From Danish champions Midtjylland, to English sides Brentford and Coventry City to the A-League’s Central Coast Mariners, teams from all over the globe are looking to use statistical analysis to gain an advantage over their opponents.
While it’s commendable to try and think of innovative ways to succeed in football, it is important to understand these methods and question if they would actually work.
The evidence for how analytics can be used successfully in sport was famously captured in Moneyball. The Oakland Athletics used statistical analysis to make the most out of their meagre resources and give Major League Baseball’s financial big hitters a major run for their money.
But can this system work in football? There are many that believe it can and have been searching to find the right algorithm for success.
Like Everton manager Roberto Martinez, I am quite sceptical about the reliance on statistics to judge performance in football, for one vital reason - the most important elements of football success are very difficult to measure.
Football is a fluent game where creativity and decision making, as well as technical skill and tactical intelligence, make the difference between success and failure.
A top player has to have excellent technical ability, which has the potential to be measured, but more importantly the very best possess the ability to create, which is much far more difficult if not impossible to measure.
The world’s best are also excellent decision makers, taking the right option at the right moment in the right part of the field. How can that be measured objectively with a numerical value?
The game’s greatest defenders are excellent at reading the situation and positioning themselves to deny their opponents. Again how can that be rated numerically?
There’s no doubt that statistics can be used to measure pass completion, tackles or dribbles, but because no context is provided as to how this took place during the game, these other more important elements are not taken into account.
Often central defenders have the best pass completion statistics in a match, but most would realise that doesn’t mean they are the best passers of the ball in the team. They achieve this ‘stat’ because they usually receive the ball under less pressure than midfielders and attackers, and typically have more options available to them (unless you’re getting pressed by Bayern Munich).
So how is a good pass measured? It’s possible that someone can come up with a formula to do so, but then isn’t that defeating the idea or removing subjectivity from the performance judgement process?
What is ‘good’ is a subjective matter. A good pass for one coach might completely different to what a good pass is for another.
In fact, choosing what to measure or how to weight that measurement is another subjective process. For instance, if an analyst decides to weight the measurement for tackling more heavily than the stat for interception, doesn’t that suggest that the analyst believes tackling and intercepting are important enough to measure and then that one is more vital than the other?
Yet a defender could be forced to tackle an opponent because they themselves made a mistake by being out of position, while a defender making an interception could just be the result of the excellent positioning and pressuring of their team-mates.
These are all factors that must be taken into account when judging the performance of the individual as well as the team and to me that means understanding the actions that cannot be measured.
I think it’s noteworthy that in the book Pep Confidential that Pep Guardiola, who I’m sure most can agree is one of the world’s great football minds, pays little attention when the author mentions Bayern Munich’s phenomenal possession statistic.
Instead Guardiola asks the journalist if he saw that excellent movement or that fantastic pass. This is the key to a successful football team.
That is not to say that statistics cannot play a role. When used by knowledgeable coaches, scouts or other performance judges they can be useful to confirm what is believed or even question that belief.
But to completely rely on ‘football analytics’ to make the decisions seems to me quite problematic.
