Finance, politics—and sport? From the perspective of science and technology studies (STS), these industries are more similar than they might first appear; all are high-stakes sectors that produce masses of data but are often run as much by gut feeling as analysis. The rise of “quants” in each sector has pitted traditional punditry against new forms of expertise, with resultant disputes over meaning and authority that have been of great interest to STS scholars, at least when it comes to economics and politics. In the case of professional sport, sustained attention from STS scholars is only just emerging.1

While this lag may not be surprising, sport is actually the paradigm of the challenges posed by data analysts to traditional experts—a paradigm embodied in Moneyball, the book and subsequent film that put the power of statistics in general and “sabermetrics” in particular on the map for many.2 Of course, quantitative analysis of sport (professional and otherwise) is a longstanding practice, but Michael Lewis’s account of the Oakland A’s helped frame it as a matter of competing visions of expertise. The decades since have seen a similar contest play out in nearly every sport, including football—soccer in the United States—where data-based approaches to ball events and individual player movements have entered into particular tension with more traditional forms of authority embodied in coaches. This makes football an ideal case through which to examine what it means to be an expert in the world of professional sport.3

There is reason for caution for data enthusiasts. Data analysis in other high-velocity environments (such as business) need not take into account the role of intuition and confidence so central to elite performance and sport.4 More data are not always better, especially if we consider the role that overthinking plays in the widely recognized phenomenon of choking under pressure across a range of sports. In the case of football, the coaching staff is tasked with weeding out from their data analyses any information players do not need, in the interest of imparting the kind of precise, simple instructions that will improve performance.

This form of translation may well be the primary epistemic virtue of good coaching, as hours of video footage are condensed into fifteen-minute videos for player training and the analysis of assistants is turned into precise midgame instructions on the fly. Having (typically) been players themselves, coaches know that the “superior” bird’s-eye view afforded by video analysis is only part of the story. To translate such analysis into wins, the “inferior” view from the pitch itself must be considered. The result is a delicate balancing act, conducted live, in which expertise means both marshalling data and knowing when not to marshal it. Data are only useful insofar as they help produce wins on the pitch, by the players.5

Is this enough to dissuade the football data analyst? Not at all. The partial knowledge of players is exactly the type of ill-structured problem that excites such analysts. The result is an ongoing turf war between traditional and novel forms of expertise, pitting the “eye test” of seasoned football men against those who see—behind the passes and goals of a match—the permanent laws of mathematics. The result is a war not only over the substance of expertise, but over its subjectivity—must one have been a part of the football world for life, or is a math degree just as validating as a playing career? These social tensions produce huge amounts of distrust, with traditional experts doubling down on their experience and analysts describing such experience as not just irrelevant but dangerous. Indeed, analysts have gone so far as to eliminate human judgment from the consideration of data themselves!6

Of course, sectarianism has always been a feature of football expertise.7 But unlike the use of goal-line technology and video assisted referee (VAR), which appear as part of televised broadcasts, the algorithms of in-house data analysts are hidden from sight, so as to confer a tactical advantage on the pitch.8 The hidden nature of this new expertise mirrors the extreme black-boxing of the mechanisms on which it is based; both the analysts and their algorithms are kept out of sight. Many take this commitment to inscrutability so far that even the scientists responsible for building football algorithms find it impossible to predict how an individual result will be computed or an individual decision will be produced. This is, in a sense, a form of iconoclasm, in which the idols of football wisdom have been so thoroughly destroyed that a human understanding of sport becomes impossible, with decision making reduced to trial-and-error experiments conducted by the machine.9

What can we, as spectators and commentators, make of this battle of the experts? Will data analysts succeed in taking over for traditional football experts? And if they do, will they be positioned as the new experts—or will their algorithms own that expertise, occluding the nature of decision making in football not only from players, coaches, and fans but from the analysts themselves? It is possible to see this form of black-boxing not as a displacement of the genius of traditional coaches but as an extension of it: the algorithms now occupy the old role of the side-line manager, shielding players—or in this case, all humans—from data that might endanger their performance, and which they are incapable of understanding anyway.


For a discussion on the role of quantitative expertise in finance, see Donald MacKenzie, Trading at the Speed of Light (Princeton, NJ: Princeton University Press, 2021). For emerging STS scholarship on sport analytics, see Christopher J. Phillips, Scouting and Scoring: How We Know What We Know about Baseball (Princeton, NJ: Princeton University Press, 2019); and this post on STS and sport in Europe by Michiel Van Oudheusden and Gian Marco Campagnolo: “A Question of Sport: Opening a New Research Agenda in Science and Technology Studies,” EEAST Review 39, no. 2.


Michael Lewis, Moneyball: The Art of Winning an Unfair Game (New York: Norton, 2003). The subsequent movie Moneyball (Columbia Pictures, 2011) is directed by Bennett Miller.


Simon Kuper and Stefan Szymanski, Soccernomics (New York: Nation Books, 2014). For a recent account on how the “March of the Geeks” has advanced in football in the last decade, see Christoph Biermann, Football Hackers: The Science and Art of a Data Revolution (London: Blink Publishing, 2019).


Neil Pollock and Gian Marco Campagnolo, “Subitizing Practices and Market Decisions: The Role of Simple Graphs in Business Valuations,” in Making Things Valuable, ed. Martin Kornberger, Lise Justesen, Anders Koed Madsen, and Jan Mouritsen (Oxford: Oxford University Press: 2015), 89–108.


Giolo Fele and Gian Marco Campagnolo, “Expertise and the Work of Football Match Analysts in TV broadcasts,” Discourse Studies 23, no. 5 (2021): 616–35.


Adrian MacKenzie, Machine Learners: Archaeology of a Data Practice (Cambridge, MA: MIT Press, 2017).


Stephen Turner, “What Is the Problem with Experts?” Social Studies of Science 31, no. 1 (2001): 123–49.


Harry Collins, “The Philosophy of Umpiring and the Introduction of Decision-Aid Technology,” Journal of the Philosophy of Sport 37, no. 2 (2010): 135–46.


Paul Di Maggio, “Adapting Computational Text Analysis to Social Science (and Vice Versa),” Big Data & Society 2, no. 2 (2015): 1–5.