Football can’t be reduced to number-crunching
Moneyball shows how statistics helped to change baseball. But maths won’t deliver football silverware.
What on earth is Aaron Sorkin playing at? First he succeeded in glamorising IT nerds in his screenplay for The Social Network. Now he’s at it again, glorifying nerds in another movie, Moneyball, which tells the story of how statistical analysis revolutionised baseball. The moneyball concept – using performance metrics to gain a competitive sporting advantage – is relatively well known on this side of the pond. The million-dollar question is whether Moneyball could be successfully transplanted to football.
The film, which opened in the UK last week, is based on Michael Lewis’ bestselling book Moneyball: The Art of Winning an Unfair Game. The hero of the story is Oakland Athletics baseball coach Billy Beane, who successfully used ‘outsider’ performance data or sabermetrics – a term derived from the acronym Society for American Baseball Research – to recruit players who were overlooked by old-fashioned ‘insider’ scouting knowledge. It wasn’t that baseball had completely eschewed statistics. Rather, the new metrics deployed by Beane, such as on-base percentage, proved more effective than traditional indicators, like batting averages. Beane’s use of performance data to unearth gems from college baseball paid considerable dividends. In 2002, the Oakland Athletics were able to compete with powerful rivals, such as the New York Yankees, whose salary budgets were three times as high. The moneyball approach was game changing. Today, most of the big baseball teams employ full-time sabermetric analysts.
In the film, Brad Pitt performs the part of Billy Beane. It’s a cynical piece of casting, a blow for nerd supremacy. Will it wow British audiences? Judging by my wife’s point-blank refusal even to consider a trip to the cinema, this might not be a film for the ladies. For her, even top-drawer Hollywood eye-candy isn’t sufficient to sweeten the unpalatable mix of sport plus statistics. Maybe I should have pretended that the film was based on a long-lost Akira Kurosawa screenplay about a group of unheralded samurai warriors recruited by a number-crunching monk. Anyhow, I digress. Although the missus wasn’t interested, Moneyball will appeal to the swelling ranks of statistic-obsessed football fans. The successes of Fantasy Football Leagues and the computer game Championship Manager attest to the popularity among geeky men – and it is predominantly a bloke thing – of performance data as a tool for predicting footballing success.
But, while Opta stats might enable laptop managers to achieve virtual success on Championship Manager, could data analysis be a recipe for winning real silverware? Billy Beane believes that football’s version of sabermetrics is out there waiting to be discovered. ‘Every business has metrics that correlate to success’, Beane recently told the BBC. ‘It’s just finding them and which ones are the most valuable and which ones do you invest in and which ones you get a return on.’ It’s certainly true that some performance data is now considered indispensible in professional football. Most top clubs now use Prozone analytical tools to evaluate player performance and design individually tailored training regimes. Damien Comolli, Liverpool’s director of football, is both a friend of Beane and a fan of his methods – though, as Gabriele Marcotti recently observed in The Times (London), ‘it’s hard to see how Liverpool’s summer signings fall into the moneyball category’. Marcotti has a point. Is splashing £35million on Andy Carroll evidence of smart stats or just an old-fashioned hunch?
Arsene Wenger is another well-known stats obsessive. ‘Technical superiority can be measured’, Wenger has said. ‘In the past, it was just about feelings, opinions. I wanted to know a little bit more. I am always in the situation where I have to judge people, and the more concrete, objective numbers you have the better you can achieve that.’ Wenger places a great premium on passing speed and, consequently, he measures the average time players spend on the ball. Less time on the ball indicates better technique.
Prozone or Opta can be useful for measuring some aspects of performance, such as fitness or passing accuracy. But, unlike baseball or cricket, which are comprised of a series of individual episodes of play, football is a relatively fluid and complex game which makes metrics for individual player performance much more difficult to devise. How do you measure movement off the ball, positional awareness, defensive concentration, good decision-making or tactical nous? Fabio Capello recently said that he could tell from the England warm-up that his players weren’t in the right frame of mind. Can Prozone tell you that? It’s easy to ridicule conventional scouting wisdom but, while there are metrics for fitness or passing accuracy, the old methods are still better for assessing whether a player has the hunger or drive to succeed, whether he has a good football brain, whether he is a leader on the pitch or whether he is mentally tough. There are too many subjective success factors such as these which are nigh on impossible to quantify.
I’m all for using analytical tools to aid coaching. However, I find it reassuring that no one has yet discovered a moneyball-style winning formula. Success in football is determined by attributes such as character or creativity and this can’t be measured so easily. Football is still more art than science. The title will never be won by algorithms. The nerds will never take over.
Duleep Allirajah is spiked’s sports columnist. Follow him on Twitter @DuleepOffside
To enquire about republishing spiked’s content, a right to reply or to request a correction, please contact the managing editor, Viv Regan.