so traumatic to those who would quote him as scripture because it upends their notion of a society and a human nature whose interactions can be easily reduced, predicted and thus controlled.