Towards a Techno-Scientific, Socio-Cultural, Tactico-Strategic History of Basketball Analytics

I know, I keep rebooting. In my last post on this topic, I referred to a larger (academic) research project on the rise of basketball analytics.  The first step in the project is a 5000-7000 word article for a workshop organized by and special issue of the Journal of Sport History dedicated to “Doing Sport History in the Digital Era.”

However, I’m also realizing that there’s enough material, and enough curiosity on my part, to make a book out of this.  Moreover, I discovered that there are no comprehensive histories of statistical thinking in basketball (or in sports, period) of the sort that exist for statistical thinking in general. So I think there’s a need; perhaps for a companion volume to Ball Don’t Lie! called Numbers Don’t Lie! The Quantification of Basketball. Who knows? Perhaps it’ll be volume 2 of a Hoops Quartet, I’m beginning to visualize.

As I learn more, I’m finding I’m more interested in understanding, describing, and offering interpretations of the various facets of the rise of basketball analytics and less interested in making judgments about it, as I did too hastily here.

To that end, today I formulated two basic questions guiding my thinking and research.

Screenshot 2016-02-06 13.21.22

That is, I have questions about the causes or conditions of possibility for the rise of analytics and questions about the effects of this rise.

Next I tried to turn these questions into a concept map, to which I added a branch for milestone events.

Screenshot 2016-02-06 16.27.12

I then started filling in some of the first conditions of possibility I could think off the top of my head, or that basketball people on Twitter suggested. There are in no particular order.

Screenshot 2016-02-06 16.31.15

To this, I added just a few obvious effects, as placeholders.

Screenshot 2016-02-06 16.40.52


Lastly, I begin to fill in some of the most obvious milestone events.

Screenshot 2016-02-06 16.42.52

When all these nodes in the concept map are expanded, it looks like this:

Screenshot 2016-02-06 16.45.39

I’m still working out how best to use this concept mapping platform, and welcome suggestions in that regard. But most of all, I’d like to draw on the varied, collective expertise of readers to help fill in facts, refine conceptual distinctions, and suggest sources and specific avenues of research.




  • I think this is a really interesting idea. Two (or maybe three) things occur to me. First is the Moneyball connection, but specifically the idea that Billy Beane, et al. took the approach they did as a way to be successful under fiscal constraints. I don’t know that basketball teams have relied on analytics for the same reasons, in part because the league salary and spending structures are different. I think this is an important part of the story because “analytics users” seemingly have gone from identifying “market inefficiencies,” i.e. guys they can get for cheap because other teams are overpaying for players with different skills, to trying to take existing players and mold them to fit specific roles – in Washington, for example, there was some hand-wringing over the fact that Otto Porter wasn’t fitting the 3-and-D type, even though he was the number 3 pick in the draft and so wouldn’t fit the market inefficiency paradigm.

    This relates to my second thought, that we’ve seen a kind of (organizational) isomorphism w/r/t hiring front office people and coaches. I don’t think that’s radically different from the past in the abstract – the idea of a coaching tree has been around for a long time – but I’d be interested to see more data on which teams have tried to adopt an analytics-heavy approach. The teams that come to mind (Houston, Philadelphia, Dallas, maybe Atlanta and Boston) aren’t exactly the kind of small markets that (though I guess this also mirrors baseball, where the Dodgers and Red Sox went for “numbers guys” even though they’re in huge markets and can spend a ton of money on whomever they want)

    Third, I wonder to what extent the rise of a particular blogging model contributes to this analytics-positive atmosphere. By that I mean, SB Nation and ESPN very quickly ramped up efforts to build entire “networks” of blogs that ensured that every team was covered. Particularly early on but even continuing today, many of the bloggers they hired work for either no money or very little money; they were either college students or people for whom blogging was a second (or third) job. Over time these blogs have become legitimized and incorporated into the larger NBA media, but I don’t think their economic model has changed much. Using analytics (usually statistics developed by others) to help tell a story about a game or a player not only costs less but is less time consuming for someone who spends most of their day in class or working another job and can’t always attend games or spend time on the phone trying to connect with sources. And, of course, there has been increased demand for a greater number of stories each day, and analytics can facilitate quick-hit blog posts about why a team is better off with player X in the rotation (for example).

  • Pingback: On Steph Curry: A Reply And A Clarification | Between The Lines

Leave a Reply

Your email address will not be published. Required fields are marked *