Keeping Score

In late July 1846, the Knickerbocker baseball club won a game by the score of 31–22. The contest was largely meaningless, played between members of the New York club to hone their skills. The Knickerbockers did, however, keep score. Not just tallying runs, but also creating statistics: when players batted and which ones did so successfully, when runs were scored and the players who scored them, when outs were made and the players who made them. Playing baseball meant creating data.

Baseball fans—and sport fans more generally—know this story, even if the particulars are unfamiliar. Baseball is a numbers game. And while the Knickerbockers kept paltry records compared to those that emerge from Major League Baseball games now, they were acting in the same vein. To evaluate ability, understand strategy, and improve performance, you need data. As has been claimed for decades, one essential feature, if not the defining feature, of modern sports is that they are quantified and measured. Baseball in particular is well suited to numbers. It’s amenable to mathematical modeling. It’s a game of inches. It’s a game of statistics.^[1]

This conclusion is doubly misleading. It implies that there is something natural about statistics in sports like baseball. As if keeping score were just part of playing the game. True, participants do often care who wins. But data aren’t simply the intellectual residue of sporting contests, collected after the fact as a matter of course. They are also physical, material things that must be manufactured, curated, and maintained.

The conclusion also belies the historicity of numbers: baseball’s statistical origins were inseparable from nineteenth-century ideas about scientific progress, about attempts to make games and exercise scientific and “manly.” Numbers are nothing if not fungible and flexible, able to travel efficiently across time and space, but numbers don’t carry their interpretation along with them. Over the twentieth century, quasi-legal gambling and the proliferation of fantasy leagues have made sport statistics into big business. In the twenty-first century, the “moneyball” phenomenon has made baseball statistics emblematic of the triumph of objective, data-driven knowledge over tradition-bound expertise.^[2] As is the case with so much of the bluster surrounding big data and informatics, however, there’s been little thought given to the physicality of statistical data, and to the people and material technologies by which they were made. Every time fantasy league participants check the standings, they unknowingly draw on this long history of keeping score.

• • •