On a cloudy and cool autumn Monday morning following a physical draw with a conference rival, members of the Virginia Tech women’s soccer team gathered on the Moseley Practice Field for a training session. There, amid the collection of student-athletes performing light stretching and the coaching staff setting up cones for upcoming drills, David Tegarden and Jay Williams prepared for practice.

Accelerometer–check. Gyroscope–check. Heart monitor–check. Global positioning system (GPS) receiver–check.

Before practice began in earnest, each athlete walked over to Sam, an undergraduate working with Williams, and affixed a unit containing each of the four instruments to their custom vest, which is similar in design to that of a women’s sports bra. Each unit noninvasively records and transmits, in real-time, information to a receiver located adjacent to the playing field using two-way wireless encryption.

For Tegarden, an associate professor in the Department of Accounting and Information Systems, and Williams, a professor in the College of Agriculture and Life Sciences’ Department of Human Nutrition, Foods, and Exercise, this data collection is an invaluable step in their research goal of maximizing athletic performance while minimizing injury risk.

By using fractal analysis, a concept developed from business analytics that applies nontraditional mathematics to patterns – Tegarden and Williams are investigating whether a player’s training load leaves evidence, or markers, that are predictive of potential future injury.

“Fractal analysis is used in finance to determine stock trends or detect fraud,” Tegarden explained. “Injury can be seen as a type of fraud.”

Video analysis system data display
Video analysis system data display

Why women’s soccer?

“Female soccer players, and female athletes in general, suffer injuries at a higher rate than their male counterparts,” Williams explained. “Specifically, they are eight times more prone to knee injuries than male athletes.”

For both Tegarden and Williams, the experience of watching their respective children get injured playing the sport made the research especially personal. They began to think, “What if there was a way to help predict – and ultimately prevent–these all-to-common injuries?”

Training load, or the cumulative amount of stress placed on an individual from training sessions and competition over time, and its relationship to injury risk is not yet completely understood.

While not every injury can be predicted via data points and heart rate–think player-on-player collisions or a cleat getting hung up in the turf–there may be a correlation between the physical demands of training and competition and injury.

“There is evidence that fatigue and over-training can impact injuries,” Williams said.

Before each practice and competition, players are asked to rate how they feel physically on a 10-point scale. The players are then outfitted with their units, which tracks and transmits their heart rate, geo-location, acceleration, distance traveled, and gait throughout the session. After practice or competition, the players are then asked to rate their perceived level of exertion on a 10-point scale (10 being an extremely hard session, zero being a very light session).

The data gathered is imported into a database and then run through different programs, including one written by Tegarden. The information is then merged with a video of the session from the coaching staff, information from Virginia Tech Sports Medicine regarding any injuries sustained, and the perceptual ratings provided by the players.

“We built a video analysis system to tag videos so that we can then tie data to the video tags,” Tegarden said. “We combine the data with video to see if there are correlations between the information we are tracking and injuries sustained.”

Heat maps, line maps, and field maps are constructed to help visualize the data for the coaches and players.

Riley McCarthy is a defender for the Virginia Tech’s women’s soccer team, shown kicking a soccer ball.
FIELD WORK: Riley McCarthy is a defender for the Virginia Tech’s women’s soccer team. (Photo by Tomas Williamson)

What to do with all of that data

Despite the massive amounts of data collected, which, according to Tegarden and Williams, is well over 7 million records for a 90-minute game, the turnaround time for personalized data is as fast as Lionel Messi cutting in from the wing.

“Given the software we are using,” Tegarden said, “we can provide some personalized results to the coaches and players immediately after a game or practice.”

The speed at which the data is available is extremely important, as it allows the coaching staff to alter changes to training on a personalized level.

“For example, coaches are using distances run during a match to tailor recovery sessions for individual players the day after competition,” added Tegarden.

Williams stated that their analysis can pick up such changes as the manner in which a player runs over the course of a season. “We also see changes over the week when the team plays multiple games,” he said. “Those changes may be indicative of fatigue.”

These physical fluctuations, from the nearly indetectable, such as a change in gait, to the more obvious, such as a loss of acceleration, are analyzed to see if any markers can be picked up on that show when a player is at higher risk of injury.

“We’ve observed, unfortunately, ACL injuries during our research,” said Tegarden. “Because of this, we can also examine the trunk movement of the players at the time of the injuries. This may help us understand the biomechanics of how injuries occur.”

Assistance from the coaching staff and the players has been key to the success of their research.

“What has helped with the buy-in is the idea that we are not here to tell anyone what they should do, we are here to show them what they did do,” Williams said.

“I’m grateful to the coaches and staff for their assistance and accommodations,” Tegarden added. “The players are great to work with. They are interested in their data and appreciative of the work.”

After a match or training, the players want to look at their data to see how much they ran and how hard they worked, Tegarden said.

Aside from the direct benefit to women’s soccer, the research and analysis performed are providing vital hands-on experience to graduate and undergraduate students. Williams estimates that he has had research assistance from about 15 to 20 students, while Tegarden estimates that he has worked with well over 50 students.

“So much of what we do involves our undergraduate students,” Tegarden said. “They volunteer to collect data when we cannot. BIT students help with programming. They are very good at ‘prototyping’ ideas to see if it makes sense and can be used. The BIT students are learning a lot of programming and other skills in real-world scenarios. Their involvement is positive for us as well as them.”

“We’ve had dissertations and master’s theses based upon the work we are doing,” said Williams. “Our research is about as transdisciplinary as you can get. Our transdisciplinary relationship between teaching, research and athletics is quite unique. It’s truly a blend that promotes translational science.”

diagram of a soccer field showing tracks of players
A diagram of a soccer field showing tracks of players.

Future of their research

“I foresee a future where trainers utilize data in real-time during games,” Tegarden said. “We have already utilized data during rehab and training.”

Within the next year or two, researchers will have enough data to track specific players from the beginning of their career at Virginia Tech until they graduate, Tegarden said. That data is invaluable to see how an athlete’s performance improves or regresses over time-based upon their injury history–or lack thereof–and training regimen.

“We will have access to year-to-year analysis to see what has worked and what hasn’t,” Tegarden added.

Williams said that both he and Tegarden are developing courses around their research.

“We want to continue to merge this research with the academic side,” Williams said. “We want to give our students the ability to analyze real-world data.”

Tegarden explained that there is use for their research beyond the pitch.

“In accounting, we use control charts to find fraud,” he said. “We took the same idea, adapted it, and applied it to the playing field. With these insights, we can adapt it again and apply it to business.

“We are taking information from one field, literally, and applying it to another field.”

Share this story