Photo credit: ISI Photos / San Jose Earthquakes
Last year I wrote about six things the Earthquakes needed from their new general manager, whoever he or she would be. The final item in that list was that the individual needed to take data analysis seriously.
Unbeknownst to me, there was an individual in the San Jose Earthquakes sporting side who has been thinking about all the things that “he didn’t know that he didn’t know” about the sport after he took his Elite Formation Coaching License course from the French Football Federation via an MLS partnership created back in 2013. That person was Chris Leitch: “I wanted a way to ‘pressure check’ my subjective mind with something more objective,” Leitch reflected.
When Leitch interviewed for the club’s general manager position, the creation of a Soccer Analytics department was a cornerstone of his presentation, “In an industry where the margins are really thin, even the slimmest of marginal gain that can be accomplished is really important,” Leitch told me. As the front office leadership discussed the creation of the team, there was a universal agreement that even a small competitive advantage gained would be worth the investment for the San Jose Earthquakes.
Meet the Team
Unlike the vast majority of MLS clubs that have already made a move in this area prior to the 2022 season, the Earthquakes didn’t go out and get just one director to take on all the responsibility for competitive analysis, tactical analysis, match reports, second team and academy reports, video and data sync, performance data integration, backend data store creation, prospective player identification, and the dozens of other things that one individual is expected to do in this era of the league.
Instead, Chris Leitch and Vice President of Strategy, Ian Anderson, built possibly the biggest data analytics department in Major League Soccer with three full-time hires. These aren’t “computer science college freshers” without real-world experience, either. This team is expected to be among the most experienced teams in the league, with all three hires working in the professional soccer ranks since 2017. For a very young industry, 15 years of combined experience in soccer data analytics is a lot.
Here’s a bit more about the three team members in the Earthquakes Soccer Analytics Team:
Lucy Rowland
Lucy was the lead analyst for the Canadian Men’s National Team up until the start of the MLS season. Yes, that Canadian Men’s National Team. Yes, the one that beat both Mexico and the USMNT convincingly at the “Iceteca” and won the Concacaf group in World Cup Qualifying.
In her time in Canada Soccer, she created the first-ever data department and brought in innovative methods for match analysis, player analysis, and opposition profiling for the men’s side. Doing so gave her extensive experience working side-by-side with the coaches and players during the course of the 2022 Concacaf World Cup Qualifying campaign.
Lucy is proficient with languages such as Python and with many key technologies like machine learning, storage solutions, and BI solutions. She got her start in analytics as a video analyst for the Canadian Women’s National Team from 2017-2019. She also had a stint with Electronic Arts as a data analyst on the FIFA team.
Lucy has a Bachelor of Computing degree from Queen’s University in Kingston, Ontario.
Grant Wenzinger
Grant brings crucial experience directly from his time with Second Spectrum in the Artificial Intelligence team.
Second Spectrum is one of two providers of match data in Major League Soccer. Stats Perform (Opta) is used for “event data” by the league. Event data is the action being performed with the ball at any time and is manually logged by match analysts working from television angles, producing a few thousand data points. Instead, Second Spectrum uses special camera technology to track all 22 players on the pitch and the ball, generating millions of data points every match. This is called “tracking data.”
In his time at Second Spectrum, Grant developed novel techniques for quantifying and visualizing team shape, team field control, and player acceleration using their tracking data. He trained machine learning and artificial intelligence models and built data frameworks for querying information quickly.
Grant has a Bachelor of Science in Computer Science from Cornell University and has a US Soccer National “D” License as well as youth coaching experience.
Aditya Nag
Aditya brings key MLS experience to the team having worked recently as a data and performance analyst for the Philadelphia Union for multiple seasons and previously with FC Pune City in the Indian Super League.
While at the Union, Aditya used many methods to help with team analysis, player and positional fit modeling, opposition scouting reports, and player acquisition models. Prior to the Major League Soccer deal with Second Spectrum, he used tracking data provided by ChyronHego to create various reports for the technical staff.
Aditya also brings in critical video analysis skills to the team from his experience at the Union and FC Pune City.
Aditya has a Bachelor of Science in Economics & Mathematics from UC San Diego and is working towards his US Soccer National “D” License.
Building the Runway
Despite the experience of this team, let Chris and I both slow your expectations. This is going to take some time. I can tell you from my own experience with American Soccer Analysis where we have built metrics like Goals Added (g+) and, more recently, the Where Goals Come From Framework, that it takes thousands of hours to build something that can provide a distinct advantage against everything else out there. It’s very normal to head down a particular path for a couple of months that ends up being discarded when it doesn’t work or provides less value than existing machine learning models and methods.
“This is a team of really smart people, but to go from a team of zero to a team of three — and getting different answers than what you already had a month or two ago — it doesn’t just happen overnight,” Leitch said. “There needs to be some runway for those capabilities to be built.”
That doesn’t mean it will take years to see fruit for the Earthquakes. The club was already using partnerships and various data sources they had access to in evaluating the 2021 roster and prospective players this past offseason. The Earthquakes were the first team in the league to partner with Second Spectrum and have done some work with it in the past four years through consultants and their work with the German Football Association. Now that groundwork will begin to pick up speed with a dedicated team focused on event data, tracking data, and performance data. Over time, the part that data will play in decision-making will increase, particularly as the specific competitive advantages the club identifies from its primary research are built into various machine learning and artificial intelligence models.
From Playing Catch-up to Gaining an Advantage
And let’s be honest: the Quakes are late to the game to make full-time hires compared to most clubs in the league and have a bit of catching up to do, but there is an opportunity for the Earthquakes in hiring someone with direct knowledge of Second Spectrum’s data structures. Most clubs in MLS and in Europe struggle with getting an advantage out of Second Spectrum and other tracking data sources because the sheer volume and complexity of the data is to the point that it would take a good-sized department in a corporate IT organization to mine for insights in that world. Multiply millions of data points each match times the number of matches over several seasons, and you effectively have a “big data” problem for clubs to solve that they aren’t really equipped to solve. It’s a growing concern within the league, and many clubs throw up their hands at the enormity of the task.
The temptation is to think that the association of the Earthquakes with the Oakland A’s and majority owner John Fisher means the point of implementing data analytics is to save even more money and try to “squeeze blood from a stone”, but multiple discussions with several front office members have emphasized this is anything but true. They point to clubs that have been the most successful with data analytics, such as the Seattle Sounders and Toronto FC, who won multiple championships with good-sized budgets driven by analytics-backed decision-making. I wrote in the aforementioned “six things” article that true “Moneyball” is based on running a club by key performance indicators and not cheaper spending, and I feel good about the direction I’m hearing from the front office even if the full vision is early in its implementation. For an understanding on how “smart spending” in MLS works, see my preseason article on this topic.
One area fans may not think about as much is how these analytical capabilities can be used within the new Quakes II team and the Quakes Academy. They can be used to identify how to correct a player’s development or for a Quakes Academy team to be more effective in adopting the system from Director of Methodology and now interim Earthquakes head coach, Alex Covelo. Leitch identified data and analytical methods can play a very important role in player development and in the decisions around promoting a player to the next level and provide more certainty when projecting success.
Like the new Quakes general manager discovered in the French Football Federation course: the more you learn, the more you realize how much more there is left to learn. “This is iterative. We are going to constantly be searching for marginal gains and competitive advantage. I don’t know that we’re ever going to find the ‘silver bullet’ that’s going to give us the key to all the answers.”