Malvern, Pa. — Two graduate students in Penn State Great Valley’s data analytics program, Praneeth Sunkavalli and Jainil Kakka, won second place for their research poster at a recent symposium hosted by the Penn State Institute for Computational and Data Sciences.
For their research project, they used machine learning to analyze event data from soccer games to measure the success rates of a defensive tactic called "pressing," when players pressure their opponents in an attempt to regain the ball.
Sunkavalli and Kakka combined their love of soccer, especially the Arsenal Football Club in England, with their expertise in analyzing data, they said.
“It’s a mix of passion and profession,” Kakka said, explaining why he and Sunkavalli chose to research this topic. They noticed plenty of sports analytics research on offensive tactics aimed at scoring goals, but few studies of defensive tactics like pressing. “Measuring pressing success with metrics like press recoveries and pressure success rates is essential for understanding and improving defensive performance,” they wrote.
The two examined a range of information from soccer games, such as player positioning and movement patterns, to quantify the intensity of pressing strategies and predict success rates in critical situations.
“This data empowers coaches to make informed decisions, while analysts can identify key moments that influence match outcomes,” they wrote. They identified several implications of their research for training players in pressing skills and optimizing pressing tactics during soccer matches. Their analysis can also be applied to other sports, they said.
Sunkavalli and Kakka thanked their mentors for this project, Penn State Great Valley’s assistant professor of software engineering Dusan Ramljak and associate professor of information science Satish M. Srinivasan, as well as Penn State Scranton’s assistant professor of business Nonna Sorokina. “Your guidance and support have been instrumental in shaping not just this project, but our development as researchers,” Sunkavalli said.
Reflecting on the award from the Penn State Institute for Computational and Data Sciences, Sunkavalli said, “This recognition marks an incredible milestone in our research careers.”
He and Kakka have presented their research at two other recent data science conferences, and they have submitted a paper to be considered for publication in a sports analytics journal.
“Having our work recognized at such a prestigious event gives us the confidence to aim even higher,” Sunkavalli said.