Leeds United follow Liverpool and turn to a 'perfect partner' to address 'criminal' pain point

Leeds United will harness artificial intelligence to try to reduce the number of injuries to Jesse Marsch’s players.
Watch more of our videos on Shots! 
and live on Freeview channel 276
Visit Shots! now

Just a few weeks after the end of a season that was ravaged by player unavailability and muscle tears, the Whites have struck up a partnership with Zone7, an ‘AI-driven human performance platform’, already used by Liverpool, Glasgow Rangers and a host of European clubs.

“As we prepare for a third season in the Premier League, we are consciously improving and moving towards a more holistic approach to fitness that benefits our players’ recovery between fixtures,” said director of football Victor Orta

Hide Ad
Hide Ad

“Zone7 is the perfect partner to help us harmonize the data we’ve been collecting, optimize player performance, and lower injury incident rates.”

The club’s head of medicine and performance, Rob Price, will spearhead Leeds’ use of the web-based platform, plugging in all the information accrued by the various bits of technology already in existence at Thorp Arch to track player performance.

Zone7 will then analyse the data and alert Price and his team to risks for individual players, with the aim of cutting down the number of days lost to injuries, which analyst Ben Dinnery calculated at a division-worst 1,542 last season.

Rich Buchanan, performance director at Zone7, hopes that, with their help, Leeds will record far better numbers next season.

Hide Ad
Hide Ad

A qualified physio, sports scientist and UEFA A licence coach, Buchanan has spent 15 years working with Swansea City, the Premier League itself and the Football Association of Wales.

INJURY CRISIS - Kalvin Phillips was one of three Leeds United players to undergo surgery for torn hamstrings last season. Pic: GettyINJURY CRISIS - Kalvin Phillips was one of three Leeds United players to undergo surgery for torn hamstrings last season. Pic: Getty
INJURY CRISIS - Kalvin Phillips was one of three Leeds United players to undergo surgery for torn hamstrings last season. Pic: Getty

He brings football-environment knowledge to a firm founded by Tal Brown and Eyal Eliakim, two men with backgrounds in data science, cyber security and machine learning.

“If you think about what exists in a typical football club, you have a tech stack of data-generating hardware, like the GPS devices that you often see the players wearing and then, on top of that, you often have digital filing cabinets where this data gets stored, and it’s left then to human practitioners to try and make sense of it all, which is really difficult to do consistently well, particularly as there’s so much data available on players now,” he told The YEP.

“What we’re doing is we’re applying AI or machine learning or deep learning processes to this to try and do it in a much more accurate and time efficient manner.

Hide Ad
Hide Ad

“The evolution of the game means now that I guess data is at the forefront of the arms race to be successful. And it’s not about making players fitter, faster, stronger anymore because they are optimised athletes. It’s about making sense of how we manage them more effectively.”

Zone7 already works with the NBA, NFL teams, baseball teams and rugby clubs - sports that are all embracing the idea of injury risk management. Football has historically had a blind spot in that area according to Buchanan.

“There’s a historical pain point in football that’s just kind of accepted that injuries happen,” he said.

“But there was a massive research project by some really highly respected researchers that often do stuff with UEFA, which showed that, in football, muscle-related injuries hadn’t changed in terms of frequency from 2001 up to 2019, which is criminal really. The game has probably got fitter, faster, stronger as the players have got fitter, faster, stronger but the actual injury rates have flatlined.”

Injuries are costly for clubs like Leeds on two fronts.

Hide Ad
Hide Ad

“Certain amounts of days lost have been shown to equate to losses in league position,” said Buchanan.

“We all know that when we’re missing our best players, our results will suffer. And arguably that may have been the case for Leeds last year, but there is some research that backs that up.

“Another vertical as to why we do what we do is that every injury a Premier League team suffers has a cost implication and again, there’s research that shows that, on average, in the Premier League, each injury is about £200k, which is an astronomical amount of money.

“That’s why Zone7 exists. It’s about forecasting injury risk to the practitioners at the club, to give them as much kind of data-derived insight as possible, to help them manage their players in a better way.”

Hide Ad
Hide Ad

Zone7 allows clubs to examine amounts of data that Buchanan says would be “otherwise humanly impossible” because for every player, every day, GPS data generates vast quantities of information.

Applying machine-learning enables patterns to be identified in data the machine has seen before, associated with injuries. Zone7 has the benefit of 200 million hours’ worth of training data in its database.

“It’s a bit like getting older and wiser - the more good data Zone7’s algorithms have seen, the more wisdom it’s got so we can see these patterns in our database being matched up with perhaps Player A at Leeds, therefore we flag him as a player at risk,” said Buchanan.

“We’re not saying he will definitely get injured; it’s probabilistic based on everything we know. We feel there’s a strong possibility that that might happen. How then Rob and his multidisciplinary team of staff act upon that is entirely up to them.”

Hide Ad
Hide Ad

Whether it works at Leeds remains to be seen, but Buchanan says it has worked elsewhere.

“Liverpool were obviously one of the early adopters, very innovative as a football club and as a club that utilises data science,” he said.

“Much like Rob at Leeds, there’s a high calibre of practitioner there, working with the players day to day and, obviously, the challenges for them are very similar.

“We worked with Rangers the [title-winning] season before and continue to work with them, as well as with Hull City during their League One title-winning season and again last season. We work with clubs in Germany, Spain, the MLS, Italy, Turkey, Holland, various different leagues around the world and that’s how we’ve been able to create this beautiful database, essentially.”

How it will work at Leeds is a call for Price to make.

Hide Ad
Hide Ad

“He’s obviously somebody who has been in the game for a long time, extremely knowledgeable in his field, but also highly, what I would call responsibly innovative,” said Buchanan.

“He understands what machine learning, deep learning, artificial intelligence can offer. He understands it’s not the be all and end all but he understands that there’s value to be had when applied appropriately.

“It’s a web-based programme so he can log in as and when he feels, he can access it via his phone if he wants to. Essentially they upload their data, we automatically ingest it and analyse it and then we push out things back to Rob and his team and they determine how they want to use it.

“It’s a tool in their armoury essentially; it still requires the human practitioner to take that information and apply it as they deem appropriate. They’re really smart individuals and we hope to, in some small way, contribute to their success.

Hide Ad
Hide Ad

“The fundamental success metric is that Leeds United’s injuries reduce next season, compared to the season that they’ve just had.”

Leeds fans will argue that next season couldn't possibly be worse than the last one, on the injury front, but any measures that help ensure that is the case will be as good as any business the Whites do this summer.

Comment Guidelines

National World encourages reader discussion on our stories. User feedback, insights and back-and-forth exchanges add a rich layer of context to reporting. Please review our Community Guidelines before commenting.