Engineering a smarter game for Michigan’s soccer teams
A unique collaboration between the University of Michigan College of Engineering and its soccer teams is transforming the game, one data-driven decision at a time.
University of Michigan (U-M) Industrial and Operations Engineering (IOE) Assistant Professor Albert S. Berahas has only missed four U-M soccer games since joining the department in 2020.
Before becoming an engineer, he played semi-pro soccer in Greece. As he continued to play and his interest in data analytics grew, he realized that analyzing past games could help improve decision-making and performance. In 2022, while cheering on the U-M women’s soccer team alongside U-M IOE Associate Professor Raed Al Kontar – who played semi-pro soccer in Lebanon and also has expertise in data analytics and statistics – they saw the potential to merge both of their passions.
“We thought, the team must have a ton of data and we have a ton of students that have expertise in data analytics,” said Berahas. “This seems like a natural partnership.”
Using data analytics in sports isn’t new – sports analytics was popularized in the early 2000s, but the technology used to capture data has expanded and transformed immensely since then. Software like Catapult and Wyscout allow players to track their in-game health and performance.
Erin Kirkland/IOE
A partnership between athletics and engineering
Today their vision has come to fruition as U-M IOE works closely with the U-M men’s and women’s soccer teams through a project titled “WinAI: Propelling UM Soccer with Data-Driven AI.” This progressive, interdisciplinary venture aims to harness untapped data to enhance competitive strategy and player development, ultimately contributing to the U-M data science, AI and athletics ecosystems.
After the 2022 soccer game, former U-M IOE Department Chair Brian Denton connected Al Kontar with U-M IOE and women’s soccer alum Lauren Brideau, while U-M IOE Professor Amy M. Cohn connected Berahas with U-M IOE student and women’s soccer alum Claire Dawson, who joined the project in Fall 2023.
Al Kontar and Brideau, who tackled the first iteration as a part of her senior capstone project in the Winter of 2023, realized the expanded effort and resources needed to fuel the project. The U-M IOE team then strengthened their partnership with the U-M women’s soccer team through head coach Jennifer Klein.
“We are so excited to be partnering with U-M IOE on this project,” said Klein. “Analytics is becoming a major part of the soccer landscape. It is very helpful in analyzing our performances, assisting in player development, and helping to provide trends to inform our decision-making. The work Albert and Raed are doing with their team is helping us make sense of all of the data.”
Tackling the data
As stated in WinAI’s proposal, “Every pass, tackle and goal contains insights.”
The group seeks to understand this data, build predictive models and propel the teams forward with personalized, data-driven AI.
The engineers spent the 2023 season trying to understand the vast amounts of raw data that were previously viewed on Excel spreadsheets. They were in constant communication with the team, asking questions but also listening to their needs and curiosities.
“One of the most surprising and exciting parts of this project has been discovering how data can reveal insights that aren’t obvious on the surface,” said U-M IOE student and U-M men’s soccer alum Simon Vasquez who joined the project in Fall 2023. “like finding patterns in player performance or metrics that directly influence game outcomes.”
WinAI expanded to the U-M men’s soccer team in 2024, and the engineers are now presenting the data in easy-to-digest radar plots which allow the coaches and players to visually compare multiple variables of a data set simultaneously. Here, players can see their own – and their opponent’s – keys to the game, ranging from successful clearances to pressing duels. The WinAI team has customized the Wolverines’ plots by position and player.
“The real statistical analysis that we did was trying to figure out which factors in the game were the most important in terms of winning,” said Berahas. “Both for us and for the other team. For us, because we wanted to score goals. For them, because we wanted to stop them.”
In 2024 Berahas and Al Kontar won a Michigan Institute for Data & AI in Society (MIDAS) Propelling Original Data Science grant which has allowed them to expand student participation. The team of students and professors sent pre-game reports three days before each match. Between the men’s and women’s teams, there were close to 40 games in three months.
“It’s incredible to see how this project has grown from an engineering capstone to a game-changing tool,” said Brideau. “As a former player, it’s been rewarding to be able to continue to contribute to the program through the support of the U-M IOE department.”
Looking ahead
After a jam-packed season, the WinAI team is gearing up for the next one. They’ve grown to nine students and are currently preparing postseason reports for both teams. They are also focusing on building a private website, and Al Kontar is working with his students to write a research paper on their findings. Eventually, they plan to harness AI to deliver real-time insights and predictive intelligence, enhancing pre-game preparation and in-game strategy optimization.
“AI has come a long way, and it is only natural to harness its potential to advance soccer,” said Al Kontar. “I envision a future where reports are automatically generated, tailored to the specific requests of coaches, providing real-time insights that seamlessly enhance game strategies and performance.”
With their eyes on the future, the team is also taking the time to reflect.
“When I started working at the U-M IOE Department I didn’t know that I would be able to put together my passion for soccer and my love for IOE,” said Berahas. “The students have echoed this same feeling.”
At a recent meeting, Berahas went around the room and asked the students why they wanted to work on this project. Almost every student had the same response.
“I didn’t know that I could do something I love with the skills that I have.”
Students who have worked on WinAI
- Colin Badgero (current U-M IOE student)
- Lauren Bridaeu (U-M IOE and women’s soccer alum)
- Nicholas Bourgeois (U-M IOE and men’s soccer alum)
- Claire Dawson (U-M IOE and women’s soccer alum)
- David Lee (Princeton University ECE undergraduate student)
- Carly Newman (U-M IOE and WinAI alum)
- Selina Rey (U-M IOE undergraduate alum; current U-M ME master’s student)
- Nikhil Satpathy (current U-M IOE student)
- Simon Vasquez (U-M IOE and men’s soccer alum)
- Aubree Xia (current U-M IOE student)
- Megan Wu (current U-M IOE student)