By Dr. Liz Williams, SAP Customer Co-Innovation and Strategic Projects.
The Digital Athlete
Athlete monitoring capabilities are evolving at a rapid pace due to the explosion of wearable sensor technology. It is now possible to monitor practically every facet of an athlete’s physical performance from the sidelines, in real time. We are in the midst of a digital transformation where coaches, athletes and sports medicine practitioners are virtually swimming in data. Recently, five leading experts from across the sports spectrum were invited to partake in a panel discussion as part of the SAP’s Spotlight Tour in New York City. The panelists shared their experiences and insights about wearable devices, the role of IoT, the benefits and pitfalls of the digital athlete and the role of SAP HANA.
Never miss the latest sports analytics news with the SportTechie newsletter.
For the panel, Kurt Kendall; Global Head of Consumer Engagement from Under Armour was joined by Kevin Abrams; Assistant General Manager of the New York Giants, Jack Swarbrick Jr.; Vice President and Director of Athletics at Notre Dame, Kathleen Stroia; Senior Vice President, Sport Sciences and Medicine at the Women’s Tennis Association and Dr. Bill Seringer; Physical Therapist and Clinical Supervisor, Outpatient Orthopedics and Sports Medicine, Stanford Health. SAP North America’s Chief Innovation Officer, dedicated cyclist Martin Mrugal moderated a stimulating and informative discussion and here is a recap.
Improving Athletic Performance
There is no denying that the use of wearable technology is revolutionizing the nature of sport and exercise science. During a football game, coaches and sports physicians on the side lines can view real-time sensor data from every athlete. They can monitor heart rate, temperature, measure distance covered, maximum speed and the number of tackles made.
This data helps them monitor all aspects of performance, signs of fatigue and improves their situational awareness. This enables more informed tactical decision making such as when to remove an athlete from the field.
The benefits of real-time athlete monitoring are also abundantly clear in the world of women’s tennis. Kathleen Stroia is dedicated to giving players maximum performance capability. In conjunction with video analysis, wearables data is enabling a new level of analytical tools. “We can objectively measure differences in stroke mechanics as the match progresses, using data right there on the court”.
The experts agreed that a holistic solution using advanced analytics is necessary in order to derive meaningful conclusions about athletic performance, health and injury prevention. The consensus was that the insights drawn from real-time sensor data have the greatest value when considered in conjunction with background information. An athlete’s performance on a given day is what they and their team are judged on, but it is only the tip of the iceberg. Their performance in competition is a result of their preparation; their training load in the weeks leading up to the match and over their career. Their fatigue level which is affected by health and injury status, nutrition, recovery and psychological state plays a significant role. Then there is tactics, skills and the list goes on.
“Sleep is by far the most important factor” said Swarbrick. “Our athletes even have customized pillows”.
The real game-changer comes in the form of understanding the complex interactions of physical, behavioural and environmental factors which affect performance. Swarbrick summed it up perfectly. “You can capture all the data during a game, but if you don’t know that the athlete had the flu the night before, then that data is meaningless. You need to understand the connectivity between the raw data and the subjective data. We have to understand that or we are drawing false conclusions from it.”
“The Giants are very good at collecting data” said Abrams, “but the real insights stem from good analytics. The question is what do we do with the data to make it meaningful and how do we make that easier”?
The holistic approach is also relevant for recruitment. The process of selecting the best athletes for particular sports involves identifying applicants with favorable attributes for a given role.
Both Abrams and Swarbrick agreed that the process of determining what those relevant attributes are, or combinations of attributes can be challenging. An individual with a phenomenal natural fitness level or superior strength may not possess the mental toughness to excel in the demanding world of professional American Football. If you had one college scholarship to offer and two promising 17 year old quarterbacks, what are the best predictors of how they will perform in five years’ time? A big set of longitudinal data collected from elite and average football players may help.
“You can see what doesn’t work, and what is successful,” added Kendall. “That way you can look at somebody who is similar and leverage that information. You can provide that individual with very personalized guidelines of how they can be successful.”
Perhaps a challenge just as great as the analytics is presenting the data to end users in a digestible form. “Most people don’t have a PhD in hard science,” offered Kendall. “Our job is to present the data in a form they can use”. Bill Seringer added “I think it’s a matter of can we get the information to the person who needs it and can we do it in a way that is very simple for the user – that’s key.”
Injury Prevention
Dr. Seringer provided a clinician’s perspective on the prevention and management of sports injuries. “The data we can now collect has the potential to be a big game changer for injury prediction and injury prevention, but how do we get the right data to the right people and how do we connect it?”
He made the vital statement that physiotherapists are great at collecting data, “but if you don’t get the information that’s important to make the decisions that you need to make, then why are you collecting the data?” We have so much information going into our electronic health records that we have no idea what to do with it.”
Advanced bioinformatics and analytics has provided invaluable insights for medical research. There have been cases where factors which were not thought to have any bearing on particular illnesses have been shown by big data analytics to be key indicators. The same could be applied to sports science.
Bill Seringer is a great advocate for this:
“We know some major risk factors are for ACL tears (a common knee injury in many sports), we know other things which may play a role, but you know what, let’s combine all that information. Let’s look to the future and figure out if we can predict those injuries. If we have a big dataset combining information on all measurable factors, advanced analytics of that data could make a big, big difference.”
Inevitably, Dr. Seringer made mention to the most controversial issue in sports; concussion and long term brain damage. Arguably this would benefit most from big data sets and advanced analytics. He stressed the importance of collecting the medical history of high school athletes, particularly those involved in high impact sports (such as football, hockey and boxing).
The effects of head injuries are cumulative, but what remains unknown is what short term and long term head impact force thresholds result in neurological impairment. Also unknown is why some former football players suffer from the long term degenerative condition known as CTE, and others with a similar playing history do not.
“We’d love to get that concussion information and we’d like to know more about the medical history of high school athletes” said Swarbrick, “but we can’t get hold of that data. People with access to that information won’t or can’t share it.”
Swarbrick is all about athlete development. “There are performance reasons to engage in this (big data analytics), but it’s much more about health and safety,” he later said. “The discussion that’s going on today about concussions and long term consequences to joints… I want to make sure we are at the forefront of that. The experience of having been a student athlete at Notre Dame, I don’t want that to have physical consequences for you down the road. I want it to be a positive which has impacted your lifestyle and made you healthier for the next 40 years of your life, not created challenges for you.”
With regard to the monitoring of head impacts in American Football, Dr. Seringer replied “the only way to do this is to create a massive database that everyone has access to.”
This is where things get very complicated.
Pitfalls and Ethical Dilemmas
Concerns were raised around protecting athlete privacy, ensuring the security of very personal information and whether making all aspects of performance measurable was necessarily a good thing.
The security of personal information is a problem not restricted to athletes. Teams may be asking athletes to put a high degree of trust in their data infrastructure and assurances that their data will be secure. Kevin Abrams is concerned that the collection and usage of athlete data is handled very carefully. Kathleen Storia was in full support of this:
“Some of the reasons that we are not moving forward as fast as we could in that area is because of privacy concerns,” she said. “We need to make sure that data is secure. We are concerned about the person first, then the athlete”. At the suggestion of a nation-wide head injury monitoring system for American Football player safety, the privacy concerns were paramount.
Ensuring athlete’s privacy is also a major factor in determining what constitutes a ‘measurable’ quantity for athletic performance and what crosses the line. While quantitative data from devices like sleep monitors are an invaluable addition to athlete performance management, this may be considered too invasive by many individuals.
“I hope sport doesn’t become too scientific,” Swarbrick concluded. “We want to make it safer and keep athletes healthier, but we want kids to be kids. We want them to have the experience on the field and figure it out for themselves.”