Twenty years ago, on September 27, 2004, the front section of the New York Times included a story about John Kerry, then the Democratic nominee for President, and a Chinese assault rifle. During an interview with Outdoor Life, Kerry had suggested that he owned one, but the Times was reporting that Kerry, in fact, did not. Elsewhere in the paper that day, there was a preview of the upcoming debate between Kerry and George W. Bush and a brief check-in with Howard Dean, who, a few months before, had screeched at the end of a speech in Iowa and squandered his early polling lead in the primaries. There were also seven letters to the editor about a Stanley Fish op-ed titled “The Candidates, Seen from the Classroom.” Fish had polled his freshman writing class at the University of Illinois Chicago about who had won an earlier debate between Bush and Kerry. An overwhelming majority—13–2—believed Bush had been the more effective communicator. Fish provided their “devastating” analysis, which included assessments on who had used topic sentences to begin their answers, and his own aside about how Bush, at a pivotal moment in the debate, had listed off countries that contain the letter “A.” “He and his speechwriters deserve credit for using the accident of euphony to give the argument cohesiveness and force,” Fish wrote.
There were no stories about the polls. In contrast, on September 23, 2024, the day I am typing this column, the Times has stories on the polls in Sunbelt states, a “State of the Race” about Harris’s small debate bounce, and another story titled “What’s Behind Trump’s Best Poll Results in Weeks.” I have long been quite critical about the amount of space the polls take up in our political discourse. The reasons are pretty simple: while polls certainly have a place in assessing the state of the election, they’ve inspired a type of sophistry in which the pundit or the politician flashes the results of some fallible poll and treats it as irrefutable proof of the will of the electorate. The result is a tower of bad takes, built upon a foundation of solid polls and good pollsters. The question is not whether we should “trust the polls.” It’s whether the onslaught of analysis and extrapolation that invariably follows them actually holds any predictive or explanatory power.
We, the punditry, offer up these polling takes because you, the public, want to read them. You want to read a poll that says your preferred candidate is up five points nationally and then, two days later, when a different poll says your side is down in swing states, you want to feel the panic of needing to do something to right the ship. In those moments of dread, we offer up possible solutions to make the polls look like they did two days ago. An assembly line has been constructed: one of hundreds of polling outfits generates a number, which then goes to someone who tells the public about the number. Then the number is delivered to a finishing department that tells you how to feel about the number. What we don’t really know is whether this assembly line of discourse is better now than it was twenty years ago when Stanley Fish was writing about all the countries that contain the letter “A.”
This polls-first approach to political commentary came, in part, out of fantasy sports. If you look at today’s polling gurus, you’d be hard pressed to find one who didn’t at least dabble in fantasy sports. Many of them—including Split Ticket’s Lakshya Jain, whom I interviewed last year, and Nate Silver, the founder of FiveThirtyEight—got their starts on fantasy-sports sites. The overlap is obvious enough—if you can craft an argument around one set of numbers, the skill will likely transfer to another set of numbers. In some sense, this was a positive development, given how gate-kept, sclerotic, and obscurantist the field of political science can be. But the move to fantasy politics also encouraged a type of thinking that distracts from the actual show—the candidates and their promises—and promotes, instead, the belief that a few tweaks here and there can lead a campaign to the promised land.
Fantasy sports turned the fan into a manager. Sports talk increasingly centered on transactions and contracts. Players were discussed as value-producing machines. Today, thanks to the explosion of sports-gambling content and the continued dominance of fantasy leagues, sports outlets pump out all sorts of silly numbers, many of which, at least in the gambling sense, are misleading or just flat-out nonsense. Is this worse than the days when some bourbon-soaked dilettante would write six-thousand-word profiles of an eccentric country-club legend for Sports Illustrated? I don’t feel much nostalgia for stuffy, overly indulgent prose and the metaphors of sport, but every time I turn on a game or read a sports Web site I feel assaulted by a flood of odds, advice on same-game parlays, and what I find to be a mostly thoughtless fealty to numbers. While watching Florida State play against the University of California, Berkeley, last Saturday, I noticed that the announcer kept reciting what “analytics” would do in any given fourth-down situation. It was striking to hear “analytics” not only as a monolith—certainly different analytics would offer up different answers—but also as an anthropomorphized entity whose conclusions cannot be questioned. Whose analytics are we even talking about? Are we so sure they’re correct?
Fantasy-politics pundits similarly see themselves as managerial consultants for one party or another (usually the Democrats). I would provide you with a list of perpetrators, but, frankly, almost everyone in the commentariat, including me, does a lot of this these days. We look through polls, peer in at undecided voters in swing states, and then tell you, with all the nerdy confidence of an N.F.L. quant who says you should bench Matthew Stafford in week four and pick up Baker Mayfield off the waiver wire, that Kamala Harris should absolutely tack to the ride on immigration or that Donald Trump is trailing in the election because he can’t control his erratic impulses. The problem is that, while fantasy sports more or less relies upon an agreed-upon foundation of statistics, fantasy politics usually bases its conclusions on much more suspect polling data. The N.F.L. quant, at least, is giving you an actionable set of instructions that will lead to a result. If you pick up Mayfield in fantasy football, you will either score more, less, or the same points than you would have with Stafford. This isn’t really true of the fantasy pundit. We don’t really know why elections were won or lost outside of spotty exit-poll data and anecdotes. Even when we can properly identify issues that voters care about—whether it’s the economy or how honest a candidate is or isn’t—there are still missing connections in logic between the fantasy-politics advice and the results of an election. We can make an educated guess, with a whole host of appropriate disclaimers about sample sizes and the quality of one poll or another. But that type of wishy-washy, almost academic talk doesn’t sell all that well, so we shellac the whole thing with a bunch of confidence and pray that the world proves us correct, perhaps by random chance.
The influence of fantasy-politics players comes, in no small part, from a near-existential insecurity about all we don’t actually know. We want to believe that the numbers can alleviate our uncertainty, and that a bunch of nerdy wizards hold the algorithm that will unlock the bounties of the world. But the assembly line of number production and analysis has come at the expense of informing voters about the candidates. We do not have to abandon polls as a tool, nor do we need to return to an era when much was made about the height of each candidate and the tie colors they wore. But, when so many stories are about how Trump and Harris are doing in swing states and when the actual issues are mostly discussed through the effect they would have on those polls, there’s little space left to actually make a moral or even a real-world case for how a candidate’s stance on an issue would change the country. It is true that one candidate winning over another will yield differing policies, and if I believed more in the predictive power of the last part of the assembly line, in which the numbers get used to tell the candidate what to do, I might be more willing to tolerate this type of discourse, but, when Donald Trump ramps up his attack on Haitian residents of Springfield, Ohio, and screams “you have to get them the hell out,” the game of fantasy politics feels both unserious and beside the point. ♦
