When the athletes dramatised in the 1981 Oscar-winning film Chariots of Fire were competing nearly a century ago, a stopwatch was one of the few devices producing data to measure their sporting achievement. But these days sport is all about measurement and analysing the data it produces. Whether it’s tracking your location, heart rate, oxygen saturation, or nutrition, huge amounts of information are being collected from athletes, including amateurs. But professionals in particular are finding that data collection and analysis could be what gives them the edge in competition.
Sports sensors sit where two of the biggest trends in contemporary computing meet: big data and the Internet of Things. The latter is a significant driver of the former. On the one hand, you need the connected devices that can track the relevant parameters to assess the factors behind sporting excellence and measure improvements. On the other hand, you need powerful data analytics to take the information produced, make sense of it, find trends, and help inform how athletes can do better.
Sales of running watches alone are growing five per cent every year, according to MarketWatch. These have now gone well beyond pedometer wristbands like the original Fitbit, and can include a GPS, heart rate monitor, and even a pulse oximeter to measure blood oxygen levels. They can also link wirelessly to cadence sensors in running shoes and on bikes, monitor your sleep patterns, and then automatically transfer the data collected to the internet. Some sports watches can use an accelerometer to detect which stroke you are using when swimming and when you push off at the beginning of a length, so the number of lengths you swim can be counted automatically.
Even consumer-grade systems can tell you useful things about your exercise ability that can guide how you train, such as VO2 Max, which measures the maximum amount of oxygen a person can utilise during intense exercise. This provides an assessment of cardiovascular fitness and can help you track your progress getting fit as you implement a distance-running programme. However, whilst the amateur trend for tracking exercise is driving device ubiquity and the sheer volume of data, professionals have access to systems that can provide much greater levels of detail, and with it a real edge in performance.
Data gets results in the beautiful game
For example, data is being used to improve football performance by analysing opposing team strategy and finding ways to combat it. Scottish football team Hearts used information from the InStat database to predict that a high-pressing game using players who could keep the pressure on for many kilometres of running would help them beat Celtic – and it worked. They are not alone, as more than 1,500 clubs and national teams use InStat, giving the company information on more than 400,000 players.
But clubs also build up data on their own players using sophisticated devices from companies such as Catapult Sports that can collect up to 1,000 data points per second. Similarly, the STATsports Apex can calculate more than 50 metrics at once, such as max speed, heart rate, step balance, high metabolic distance and dynamic stress load. This goes beyond the pure numbers but adds interpretation about how this is affecting an individual athlete. The data is collected in the cloud for historical comparative use. Teams now use this information to help decide which players to purchase to achieve their objectives for the season, employing services like InStat and Opta to provide the details they need.
InStat collects data for football, ice hockey, and basketball, whilst Opta includes these plus cricket, rugby union, baseball, golf, motorsport, tennis and handball amongst others. Although team games have many variables that can make player statistics only part of the picture, sports that focus on individual performance such as athletics can rely heavily on data to provide clear insights on how to aid improvement. This goes well beyond GPS tracking of outdoor events. Wearable devices with accelerometers, magnetometers and gyroscopes can track hundreds of data points to describe an athlete’s physical motion.
Stryd’s running shoe attachment can capture cadence (steps or cycles per minute) and ground contact. This can be used to analyse running style, which can be compared to previous sessions and other athletes. This information can spot nascent talent or help an athlete hone their style so they can emulate what makes the most successful sportspeople win. It can also detect warning factors like asymmetric movements that might cause a future injury or imply an impending one. EliteHRV’s sensors can provide high levels of detail on heart rate variability to see the physical effects of different levels of performance, so that athletes can recover adequately from their sessions and not over-train.
The secrets of the perfect golf swing
Another individual sport that is gaining considerable benefit from wearable sensors and data analytics is golf. Any sport using a bat, racquet or club can gain benefit from analysing a player’s swing, but in golf, the swing and body posture are particularly constrained, without also having to take into account additional factors like cross-court movement or ball spin, although atmospheric conditions will have an effect. Systems collecting golf performance data include Opta and ShotLink. The latter has results data dating back to 1983 and tracks 93 events a year.
GolfTEC, in contrast, is more focused on how an individual achieves their performance. The company has developed a SwingTRU Motion Study database of 13,000 pro and amateur golfers that includes information on 48 different body motions per swing. The analysis found six key areas that indicate an excellent player. These factors were discovered by correlating swing data with performance. Similarly, TrackMan uses cameras to analyse a player’s swing to aid training.
Rather than just analysing the past, predicting the future is where the application of big data analytics to sport will prove particularly valuable. As with every area of big data analytics, this will be dramatically affected by the application of AI and machine learning. Any massive store of unstructured data can potentially benefit from AI technology, which can help structure the data and find patterns in it proactively. First you feed in performance data, physical metrics during the activities, nutrition, sleep, atmospherics, plus anything else available. Then AI-empowered analytics will look for patterns that could provide strategies that make a difference, particularly when the margins for winning can be so small.
The systems we’ve discussed here are just the beginning. Data-driven sports are still only in their infancy, with much more to come in the next few years to help athletes find a competitive edge. Sport is just one area where big data and analytics are having a major influence, too. Healthcare, smart cities, and our understanding of the natural world are all seeing dramatic contributions.