Arsenal FC’s Tom Allen addressed the IDTechEx event in Berlin, explaining how technology is used to improve player performance.
Speaking at IDTechEx in Berlin earlier this year, which SciTech Europa attended, Tom Allen, Lead Sports Scientist at Arsenal Football Club, discussed how the club is using technology to improve player performance, and also provided his views on what new wearable technologies need to offer for them to be used.
He began by explaining that the club collects a lot of data; essentially, he said. “If you think we can measure it, then we probably have done.”
Arsenal’s first team alone contains almost 30 players, and data is collected on all of their activities, including match days, training, and even their time in the gym. And, for Allen, the main challenge here is “finding the signal” in this mass of information.
He explained that the Arsenal manager approached him recently with the task of enhancing player performance and reducing injury, and that this needed to be done in such a way that the manager was not given multiple pages of data which he would then need to go through in order to glean the relevant information. As such, Allen said, he looked to other sectors, including banking and retail, for methodologies which could be used for sport.
“I settled on a way that bankers look at the stock markets,” he explained, which meant he was able to then list the Arsenal players, colour-coding them as red, orange, or green depending on their risk of injury.
Telling a story with data
However, this did not meet all the requirements of his manager, so, Allen said, he went away and took this further. “Everyone loves to see a story,” he said. “They want to see why we are saying what we are saying.” And he knew he needed to transform the story he needed to tell into a single page narrative that the manager could view and digest with ease.
He went on to explain that there is a copious amount of data gathered for each player at Arsenal, much of which is GPS data, and this information needs to be used in order to enhance performance and reduce the player’s risk of injury.
“In sport, research suggests that stress is linked to performance and injury – more stress equals poorer performance and more injury. We also need to look at how we can implement this information into a more condensed form; I can’t put all of this data into a single algorithm, and so I need to have a way of breaking this information,” he explained.
To do this, Allen detailed the five key areas he used: skeletal (“how the system responds when they are doing high speed running or distance running); metabolic (“looking at acceleration and deceleration,”); cardiovascular (“heart rate analysis so how they are responding during training”); muscular (“looking at how the muscles respond,”); and wellness.
These areas are measured against the level of stress being experienced by the player, with Allen defining the most stressful situations for a football player as being match day, where they can be playing a game in front of thousands of people. The players work with the highest intensity during this time, he said. And “the higher the intensity, the more stress.”
The data that Allen gathers is used in a longitudinal study that allows him to see over time how stressed an individual has been.
Of course, this information still needs to be implemented into a single page ‘story’ for the manager and, Allen said, the simplest way to do this was to create a graph, which is made up of two key measurements. The first of these is what Allen termed ‘load’. This is essentially the amount of work given to an individual, and much of the data that is collected is from GPS, and is tracked by Allen and his team over time.
He went on: “The element of the graph that is statistically significant is where it shows that a player is under more stress than they have been previous exposed to. Now, that isn’t bad, but it is bad but if it is prolonged.”
This amount is stress is therefore kept below a certain level, although it is necessary to include it to some degree because in the long run it will enable the player in question to increase their fitness and to tolerate more stress moving forwards.
Allen then explained that alongside that, it is necessary to measure how the players are responding, and this is made up of 35 different measurements, including a psychological and a physical analysis. The players, he said, are therefore tested each day before they start training, and the data from this is all brought together into a more coherent whole.
Using a specific, though anonymous, player as an example, Allen explained that he has experienced issue with his ankle, meaning that he had been unable to train as much as normal and, as such, the available data demonstrates a reduction in his training load, while his wellness also decreased as a result of this relative inactivity, with the player also reporting more soreness.
Once the player had received some rehabilitation, his training load was increased, and the data that Allen showed to is audience illustrated how, as a result of this, he went into a “state of overload”. “If we put these two things together,” Allen continued, “we see the wellness drop considerably as we prolong his overload. The more stress you get the worse you are going to feel. So the player is reporting soreness and fatigue…and he is also complaining of a lack of sleep – we play a lot of night games, so his sleep has been affected.
This information, and the subsequent graph, was then recreated for each player in the Arsenal squad. Allen also created an overview for the team’s manager.
He also further grouped the players into categories for the coming week: those who need to be “protected”, those whose workload needs to be maintained, and, finally, those players whose workload needs to be increased for the next seven days. He said: “Regarding the players in the protect group…as soon as a player shows that an increase in training load equals poor wellness, I would be stupid to push this player; I would increase the risk of injury. I therefore take him out of that and let him adapt, and then we can hit him again later on.”
Regarding the maintain group, Allen explained that these individuals have recently experienced some overload, so have been removed from that, but the level they are now at is to be maintained.
Finally, he continued, the players in the last group are those whose wellness and overload levels have been measured and who are now deemed “fresh” enough for their workload to be increased.
Taking another anonymous individual as an example, this time a centre back, Allen explained that the five key areas he currently outlined (skeletal, metabolic, cardiovascular, muscular, and wellness) are measured on a scale of one-to-five, and when this particular player’s data is analysed, Allen is able to see that the skeletal and cardiovascular sections need attention, meaning that he would go on to use “high speed runs” and to “create a drill that is based on 40m, and the player has to reach that in 65 seconds. And because he is also lacking in the cardiovascular area, the ratio of exercises will be reduced – he will be allowed larger rests in between.”
Discussing the data, Allen explained: “This is all designed to make better decisions and have a sound rationale for what we are doing,” adding that this has not always been the case in football, but that now, the availability of more robust data is making that happen. “We have all the data we want to make actual decisions,” he said.
Allen then turned his attention to the way in which wearable technologies are changing the way this data gathering is approached. He said: “In sport, we need to add to that story; we have a story that we try to tell through our signals, and every team will have a different one. For me, any kind of wearable technology has to have evidence based on it,” and while this evidence doesn’t have to be on the product itself, he added, the evidence of what the new technology can achieve has to be there, and, essentially, any new technology has to “improve on what is already out there – make it more reliable and more accurate, and provide efficient feedback.”
Allen also informed his audience in Berlin of some of the things he looks for when evaluating new technologies. He said: “Most important… is player focus; these are our assets, and if the player isn’t going to wear it, then we won’t be using it.
He added: “If it is between two companies, then we are going to look for the most bang for our buck. If there is one that can provide us with reasonable information that we can add to the story and it is quick, then that will win over another which can give us really good info but takes a day to get it….We need to be efficient as possible, and we are playing a game every two-to-three days.”
Concluding his presentation, Allen explained that Arsenal are always looking out for new technologies that can help them enhance their performance, and he mentioned the potential of innovations such as graphene tattoos, which have been developed as a non-invasive way to monitor a person’s health . “That may add something we don’t already have, thus adding to the story,” he said.
His final point in Berlin was that no amount of data, no amount of technologically innovative equipment, will be an answer in and for itself, as it is “always the person interpreting that data” who will be making the decisions, and so any new technologies need to be developed as an enabler, helping practitioners to “find their story”.
This article will appear in SciTech Europa Quarterly issue 27, which will be published in June, 2018.