Lucy Rushton, Atlanta United’s Head of Technical Recruitment and Analysis, discusses how the MLS expansion franchise is utilising statistical analysis to build its roster from scratch. She talks risk management, the Atlanta recruitment process and being both subjective and objective when analysing.
Recruiting an entire team of players is a difficult process in any league and at any level. Doing so in Major League Soccer (MLS) is particularly challenging because the league itself is unique within the soccer landscape. In addition to operating within a salary cap environment, there are added complexities including the drafts (Expansion and SuperDraft), discovery lists, allocation monies and regulations that impact the make-up of the roster. Nationality and player status (i.e. ‘designated players’ or ‘home-grown‘) are also factors specific to this league.
At Atlanta United, we aim to be at the forefront of modern recruiting not only in MLS but in the global game by utilizing the latest statistical analysis methods in player identification. For most clubs, the primary considerations throughout the recruitment process are ‘can we afford the player’? and ‘is this player right for us’? Our approach is to take the process a step further and dive into the careful consideration of a player’s ‘fit’ within the league regulations. This process is almost an analysis of its own, and one which takes an immense level of planning, discussion and deliberation.
This level of specialized recruitment cannot be achieved through analytics alone, and requires considerable contributions from the technical staff. Atlanta United has a close knit team of front office experts each of whom have high levels of expertise in specific areas of the game. I am in constant communication with our club President Darren Eales, our Director of Soccer Operations, Paul McDonough, and our Technical Director, Carlos Bocanegra, throughout the identification process to ensure that all angles have been covered when analysing a target.
Ultimately, the signing of any one player can have notable implications on the ‘fit’ of other potential targets. Each decision ultimately impacts the next. For example, a previous top target can become unviable in order for us to comply with league regulations such as the ‘senior roster’ (roster positions 1-20) rule. Or, perhaps we have exceeded our allocation of international players, or an acquisition is no longer viable within the salary cap. All of these factors have to be prudently assessed when deciding to move towards acquisition. It is a careful process of fitting together all of the pieces of the puzzle. As an expansion team, we are starting with a blank canvas, so the process is even more complex when choosing which pieces to place and in what order of significance.
Whilst at times challenging, the confines and complexities such as the cap and the drafts ultimately makes statistical analysis during player recruitment in MLS even more critical compared to other domestic leagues. How can we maximise our potential for success through a roster make-up which also satisfies league requirements? What has been effective for past teams? What strategies and approaches, if any, have been most strongly correlated to the elusive play-off finish? How can we make better use of our allocation monies? For our club, in the unusual position of recruiting from scratch, what is it about previous expansion club rosters that have aided or impeded success in their inaugural campaign? What can we learn from their experiences and outcomes? These are all questions which have been asked in the early stages of roster formation, and will continue to provide direction in our decision-making.
The release of information pertaining to player salaries is another of the league’s unique practices, and one which, as an analyst, I welcome with open arms! When assessing the suitability of potential targets, such data enables player ‘efficiency’ to be measured in terms of monetary value and return, with more valid comparisons being drawn between players of different quality and/or value as a result. This information has undoubtedly been of use to clubs when considering players with a playing history in MLS.
“It is a careful process of fitting together all of the pieces of the puzzle. As an expansion team, we are starting with a blank canvas, so the process is even more complex when choosing which pieces to place and in what order of significance.”
As of 2010, analysts have been able to take advantage of a full MLS Opta data set which has been another welcome tool especially as we prepare for the expansion draft. Opta provide high definition technical data on every match event in a game of soccer. We are very fortunate to be working alongside professors at Atlanta’s Emory University, where we have been able to engage in projects on player profiling, player ‘value’ and future performance-forecasting all of which will help add further weight to our decisions when selecting players in the draft. Undoubtedly the scope of such research stretches far beyond the current recruitment phase and expansion draft to upcoming transfer windows and other player acquisitions.
Such analysis is both subjective and objective in nature (as is all of our analysis), with indepth Opta data being utilised alongside the more traditional video and scouting processes. Far from being used only at the ‘tail end’ to supplement visual/subjective perceptions, statistical analysis has also been used in the earlier stages of player identification. Pinpointing players who best fit our profile statistically helps us close the net slightly on what is an otherwise overwhelming market. This analysis is also used to guide our scouting to areas and players which may have otherwise gone unnoticed.
With up to eight international roster spots to fill, our market is very much open outside of the US, and as a result, we are consistently performing extensive analysis of foreign leagues. Since salary information is unavailable outside the US, for the most part, the analysis of players from these leagues remains more restricted in terms of performance related data and information.
As with all clubs around the world, we too face the key question, ‘How does/will performance in one league translate to another?’ when scouting abroad. This is a question we must factor into our analysis and decision-making at all times. Statistics can only go so far in identifying trends and/or strengths of individual leagues, but at present is still elusive in terms of its ability to portray more direct league-to-league links. This is where we must draw on the expertise and experience of our core scouting team in order to bridge the gap.
To talk only about our recruitment processes at the ‘higher’ (i.e. professional) level would be inaccurate and unjust, for a notable amount of our work will also focus on ‘lower’ (i.e. amateur and minor league) levels. Scouting and analysis of college soccer, for example, is imperative to ensure that we are fully prepared for the upcoming MLS combine and SuperDraft in January 2017, where we could draw a percentage of our final roster. Constant review of our affiliated USL team, Charleston Battery, where we have players out on loan, and of our Academy teams will also be critical as it is from these places that our vital ‘supplemental’ and ‘reserve’ rosters (roster positions 21-28) will be formed. In addition to adding obvious strength in depth to the senior roster, these players may also have ‘home-grown’ status; a status of added value to the roster/club. Among the many beneficial implications of signing a homegrown player, the player’s salary does not count toward the club’s salary cap.
“As with all clubs around the world, we too face the key question, ‘How does/will performance in one league translate to another?’ when scouting abroad. This is a question we must factor into our analysis and decision-making at all times. Statistics can only go so far in identifying trends and/or strengths of individual leagues, but at present is still elusive in terms of its ability to portray more direct league-to-league links.”
Tying all of these factors together, in order to maximise our recruitment outcomes at Atlanta United, we apply a multi-dimensional approach combining ‘subjective’ and ‘objective’ analysis. This approach is achieved primarily through the implementation of a vigorous five-stage recruitment process which is applied at all levels (from the senior to the supplemental roster). When a player reaches the end of the process, which ultimately results in their acquisition, we, as a team, are fully ‘educated’ on him and enter into recruitment with confidence that he will be a success.
The key in analytics is to realise the potential of analysis to assist in risk management during player recruitment. Through the careful combination of statistical analysis alongside subjective opinion, our goal is to reduce the risk of making an unsuitable signing. We can never make the decision risk-free, but we can certainly help to minimise it. With that one important part of the jigsaw puzzle confidently in place, we can comfortably turn our attention to the next player.