“They can lead you the wrong way. You have to be very careful.” This is how Arsenal manager Arsene Wengerspoke of statistics in a recent interview with Bein Sports.
Wenger is right. In 2001, Sir Alex Ferguson sanctioned the sale of Jaap Stam to Lazio. After reportedly looking into the defender’s tackling stats, Ferguson had attributed a decline in volume to the beginning of a regression in playing ability. Stam went on to shine in Italy, in the process becoming one of the Scotsman’s most notable mistakes.
Although the use of data within football has advanced considerably since the Stam transfer, its spread has been largely restricted to player recruitment. This is a natural fit. Even the increasingly wealthy Premier League clubs have finite resources, and so for now, statistics in football have been most effective in creating shortlists of players based on a variety of filtering processes before being further evaluated by more traditional methods.
This balance between tradition and modernity varies from club to club. At some teams, like Liverpool or Aston Villa, their publicised movement towards the use of stats in recruitment has made them punching bags for various pundits when signings fail. British football has been turned into a tug of war between straw men – on one hand, the ‘real football men’, on the other, the ‘number wizards’ – with little crossover.
That is, apart from Arsene Wenger. But the 66 year old’s embracing of football analytics is not simply the case of an old dog learning new tricks. In 2002, he began to substitute Dennis Bergkamp late in matches. Annoyed, the non-flying Dutchman initially tried to complain, only to be shown his running stats, clearly illustrating a slow-down towards the end of games.
Wenger’s unusual hybrid approach, employing both the ability of impartial analytics and his own instincts, gives an empirical edge to his instinctual tactical decisions. When Aaron Ramsey was recently moved back to central midfield from the wing, Wenger defended the move: “If you look at his expected goals when he is in a central position, it is among the best in the Premier League”. Evidence-based analysis helps Wenger translate his findings to players too: “The weight is greater if you can say, “Look, this team has conceded 70 per cent of their goals from the left.” The players will believe you much more.”
When it comes to taking concrete action based on sound statistical analysis, Arsenal are ahead of the curve. This may be related to their purchase of the football statistics company StatDNA in 2013, which gave the club the unique capability to source, compile and analyse their own data internally. This has given them another edge; Arsenal recently signed Mohamed Elneny from the Swiss Super League, notably one of few competitions not covered comprehensively by Opta, who are the data providers for almost all Premier League clubs.
Yet while it may be tempting to see Wenger’s use of stats as a trump card in the club’s push for their first title since 2004, the reality is less black and white. In a sport as immensely complicated as football, particularly when it comes to the competitive position that Arsenal are in, much of what they are do with stats is in pursuit of marginal gains – small but cumulatively important advantages over opposition. The intelligent use of analytics provides a host of these marginal gains, like improving tactical choices, scouting leagues for which your competition doesn’t have data, or getting a small edge number of other low-hanging-fruit like free kicks, corners, etc. Yet the combined advantage of these remains to be seen, and there are a multitude of other factors to consider.
Arsenal’s consistency under Wenger is polarising. To some, Wenger’s Arsenal are always the bridesmaid, never the bride; to others, the Gunners are a stable and somewhat unlucky dynasty. A lack of spending is deplored by the very same pundits who will tear apart another team for spending money “inefficiently,” but, as some analysts have pointed out, what Wenger seems to have been aiming to do at Arsenal with the use of stats is limit variance and build the team slowly, aiming for small wins rather than make big, risky gambles. By contrast, Liverpool have a similar mean finishing position in recent times, but their swings – up to 2nd from 7th at one point – are more volatile. Manchester City may have won the league more recently, but they have also spent much more to try and achieve that, which in turn increases the perceived magnitude of their failure when they don’t.
Wenger’s attempt to gain an edge with statistical analysis is commendable regardless of what happens this season, but football is outcome obsessed. Were Arsenal to win the league this season, the nerds would, by sharing the limelight with Wenger, win the straw man tug-of-war – at least temporarily. Should the Gunners fail, in a campaign as remarkably open as this one and with a gradually accumulated burden of expectation, public opinion could bury them both.