Why they haven’t made a film called Long Ball

The film Moneyball, ed based on the book by Michael Lewis, health has opened to great acclaim and commercial success in the US and will be coming to Europe later this year.  It celebrates how the general manager of the Oakland A’s used statistical techniques to put together a successful team on a minimal budget relative to other Major League Baseball franchises by employing statistical techniques to identify which players he should bring in.  The question this prompts is why haven’t British film producers made a film called Long Ball to celebrate the statistically-based eponymous theory that captivated, seek among others, Charles Hughes, the England Football Association’s Director of Coaching in the 1990s?

In both cases there was short term success until other teams caught on.  Two exemplars of Hughes philosophy were Graham Taylor, who took Watford from the fourth division in 1977 to the first in 1982 using the long ball approach, and Howard Wilkinson, who managed Leeds to winning the final old First Division championship in 1992.  Taylor even employed Charles Reep, a retired RAF Wing Commander, accountant and amateur statistician, on whose pioneering work in the 1950s Hughes’ philosophy was based.  Reep was football’s first performance analyst and his copious short-hand notes – nearly 2,500 of which survive – provided coaches with comprehensive match data at least 20 years before any similar process emerged.  He analyzed the number of passes that led to a goal and found that a higher percentage of goals involved just three.

In similar vein, Hughes analyzed over one hundred games from all levels (including those from the World Cup) and determined the majority of goals were scored with 5 or less passes of the ball.  Hughes deduced from this that teams should play ‘direct’ football and emphasized the importance of particular areas of the field from where goals were most often scored, which he called POMO – Positions of Maximum Opportunity.  He asserted that players would score if the ball was played into the POMO enough times and attributed great importance to set plays and crosses into the opposition’s penalty area.

The real difference, of course, was in how other teams caught on.  The Oakland A’s success was short-lived because other teams copied their approach whereas direct football proved to have no lasting efficacy because teams worked out how to defend against a team dedicated to playing direct football.  The world’s most successful team in recent years, Barcelona, exemplifies the potency of the combination of skillful players, one-touch short passing and fluid movement (also working hard to win the ball back when it is lost).   Whereas teams using a long ball approach are much easier to defend against – defenders can be clustered around the target man either to beat him to the ball and head it away or to intercept his headed pass before it reaches a team mate.

The failure of the long ball theory highlights what can happen when there is over-reliance on statistics.  There is a belief that because certain events are statistically associated with particular outcomes, the way to achieve that outcome is to keep repeating the statistically associated event.  In essence correlation is confused with causation and what happens naturally – a team scoring from a swift counter-attack involving few passes – cannot be forced when the circumstances aren’t right and the opposition is organized rather than spread out in disarray.

Underlying such errors is a common human fallibility that stems from our innate dislike of uncertainty.  We project causality onto a relationship because if it is causal, we can impact it; but if it is just a correlation, we can’t.  And in our desire to control our environment, for reasons of psychological comfort, we conspire in our own confusion.

As a result we can see similar examples from the worlds of economics and business.  From the former there is Goodhart’s Law, named after the adviser to the Bank of England who identified that once a measure (for example a measure of money supply such as M3) became a target (with the aim of reducing inflation) it lost its value as a measure – the previously observed relationship between M3 and inflation broke down.  Similarly from strategy there is the relationship between relative market share and profitability which led business executives to chase market leadership positions in the hope that it would deliver superior profitability, only to find that it didn’t.

This is not to deny there is huge business value in statistical analysis – there clearly is as companies such as Tesco and Progressive Insurance among others have shown.  But there are also black holes.  Data mining and other advanced techniques can tell you whether there is a relationship between variables, not whether one causes the other or both are the effect of some other cause.  The best that you have is a hypothesis which needs testing.

Remembering this is going to be even more important in the future as data generated are increasing at an exponential rate that not even a liquidity crisis followed by a debt crisis can slow.  More data means more relationships will be identified resulting in both greater insight but also greater scope for misinterpretation.   Statistical smarts need to be matched by business smarts.  Businesses that forget that will be stuck playing a long ball game.

Share the knowledge:
  • Twitter
  • Tumblr
  • Digg
  • StumbleUpon
  • Reddit
  • del.icio.us
  • Slashdot
  • Facebook
  • LinkedIn
  • Yahoo! Buzz
  • Google Buzz
  • Google Bookmarks
  • Print

Related Posts:

About Jack Springman

I am a consultant with experience in business strategy and customer strategy development, customer management and customer service transformation.

Leave a Comment