This is the fourth year of a dedicated Soccer Analytics panel at the SSAC, and the fifth time that a panel has discussed soccer at the conference (in 2010, it was lumped with American football in an Emerging Analytics session). The session has experienced a variety of participants since 2011, but with two exceptions — Bruno Aziza in 2011 and Chris Anderson last year — the participants have been representatives of clubs, media organizations, or data suppliers, and all of them have represented organizations in the USA or Great Britain.
It is difficult to present a complete picture of analytics in any sport in a 60-minute panel discussion, but the Soccer Analytics panel in past years has struggled to do so more than most sport sessions. If the panel at the SSAC is your only view into the world of soccer analytics, as venture capitalist and soccer fan Mitch Lasky argues in a bracing essay, you would think that the main actors in the field are struggling to make any kind of progress toward making sense of the spatiotemporal data available, that those on the outside with the skills to make contributions can’t because of lack of access to proprietary data, and that there is an overall lack of sophistication in the approach to data in the world’s most popular sport. But there has been tremendous progress in some technical problems in soccer analytics, as indicated by the OptaPro Analytics Forum and the work presented at StatsBomb, OptaPro, and, well, here. So why has the knowledge transfer from the analytics research community to the clubs been so difficult?
This year’s panel contains representatives from clubs, the media, and the data companies, but no independent blogger this time. Steven Houston, a technical scout formerly of Chelsea and Hamburg, is a second-time participant and not independent of any club, but as far as I know he’s not a football analytics blogger. (Point me to your site if I’m wrong, Steve!) Robbie Mustoe is a studio analyst for NBC’s English Premier League coverage, Dr Paul Nielson is senior business manager at Prozone Sports (who I believe has taken over Blake Wooster’s role), and Jim Pallotta is president and chairman of AS Roma. So even though there is a representative of a football club based on the European Continent, all of the panelists are either American or British. Moderating the panel is Taylor Twellman, former New England Revolution star, current match analyst for ESPN, and an excellent choice to lead the discussion.
I’m not sure what to expect from this year’s conference. The objective of the panel is to “go deeper into the strategies and analytical tools used on the pitch, the changing room, and the manager’s office” but how deep will the president of Roma go with his statements? I hope Steven Houston, as someone familiar with the approaches to analytics in two major European leagues, can share his perspective on the analyses that are used to support decisionmaking. Hopefully he can mention the problems with transferring the research presented at venues like OptaPro Analytics Forum to the technical staffs at the clubs. How much detail from match data will Paul Nielsen give to support his statements, knowing that there are some data, such as fitness data, that are considered confidential? And what kind of analytics, if any, does Robbie Mustoe look at as preparation for broadcast, and how different is it from what is shown on TV? I think that Taylor Twellman will have a challenge in posing questions that encourage substantive answers and force panelists to provide them.
The Soccer Analytics panel will be Friday at 1pm EST / 6pm GMT. See you then.