Do Soccer Teams Need A Manager Or A Data Scientist?

Arsenal’s Arsene Wenger has finally bought into the numbers


When early analytics in football is discussed it seems it’s impossible to avoid referencing Sir Alex Ferguson’s misjudged sale of Jaap Stam to Lazio. After three indomitable years at Old Trafford, Stam was offloaded for a handsome £15.3 million. Ferguson looked at Stam’s tackling statistics, saw that they were in decline, and couldn’t refuse the chance to bring in that kind of money for a 29-year-old. United would ultimately regret the decision, though, given Stam’s career lasted another four years, at a very high level.

Ferguson is one of the most celebrated managers in soccer’s history, and rightly so, but he had a fair few missteps in his time in Manchester. Stam’s tackling stats were in decline, yes, but this was by no means an indication of the defender’s waning form. Rather, it was an indication of the Dutchman’s positional sense improving - for some defenders, the more experienced and positionally aware they get, the fewer tackles they’re drawn into. The tale is somewhat apocryphal, but it addresses one of the key detractions from data in soccer - the idea that there is no substitute for gut instinct and managerial experience.

If you’ve read Simon Kuper and Stefan Szymanski’s Soccernomics, you’ll know that the introduction of a new manager into a club can give a club a boost. Similarly, a sour atmosphere around the dressing room can negatively affect performances. When numerous seasons are tallied and the effect of a different manager on a team’s fortunes is calculated, though, it is shown to be minimal. Ultimately, a team will perform about as well or as poorly as expected. A new tactical system can outsmart opponents, an astute signing can give a team a boost, but as much as we like to eulogise and hypothesise about the effect of a soccer manager, there are bigger elements at play.

Club finances and luck are two of the biggest elements in the sport. The latter comes and goes, but accounts for a large points differentiation at the close of a season. The former is more difficult to change, and Soccernomics goes some way to proving that over an extended period of time, money wins out. Data might be about to muscle its way in as an important factor, though. There’s a reason clubs have been putting so much money into building analytics teams. Managers will change, luck will come and go, but if the club has sound data on not just its own team but its transfer targets and its opponents, it can lay a foundation for success.

In 2014, the news broke that Arsenal had bought data company StatDNA. The story goes that in a pitch to the otherwise unsold Arsene Wenger, the company explained that an empirical approach to recruitment could have gone some way to avoiding poor signings like Marouane Chamakh or Park Chu-Young. The manager was sold, and the company was bought outright for around $4 million. The data arms race in soccer has been raging for years - Liverpool’s sporting director is its former head of analytics, Manchester City use data to inform every significant decision, and Swansea City appointed the founder of North Yard Analytics, Daniel Altman, as a transfer consultant.

According to the New York Times, Arsenal’s transfer policy could’ve already been given a facelift by data. After years of being accused of excessive frugality in the transfer market, both Manchester City’s Kevin de Bruyne and Napoli’s Gonzalo Higuain were flagged cup as potential impact signings. The failure to sign the two hugely successful players falls unfortunately to Wenger. A poor run of results has seen Arsenal fans fighting in the stands over the future of their manager, with prominent figures in the fan base calling for his sacking, a situation that could’ve perhaps been avoided by more decisive and more shrewd transfer activity backed up by data analytics.

Having said that, the work being put into analytics at Arsenal reportedly runs far deeper than transfer advice. Because the club have the luxury of having an in-house data team in StatDNA - where most clubs source their data from companies like Opta - their data is more thorough. The New York Times called it ‘cleaner’ and ‘more in depth’, with specific measures in place to record player performance down to incredibly specific actions. The club is working to evaluate certain partnerships on the pitch through their data, to improve fitness, to catalog the severity of errors made and, of course, to make more intelligent transfers. Wenger has shown that his interest in numbers extends far beyond the financial.

Soccer has long been seen as a key area of opportunity for analytics to flourish, and despite some opposition in its early years it’s now a fundamental. It’s argued by many that Leicester City’s incredible title triumph in the 2015-16 season was down to the signings of Riyad Mahrez and N’Golo Kante, who were both signed at least in part on the back of their numbers. If the bigger clubs, like Arsenal, can build a data analytics strategy that works, the true power of exploiting the numbers might just come into focus. 

Vision small

Read next:

Big Data Forecasting In Pharma