In collegiate athletics, the numbers matter, but not just the ones on the scoreboard. Behind the scenes, injury surveillance data plays a critical role in keeping athletes healthy and programs accountable. The Datalys Center, a national leader in sports injury research, manages an expansive injury surveillance program that tracks participation across hundreds of NCAA institutions and sports.
While universities’ data involving their participation in the injury surveillance system is collected and stored, the challenge lies in the ability to explore, compare, and act on the data quickly. Institutional data was locked inside restricted spreadsheets that required manual digging and expertise to interpret and determine what is allowable to share. This created a barrier for researchers, administrators, and decision-makers who need timely insights to improve athlete safety and program performance.
The Datalys center needed a tool or application which could automatically transform participation data into a dynamic, user-friendly interface. This would seamlessly put reports and insights right at the fingertips of any stakeholder.
How We Solved the Problem
The Datalys Center tasked us with creating an executable, portable, and reusable dashboard that could filter the data by variables like sport, division, conference, or school and could present visualizations instantly without the need for manual reporting or changing of the underlying code.
Before we started developing the application, we had to clean the data. This involved creating additional synthetic data and wrangling it so it was in an effective format. We also transformed the data into two efficient tables so filters can be rapidly executed when called upon. To ensure our transformations were sound, we created a comprehensive data dictionary for Datalys which defines each column, possible values, and relationships.
To create the dashboard, we used Plotly and Dash – two tools available in Python. Dash creates the framework for a dynamic dashboard that can be filtered and interacted with by users. We used a feature called callback functions in Dash to allow graphs to update based on user selected values such as if the division filter was selected to “1” (this would only include visualization for Division 1 schools) or if the conference filter selected to “Big 10” (this would only include visualizations for schools in the Big 10 conference). Plotly is a library which is used to create interactive visualizations, and we utilized this for creating customized bar charts, line charts, heat maps, and pie charts. A few of these you can see in our image below. Additional statistics were shown alongside the visualization, such as current season participation rates expressed as a percentage, for easy references.
The Value We Created
If conference executives, for example, want to know which schools in their conference haven’t submitted injury data this year, they would need to ask Datalys for a manually created report. That takes time on both party’s ends. With our dashboard application, conference executives can immediately answer these questions and many others such as “which sports have the highest participation rates?” or “is men’s injury surveillance participation higher than women’s?” There is no longer a need to interact with, and wait on, additional parties to help answer these questions – the data has moved from the stakeholders’ inbox to their fingertips.
“The data has moved from the stakeholders’ inbox to their fingertips”
It takes a lot of time to create custom reporting and consistently update relevant parties. Program administrators, researchers, and oversight bodies will now get to spend more of their day discussing insights from the data rather than hunting down and waiting on the relevant data to arrive. The same applies for Datalys researchers who can now spend more time doing research and finding new ways to deliver valuable insights to stakeholders.
These newly unlocked opportunities are what make a project like this so special. There are no data and dashboard developer experts dedicated to the Datalys Center, so creating this automated reporting system may not have been feasible given their workforce. Once introduced to the benefits this product yields, the doors will open to further data projects which will allow Datalys to deliver more value to stakeholders as higher quality insights can be generated efficiently.
Our dashboard can easily be modified to include geographic data and actual injury data in the future. Similar to allowing program managers to observe their respective conferences or schools, we can let politicians and policy makers observe their state’s or district’s participation data. Instead of just seeing participation rates, which is the only data available currently, if injury data was integrated we can immediately see what injuries come from certain schools, sports, and geographic areas. These additional insights could allow for greater transparency and accountability for schools, and also aid in making data accessible for epidemiological research of injuries both for professionals and those with vested interests in athlete health.
#SportsInnovation #SportsAnalytics #Indy4Sports
Leave a Reply