How computers could be England’s best chance
England claimed an against-the-odds win against world champions Spain on Saturday. The reception to the victory, however, was muted, following as it did a defensive display Bolton Wanderers would have been proud of. Not exactly the stuff as footballing dreams are made on.
But manager Fabio Capello was simply applying statistics. Spain were clear favourites to win (England were ranked as underdogs by every bookmaker in the country). So Capello flooded the midfield, massively decreasing the chance of either side scoring and therefore reducing the chance of England being on the receiving end of a hammering. In the event, the tactics payed off and England grabbed the game’s only goal.
This is a very simplistic application of statistics to football. But many clubs are going far deeper, using complex algorithms to help them gain an edge.
Manchester City scored only 10 set piece goals last season, equating to 22 per cent of their total (contrast this to Stoke, with 44 per cent). They are on course to dramatically increase this, based on their games so-far this term. The difference? Statistics, apparently. The club is one of the increasing number to use computers to analyse games, including successful corners and free kicks, noting factors including the speed, height, angle and spin of the ball. Players can then be trained accordingly: to take the “perfect” free kick, for example.
Of course, statistics alone are not responsible – City have spent the GDP of a small country on players this year, presumably increasing their set piece acumen.
But analysing these stats can give a club an edge. During his time as manager of Liverpool, Rafa Benítez used controversial zonal marking for set pieces, in which a player is tasked with policing an area of pitch, rather than a player. When this goes wrong, it can make a team look incredibly inept, leading some managers to dismiss it. But over the six seasons Benítez was in charge, Liverpool twice conceded fewer set piece goals than any other club in the league.
Statistics make it easier for managers to overlook the odd, albeit conspicuous, howler and employ the most consistent strategy. Top players use statistics every match they play, although they would probably give you a blank stare (blanker than usual) if you told them. While commentators talk of “intuition” when a player makes a perfect run into the box, in reality he is just playing the odds, and has probably made the same run unsuccessfully several times before. Computers just help us understand what these odds mean and how we can take advantage of them.
Americans have been doing the same thing for years. Their sports fans revere statistics in a way that would make Fantasy Football’s Statto look ill informed.
One of the first proponents of the application of statistics to sport was Oakland A baseball coach Billy Beane. He shunned big name (and big money) players, instead relying on computer programs to pinpoint the most consistent players, regardless of stature or reputation.
They went on to win everything and Beane’s success has been made into a biopic starring Brad Pitt (Moneyball, out later this month).
England’s Under-21 manager Stuart Pearce is known to be a fan of advanced analytics, sifting through comprehensive breakdowns of his sides’ performances. Fabio Capello is thought to be less convinced. Perhaps he should reconsider – looking at England’s chances of winning a major trophy, I think we need all the help we can get.
By way of full disclosure, I owe credit for this column to the guy sitting next to me at the pre-match dinner on Saturday, whose name I forget (or perhaps never knew), who very eloquently explained large parts of this.