GAINESVILLE, Fla. — The University of Florida men’s basketball team thrilled Gator Nation last weekend by winning three games in three days to claim the Southeastern Conference Tournament championship for the first time in more than a decade.
The Gators earned a No. 1 seed in the NCAA Tournament, and by virtue of resounding victories over Missouri, Alabama and Tennessee, they enter March Madness as the favorite to win it all in the eyes of many pundits.
A constant theme on ESPN telecasts of Florida’s games was how physically fresh the Gators looked late in contests. The Gators continued to run up and down the court, chase loose balls and grab rebounds as their opponents appeared to slow down.
That was the plan.
“It worked out like a charm,” said Victor Lopez, director of strength and conditioning for UF men’s basketball. “Our biggest metric we’re using right now is accumulated workload.”
Workload has been a buzzword in the NBA for a few years, primarily as a hot-take item when NBA stars take a night off for body maintenance. However, as sports analytics and artificial intelligence expand on a larger scale in college athletics, the term is increasingly used in college basketball.
UF coach Todd Golden has embraced modern statistical analysis since well before he took over the Gators, using analytics for everything from in-game decisions to roster construction and recruiting strategy.
In preparation for the SEC Tournament, Lopez reported workload data collected throughout the season to Golden and his staff to help prepare practice plans. The numbers informed Florida’s decision to turn up the intensity volume at practices to ensure optimum stamina should the Gators have to play three games in three days in Nashville.
“Really precise, so that our guys didn’t feel that workload for the first time,” Lopez said. “To make sure that we’ve been there before and we could operate at a high level.”
The men’s basketball team’s usage of analytical data is just one example of how the UF & Sport Collaborative, a $2.5 million initiative launched in January 2024, aims to optimize student-athlete performance through AI-powered decisions based on data collected through wearable sensors and monitoring devices.
The five-part initiative is a collaborative effort between the University Athletic Association and five UF colleges, and has a holistic approach that involves working with every Gators team on campus. The AI-Powered Athletics piece is a partnership between the UAA and the Wertheim College of Engineering that has grown since its inception to include research projects for coaches and data responses from student-athletes.
For more than a year, the alliance has worked to create a database of student-athlete information that players and coaches can use to maximize performance, help reduce risk of injury and an array of other potential uses.
“The working relationship between the University and the UAA on this initiative is natural considering our resources,” said UF Associate Athletics Director Dave Werner. “The collaboration seeks to produce tangible results that help our student-athletes and coaches achieve success and serve as a benchmark on a much larger scale to understand better how AI and sports analytics can benefit athletic performance.”
Mollie Brewer, a doctoral student at the Wertheim College of Engineering, has been on the frontline of the initiative since its early days. Brewer has a background in athletics as a cyclist and extensive sports science knowledge. She and fellow PH.D student Kevin Childs are traveling in April to present a research paper at the prestigious CHI (Computer-Human Interaction) Conference in Yokohama, Japan.
“This is our first big outcome,” she said.
In Japan, they will share insights from the first year of the initiative and offer a variety of discernments, from meaningful ways coaches engage with the data to designing tools that align with the workflows in high-performance sports environments.
“The first thing I was tasked with is in sports, we do something called a ‘needs analysis’ when we want to improve performance. I took that on,” Brewer said. “How are the coaches using technology? Let’s do an inventory. Let’s talk about how each sport is using the sport’s technologies. How are they being implemented? What do they like? What do they not like? What do they want out of them?
“Before we as a project decide on how to allocate our money and what our next steps are, we have to hear from the most important voices, those that are going to be implementing and using it and making decisions.”
Spencer Thomas, the UAA’s director of sports performance and analytics, and Jennifer Nichols, Ph.D. and a UF biomedical associate professor, work closely together to lead the AI-Powered Athletics study that now includes more than a dozen professionals. Celeste Wilkins is one of the most significant additions to the initiative since UF launched it.
Wilkins is a biomechanist and research scientist who came on board in July following a stint working to provide biomechanical and technological insights to the NBA’s Toronto Raptors. In Toronto, one of Wilkins’ tasks was to analyze tracking coordinates of 26 body parts during games, and then she would calculate analytics such as how many jumps a player took in a game, how they landed, and their joint angles to learn more about body mechanics.
She came to UF to help make sense of the massive data collection over the past 14 months.
“I’m the boots-on-the-ground doer,” Wilkins said. “The aim of the project is to create a data bank for both research purposes and for the UAA to leverage with advanced analytics. My role is coordinating all the data. We make some dashboards for the teams. We’ll look at how they can better use the data they’re already collecting and see if we can ensure the athletes are getting the best experience.”
While the use of analytics in sports has grown from infancy to adolescence over the past decade, the AI component remains a relatively new dynamic. UF is at the forefront in that area with HiPerGator, the most potent university-owned supercomputer in the nation. The AI-Powered Athletics group is using HiPerGator to store the data collected, analyze it and produce valuable information that coaches and players can use in real-world situations, such as the way the men’s basketball team created intense workload practices to prepare for the potential of three games in a 50-hour span.
“The bridge between academia and athletics can be so resourceful,” Thomas said. “Not everyone has these resources. It enables us to do much more to uncover different technologies, data points and systems to improve performance.”
Nichols is one of three engineering leads in the group, and her primary task is allocating resources. She said the collaborative effort has sped up the process and made important strides in the availability and interpretation of the data.
“We’re bringing it all together,” Nichols said. “And as you might expect, they want different things from the data. I mean, the musculoskeletal system of a gymnast is probably going to be different than that of a sprinter. It’s going to be different for your center than your point guard.”
Meanwhile, through conversations with coaches, student-athletes and members of the sports science community, the UF & Sport Collaborative seeks to be a leader in managing privacy concerns with so much data available. How much data is too much? How much do student-athletes want to know? How much do coaches want to share? What impact does the data have on the mental well-being of student-athletes already under immense public scrutiny and pressure to perform?
Thomas said the collaborators discuss those questions regularly when they meet each Monday with a shared vision.
“We’re here for student-athletes,” Thomas said. “We’re actually helping student-athlete development by reducing injury and optimizing performances. That’s the main goal: student-athlete wellness. It’s a study hopefully to pave the path for understanding technology for our athletic departments more precisely.”
Brewer specializes in human connection to technology, which is one of the reasons she jumped at the opportunity to join the group after working in professional cycling.
“I’m in human-centered computing,” she said. “I’m on the human side of technology. I’m trying to bridge the gap between how it is being designed and how it is being used. It takes someone who can speak both languages. I can get in with the coaches and the staff and earn their trust, but I also understand what the computer science community thinks and how they design technologies.”
Wilkins, a veteran research scientist who has studied subjects ranging from NBA players to horse-rider coordination during equestrian events, considers emerging privacy concerns one of the top priorities of AI in sports analytics.
“Something really important to us is how we’re handling the data for the safety and security of really valuable information,” she said. “Hopefully, by the time we’re done with this project, we will be able to create a standard code of practice for how teams and individuals deal with student-athlete data.”
What is clear for now is that the data age is in full bloom.
There is more information than ever in sports, and people use it in many ways, from injury prevention to devising practice schedules. The UF & Sport Collaborative has staked its place on the front lines of exploring this new world and designing systems to collect, interpret, and secure data.
AI can be a valuable teammate in the process. Still, humans are the most important players.
“One of our findings is that [coaches and support staff] are still manually doing a ton of work to make sure these technologies work for their athletes,” Brewer said. “Despite the promise of automated systems, they are still manually connecting the dots. It’s not being applied for them. They are active users. Technology is not replacing coaching.”
The Gators men’s basketball team’s performance in Nashville is proof.
Lopez accessed the data. Golden and his staff interpreted it and devised a plan as the Gators prepared for the potential of three games in three days. They won all three and returned home with an SEC Tournament championship trophy.
“Our approach to every game is extremely methodical,” Lopez said.