When Leicester City won the English Premier League in 2016 the rest of the world looked on in disbelief.
The odds of them winning the competition at the start of that season were 5,000-1. Bookmakers reckoned that Elvis Presley being found alive or the discovery of the Loch Ness Monster was more likely to happen. But it was accumulated data that became the secret weapon that helped them to achieve the seemingly impossible, a Dubai summit on artificial intelligence heard on Tuesday.
Paul Balsom, head of performance innovation at the club, revealed how they used data to monitor the positioning of players up to 25 times a second, with each game producing hundreds of millions of unique data points.
“We were able to measure the volume, frequency and the intensity of the training for each player,” said Dr Balsom.
“It meant we were not only able to optimise their physical performances but we could drastically reduce the risk of injuries.
“We could tell if they were training at too high or low a level.”
Dr Balsom, who is also the performance manager for the Swedish Football Association, said that Leicester’s famous achievement would never have been possible without the advent of player tracking data, which helped level the playing field.
His comments came on the sidelines of the AI Everything Summit taking place in Dubai on Tuesday and Wednesday.
The world of sports analytics has been moving so fast that Leicester have not had time to rest on their laurels.
Dr Balsom revealed that the club has appointed its first ever head of analytics this week.
“He’s joined us to bring all the data into one platform,” he said.
“There are millions of data points being generated each week and if it is not all in one place then it is difficult to manage.
“Now we will be able to interrogate all the data for trends and patterns.”
It is a far cry from the days when the only statistics available to football clubs were how many shots they had or how many tackles they had made in a game.
Dr Balsom added that AI is only going to have a bigger role in the world of sport in the future.
“If we’re analysing opponents, we may want to take a look through the last five or 10 games,” he said.
“That is hours of work for an analyst but AI could do that job in seconds.”
However, he believes the human factor will never become obsolete in football, despite the increased of big data and AI.
“Let’s say a player is looking tired in the middle of a game, there are many possible reasons for this,” he said.
“Before the manager considers substituting them they have to look at other factors such as if it would have a negative impact on the rest of the team.
“They could be a star player who has the ability to score a late goal despite being tired.
“There will always be a lot of information available but ultimately these decisions will always come down to a manager or coach.”
Updated: May 1, 2019 04:53 PM