Skip to main content

Viewing stats in the wrong light: Why context matters so much in football

Trent and Robbo
(Image credit: Getty)

The fallibility of the human eye. They are sensory marvels, in that they allow us to perceive our surroundings but there limitations in seeing, illuminated by the fact that our eyes can only see 1% of what is around us. The other 99% sits on a spectrum untouched by the human eye. 

Once you get into football though, this sensation is reversed in a sense. What scientific data is unable to elucidate, the watching eye of a supporter, scout, manager, any onlooker can be seen by what is on the screen. 

Statistics in football have always been a sticking point for many. Its quick proliferation in popularity and use within recent years has only made it ever-more contentious. In 2017, Match of the Day began to use xG (Expected Goals) in the usual stats round-up they show during the post-match interviews. It had come a long way from being sneered at as something unuseful in the evaluating performances of players and clubs.

Alas, the opposition to these types of stats wages on till this day. In essence, xG is statistical proof for “he should have scored that” or “that chance is harder than it looks”. There was never an issue with saying that but having a value to ascertain this truth manages to rub people the wrong way. Perhaps it is the idea that it rules out the proverbial ‘discussion in the pub after the game’ that supporters seem so wedded to. Maybe it is the fact that a new stat represents a microcosm of the cold, mechanical world that football seemingly changes more into. 


Juan Cuadrado

(Image credit: PA)

RANKED! The 10 worst January transfers EVER

Because it is a statistic that has worked its way into clubs’ judgement of performances. One of the biggest effects that xG has had on the game has been its reduction of long potshots (opens in new tab). Teams are electing to move the ball to more advantageous areas to work a better shot location with better xG value. Smarter? Yes. As good to watch? No. It's comparable to the shunning of the mid-range shot in basketball for similar reasons. It is not just a problem with long shots. 

It is even impacting stylistic ways of playing football. Crossing and its efficacy has been lauded with the Trent Alexander-Arnold and Andy Robertson’s great use of it for Liverpool these past few seasons. Just on probability, eventually, the ball is likely to break in a favourable way for the attackers. Even if it is headed out, there is still a good possibility that you regain possession. 

RICH JOLLY Can Jesse Lingard shine again at West Ham?

It isn’t a great watch. It does not evoke the same feeling as beautiful passing moves or long-range stunners. But they are more likely to get you the goals and the result. No matter how they are scored, a goal is a goal. 

Stats are not just focused on goals. xA and xGBuildUp also make up the Expected Stats family, looking at the creators and progressors on the pitch. Progressive passes, passes into the final third, passes into the penalty area are also more recent iterations of creator statistics. As aforementioned, the value of the deep-lying playmakers or the value of creators, who may have their numbers stunted by poor finishers, are seen clearly within these metrics. Progressive carries show those that take their team up the pitch, one version showing the distance and the other the number made. There isn’t much that stats are unable to cover these days.

But they are unable to cover everything. You see, the problem with all this is that stats are not infallible. Whatever the numbers show you on the screen, contextual support will always be paramount. In fact, it is more important. Even in lab reports, when you compare the sections with the stats to the other sections, it is considerably smaller. 

The methodology, the hypothesis and, most importantly, the conclusion is where the true value of any report or study is held. What is the stat trying to tell you? How did the result come about? Is the stat truly showing what it is measuring and is there something unmeasurable swaying the results?

Before xG became a thing, and thus xA, the measure of a creator was a chance created or a key pass. It was a relatively new stat. A key pass is an assist of a shot. A chance created is the number of key passes and assists added together. There was opposition from many. These days, they have become widely accepted as fine to use. There is a bigger bad in town, now there is xG and xA.

“Most people use statistics like a drunk man uses a lamppost,” Mark Twain once said. “More for support than illumination.”

And therein lies the issue with how people use not just these inferential stats but descriptive ones, like pass completion, key passes, interceptions, tackles won. It is to support an argument or a viewpoint, rather than seeing something new. Like the lamppost though, the stats will only light part of the picture. We figure out the rest with our other senses. 

On the wave spectrum, the 1% that we are able to see is the light section. Stats and the ‘eye test’ needs to be viewed in a similar vein.

While you're here, subscribe to FourFourTwo today and save 37% (opens in new tab). All the exclusive interviews, long reads, quizzes and more but with more than a third-off normal price.


RANKED! The 10 best centre-backs in the world

MANCHESTER UNITED VS ARSENAL Buffets brawling and the Busby Babes' last hurrah: The 12 best games between the pair

LIVE STREAMS How to watch every Premier League game this week, from anywhere in the world

Thank you for reading 5 articles this month* Join now for unlimited access

Enjoy your first month for just £1 / $1 / €1

*Read 5 free articles per month without a subscription

Join now for unlimited access

Try first month for just £1 / $1 / €1

Elijah Sofoluke