Analytics changing way we interact with soccer

Stewart Downing, middle, in action for West Ham United. (Tim Ireland/AP)

Statistical analysis is commonplace in almost all sports today.

In baseball statistical analysis has its own name, sabermetrics, and is used almost universally. In hockey the analytics debate has taken on a life of its own with many teams hiring independent bloggers as consultants. In soccer, however, the growth and development of analytics has been a bit slower.

Soccer by its very nature is an extremely conservative sport. The game itself only has 17 laws and these laws have changed very little over the years. Soccer isn’t a sport that accepts change easily, so it is no surprise that statistical analysis has taken longer to catch on.

Now that analytics is just starting to gain some mainstream attention in the soccer world it inevitably has its detractors, and with detractors comes misinformation and misunderstanding. So it is important to answer the questions of what soccer analytics is and how should it be used? A good starting point is with Thomas Bayes.


Soccer Central podcast: SPORTSNET.CA's Soccer Central podcast, hosted by John Molinaro and James Sharman, takes an in-depth look at the beautiful game and offers timely and thoughtful analysis on the sport's biggest issues. To listen and subscribe to the podcast, CLICK HERE.


Bayes, an eighteenth century English statistician and philosopher, developed one of the most influential theorems in the history of statistics. Bayes’ Theorem is a way of updating the probability of an event occurring by combining what you initially thought the probability was with new evidence that might influence the probability.

What does this all have to do with soccer? Bayes’ Theorem perfectly lays out how we should use analytics to update our opinions on players and teams. We have our preconceived notions that we get from watching the game, and then we look at the numbers to see what more we can learn. Finally we can make an opinion based on the combination of these two things.

Consider a scout trying to find a winger for his team in the summer of 2012. The scout is told to give his club a recommendation of whether to sign Stewart Downing then of Liverpool or Adel Taarabt of Queens Park Rangers.

Both Downing and Taarabt were coming off poor seasons in 2011-12 when they were both heavily maligned by supporters of their respective clubs. The scout watches their performances and has a rather negative view of both. Thomas Bayes would call this initial evaluation of the two players the scout’s “priors.”

The scout then looks at the underlying numbers of the two players. Suddenly both players look much more appealing. Downing averaged 2.6 shots and two key passes per 90 minutes. Adel Taarabt averaged 3.9 shots and 2.2 key passes per 90 minutes. These statistics adjusted per 90 minutes have been shown to be highly indicative of future performances and team goal scoring.

Based on this new information the scout updates his view of the two players. His opinion is now based on both what he saw on the pitch, his priors, as well as the new information that comes from their statistics.

The scout waits another season to make a judgement call. Downing’s performances with Liverpool start to look a little better. Taarabt on the other hand still gives the ball away far too often and shows himself to be a poor decision maker.

With two years of information the scout has updated his original opinion by analyzing the statistics of the two players and by watching them play another season. Based on this analysis the scout would probably suggest to his club that they should buy Downing.

Downing did, in fact, move that summer to West Ham, and is now on pace for a career season. Taarabt has continued to struggle with QPR after a series of loan moves. It’s no surprise that West Ham manager Sam Allardyce is one of the biggest proponents of analytics in soccer.

If the scout had only watched the two players without any statistical background he probably wouldn’t have suggested buying either one. Had the scout only looked at the players’ numbers he probably would have suggested signing Taarabt. Using a combination of both, watching the players and looking at their underlying statistics, it is clear that Downing was the better option.

This is exactly what analytics is. Statistics are the raw numbers that are reported on in a match report or by data companies such as Opta or Prozone. Analytics is the process of putting these numbers into context, analyzing them and trying to understand them. The Oxford dictionary definition of analytics is, “the systematic computational analysis of data or statistics.”

When people attack analytics they often just attack statistics on their own. They point to teams that have low possession numbers but still win. This shows a misunderstanding of what analytics is. The biggest proponents of soccer analytics would actually probably agree with most of these critiques of using statistics in isolation.

Given this broad definition of analytics, how can people in soccer use analytics?

Firstly, clubs should be using analytics in a similar manner to the Taarabt-Downing example above. Analyzing statistics to help understand what is sustainable and what isn’t, which players are actually good and which are just getting lucky, etcetera. Some clubs such as West Ham appear to have “bought in” to the system, while others are still languishing behind.

From a journalistic standpoint analytics can be incredibly useful in distinguishing between actual narratives and fabricated ones brought on by randomness or luck. Using analytics can also help with predictions and testing various theories about a team or player. In general, analytics can bring a fresh perspective to an issue in the sport that you couldn’t necessarily get from just watching the game itself.

Finally, what should fans make of analytics? Obviously people are fans of the game because they love watching it, and it isn’t my place or anyone else’s to tell someone how they should enjoy the sport. If you hate the idea of using numbers in sports then it is your prerogative to ignore analytics and go on enjoying soccer as you always have.

However, if after watching a game you are still craving for more soccer, more ways to understand and interpret the game, then analytics will open up a whole new world. Using analytics can launch you into a new series of questions and debates about the sport that watching the game alone never could.

Personally as a fan I love nothing more after a game than going to the numbers and learning more about what I just watched and what it means going forward.

The growth of soccer analytics may be slightly behind that in baseball or hockey, but like other sports analytics will start to gain more traction and change the way we appreciate and interact with the game.


Sam Gregory is soccer analytics writer based in Montreal. Follow him on Twitter