MIT’s Sports Analytics Conference: How to make it better

I've already written a trio of review pieces on the Sloan Sports Analytics Conference earlier this month, but I have not written anything on my own impressions of the event and the areas in which the conference could improve.  I just answered an evaluation form that the conference sent me a few days ago, and the effort gave me a chance to organize my thoughts.

This was my first SSAC — I had found out about it in the late summer of 2009, and I immediately thought that it was a conference that I needed to attend.  In that respect the conference did not disappoint.  This year the conference reached critical mass with an influx of attendees from academia, the sports world, and private industry (whether consulting firms, think tanks, or individual entrepreneurs).  To me the most valuable part of the conference was the networking and exchange of ideas with people in the same field and in related fields as well.  It was great to connect a face to the names of people with whom I have corresponded over previous months.  I learned a great deal about the state of analytics in several sports that could have some useful applications to soccer.  

It seemed to me that the SSAC was genuinely surprised at how big the conference had become in the space of a year.  The conference had been staged at the Sloan School of Business at MIT, but had become so oversubscribed that it was moved to the Massachusetts Convention and Exhibition Center.  I don't know when the decision was made to move the conference, but the fact that they had to move venues during the preparation for the meeting indicates to me that they had to scramble to satisfy the increased demand for the conference.  Now, convention halls are not easy to reserve on short notice, and they're also very expensive to reserve as well, which might explain why the conference only occupied one long section of the convention center.  There was a section reserved for exhibitors and vendors, but it seemed very ad-hoc in its organization.  I was chatting with one of the vendor representatives afterward, and she felt that the conference could have easily doubled the number of exhibitors if it had planned for it better.

The conference has done quite well with its format of keynote speakers and panel discussions, which have provided a high-level view of the state of sports analytics.  The panel discussions were 80 minutes long with maybe 5-10 minutes of questions from the audience (although I attended sessions where the session chair didn't take any questions).  This year there was a conference paper track, limited to eight papers (the contest in the morning, and invited papers in the afternoon).  I would like to see that track expanded, and perhaps with multiple tracks in different areas of sports analytics.  And this gets into my second suggestion: this conference should be expanded to two days.  I think the conference has gotten to the point where there are more than enough topics for a two-day event.  The downside is that the conference fee would almost certainly go up and the planning could become overwhelming for the planning committee, which, lest we forget, is made of up of business school students who are trying to complete a one- or two-year MBA program.

The other thing I noticed with the conference was that it was almost entirely composed of British and American panelists and attendees.  That is not too surprising; the Americans and to some extent the British has shown the greatest openness to the use of statistical methods and other metrics in sport, and there remains a high level of resistance to these analytical approaches in sport in the rest of the world.  That said, there are some people outside of the Anglosphere who are doing interesting work in sport analytics, and the SSAC should try to bring them to the conference.  One person who I would like to see is Bernard Lacombe, who is technical director at Olympique Lyon and was highlighted in Stefan Szymanski and Simon Kuper's Soccernomics.  I don't know who is the director of the Milan Lab, but they should also be invited to future conferences; after all, they have research connections with MIT!  The representatives from the Australian Institute of Sport would have some interesting insights in the use of analytics in a wide range of sports.

So those are my suggestions for improvement: larger meeting areas for the panel sessions, more exhibitors, two-day conference, and more panelists outside of USA/UK.