SSA Blog: Home Surveillance
While the football authorities across Europe discuss how and when professional competition will return, the players have been trying to remain in the best condition possible. With return dates now confirmed in Germany and squads in England beginning to head back to the training ground, we take a look at what methods the pros have been using to train during lockdown, how they are being monitored by the clubs and whether we should believe some of the impressive stats we’ve seen on social media.
On the 13th March the Premier League was suspended due to the coronavirus pandemic. Currently, the Premier League’s plans, dubbed Project Restart, to get things going again are working towards a return on 12th June. While Championship club Blackburn Rovers opted to use it as a chance to give their players three weeks off, most clubs in the top-flight sought to find a way to continue training as best they could. This has typically fallen into one of two categories, group training (conducted virtually) or individual football, running or strength training.
Ross Barkley posted an impressive Strava session that set twitter alight
While group training, as we have seen with Liverpool’s Instagram Live yoga classes is great for fan engagement and, no doubt, team camaraderie, it is likely the individual sessions are where clubs are hoping for greater fitness benefits. Of these, bespoke running sessions will be of great value to the players and likely of most interest to coaches. Research from Barnes et al. (2014) shows that the Premier League is getting faster each year with players covering greater distances than ever before. During normal training, in an effort to ensure training prepares for this, clubs access session information via technological advances, such as the use of GPS systems including those produced by Catapult and STATSports.
It is unclear if all clubs have sent players home with the GPS units they would typically use for training though it is likely some have. These are expensive pieces of kit that require a level of expertise from staff to download and synthesise the data once a session has been completed. As such, it seems most are using easier to come by GPS devices such as Apple Watches, Fitbits or other wearables, and sharing on, or even recording sessions directly to, Strava. Strava outputs especially have been the source of a few amusing social media posts with Chelsea FC sharing Ross Barkley’s particularly impressive 5 km time. This was subsequently ridiculed by those with an eagle eye who may have spotted that Ross had rested, and paused his clock, a few times along the way… Without context, no one knows what the session was supposed to look like of course but in these lockdown days, the public certainly enjoyed getting at Barkley over it.
Barkley’s questionable 5 km time of 16:11 was fast, but by no means world class. Eluid Kipchogee, when he recorded 1:59:40 for the first ever sub two hour marathon, averaged almost two minutes per 5 km faster than Barkley’s effort. We know that professional footballers do not train to run marathons, or 5 km for that matter, so the performance teams at top clubs will analyse player training and match data to make sure that they are getting the best from each individual. Ensuring an understanding of performance data, how it is captured, used and the context within which it is meaningful helps players to improve. If harnessed by clubs, it can also become a valuable engagement tool. Fans are increasingly interested in player performance and training. Social platforms are the perfect environment to bring behind the scenes performances to life. However, when mishandled or presented in complicated terms data can be seen as misleading, undermining the message and causing embarrassment. In the Ross Barkley example, the post on Chelsea FC’s social media feed is unlikely to have passed by any of the performance team, or apparently anyone familiar with Strava and the difference between elapsed time and moving time. If it had, the club, and Barkley, could have avoided some embarrassment. Most fans though will be less concerned about a social media faux pas than an on-pitch capitulation and in order for coaching staff to optimise training adaptation they need confidence in their measures.
A conceivable difficulty around the use of technologies like Strava is the accuracy of the data it records. Coaches are interested in key running metrics such as total distance covered and how much of that distance is covered while running and sprinting. This helps to plan training to ensure players are doing enough to obtain an adaptive response but not too much, which could be problematic. The GPS units clubs use, collect a piece of GPS information ten times a second (we’d say the units have a sampling frequency of 10 Hz). The Strava App, and most common wearable running watches, are likely to sample at 1 Hz at best (Strava does not publicise specifics but the highest sample rate on most Garmin sports watches is 1 Hz). This means that whatever happens between these data collections must be estimated. Although this might not seem like it would make much difference Rampinini et al (2015) found that even a 5 Hz device showed 7.5 to 23.2 % error in the faster running speeds you would see during typical player training. The 10 Hz device they looked at had around half that error. These errors would result in players falling short or going beyond the aims of the session. No data currently exists to compare against the 1 Hz devices but you would expect the error to be magnified further. For those familiar with Strava, you’ll know that the key output metrics are total distance and pace, with specific high-speed running metrics not offered (though roughly calculable). It also does not provide metrics such as the number or intensity of accelerations or decelerations performed, which are also key to coaches. As a result, the more proactive teams will have sent their players home for lockdown with some pretty impressive GPS kit and will have been analysing their data via those tools. We have probably even seen some players recording to Strava whilst also recording their data with a Catapult or STATSport device.
As clubs begin to talk of a return to normal training, we have already heard of teams scheduling pre-season-style fitness tests to assess current fitness levels. It will be interesting to see what state the players return to their clubs from lockdown in. Who will have religiously followed the sessions and who might have been enjoying the lack of visibility to take a break from training. Furthermore, who might have rigidly followed training plans but may not be where they expect due to the use of less accurate technology. That is not to say that tools such as Strava do not play a part, far from it, players sharing their training via tools like Strava during the last few weeks have done more to engage fans in their fitness training than clubs have managed in the last few seasons. Whilst the data might not be overly useful for the performance teams, it has certainly been good for fans to learn more about their favourite players.
Sport Science Agency uses its insight and expertise to tell performance stories and unlocks their value for brands, broadcasters and rights holders. If you want to know more about what we can do for you, drop us a note via info@sportscienceagency.com and we can arrange to go for a healthy vitamin packed drink.