The Currency of Numbers in Football: The Role of Data in the Evolution of Player Recruitment

Since the very dawn of the first professional domestic league competition 125 years ago, it has been widely acknowledged that good player recruitment is a vital element, if not the single most important factor, in a club securing long-term on-field success.

But like the game itself, the business of player recruitment has evolved since that inaugural Football League campaign, throughout the generations and into the present day, as finance, sporting regulations, competition, employment rules and social demographics have impacted on how a scouting operation can operate, as have the parameters to what type of player a club can feasibly recruit.

Now well into the 21st Century, it would appear that this evolution has accelerated rapidly in the current digital age.

If you look into a leading European club today, you will find that they employ a blend of traditional scouts, technical scouts and performance analysts, who have all been embracing a wide variety of new information sources and technologies that are now available at their disposal.

Indeed in a world where detailed match information is available at your fingertips from the latest mobile app, it is hard to believe that as little as 18 years ago, when the Bosman ruling was revolutionising the way the transfer market operated, many clubs were still relying on hand-written reports posted from its scouts and standalone recommendations from agents to manage its operation.

In contrast, in today’s world a professional club will look at every source of information that is available to them in order to enhance its decision making, which in itself is a process which can stretch back many years to the moment when a player was first initially on the scouting department’s radar, to the moment they were considered ready to be recruited.

Several aspects of this evolution have been documented in a new book which was published earlier this summer, titled ‘The Numbers Game’, authored by Chris Anderson and David Sally.

David Sally (left) and Chris Anderson, co-authors of 'The Numbers Game'

Anderson is a former semi-professional player and now a professor at Cornell University in Ithaca, New York and is widely considered to be a pioneer in sports analytics, whereas Sally is a former baseball pitcher who now works as a behavioural economist and consultant, as well as being a visiting professor at the Tuck School of Business at Dartmouth.

Whilst writing the book they visited professional clubs around a world to attain a first-hand insight into how elite clubs operate with its match preparation, analysis and recruitment. Subsequently they have put together a series of arguments, which amongst many things consider the importance of luck, the value of a prolific striker and the challenges of how the performance of a team’s defence can be valued to measure defensive performance.

More significantly, despite the book’s title hinting at a focus on quantitative soccer analytics, ‘The Numbers Game’ has been written in the context of football as a sport, and an unpredictable one at that. It is not a book about revolution, but one that is more about natural evolution and how practices can be implemented to improve decisions to support a Manager and a Club. It is also very pragmatic about the impact of money, or indeed the lack of it, and the correlation between resources and success.

It also exposes a number of myths – notably how data-driven analysis is not a new phenomenon to hit the professional game. Instead, the difference in how this data is used today lies in how a club understands and applies the numbers as part of its wider analysis.

Following its publication, Anderson has spoken to Scout7 about how clubs can potentially structure its recruitment to embrace new approaches and skills that can further enhance the recruitment process, including the application of analytics to assist a club in making key decisions when signing a player.

In the first instance, he believes that if a recruitment department is going to embrace analytics they need to fully understand the value it brings to the process, and then utilise it in the correct way.

“As the industry has evolved I think the job description of the Head of Recruitment has evolved, and I think that is true of Managers as well,” says Anderson.

“With new tools and with new technologies coming online, Heads of Recruitment need to find a way to take them on board. That does not mean that they have to become analysts or become technologically advanced with a PHD in maths, but I think what it does mean is that they have to skill up a little bit to understand how these new technologies and new types of information can help create value for the club.

“What is important as part of that is not just simply to have an understanding of what those tools are, but to find a way to hire the right kinds of people into a scouting department that have the ability to then make use of those tools in a very effective way.

"There is nothing worse than spending a lot of money on tools that are going to sit around unused and I think that still happens too much, in part because there are established work routines that clubs have and a new toy often sits in the corner because it has not been fully integrated into the decision making process.

“So I think Heads of Recruitment need to understand both the hardware and the software, and by software I mean the human capital, the people involved in using new tools to make then fully functional and effective.”

Since the mid-1990’s, several data companies have been collating various types of information captured from leagues around the world, ranging from physical performance data through to detailed match events.

The availability of this information can aid a club with its analysis of a player, however with such a broad dataset, how easy is it for a club scout or analyst to contextualise events and separate the meaningful data from the ‘noise’?

Again, Anderson believes that the key lies in the recruitment of the right people within the club to evaluate the information intelligently.

“Data analysis requires patience, trial and error and very importantly, it requires hiring the right kinds of people, who are able to employ that information in meaningful ways,” he explains.

“What I mean by that is you need people who have both the knowledge about football as a game and who understand how football clubs work internally, but they also have to have the technical and analytical skills to make sense of the information.

“You want to ask good questions about the data and you can’t ask good questions about the quantitative information unless you understand the game and understand what the club’s needs are and what the club’s limitations or aspirations are.

“So ultimately it has to be driven by one single question – how do we win football games? That has to be the driving analytical question behind all of it. You can split that up into any number of smaller parts – how do we produce goals? How do we not concede them?

"We can split that up but ultimately there has to be a question that underlies any kind of analysis and that question has to be ‘how to win?’ and you need the right people in place to do that.”

Despite all the match information available, one area of the game which is difficult to quantify based solely on data is defensive performance. How can attributes such as positional intelligence be measured in a way that tackles, blocks and interceptions can?

In a game dominated by turnovers, exploiting weaknesses and punishing individual mistakes, being able to identify outstanding defenders has always been a cornerstone in building a great side. So moving forward, how will data be applied to support the assessments made by a scout sitting in the stand?

Anderson believes that in time, the answer will lie in positional grid technology, which will inevitably become available to football clubs at some point in the future.

I think, given the current availability and nature of the data that we are working with, it’s inherently more difficult to value defence and a defender’s contributions to the game accurately or appropriately” he says.

“So what I would say is the more traditional scouting tools at the moment are more valuable on the defensive end of the pitch than on the offensive end of the pitch.

“Now I think as we are moving forward, one of the quantitative tools that will help us a little bit with that is positional data, XY data, because I think the value of good defence or a good defender comes out when we understand things like positioning and coordination among a set of players, among the back four, a goalkeeper or centre backs or whatever we are thinking about.

“Positional data will help us with that but I think we’re far from a point where that can become a standard tool for clubs, I don’t think that will happen for a few more years.”

Another trend that has appeared in recent years has been the ranking of players and teams, based on results and individual player performance. UEFA has its own club coefficient rankings for each season based on past Champions League and Europa League performance, whilst the London-based consultancy company Decision Technology have developed various performance models that power amongst other things the Castrol Player Rankings.

A now-defunct website called Eurofootsie also tried to establish a team ranking model for European teams, based on each club’s previous 50 results and current form.

Whilst the make-up of any ranking model is open to individual opinion as to whether or not they capture the essential performance indicators required to provide an accurate reflection of a club or player in the world order, can models, whether devised internally or available within the public domain, offer a viable tool to assist a club with the focusing of its recruitment?

Whilst Anderson thinks that some degree of benchmarking is important when evaluating key recruitment markets, he feels that clubs should also adopt a degree of caution when placing a value on a ranking mechanism.

“It is a really interesting and challenging analytical question and it has to do with separating out the value and the contribution of a player to the team and the team affects overall,” he explains.

“Strikers will look better on good teams, as they will get more chances and they will have their team mates contribute to their ability to take high-quality shots, so separating out the unique contribution of the player from the team’s strengths or weaknesses is a hard one.

“Think about the Dutch league for instance. It is a league that is relativelyhigh scoring, so strikers almost inevitably look good, but it doesn’t mean that Dutch strikers are particularly good necessarily. So we have to be very careful making sure that what we see isn’t a function of the quality of the league or the nature of the league or the quality of the team that’s someone’s playing for.

“So to cut a long story short, I think any effort you can make to go down the path of benchmarking, measuring and comparing along those lines is a very important thing, whilst still keeping in mind that it’s fraught with potentially faulty conclusions.

“Another good example for me is Dutch defenders. Just because the Dutch league is a high scoring league doesn’t mean that Dutch defenders are all rubbish. So it’s a good first step, but I think it’s also then important to calibrate the league in terms of quality, the nature of the competition and then the team and the style of play that they have.”

A key requisite of any recruitment department is managing the constant turnover of players within the club’s playing squad.

Clubs now have information at their disposal which allows them to identify key recruitment trends and mobility rates over a medium-to-long term period around major leagues in world football, offering context if a club elects to structure its recruitment in cycles or put a greater emphasis on ensuring that they extract the peak performance of an individual player during their time at the club.

Based on his observations, Anderson is in little doubt that intelligent recruitment departments are already addressing these key objectives and that moving forward, we will see players being given big contracts before they hit their peak, rather than at the point where they have reached their ultimate performance.

“I think managing what you would call the life cycle of a player has to be part of any good recruitment department’s remit,” he says.

“I think the trick or the temptation is always to react too strongly to short-term events and recent results and often those well laid plans for the management of player’s life cycle go out the window as a function of a bad run of results or a bad set of performances.

“I think here data can help understand what the true underlying abilities and performance levels are of a player, to then help manage that cycle more effectively. It’s also an important part of helping clubs manage their budgets in a more coherent way over time. If you understand the life cycle of a player, then that allows you to be more coherent in your budget area or your financial model.

“What you also want to understand is at what point is it worth paying a player a lot of money and at what point does it not make sense any longer to pay him a lot money. I think contracts that can be structured to be reflective of expected actual levels of performance are better for clubs. To give you an example from North American sports, quite often players in baseball or basketball become eligible for big contracts when they’re already almost past their prime. They become free agents and can extract a lot of money from the market when their performance has actually already peaked.

“That’s very, very inefficient for clubs. You want to buy them when they haven’t peaked yet and you want to pay them what they’re worth when they’re peaking, not afterwards.”

One key observation made within ‘The Numbers Game’ is the fact that clubs with greater resources benefit from a larger range of tools to assist with player recruitment, both in terms of the number of personnel and the tools available to them, to help with their evaluation and decision-making.

Considering the importance of recruiting the right players in terms in maximising your chances of achieving on-field goals, Anderson believes that it is the clubs working on more modest budgets that can get the most value out of employing a larger pool of recruitment staff and supporting technologies, as arguably more hinges on their decisions as they may not have the luxury of bringing in additional players if a financial investment in signing a player does not work out as a club envisaged.

He explains: “We live in a perverse industry where I think the clubs that have more resources devote more resources to employing more advanced and more modern technologies and sources of information to make decisions.

“In a way I think the better value for money is for those clubs that have fewer resources because there I think the technology offers a way of making the pound stretch a little further.

“I understand why clubs with fewer resources are sometimes loath to commit resources to these technologies because it takes money away from somewhere else, but I think for them the ultimate pay-off of those resources is going to be disproportionate relative to a bigger club with more resources.

“Of course such an investment as a proportion of the club’s budget seems like a bigger deal and so therefore there is an inclination not to spend as much or devote as many man hours to it in resources, but I think the pay-off is also disproportionately high if it is done and implemented correctly.

“At the end of the day it is often about the guys you don’t sign, as much as it is about the guys you do sign and I think what you don’t want to do is sign the wrong player and I think having more information helps you avoid those pitfalls.”

With so much information now available from industry suppliers and public domain sources, questions can be made as to how much further a club can take recruitment analysis in the future.

Anderson believes that club’s will always be striving to secure any small advantage over its rivals, but in returning to his initial observation, he stresses it will be less about tapping into new information sources and more about how existing data can be interrogated more intelligently.

“You don’t necessarily have to try and pick the pockets of your competitors to find out exactly what they’re doing in order to compete,” he points out.

“I think there is enough information out there in the world now that is provided through a number of platforms and data sources that means you cannot create competitive advantage simply by making use of that information.

“I think it is not the access to the information that’s the big differentiator. I think the big differentiator is using that information productively. And there are also good things to learn from other sports so you don’t have to go fishing for good practices necessarily in football.

“Football clubs by their very nature are competitive with one another, football as an industry is a competitive industry, but I think there are analogs and similarities in other kinds of sports that could be useful for other clubs to take a look at. I’m thinking of rugby, I’m think of basketball, I’m thinking of American football, those are the kind of sports where the analysis is happening and analytics is being driven forward and I think there are some good lessons to be learned. It all doesn’t have to come from within football.

“But again, I must stress that any such initiative has to be in support of the clubs existing processes; if it isn’t, then it’s just a luxury you can’t afford. It has to be all in support of the team and it has to be all geared towards winning football games.

“If it’s just a personal hobby of some guys working in the office then that’s not a good enough justification. It has to support what you’re trying to do.”

The Numbers Game: Why Everything You Know About Football is Wrong by Chris Anderson and David Sally is published by Penguin and is available at bookstores or online now at

by Andy Cooper PR & Project Manager

Published 10 December 2013