Kobe Bryant and the Digital Humanities

What does one of the most successful and polarizing basketball players in history have to do with the digital humanities?

For those that don’t follow the NBA, Kobe Bryant is famous for a host of accomplishments: winning five championships, league MVP honors, and an Olympic gold medal, leading the league in scoring twice, winning the All-Star dunk contest, and scoring the second most points in a single game in history. He has also been accused over the years of placing personal success ahead of the team, undermining teammates and coaches, and most notoriously, of sexual assault in 2003. From a basketball standpoint, however, one of the most enduring aspects of Bryant’s career has been an overwhelming consensus of his ability as a “clutch” player. There exists a widespread perception that no other basketball player on earth is better at the end of close games. Both NBA players and general managers have repeatedly and overwhelmingly voted Bryant as the player they would want taking a shot with the game on the line. Bryant’s name and legacy have become entwined with the word “clutch.”


Unfortunately, this is a flawed narrative. Henry Abbott recently wrote a blistering (and persuasive) analysis of Bryant’s abilities as a “clutch” player. Abbott concludes that, by nearly every statistical measure he examined, Bryant is not the best in the world at scoring points at the end of close games. Depending on the metric, Bryant is somewhere between decent and very good, but nowhere close to the best. Perhaps most damningly, the effectiveness of his team’s offense (the best in the league during Bryant’s tenure) plummets at the end of games.

So the question remains: what does Kobe Bryant have to do with the digital humanities?

The fault line in the basketball world over Kobe Bryant’s “clutchness” largely falls between those that evaluate Bryant’s ability by what they see and those that evaluate his ability by what they measure. For someone watching Bryant, no other player has as many breathtaking, memorable game-winning shots and no other player looks as graceful and impressive while doing it. I draw a parallel between this qualitative analysis with more traditional humanistic research: we read our sources and look for meaningful or interesting patterns that jump out at us. On the other side of the basketball fault-line stands a young but growing movement that advocates for more rigorous statistical analysis of basketball, in the same vein as the sabermetric “Moneyball“ movement in baseball. For these stat-heads, the seductive aesthetic appeal of Bryant’s game-winning shots hides the less glamorous reality: that Bryant misses those game-winning shot attempts at an extremely high rate. And this is the side of the debate that I would compare to the digital humanities.

The analogy isn’t perfect. Much of the work being done in the digital humanities field is not, in fact, quantitative (and making the comparison brings to mind the less-successful turn towards quantitative history in the 1960s and 1970s). But the analogy does have  some useful parallels. Like the stats movement in the basketball world, digital humanities has a lengthy history but has only recently begun to gain traction across the wider academy. Like the stats movement in the basketball world, digital humanities is occasionally seen as threatening or, at the very least, promising too much. Like the stats movement in the basketball world, there are those in digital humanities that revel in revisionism and using new techniques to challenge conventional narratives. And like the stats movement in the basketball world, there are divisions within the digital humanities over method, approach, and emphasis.

One of the most important parallels to be drawn is how the digital humanities are increasingly being used to strengthen (rather than replace) traditional humanistic study, just as advanced statistics are being used in the NBA to strengthen analysis. In the past, a basketball player would be evaluated by a handful of traditional statistics, perhaps most importantly: how many points do they score? Today, teams and scouts are looking at more advanced metrics: for instance, how efficiently do they score those points? In the same vein, traditional literary history might look at a handful of canonical works in order to draw broad conclusions about, say, early-19th century British fiction. Today, advocates of distant reading are measuring trends across hundreds or thousands of early-19th century British novels beyond the canonical authors. Most of these digital researchers would continue to acknowledge the literary importance of Charles Dickens over a barely-published contemporary novelist, just as most stat-heads would acknowledge the importance of a player that scores a moderately-efficient 30 points per game over a player that scores a hyper-efficient 5 points per game.

Comparing the two also highlights their limitations. Some aspects of basketball can’t be measured, such as whether or not a player is a good teammate or how likely they are to stay motivated after receiving a contract or whether they’re likely to end up injured. Similarly, human experience can be an elusive target to study with technology. Charting the prevalence of certain phrases across time using Google NGrams offers, at best, a largely superficial indicator that requires careful and more extensive investigation, while cataloging every slave ship voyage might serve to mute and depersonalize the particularities of individual slaves.

In both the statistical movement in basketball and the digital turn in the humanities, new approaches allow for new questions. Henry Abbott and others have not “proven” that Kobe Bryant shouldn’t take the last shot of a game, but they have raised important questions: would Bryant’s team be better served by using him as a decoy? More broadly, is the long-standing convention of putting the ball into the hands of your best player in an isolation situation at the end of the game even a good idea? Using digital methodologies in the humanities can also serve to pose new kinds of questions, but I think the field should model itself more explicitly after the statistical basketball community in having specific questions drive those methodologies. There is a tendency to build tools and ask research questions later. This is useful, but I’d also like to see more focused questions along the lines of “Is Kobe Bryant a clutch player?” Those of us who advocate for the use of digital tools and techniques in the humanities could benefit from taking a break from the library and turning towards the basketball court.