Lies, Damn Lies and Statistics

It’s tempting to blame Michael Lewis, or even Billy Bean. Since the publication of Moneyball in 2003, football coverage has become awash with statistics. Newspapers fill their match previews and reports with them, television companies make great on-screen play of them although, interestingly, the pundits pretty much ignore them. The internet is full of message boards in which fans use them to support their arguments about players. Daniel Finkelstein, writing in The Times seems to be trying to create a pseudo-science from the analysis of statistics and Sam Allardyce seems to think that his obsession with measurement makes him a realistic candidate for the job of managing Real Madrid. Unfortunately, all of this thinking is flawed, for two main reasons.

The first is the very obvious difference between football and baseball. Baseball is a game made up of a series of short, discrete actions that have a finite range of reasonably definable outcomes. Of course, hard core statisticians are delving deeper and deeper into the cause and effect of these outcomes, but, nonetheless, they lend themselves to fairly simple statistical analysis. Football, on the other hand, is a game of ebb and flow, with the ball remaining in play for longer, less definable periods, and with numerous actions and interactions involved in each passage of play. At a simple level, a batter receiving a pitch in baseball has a handful of options open to him – at the simplest level, he can choose to swing or not. A footballer receiving the ball has a much wider range of options open to him, each of which would have a very different outcome and, therefore, impact upon the game. This, obviously, makes the actions of the players far harder to measure, both in terms of effectiveness of decision making and of outcome.

The second problem is what seems to have been a widespread misinterpretation of the message of Moneyball. Managers such as Allardyce seem to have interpreted Billy Bean’s actions as ‘statistics are good’. The true message, however, is rather different – what made Bean a revolutionary was not that he used statistics to measure the effectiveness of players but that he challenged the accepted ideas as to what information should be studied. This is a lesson that football really should learn – statistics such as Pass Completion Rate are interesting but, stripped of their context, largely meaningless, for reasons that I shall discuss in more detail on another day. What would be more interesting would be measurements that challenged received wisdom and which dealt with quantifiable actions that have an easily defined range of outcomes. An example of this might, for example, be the measurement of the effectiveness of shooting from free kicks around the penalty area – the received wisdom is that any free kick deemed to be within range is an opportunity to shoot, but the likelihood is that the statistics would show that it is an ineffective way of using good attacking possession. This analysis could also be used to determine the optimum range, angle etc. for certain players, which may, in turn, help to inform the decision making process as to what should be done with a free kick and who should take it. This would be closer to the Moneyball example, in that it would have the potential to positively change behaviour in a way that might be unexpected although, of course, the data would need to be re-evaluated on a regular basis – if it was found that trying to do something more creative than simply shooting was less effective then it would need to be re-considered.

The issue with this, as Billy Bean found in baseball, is that football is inherently conservative and that people have long held opinions based on tradition and perception rather than objective analysis. While Mark Lawrenson dismisses a free kick with greater ambition than simply blasting the ball at the wall as ‘rinky dink’, while Steve Claridge continues to dismiss the opinion of anyone who hasn’t played the game professionally and while pundits persist in the fiction that ‘no-one can teach Michael Owen anything about finishing’ then true Moneyball style analysis will be a long time coming.

Some Belated Thoughts About Penalties

I originally wrote this just before the World Cup after listening to the smug inaniities of Messrs Shearer, Hansen and Lawrenson one too many times. Looking back, it seems rather optimistic to suggest that England should have been practising penalties - they would have been better served working out how to pass the ball to each other.

 

As the knockout stages of the World Cup draw nearer there are two things that we can be reasonably sure of: that at least one tie will be settled by a penalty shoot out and that, at some point, a television pundit will tell us that there’s no point in practising for penalties because there is no way of replicating the pressure of the real thing.

The second part of this statement is, of course, true: it is not possible to replicate the baying, or vuvuzela blowing, of tens of thousands fans in your comfortable training facility, but to make the leap from that to saying that there is no point in practising is absurd. You might just as well say that there is little point in practising any aspect of the game, since you can never replicate match conditions, something which may have been fashionable in the 1950s but which is rather less appropriate to the modern game. It’s also an idea that would be laughed out of court in just about any other sport: it is impossible to imagine Phil Mickelson saying that he doesn’t bother practising four foot putts, for example, since nothing can replicate the pressure of putting to win a major.

What can be achieved through practice is the development of a penalty routine for each player, so that when they walk up to take their kick they can focus entirely on that in the hope that it will block out the noise and, as far as possible, the nerves. A player walking up to the spot knowing that they place the ball, walk back four paces etc. is going to be more relaxed than poor David Batty was in 1998, when he walked up not only with the hopes of a nation on his shoulders but also with no idea of what he was going to do when he got there. Recent research at Exeter University has suggested that nervous players tend to focus more on the goalkeeper and, therefore, hit the ball closer to him: a well rehearsed routine will help to take that out of the equation.

In 2009, the England Under-21s, under the management of Stuart Pearce, found themselves in a penalty shoot out against Sweden. Pearce, who, let’s face it, knows a thing or two about the pressure of penalty shootouts, made sure that his team practised and ranked the entire squad from one to twenty-three, with the surprising outcome that Joe Hart took England’s second penalty, gleefully hammering it into the top corner. The other, less surprising, outcome, is that England won.

Of course, practising penalties doesn’t mean that players are guaranteed to score by any means. Players, like John Terry or, for that matter, James Milner in the 2009 shootout, can slip, or goalkeepers can make great saves, but it is naive to think that it is enough simply to turn up and trust to luck. As Gary Player once said ‘The harder you work the luckier you get’ - if England find themselves in a penalty shootout in South Africa then let’s hope that they’ve worked hard enough to get lucky.

Why Video Technology Wouldn't Work in Football

This article originally appeared on the When Saturday Comes website in December 2009, so some of the references are slightly out of date. The point still stands, though. 

In amongst some surprisingly favourable comments the piece was described by one poster as 'reading like a record review that's so determined to be different that it overlooks abundant qualities in favour of nit-picking at the fringes', which I thought was great, especially as the rest of the comment read rather like a Sun editorial. Anyway, make of it what you will.

Barely a week passes without some perceived controversy or other causing commentators, pundits, fans and just about everyone else calling for the introduction of video technology. ‘The technology exists’ they cry ‘and other sports use it, so we must too’. Callers to 606 seem convinced that the use of technology would rid the game of all refereeing errors, although they are less forthcoming as to precisely how such a system would work.

There are two key problems with applying technology to football. The first is that its proponents almost invariably fail to take into account how technology is used in other sports. For example, a rugby union television match official would have allowed William Gallas’s goal against Ireland because he is only empowered to adjudicate on the act of scoring, not on anything that precedes it. A rugby league video ref would have been able to disallow the goal but would not have been able to pass any judgement on the other controversial incident that night, the clash between Shay Given and Nicolas Anelka.

In short, the evidence from other sports is that video technology is excellent for establishing absolute facts but falls down when an element of subjectivity is introduced. If we take cricket as an example: the third umpire works excellently for line decisions, such as run outs, because there is a clearly defined ‘in’ or ‘out’ decision to be made. However, where a level of subjectivity is involved, such as judging a catch behind the wicket, it has proven to be considerably less successful.

This is where the problem for football lies, for the laws of the game are too subjective for video to be used successfully in establishing absolute fact. The Henry handball was comparatively rare in that it was absolutely clear cut and was missed by both the referee and his assistant. The Anelka penalty incident was far more typical of the kind of decision that football referees have to make and, even after seeing the video over and over again, there is still no consensus among pundits, fans and commentators as to whether or not it should have been a penalty.

The law relating to the direct free kick includes, in its first paragraph, the words ‘in a manner considered by the referee to be careless, reckless or using excessive force’. These terms are defined later in the laws, but in a way that is still highly interpretive. An example of this is Jamie Carragher’s challenge on Michael Carrick at Anfield earlier this season. Carragher played the ball but was deemed, by some, to have used excessive force (an offence which, incidentally, the laws say should result in a sending off). No matter how many replays are shown, a decision like that remains subjective and, consequently, controversial. Compare this to cricket’s LBW law: in order for a batsman to be given out LBW a small number of clearly defined criteria have to be met. If any of them are not then the batsman is given not out. It is a simple law that is now being judged by all of the available technology, and yet even then there are still incidents in which a consensus isn’t reached.

In spite of the fact we would all be happier if games were decided solely by the actions of the players and that phone-ins and post-match interviews weren’t dominated by talk of referees’ decisions rather than the game itself, video technology would, ultimately, simply serve to deflect attention from the referee to the video ref. Certainly, technology such as Hawkeye could be used to determine whether or not the ball had crossed the line, but it is highly debatable whether the level of investment required could be justified in order to clear up a handful of decisions a season. For the time being we are better off without the video and, perhaps, would be better advised to get off the referees’ backs instead.

The alternative, of course, is just to stick Alan Green in front of a monitor and ask him to referee the game. He knows best, after all.