If you don’t know where you’re going, where do you begin? With analytics, a first step is data. Data scientists often find themselves in this situation when they are assigned exploratory research tasks. For Davidson classmate Kendall Thomas ’19 and I, our internship last summer with Athlete Intelligence began this way.
Athlete Intelligence is sports technology and data analytics company headquartered in Kirkland, Washington. The company makes wearable devices for athletes, such as mouth guards and helmet sensors, that track head impacts. Their unique user platform empowers coaches and athletic trainers to access useful insights from this data to help reduce the risk of injury and improve performance.
The company provided us with reams of data collected from thousands of high school and college football players across the country. For each impact, Athlete Intelligence devices record the corresponding player, impact location on the head, impact magnitude, and time.
However, Kendall and I didn’t receive many explicit instructions regarding what we should do with this information. Our goals for the summer were to simply “make sense of the data” and “find interesting insights.” A recently created and rapidly growing company, Athlete Intelligence had not had the opportunity to analyze their data yet. Their hope was that we could find correlations that they could then display on their user platform. As CEO Jesse Harper said, “This is a mission to Mars. We’ll know what we find when we find it.”
Armed with data, we turned to our research goals. First, find coachable moments, actionable insights which aid coaches. As games progress and players tire, we found that linemen tend to drop their heads, leading to more hits on the crown of the helmet – a dynamic that can lead to neck and spinal injuries. If a coach verified fatigue’s influence, the team could emphasize conditioning and remind players to keep their heads up late in games.
Our second goal was to enrich the data. We wanted to add information that could lead to additional insights. After exploring different resources online, we augmented the data with weather, elevation, location, the player’s position, and type of event (game or practice). Insight followed. For example, we discovered that quarterbacks take harder hits in practice than they do during games. Also, while most football players tend to get hit in the front of the head, quarterbacks and special teams players are hit more frequently on the back of the head – suggesting that players at those positions should have more protection built into the backs of their helmets.
To conclude our summer, we visited the company’s office in Kirkland. We presented our findings and discussed how it could enrich the user platform. Our work met their goals and would help the company.
While we found numerous coachable moments that could impact player performance and health on the field, our research also provided Athlete Intelligence with new ideas that could improve their business. Some of our visualizations motivated them to create similar ones on their user platform. The relevance of the additional information we added to their dataset encouraged alterations to their devices to track this data automatically. Furthermore, ideas like the padded helmet for quarterbacks inspired the thought of developing additional equipment in the future.
During a time when concussions continue to plague the sport and the NFL has yet to make any significant progress to change that, Athlete Intelligence is helping to lead the charge. By creating devices that track head impacts in real time, a coach, trainer, or parent can monitor a player’s health throughout a game. With their platform, Athlete Intelligence provides pertinent information to maximize long-term player potential. Their ambitious commitment to save the sport of football is quickly gaining momentum. In time, players will be confident to step on the field, and their families will watch without concern.