Five uses of performance analytics in sport, plus three you won’t believe
Elite sport is all about performance – of the individual and the team. By using insight from data, sports teams across the world are increasing their performance and delivering success, says author Bernard Marr
There are a lot of data points in football. In fact, cameras are now capturing 10 data points per minute per footballer in a match. In pre-analytics days, coaches used to wait until the beginning of the week for a report on the Saturday match. Now they are using the data to apply real-time analytics through machine-learning algorithms. This data includes how many calories a player has used at a certain time in the game and how many are still available to that player. If they are reaching the limit, then they will be substituted. Manchester City and Barcelona are among the clubs using such analytics. Others, such as Burnley and Portsmouth, use profiling tools to scout talent to ensure the most effective decisions are made with regards to likelihood of making the professional game. Currently, less than 1.4% of all youngsters recruited into elite teams make it into the professional game.
GB Rowing wanted to understand how to ‘make the boat go faster’. It brought together an array of data and combined it with data from competition. Working with Siemens it instrumented the boat with force sensors, angle sensors and accelerometry to get individual athlete data from the boats as well as overall data from the hull for how the boat’s moving as a whole during intense training sessions. The analysis looks at the angles the rowers are rowing through, stroke lengths, forces, power they’re putting onto the water and acceleration. Added to this is individual health and medical data from the rowers. Partnering with SAS it is able to generate much more indepth and speedy analysis of the rowers (previously such data was stored in multiple spreadsheets). Data analytics is also used to spot initial signs of injury, so the training regime can be tailored accordingly, enabling athletes to miss fewer sessions.
Data modelling enables coaches and managers to identify promising young rowers that have the potential to reach the top, as well as matching up rowers in different boat combinations to maximise the performance of every crew. According to GB Rowing: “What’s exciting are the golden nuggets or 'unknown unknowns' that can emerge – those things we didn't even foresee before we started analysing the data, so we can discover more about what factors and combinations of factors affect performance."
The US women’s cycling team had never won a major event. By collecting and analysing data, it was able to gain insights that helped it to cycle its way to silver at the London 2012 Olympics. It partnered with San Francisco‐based data analytics and visualisation specialist Datameer, whose spreadsheet approach easily integrated the different types, sizes and sources of data, making it much easier to extract insights and spot patterns in the data. Sources of data included track sensors, sleeping patterns, medical data and environmental data. Facebook and Twitter data were also used to discover the emotional state of cyclists. In one case, data showed a leading cyclist performed better if she slept at a lower temperature the night before a competition. So the team supplied a heat-controlled mattress.
Viewers of Wimbledon have long been able to wonder at the power of Hawkeye, which records whether a shot was in or out. In more recent years, we have been able to see visualisations of angles, swings and how players could have improved if they had stood more to the left or right. Now tennis players are going to be able to use smart racquets, so even more data will be available. But in 2015 a new real-time notifications system was used on the first day of The Championships. Lleyton Hewitt was about to hit the 1500th winner of his Wimbledon tennis career. The system pre-warned Wimbledon’s digital and content team of his impending milestone. It offered sufficient time to build a tile of richer content to share on social media as soon as the event occurred, breaking news faster than global media organisations.
Similar technology is now being used in the minor baseball league in the US, enabling social media and news organisations to cover some 13,000 matches a week that would otherwise remain uncovered in the media.
Rugby teams are using an app that asks the player 12 questions about their emotional wellbeing, body aches and so on before they even set off for training. The coach is then aware of that person’s physical and emotional condition when they arrive. Urine samples are used to understand hydration levels and players excluded from training if they are too dehydrated, as the likelihood of injury goes up. This is all about driving the right behaviour. You know you will be excluded if you are dehydrated, so you drink more water. At rugby team London Irish, trackers in the back of a player’s shirt can help coaches track a player across the pitch and see minute-by-minute performance data. Using such data combined with self-assessments enables the creation of an RPE figure, rate of perceived exertion. In training, if players are running more and their RPE is going down, they’re getting fitter, whereas if it is going up it could suggest long-term fatigue or overtraining.
If you look at elite teams across sports, the use of analytics is becoming the norm. It’s no surprise to see US NFL teams like the Seattle Seahawks tie up with Microsoft to create a data-driven training regime, but what about college football team the Minnesota Golden Gophers? Working with digital marketing agency FusionSports, the Gophers captured behind-the-scenes footage of practices, workouts, match day, the atmosphere at a match, campus experiences and more to give a real-life experience to potential recruits (as well as to fans) using Google cardboard glasses.
Or how about US Skiing and Snowboarding? Before the South Korean Olympics they brought the mountains to Utah in a virtual reality experience to enable athletes to learn the terrain, the turns, the position of gates and to mentally prepare for their competition. Virtual reality start up STRIVR uses data gained from a 360-degree camera on the helmet of one of the coaches, who then went down the course multiple times.
Then there’s curling. Eight of the 13 national curling federations who sent teams to the Winter Olympics in PyeongChang used smart brooms, or as some have named them Frankenbrooms. These enable curlers to discover information about body positioning, force, speed to sweep and so on. It appears no sport is immune.