The Americanization of Statistics By Premier League Clubs
Was Thierry Henry lazy? Is Florent Malouda the most valuable player in the English Premier League? Are NBA analytical models being used in the EPL?
All were questions and discussion points during the “Soccer Analytics” panel held Saturday at the 5th annual MIT Sloan Sports Analytics Conference in Boston. The conference, which has grown rapidly over the last five years, drew more than 1,500 sports analysts, executives, managers, coaches, agents and owners over the weekend from more than 47 sports franchises across the world. Baseball, which was has been a central topic of the conference since its inauguration, along with basketball, were major topics once again, but international football was also a hot topic.
While European traditionalists would be quick to mock the Yankee-created panel title, the participants and discussion points gave it a very British tea and crumpets feel. With crucial, late campaign matches taking place across the Premier League over the weekend, several executives from top EPL clubs took time to cross the Atlantic to discuss analytics. The global game of football, unlike baseball and basketball in America, has yet to reach its analytical tipping point. However, it is ripe for change. With a growing American influence on the business of football across England, the revolution is coming. Steven Houston of Chelsea and Gavin Fleig of Manchester City, who participated on the panel, were hard pressed to argue with that during the day’s discussion. You’d also find them hard pressed to argue that working with American analysts within the National Basketball Association (NBA) and Major League Baseball (MLB) has been a good thing for their respective clubs and the overall future of the Premier League.
“We like working with American sports franchises,” noted Houston. “One benefit, we’re not competing with them. Secondly, they do so much analysis.” It shouldn’t be surprising to hear Houston say that. The head of technical scouting and data analysis for Chelsea has a deep understanding of statistics in American sports. He cut his teeth within the NBA for the Houston Rockets as an analyst – applying data analytics to international basketball prospects. At which time, he worked for Daryl Morey, current General Manager of the Houston Rockets, Co-Chair of the Sloan Sports Analytics Conference and the focus of a New York Times feature by the author of Moneyball on his use of analytics in basketball.
Following that impressive apprenticeship, Houston moved to Chelsea in 2009. He now works closely with senior management, including Michael Emenalo and Carlo Ancelotti, on data modeling and visualization, statistical and video analysis, and developing technologies. However, it hasn’t been a necessarily easy transition. Although he joined the English Premier League several years after the introduction of statistical analysis at clubs like Bolton, he didn’t have the data he worked with in the NBA. With the Houston Rockets he had pre-established values associated with literally hundred of points, passes and rebounds. Chelsea, like many other English Premier League clubs, didn’t have league available data and had far less scoring to attribute events to than the NBA or MLB. To solve the problem, he along with other analysts at Tottenham and Fulham – that were also in attendance at Sloan – have worked within their respective clubs to set values for connecting a pass, intercepting a pass, completing a tackle, winning a header and much more.
They’ve certainly made strides since then. Fleig, who worked with Sam Allardyce from 2004 – 2007, has made strides as well. A strong proponent of Allardyce’s use of analytics, he followed him to Newcastle and eventually made his own way to Manchester City. In fact, he credits Allardyce’s early introduction of analytics at Bolton for its long run in the top division of English football. During his time with the Wanderers, Fleig was part of a financially-driven, multi-year effort to develop a model of where and when on the pitch games were won. Specifically he analyzed the differences in statistics between clubs that got relegated versus those that escaped relegation. The approach at the time assisted Bolton in acquiring undervalued assets – like 34 year old Gary Speed. At that age, Speed appeared to be an untouchable player at his requested price, but Bolton had statistical evidence that illustrated his play was not on the downturn.
While Fleig acknowledges that the nature of football doesn’t permit the same statistical analysis of American sports, he believes there is a definite use for the analytical approach in the EPL. “It’s true, but the idea is to understand the characteristics of the team and develop a plan to make optimal use of it – driven by analytics.” With Fleig, Houston and other analysts setting the precedent for video and data analysis, the rest of the Premier League has followed suit. Today, all Premier League clubs have cameras to track match data and roughly 95 percent of Championship clubs do as well. At Manchester City, Fleig has a team of seven analysts with him working with the first team all the way through to the u-9 boy’s squad. Without Fleig and Allardyce, Bolton also continues to develop their data modeling and other top of the table clubs like Liverpool and Arsenal are catching up behind the leadership of analytical minds like John Henry and Arsene Wenger.
So, have the elite sports franchises in England caught up to the elite sports franchises in America? Not yet. Houston notes that there still is an education process required to sell analytical approaches to front-office people and scouting heads within Premier League clubs that isn’t as widespread in America. They both also cautioned analytics advocates, echoing the idea that data is just one piece of the managing process. Owners and managers still need to apply their football knowledge to decisions based on what they see on the field, with the assistance of analytical data. In addition, all the panelists stressed the need to improve their systems in order to adequately compare players across global professional leagues that are facing various levels of competition. The need will grow even more important if UEFA financial regulations are imposed in the future and clubs with seemingly endless funds are forced to finally uncover undervalued assets like more fiscally responsible clubs are today.
“You don’t need analytics to know that Messi and Rooney are great players,” chimed Houston. “Analytical systems are useful to find the best role players, or the best players for a particular team.”