Insight

Big data

The metrics, measurements and match stats that could take football to the next level

WORDS Simon Hart | ILLUSTRATION Neil Stevens

As visions of the future go, Paul Balsom has one that must sound pretty enticing for any young footballer. As head of performance for the Sweden national team and Leicester City (he also works with Belgian side – and Leicester City sister club –  OH Leuven), he envisages his players being able to step into an analysis hub and gain immediate access to personal performance data from past matches. Using a virtual-reality station, they would be able to relive situations in a specific position on the field. Or, in the case of a youth-team player, step into the shoes of an elite performer and – via the magic of VR – see how to respond in certain scenarios. A case of “What is he looking at and what do I need to look at?” as Balsom puts it.

Yet there’s nothing virtual about how soon this could become a reality, according to Balsom. “What we are most interested in is players being able to access their own data,” he says. “Even now we’re trying to find a way to give them more interaction and accessibility – and we think voice recognition is going to be the way forward.” Step aside, Sir Alex: Alexa’s arrived.

This was a vision that Balsom presented at recent Training Ground Guru conference Big Data: The Future of Football? It offered some fascinating examples of how top clubs are already crunching the numbers. For example, coaches have access to an iPad on the bench and in-game data is available (after a brief time delay). Cameras capturing 25 frames per second chart each player’s position; coaching staff can use the technology to look at precisely how deep their team are defending, among other things. One Premier League analyst says that, as things stand, it is common for such observations to be passed on at half-time – but the dawning of live feedback can’t be too far away.

The tracking of data already keeps coaches up to date on players’ physical output, which informs decisions regarding substitutions and post-match recovery work; medical iPads, meanwhile, monitor players’ heart rates. Mikhail Zhilkin, a data scientist at Arsenal, told the Big Data conference that he provided a bar chart for players showing their energy expenditure, with information about  the corresponding amount of food to put on their plates after training.

Big data’s tentacles reach far and wide within football clubs. Manchester City employed their first full-time set-piece specialist in the summer, then added a set-piece analyst.
By

According to Tom Cooper, a performance and match analyst with UEFA, another significant step forward is the type of analyst now being employed by clubs. “More and more are data scientists,” he says. “It’s not enough to have a video and provide stats: you have to interpret the stats. So you’re seeing more and more people working as analysts who are physicists and nuclear engineers.” Opta’s team, for example, includes people with PhDs in chemistry and biology.

Metrics must now measure quality, not quantity. A new one coming into vogue is possession value, as Cooper explains. “This tells you the likelihood of a certain pass leading to a goal compared with another pass. It takes into account historical data based on that location and the other players’ positioning to try, in real time, to say, ‘If you pass here you have a better likelihood of scoring than if you pass there.’ It also helps to assign credit to players who are contributing to possession but may not be recognised by more generic stats. It can tell you the percentage chance of success for a whole possession.”

Big data’s tentacles reach far and wide within football clubs. Manchester City employed their first full-time set-piece specialist in the summer, then added a set-piece analyst. Michael Edwards, sporting director at Liverpool, was initially hired as the European champions’ head analyst; now he oversees recruitment. On the latter, Cooper says that data helps clubs build positional profiles of potential signings. “Now you can analyse 1,000 games’ worth of data in an hour, rather than watch 1,000 hours of video,” he says. “If a club know the type of attributes they want for a certain position, data can easily highlight players that fit their profile.”

However, for all the virtual valleys of information now available, simplicity is important. For example, Luke Benstead, head analyst for the Belgium national team, took a less-is-more approach when creating his own focused set of metrics to make the data more meaningful for Roberto Martínez and his coaching staff. Moreover, the consensus among analysts seems to be that data must be accompanied by video footage in order for coaches to fully embrace the message.

Balsom stresses that players must be beneficiaries too: rather than looking at metrics focused on linear movement, they must be able to see their “quality of movement”. He cites the example of Leipzig playmaker Emil Forsberg’s display for Sweden against Germany at the 2018 World Cup, whereby he spent just 44 seconds on the ball. Yet, says Balsom, “he’s coming back, he’s being part of the line, he’s blocking passes. There’s so much going on without the ball and we want to give him credit.” And preferably, we might add, in a human voice – rather than that of Alexa.

Read the full story
This is an extract from an article in issue 02 of Champions Journal. Get your copy to read the full story.
Get your copy
Insight

Big data

The metrics, measurements and match stats that could take football to the next level

WORDS Simon Hart | ILLUSTRATION Neil Stevens

As visions of the future go, Paul Balsom has one that must sound pretty enticing for any young footballer. As head of performance for the Sweden national team and Leicester City (he also works with Belgian side – and Leicester City sister club –  OH Leuven), he envisages his players being able to step into an analysis hub and gain immediate access to personal performance data from past matches. Using a virtual-reality station, they would be able to relive situations in a specific position on the field. Or, in the case of a youth-team player, step into the shoes of an elite performer and – via the magic of VR – see how to respond in certain scenarios. A case of “What is he looking at and what do I need to look at?” as Balsom puts it.

Yet there’s nothing virtual about how soon this could become a reality, according to Balsom. “What we are most interested in is players being able to access their own data,” he says. “Even now we’re trying to find a way to give them more interaction and accessibility – and we think voice recognition is going to be the way forward.” Step aside, Sir Alex: Alexa’s arrived.

This was a vision that Balsom presented at recent Training Ground Guru conference Big Data: The Future of Football? It offered some fascinating examples of how top clubs are already crunching the numbers. For example, coaches have access to an iPad on the bench and in-game data is available (after a brief time delay). Cameras capturing 25 frames per second chart each player’s position; coaching staff can use the technology to look at precisely how deep their team are defending, among other things. One Premier League analyst says that, as things stand, it is common for such observations to be passed on at half-time – but the dawning of live feedback can’t be too far away.

The tracking of data already keeps coaches up to date on players’ physical output, which informs decisions regarding substitutions and post-match recovery work; medical iPads, meanwhile, monitor players’ heart rates. Mikhail Zhilkin, a data scientist at Arsenal, told the Big Data conference that he provided a bar chart for players showing their energy expenditure, with information about  the corresponding amount of food to put on their plates after training.

Read the full story
Sign up now to get access to this and every premium feature on Champions Journal. You will also get access to member-only competitions and offers. And you get all of that completely free!
Big data’s tentacles reach far and wide within football clubs. Manchester City employed their first full-time set-piece specialist in the summer, then added a set-piece analyst.
By

According to Tom Cooper, a performance and match analyst with UEFA, another significant step forward is the type of analyst now being employed by clubs. “More and more are data scientists,” he says. “It’s not enough to have a video and provide stats: you have to interpret the stats. So you’re seeing more and more people working as analysts who are physicists and nuclear engineers.” Opta’s team, for example, includes people with PhDs in chemistry and biology.

Metrics must now measure quality, not quantity. A new one coming into vogue is possession value, as Cooper explains. “This tells you the likelihood of a certain pass leading to a goal compared with another pass. It takes into account historical data based on that location and the other players’ positioning to try, in real time, to say, ‘If you pass here you have a better likelihood of scoring than if you pass there.’ It also helps to assign credit to players who are contributing to possession but may not be recognised by more generic stats. It can tell you the percentage chance of success for a whole possession.”

Big data’s tentacles reach far and wide within football clubs. Manchester City employed their first full-time set-piece specialist in the summer, then added a set-piece analyst. Michael Edwards, sporting director at Liverpool, was initially hired as the European champions’ head analyst; now he oversees recruitment. On the latter, Cooper says that data helps clubs build positional profiles of potential signings. “Now you can analyse 1,000 games’ worth of data in an hour, rather than watch 1,000 hours of video,” he says. “If a club know the type of attributes they want for a certain position, data can easily highlight players that fit their profile.”

However, for all the virtual valleys of information now available, simplicity is important. For example, Luke Benstead, head analyst for the Belgium national team, took a less-is-more approach when creating his own focused set of metrics to make the data more meaningful for Roberto Martínez and his coaching staff. Moreover, the consensus among analysts seems to be that data must be accompanied by video footage in order for coaches to fully embrace the message.

Balsom stresses that players must be beneficiaries too: rather than looking at metrics focused on linear movement, they must be able to see their “quality of movement”. He cites the example of Leipzig playmaker Emil Forsberg’s display for Sweden against Germany at the 2018 World Cup, whereby he spent just 44 seconds on the ball. Yet, says Balsom, “he’s coming back, he’s being part of the line, he’s blocking passes. There’s so much going on without the ball and we want to give him credit.” And preferably, we might add, in a human voice – rather than that of Alexa.

Insight

Big data

The metrics, measurements and match stats that could take football to the next level

WORDS Simon Hart | ILLUSTRATION Neil Stevens

As visions of the future go, Paul Balsom has one that must sound pretty enticing for any young footballer. As head of performance for the Sweden national team and Leicester City (he also works with Belgian side – and Leicester City sister club –  OH Leuven), he envisages his players being able to step into an analysis hub and gain immediate access to personal performance data from past matches. Using a virtual-reality station, they would be able to relive situations in a specific position on the field. Or, in the case of a youth-team player, step into the shoes of an elite performer and – via the magic of VR – see how to respond in certain scenarios. A case of “What is he looking at and what do I need to look at?” as Balsom puts it.

Yet there’s nothing virtual about how soon this could become a reality, according to Balsom. “What we are most interested in is players being able to access their own data,” he says. “Even now we’re trying to find a way to give them more interaction and accessibility – and we think voice recognition is going to be the way forward.” Step aside, Sir Alex: Alexa’s arrived.

This was a vision that Balsom presented at recent Training Ground Guru conference Big Data: The Future of Football? It offered some fascinating examples of how top clubs are already crunching the numbers. For example, coaches have access to an iPad on the bench and in-game data is available (after a brief time delay). Cameras capturing 25 frames per second chart each player’s position; coaching staff can use the technology to look at precisely how deep their team are defending, among other things. One Premier League analyst says that, as things stand, it is common for such observations to be passed on at half-time – but the dawning of live feedback can’t be too far away.

The tracking of data already keeps coaches up to date on players’ physical output, which informs decisions regarding substitutions and post-match recovery work; medical iPads, meanwhile, monitor players’ heart rates. Mikhail Zhilkin, a data scientist at Arsenal, told the Big Data conference that he provided a bar chart for players showing their energy expenditure, with information about  the corresponding amount of food to put on their plates after training.

Big data’s tentacles reach far and wide within football clubs. Manchester City employed their first full-time set-piece specialist in the summer, then added a set-piece analyst.
By

According to Tom Cooper, a performance and match analyst with UEFA, another significant step forward is the type of analyst now being employed by clubs. “More and more are data scientists,” he says. “It’s not enough to have a video and provide stats: you have to interpret the stats. So you’re seeing more and more people working as analysts who are physicists and nuclear engineers.” Opta’s team, for example, includes people with PhDs in chemistry and biology.

Metrics must now measure quality, not quantity. A new one coming into vogue is possession value, as Cooper explains. “This tells you the likelihood of a certain pass leading to a goal compared with another pass. It takes into account historical data based on that location and the other players’ positioning to try, in real time, to say, ‘If you pass here you have a better likelihood of scoring than if you pass there.’ It also helps to assign credit to players who are contributing to possession but may not be recognised by more generic stats. It can tell you the percentage chance of success for a whole possession.”

Big data’s tentacles reach far and wide within football clubs. Manchester City employed their first full-time set-piece specialist in the summer, then added a set-piece analyst. Michael Edwards, sporting director at Liverpool, was initially hired as the European champions’ head analyst; now he oversees recruitment. On the latter, Cooper says that data helps clubs build positional profiles of potential signings. “Now you can analyse 1,000 games’ worth of data in an hour, rather than watch 1,000 hours of video,” he says. “If a club know the type of attributes they want for a certain position, data can easily highlight players that fit their profile.”

However, for all the virtual valleys of information now available, simplicity is important. For example, Luke Benstead, head analyst for the Belgium national team, took a less-is-more approach when creating his own focused set of metrics to make the data more meaningful for Roberto Martínez and his coaching staff. Moreover, the consensus among analysts seems to be that data must be accompanied by video footage in order for coaches to fully embrace the message.

Balsom stresses that players must be beneficiaries too: rather than looking at metrics focused on linear movement, they must be able to see their “quality of movement”. He cites the example of Leipzig playmaker Emil Forsberg’s display for Sweden against Germany at the 2018 World Cup, whereby he spent just 44 seconds on the ball. Yet, says Balsom, “he’s coming back, he’s being part of the line, he’s blocking passes. There’s so much going on without the ball and we want to give him credit.” And preferably, we might add, in a human voice – rather than that of Alexa.

It's your time to decide
Choose which classic final goal you would like to see in Issue 03 of Champions Journal.