General consensus usually traces the use of statistics and data analysis in sports back to Moneyball, a now famous term coined to describe the Oakland A’s approach to building a competitive baseball team despite having one of the MLB’s most stretched payroll budgets. The concept was rather simple: Instead of relying on the traditional and often subjective knowledge of seasoned player scouts, who watch videos and attend games to judge a player’s ability, the Moneyball approach relied on statistical, objective evidence and data analytics to build a roster. This way, small-market teams would be able to compete with much wealthier teams by buying undervalued players and selling overvalued ones.
The use of data analytics has since seen a significant and widespread increase in the world of sports. There are countless success stories of teams using data analytics to improve scouting, enhance performance, or even predict outcomes. While analytics quickly caught up and brought tangible benefits to sports such as baseball, American football and basketball, the fluid and chaotic game of soccer remained the “least statistical of major sports” (as per a New York Times proclaimed in an article back in 2010 before the World Cup) until recent times. This said, for our analysis we were primarily interested in this Financial Times article about soccer clubs using data analytics to improve scouting and sign-on optimal players.
The benefits of data-supported scouting are numerous. As the article mentions, using data analytics in the early stages of scouting casts a wider net while searching for players and providing clubs access to markets they would not have previously considered or been exposed to. This is particularly relevant to small and medium sized soccer clubs with limited financial resources, which historically, have been at a competitive disadvantage. The obvious players they wanted were already known on the market and probably well out of their budget. Enter data analytics. Having a 3 person team crunching away at global player performance stats from an office is significantly cheaper, and faster, than employing an army of scouts to go around the world looking for the next big hit.
Another big pro of data-supported scouting is efficiency. There are roughly 265 million professional soccer players around the world, without even taking into account youth and amateurs. No human scout can be expected to sort through all this data, look at different metrics and find the next best batch for their soccer club. However, a competent data analytics team can handle all this data efficiently and find the right correlations and compatibility between potential recruits the metrics most relevant to the soccer club.
A drawback of data analytics, is well, that it is entirely based on objective data. Regardless of how many metrics you capture, there are soft qualities such as team dynamics, temper, motivation and even pure skill that cannot be quantified, measured or predicted upon by complex algorithms and machine learning techniques. That is the beauty of the chaotic and fluid game of soccer. Only the human eye can be trained to pick up on these qualities (so far), and that is why we believe scouts are still critical in the process. They have a keen eye that captures what machines can’t and assess the player’s congruence within a team. All they ever needed was data to direct their vision into the right field.
As such, even though we are proponents of the use of data analytics in soccer, we don’t see it as replacement but rather a complement of the traditional scouting process. Clubs can now have a credible and reliable shortlist of potential recruits that scouts can leverage to find the ultimate player-club fit. However, as more soccer clubs adopt this approach, we believe that incorporating data analytics into club operations will reach a point of marginal returns as the competitive advantage becomes the competitive norm.