Using data to predict resilience in elite sports

For elite athletes, timely physical and mental recovery from stressors plays a key role in performing well. Psychologist and sports researcher Ruud den Hartigh from the University of Groningen, together with his team, is using data to provide insights into and predictions on the resilience of individual athletes.

‘Imagine: you took part in a tough competition, and you lost. You will be physically exhausted, but your motivation and self-confidence will have also taken a hit. You wanted to win, or maybe even had to win, but instead you lost. This also has a negative mental impact. The key is to then quickly return to your level, both physically and mentally, so that you can perform well again. This recovery process after a setback is called resilience.’ These are the words of Ruud den Hartigh, associate professor of Talent Development & Creativity at the University of Groningen.

Den Hartigh is the project leader of the research project ‘Resilient athletes: A multidisciplinary personalized approach’. In this research project, daily data on the physical but also the mental states of footballers from FC Groningen, PSV Eindhoven, and SBV Vitesse are gathered and analysed automatically. The aim: to map out the resilience of individual players and to predict when each player loses it. ‘This latter part is very important’, says Den Hartigh. ‘Once a player gets injured or has a mental dip, it is too late.’

The University Medical Center Groningen (UMCG), Leiden University, Osnabrück University, HAN University of Applied Sciences, the Dutch Olympic Committee*Dutch Sports Federation (NOC*NSF), and the Royal Dutch Football Association (KNVB) are also involved in the project, which is being led by the UG and in which the three football clubs are participating. The project began in 2019 and will run until the end of 2023.

‘The players at FC Groningen, PSV, and Vitesse all wear a vest equipped with a sensor during practice’, explains Den Hartigh. This sensor registers heart rate, for example, but also GPS data. So, it tracks the distances someone has covered and how long an athlete has performed within certain heart rate and speed ranges. The intensity of the practice is articulated through this.’

In addition, the athletes complete a questionnaire twice a day. ‘After practice, the players fill in how they performed, how much they enjoyed the practice, and how difficult they found it. Physical symptoms are also enquired after: whether the players experienced injuries, illness, or health complaints, and how far this hindered their practice. The next morning, before practice, the questions revolve around aspects that are ideally at a proper level again, like whether the players feel recovered, are motivated to perform optimally again, and have confidence in their ability to perform optimally. After all, after a difficult day, you hope that the players will be able to take it again the next day.’

Just like the data from the sensor, the results from the questionnaires are automatically sent to a secure database, after which the automatic daily analysis begins. Den Hartigh: ‘We look for patterns that indicate whether a player is likely to fall into problems. To find these patterns, we need a lot of data from each individual player. In this way, we can analyse whether a player responds differently to physical or mental stressors, or whether such a pattern already previously preceded a physical or mental issue. By detecting this, it is possible to intervene and break this pattern on time.’

How such a pattern should be broken, and thus what needs to be adjusted during practice to prevent someone from getting injured or feeling mentally unwell, is up to the coach, sports psychologist, and sports scientist of the respective football club. They receive a copy of the daily analysis of each player in the form of graphics depicting the physical and mental changes in the players.

Den Hartigh opens his laptop and presents the anonymized data of a player. He points to a graphic: ‘Things were going wrong for this player at this point, but we had already noticed earlier on that the player was becoming unstable. He had plenty of confidence one day, but hardly any the next. He performed well, and then badly. This varying pattern was a sign that things were going the wrong way for him. We are exploring this at the moment: identifying general indicators that signal when a problem will arise. You can also see these sort of fluctuations in research on people who experience depressive episodes. We might, therefore, also be able to find new indicators that are interesting for other populations as well.’

After all, just like elite athletes who want to prevent a mental dip, employees, for instance, also want to avoid a burnout. Is it conceivable that in future, we will also walk around our workplaces with a sensor and fill in questionnaires? That is not so far-fetched, in Den Hartigh’s view. ‘Our working method with footballers would also fit perfectly to the workplace. If you record employees’ data daily, you might be able to predict a burnout. If people do not want their data to be recorded, you could still learn from projects in sports. In a sports context, we learn about how people experience physical and mental stress and how they respond to it, as well as potential general indicators for problems. In addition, strategies have been developed within sports psychology and sports science to remain physically and mentally fit. For that reason, FC Groningen’s elite sports development manager Wouter Frencken and I hold lectures and masterclasses on this topic for a broad public.’