Will soccer ever embrace the MIT Sloan Sports Analytics Conference?
Two weeks ago I had the pleasure of attending my fourth consecutive MIT Sloan Sports Analytics Conference. This year I experienced the conference from a slightly different perspective though. After three earlier failed attempts, I finally managed to get a paper to the conference phase of the research-papers competition and secure a poster presentation at the conference together with my SciSports colleague Lotte Bransen and KU Leuven collaborators Pieter Robberechts and Jesse Davis. Nevertheless, I arrived home with mixed feelings for the first time since the 2016 edition although I had enjoyed every second of my stay in Boston. Rather than full of energy and new ideas as I had gotten used to after the three earlier editions, I was left wondering what I had learned from this year’s edition. More than ever before, I was stuck with the feeling that the soccer analytics community had in fact made only little progress since last year’s conference.
The soccer content this year’s conference had to offer actually felt somewhat underwhelming. Prior to the conference, my hope was to gain a better understanding of the questions practitioners at soccer clubs aim to answer as well as how and where my team at SciSports could possibly provide value. Unfortunately, the soccer analytics panel did not provide many useful insights. The panel discussed the adoption of analytics tools in youth development, which is one of the least developed areas of soccer analytics, as Ted Knutson already pointed out in a recent podcast. More interesting was the panel featuring StatsBomb CEO Ted Knutson and Houston Rockets General Manager Daryl Morey in which the latter questioned some of the fundamental assumptions in soccer analytics. Unsurprisingly, the takeaway message from this panel was that the collection of soccer data is still in its infancy and needs to be lifted to a higher level to push the whole field forward.
A vicious circle
Soccer appears to struggle to really gain traction at the MIT Sloan Sports Analytics Conference. Moreover, from my perspective, soccer even seems in the middle of a vicious circle. Since the conference has little to offer in terms of soccer content, the soccer people stay away from the conference. Likewise, since the soccer people stay away from the conference, the organizers see little reason to add more soccer content. A quick glance at the attendee list learns that actually only a handful of European soccer clubs found their way to Boston. This lack of interest from soccer clubs was also apparent at the Soccer Analytics Mixer on the night before the conference started, attracting representatives of just a few organizations. Each year again, this informal event provides an ideal opportunity to catch up with like-minded conference attendees. It is for me almost reason enough on its own to travel to Boston. Like the share of soccer people at the conference, the Soccer Analytics Mixer has sadly experienced little growth in attendance over the past few years.
In contrast to the soccer content at the main conference, the soccer papers in the research-papers competition seem somewhat ahead of the curve though. Moreover, this year’s best-paper award was won by Javier Fernández who presented an expected possession value framework for soccer inspired by earlier work for basketball. With Derrick Yam’s work on evaluating goalkeepers and our own work on quantifying players’ abilities to perform under mental pressure, two more soccer papers made it to the final phase of the competition. With also more high-quality hockey papers being submitted to the competition, the other sports track is becoming increasingly competitive. This trend will hopefully inspire next year’s conference organizers to revamp the research-papers competition and grant both soccer and hockey research a more prominent role at the conference.