Question: What do you get when you cross a social scientist with a behavioural economist, and invite them to produce a tonne of data pertaining to the beautiful game? Answer: ‘The Numbers Game’, a book that is guaranteed to reduce some fans to tears of joy and despair in equal measure.
To be fair, the authors, Chris Anderson and David Sally, have certainly done their homework here, and if endless screeds of football metrics is your bag, this is just the ticket for you. Conversely, if data turns you off, leave it right there on the shelf.
Being a football fan, of course, most will at least be tempted to have a peak. Especially as the book’s testimonials include the promise that ‘The game you love will take on new depth, colour and subtlety.’
High praise indeed. But, well, will it?
At the outset, we’re reminded that various software companies have produced sufficient packages to sate any thirst for statistics. Mind you, we can go way back to the 1950s, and a chap called Charles Reep, who can lay claim to being the first football data analyst.
As a retired wing commander, Reep clearly had a lot of time on his hands, given he meticulously accounted every action during literally hundreds of matches, each of which received about eighty hours of analysis, from which a scientific paper emerged.
Of course, technology has advanced somewhat since then, and more detailed analysis can be drawn. This is where the book lifts the lid about an enormous amount of data, some of it rather obvious, but there are a few surprises too.
Here a quick one: football is 50% skill, and 50% luck. Poor old Albert Einstein will be spinning in his grave – he, after all, insisted that ‘God does not throw dice.’ Well, he evidently does with football, Albert.
We progress from there to randomness, and in that regard football can be linked to a probability equation that emerged from – wait for it – data analysis in relation to injuries sustained by soldiers from Prussian horses at the end of the 19th century.
Feel free to scoff at this, but the equation here can be applied to predict that, in terms of goals scored in the Premiership this coming season, 30 games will be goalless, 70 will be won by a single goal, and 55 will enjoy 5 goals or more.
Still with me? Good. Sticking with the randomness theme, the favourites in any football match are 65% more likely to win, compared with an average of 80% across other sports.
Conclusion? Football is more susceptible to chance than most others, no matter the odds.
When it comes to ability, a team needs to win by three clear goals before we can be certain that the best team won. ‘Bayesian statistics’ prove the point, trust me, and whether this makes you feel any better the next time your team loses 2-0 is open to conjecture.
In terms of goals, here’s the bad news: 50 years ago, there were 3.5 goals per game in the top tier. Now it’s 2.5. What does this tell us about fitter and faster players, tactics, etc?
You may well ask, but meanwhile an analysis of valuable goals (that is, those that can be classed as game winning or saving) has indicated a clear winner in English football between 2009 and 2011. Never mind Torres, Rooney, Defoe and Berbatov; no, step forward Darren Bent, who is head and shoulders above the rest when it comes to scoring meaningful goals!
Sticking with goals, it’s rather disappointing to read that the art of avoiding the concession of goals is 33% more important for success than actually scoring them! Well, thanks very much for that, regression analysis.
As for possession, having a majority of the ball should mean a team will outscore its opponents by 1.44 to 1.19. Have more of the ball, and you decrease your losses by 7.6%. Interestingly, one team has bucked the trend in recent season: Stoke City no less.
We then wander into more predictable territory with confirmation that wage bills relate directly to success. Well, you don’t say. Indeed, this is the case by a whopping 92%, although Wigan Athletic, in avoiding relegation from the Premiership for so many seasons in a row, stick two fingers up at that particular metric. That their former manager, Roberto Martinez, is credited with being a data geek might have something to do with that!
There’s lots more of this besides, including a detailed description of why the Real Madrid Galactico project was a relative failure a decade ago – the system didn’t multiply the players’ qualities exponentially, apparently. I’m not arguing with that.
Also, given that a study showed that overall spend accounted for 89% of the variation in league placing over a ten-year period, we are asked to deduce that a manager can only influence the 11% that money can’t buy.
To be fair, the book does acknowledge that the 11% can be the difference between success and failure at the highest level. A statistic that the likes of Wenger, Mourinho and Rodgers might well utilise during their contract renewal negotiations!
For those of us who find all this analytical stuff mildly depressing, let’s end on a positive note. The good news, after all this analytics, is that the potency of randomness and chance in football will guarantee that the game will not change fundamentally – there will be plenty of room for riveting drama.
Whew – thank goodness for that!