A numbers game? How data science is changing football scouting

Combining data and judgement is a balancing act

Most clubs use data analytics alongside the judgements of scouts and coaches. That's because it’s relatively simple to collect data such as pass completion or successful defensive positioning, but much harder to analyse a player’s leadership capabilities or professionalism. A centre-back who has lost some of their pace, for instance, might be a crucial influence in the dressing room. And data is backwards-looking: young players develop very rapidly – and sometimes for the worse.

But analytics are a way for clubs to cut through the noise. “We’re advisers, not the final decision-makers,” says Graham. “We give clubs a tool to make an informed decision.” It’s a balance between the efficiency of the data analytics and the human instinct of experienced coaches.

David Hicks agrees. Clubs often use Eyeball's data to put together a shortlist of players to watch at a particular tournament, he says, so that scouts don’t have to go to endless matches in the hope of spotting an unknown talent. “Data is just data,” says Hicks. “It’s up to the club how they interpret it – they can click on our videos to watch the player for themselves.”

Getting it right means doing the groundwork

There’s no point investing in data analytics unless the club is prepared to commit to it. For that, it needs to do two things: invest in in-house data facilities so the club can work with the advice and analysis it receives, and build cultural acceptance of the science throughout the organisation.

“We always recommend that clubs have their own data teams,” says Graham. “You need to be able to ingest and store and organise the data, and to translate that data into insights and reports that inform your work.”

According to Hicks, managers who are under pressure to produce short-term results might be reluctant to pick the young players that the data has suggested the club buys. “It’s a bit of a cliché, but will the manager give young players a chance?” says Hicks. “Does the club have a good academy, or B team or partner clubs, where players can develop?”

What's next: Predicting stardom from age 12

The influence of data on football and other sports continues to grow as technology and innovation accelerate. Hicks says that advances include DNA analysis that can help clubs to predict the top speed of their players, and cognitive testing that measures players’ decision-making speed. And AI is helping Eyeball to evaluate all of its analysis so far to identify which of its recommendations succeeded and which failed – and why.

“We will soon be sitting on the largest database of youth footballers in the world,” says Hicks. “The holy grail will be the ability to look at 12-year-old players and predict with a degree of accuracy which ones will make it.”