With the 2015 Rugby World Cup around the corner, South Africans have little faith in the national team to bring back the title because of their consistently poor performance this season. Having lost their past three championship games, including one against Argentina on home soil, the team has been slammed for not playing a full game of rugby, as they seem to lose steam in the last 15 minutes, all but handing the win to their opponents. Was it possible to predict the Springboks’ poor performance using advanced analytics? I think so.
While most sports teams use analytics to some degree, usually by analysing metrics like player performance and vitals when deciding which players to draft, how much to pay them, and which players to put in a game, the majority do not use advanced analytics, which can help coaches and managers make informed decisions around team selection optimisation, injury prevention and revenue generation.
All sports – rugby, football, cricket, basketball and everything in-between – experience similar challenges when it comes to data analytics. At a roundtable on sports analytics earlier this year, former Springboks coach Nick Mallett said: “As a coach, you follow the ball a lot and by necessity you are not looking at your team and the opposition. The detail and information from technology is absolutely vital. It’s not just territory or possession or the number of tackles or carries or line outs; there are always statistics that explain why you lost.”… and there are always statistics that explain why the Springboks drop the ball (no pun intended) in the last critical minutes of a game.
Challenge 1: Backwards-looking analytics
Teams have access to many different sources of data – video footage that tracks every hit, pitch and fielding move, GPS devices that capture movements of players and balls on the field, wearable health gadgets that measure players’ biometrics like heart rate, and data from broadcast channels – which can quickly become overwhelming if they don’t have the skills and technology to effectively analyse the information. Teams need to be able to manage and aggregate data so that they can quickly create highly accurate predictive and descriptive models to determine what is likely to happen in the future and make informed decisions in real time.
Traditionally, teams measured player performance using historical information along with subjective player ratings and video analysis. But this approach is backwards-looking and meant that decisions were made based on what had happened in the past, rather than on how that information could improve performance. For example, coaches could use player data to devise personalised training regimes for individual team members focusing on their weak areas. They could use the information to establish what line-ups work best in certain situations, and what the issues are in terms of play calls and training. This type of forward-looking analysis, which is available in real time on their mobile devices, enables coaches to make immediate, game-impacting decisions. By exploring the data, coaches can determine what will happen if, for example, he kept player X on the field rather than replacing him with player Y, which was his original plan.
Challenge 2: Player health and injury
Since a major component of every team’s success involves the health of its players, prevention of injury and illness is a natural focus for the use of data and analytics. Advanced analytics can help coaches and teams to recognise when a player’s safety is at risk, and to predict conditions and situations that contribute to injury. For example, using video data, coaches can assess the level of physical activity and any patterns of movement or stresses that could lead to injury. Data can also be used to gauge the fitness level of each player for contract decisions, and to predict likelihood of serious injury. Players may soon start wearing devices that measure concussive force to the head, which may affect strategies and gameplay and prevent future injuries.
Football club AC Milan started using analytics in 2003 to prevent injury and illness. In that year, it experienced a 90% reduction in injuries compared to the five previous years, and they have remained low since then. By determining the risk of injury and low performance in specific conditions, coaches can keep players safe and ensure they consistently perform at high levels. Advanced analytics have changed our understanding of sports. Teams no longer just compete on the field or on the court, but now also in the use of data and analytics. But the only way to achieve competitive advantage from data is through innovation in application and execution, and through buy-in from leadership and players.
Analytics will never replace human intuition in sports. Rather, it will help coaches to become better at their jobs so that they can help players improve. Key is the ability to easily understand data analysis, as is clear communication of the findings and buy-in from leadership and players.