Hello. It looks like you�re using an ad blocker that may prevent our website from working properly. To receive the best Tortoise experience possible, please make sure any blockers are switched off and refresh the page.

If you have any questions or need help, let us know at memberhelp@tortoisemedia.com

Lies, damned lies and statistics

Lies, damned lies and statistics


Pep Guardiola thinks football is becoming too obsessed with statistics. Is he right?

If you enjoy working with statistics, there are so many ways you can be employed by the football industry. 

A quick search revealed numerous job titles: data analyst, methodology analyst, data scientist, data engineer, I could go on.

Many clubs have entire departments dedicated to crunching the numbers. 

Make no mistake, statistical analysis of football – and footballers – is here to stay.

But amongst all these numbers, have we lost sight of the game itself?

Pep Guardiola thinks we might. 

“Statistics are just a pattern in formation that we have. But there are players that make the game or the team play good when they are not in the statistics. But the players just go how many goals I score or how many assists or how many da da da da…this kind of situation and then forget everything. This is a problem to the real world and there are footballers that are just my statistics, my statistics…never exist statistics before.”

Football Daily

But is he right?

Let’s not forget that his club – Manchester City – is very keen on stats too. 

It employs Laurie Shaw – a former astrophysicist – as the lead Artificial Intelligence scientist across its  group of teams.

The stats trend largely comes from American sports. In the US, they are widely and routinely used. To analyse performance, to sign new players, and that’s been the case for a long time.

Some think that the increase in American owners in football has led to their rise here.

So how are they used in football?

“First versus second squared off in the Premier League when Pep Guardiola’s Manchester City hosted Thomas Tuchel’s Chelsea. Tuchel’s men desperately needed the win to claw back the ten point gap against City. But in the end they couldn’t manage it, with the game ending one-nil to City thanks to an outstanding De Bruyne goal. The xG indicated it was just about a deserved win, 0.9 to 0.6. But the xG time map shows that Chelsea only really had one big chance.”

xG means expected goals. It’s used to assess the predicted number of goals a team will score..which depends on event-based variables. 

Here’s an explanation.

“Now the headline is that expected goals basically measures the quality of shots that a team takes. Now you can create a simple expected goals model just using the location of the shot, but actually, sophisticated expected goals models take a bunch of different factors into account. Whether the shot came from a foot or somebody’s head, whether a player went round the keeper, whether it was a cutback or a through ball that led to the shot, the speed of the attack that led to the shot…”

Euro Football Daily

Stats companies can also measure post-shot expected goals, expected assists and PPDA… that stands for opponent passes per defensive action. 

In simpler terms, if your team makes a “defensive action” to win the ball back, it measures how many times your opponent can make a pass straight afterwards.

If a team plays with an intense pressing style, this figure would be lower. In this case your aim is to win the ball back and play it forward by passing to a team-mate.

And there’s a new metric – xT – expected threat. It’s defined as “the chance of scoring during the current possession based on where the ball is now.”

Whether you love this kind of analysis or not, we need to focus on one part of what Pep Guardiola  was saying, that there are some things in football that can’t be measured with stats.

To understand why football stats don’t always give the full story, let’s look at a widely available statistical report of Tottenham’s Eric Dier from their FA Cup defeat to Middlesbrough.

He had three shots on goal – two of which were on target – three tackles, five clearances, completed 86.6 percent of 82 passes and made four accurate long balls.

“So it’s Eric Dier’s turn. Dier. Dipping! And beaten away by Lumley. Good effort from Eric Dier.”

Tottenham Hostspur

Even though his team lost, it sounds like a good personal contribution, right?

But Eric Dier was one of the players most at fault for the Middlesbrough goal in extra time. 

He failed to properly mark Josh Coburn in the box, who was then able to make a neat one-two before firing in the decisive goal.

Good positioning and tactical awareness are two things that can’t easily be measured by statistics.

So while Pep Guardiola might make good use of analytics in his day-to-day work, it’s worth listening when he says some things in football “are not in the statistics.” 

The numbers are here to stay, but football is an art as much as a science.

Today’s story was written by Chloe Beresford, and produced by Imy Harper.