Lucy Rushton Q&A: The role of analytics in building Atlanta United FC
With the 2017 MLS season set to start in less than three weeks, Five Aside caught up with Lucy Rushton, Atlanta United FC's head of technical recruitment and performance analysis. Rushton is a key player in the development and formation of the Atlanta side that will debut in the league on March 5. With the unique challenge of assembling a roster from scratch, statistical analysis plays a determinant role in the club's first ever season. Here is our full conversation.
How do you go about adapting your statistical work to the coach's playing philosophy?
LR: I have to be flexible and tweak the model to the needs and demands of the manager in charge. Adapting my work to the needs of the manager is the key of analysis. When I build and design statistical models, I always have that in mind. Due to the changing culture of managers, I need to be able to make alterations to the system and adapt it to the philosophy of the person in charge.
Having announced Gerardo Martino as the manager, how has Atlanta gone forward integrating statistical analysis and building a roster to a new manager's philosophy?
LR: Here in Atlanta, in the process of building a club from scratch, we put a lot of emphasis on club culture. Of course, we will have a playing style that is built around the manager's philosophies, but equally as important to us is that we adopt philosophies and principles which persist and become associated with the club for years to come. It's a collaborative process which includes many personnel: the club president, the technical director, the manager. Having established these criteria, I am then able to tailor my analysis to meet both the needs of the club and the manager. For example, if when looking at full-backs I know that we want to prioritize contribution to attack, then I may place extra emphasis on the player's action areas and his cross data when completing my analysis. This syncing of analysis to the club/manager's philosophy happens in many ways.
What is the main challenge in assembling a roster from scratch? Also, how do you navigate effectively through unique challenges in the MLS landscape such as draft picks, expansion drafts, discovery claims, allocation money, etc.
LR: This is my first time in this situation. It's a major challenge, especially in MLS: It's different to all other leagues in the world. There are so many factors that are challenging with developing and assembling a roster. In any other league, you usually have a list of targets, and then whether you sign any particular player or not is based on maybe three or four variables -- do you want them, does the parent club want to sell, does the player want to come to you, and can you afford them. In MLS, the number of variables which can affect a player signing is much vaster. The whole process is much more complicated.
I think at this point, the most challenging aspect is finding the first few pieces of the jigsaw puzzle, because any one signing can ultimately affect the players you can sign in the future. In addition to choosing players based on their immediate contribution to the team, you have to also think in strategic terms of how that may impact the roster later. If we sign an international player now, how will that affect future draft selections, for example. It is essential that we are both proactive and reactive throughout the recruitment process and that we have the flexibility to change and know all of the available options at hand. We are fortunate here at Atlanta United because we have an excellent group of staff who are well prepared to assess the ever-changing MLS landscape.
Does the nature of MLS, make you more reliant on analysis compared to other leagues?
LR: To some extent, yes. A good example of why this is the case is the salary cap. It forces you to think a little more outside the box and determine where you may gain a competitive advantage. In terms of analysis, we try to see how to best analyze players to make sure we are using the salary cap efficiently. So naturally, analysis is more useful. Given that salaries are publicized, it gives us the potential to measure how efficient we are being with our player selection.
Is there a particular strategy that Atlanta United is fond of?
LR: We look at everything and consider everyone. For example, in addition to paying attention to international players playing abroad, we also scout U.S. college players and other MLS rosters extensively. Our search covers numerous areas, as we have to stay within the league's regulations and keep in mind things like potential trades and possible SuperDraft selections. As a new team in the league, acquiring players with MLS experience is invaluable, and this will be an important source of player acquisition for us. However, there are numerous other means of acquisition which will be of equal importance.
Having lived on both sides of the Atlantic Ocean, what differences do you see between English and American soccer fans in terms of how they use and perceive statistics? Is one group more statistically inclined than the other?
LR: In the United States, the culture of analysis is well established, especially in other sports, but I think football [soccer] is now just starting to follow that path. One thing I can say for sure is that as analysis in football [soccer] keeps developing, the statistics and elements used will be very different to most sports. Soccer is such a fluid game, and it has infinite patterns, so that's a substantial difference. In regards to supporters, fans do see statistics and data in the media, but it usually tends to be a little bit raw. On the other hand, as an analyst, you are more inclined to look in much greater depth. For example, if a center-back has a high pass completion rate then we, as analysts, ask why? Are his passes less difficult?
How do you get your information and analysis into the hands of the coaches? The players? Any differences there between the English and Americans?
LR: When it comes to sharing information, it varies from coach to coach and from player to player. Every manager I worked with has a different style, so I adapt my delivery methods to their preference. Players are exactly the same; some players like the statistics, while some prefer video analysis, and others a combination of both.
Since you started your career, how has the statistical analysis of the game developed within soccer clubs?
LR: The field of statistical analysis has developed massively. Eight seasons ago when I started, the only provider was Prozone, pretty much. Now, there are numerous products, like Opta, Ortec and others. Not only has the market place grown, but also the process of analysis itself. In the past, analysis was more basic: How many shots? How many passes? Now, we have elements like expected goals where we look at data and try to derive meaning from the numbers. We ask questions such as how do we view the numbers differently to measure efficiency in strikers? Another example is sequencing; rather than assessing individual events, you look at collective actions to see how players interact with teammates on the pitch. We aren't looking so much at events in isolation anymore. Instead, we are looking at the whole and wider picture. We try to identify a player's playing patterns, not one-off movements. Additionally, now there is much more emphasis on developing players and predicting future performance than maybe there was when analysis first entered soccer.