MALVERN, Pa. — Penn State Great Valley celebrated the beginning of the spring sports season with an educational and networking event, “AI in Action: Transforming the Sports Industry” on March 24. The event brought together a panel of industry experts, including representatives from the Philadelphia Phillies and Philadelphia Union, to explore how artificial intelligence (AI), data analytics and cutting‑edge technology are changing sports, from elevating player performance to transforming the fan experience.
Amy Fisher, Great Valley’s student engagement coordinator, said she was inspired to organize the event after learning about numerous high-profile sporting events taking place in the Philadelphia area this spring, from March Madness and Major League Baseball All‑Star Week to the FIFA World Cup games and the PGA Championship.
“To make the theme even more meaningful for our students and strongly connected to their educational focus, incorporating artificial intelligence became the most logical and engaging direction,” Fisher said.
Raghu Sangwan, director of engineering programs and Great Valley’s Insights Lab, introduced the panel, moderated by Mike Gadsby, co-founder and chief innovation officer at O3XO, a Philadelphia-based AI consultancy. The panel featured five industry experts:
- Addison Hunsicker, senior manager of soccer analytics at the Philadelphia Union
- Jainil Kakka, data manager at Brown University; graduate of Great Valley’s master of data analytics program and co-author of a research poster on using machine learning for sports analytics
- Casey Thayer Luna, manager of software engineering at the Philadelphia Phillies and co-founder of Women in Sports Data Symposium & Hackathon
- Manujith Prakash, artificial intelligence/machine learning engineer at Hughes Network Systems, EchoStar; graduate of Great Valley’s master of data analytics program and former intern in marketing analytics at Major League Baseball
- Praneeth Sunkavalli, data scientist at MyRaina; graduate of Great Valley’s master of data analytics program and co-author of a research poster on using machine learning for sports analytics
Below are key highlights from the conversation, edited for length and clarity.
Gadsby: “Moneyball” came out a while ago, so we've been talking about data in sports for a long time. And now, AI has changed our decision-making by giving us deeper, faster insights into the data. Seeing AI in action and what it has done to transform organizations in a positive way has been really remarkable. So, how is AI changing sports and sports analytics?
AI helps analyze increasing amounts of data for enhanced insights
Luna: Statcast has led to an explosion of tracking data, which dovetails with advancements in machine learning. Data has continued to balloon — biomechanics data, junior league data and more. Scaling and operationalizing AI requires a lot of coordination between software engineers and data analysts as well as a lot of computing power to run robust models reliably.
The front office can get a big return on investment from AI. The machine learning and AI models that drive roster construction and player decisions can have a huge impact on the club’s success.
If you watch baseball, you might see the cards that the outfielders check to see where they’re supposed to stand. So, that's another example of how using AI to predict batter tendencies — and then providing the players with positioning guidance — can optimize their chances of getting the batter out.
Hunsicker: Most Major League Soccer clubs now have an analytics department, but that’s only happened in the past decade. At first, the big thing those analytics departments looked at was event data. Now, our data providers give us in-depth metrics derived from tracking data that we're able to use with the help of AI, which can really enhance our analysis.
The Philadelphia Union recruits from 45-50 leagues around the world, so we handle massive amounts of data and run it through our internal models for scouts and executives to analyze all the players they’re targeting. All of our player recruitment modeling is geared towards our style of play. We want the best players that fit our style of play, and AI has been great in helping us sift through all the metrics that are being developed by different data providers.
We’ve created an AI tool that allows scouts to type in queries such as, “Find me the fastest midfielders who rate well in ball-winning,” and the tool will sort through 4,000-plus midfielders and give them a list of players who fit those qualities.
Kakka: Hardware advancements allow us to process data faster, and more metrics are available now, due to advancements in computer vision. Earlier, large image and video files were difficult to process, but advancements in graphic cards and other data assets are helping research nowadays.
Prakash: AI makes a huge difference by transforming unstructured data — such as data collected in video format — into structured data that can be analyzed. It also enables analytical work to happen much faster, even during games, without the large gap that traditionally exists after a game ends.
Sunkavalli: AI can not only help analyze events but also try to simulate what might happen in the future. We can simulate matches in a virtual environment to train the models.
AI offers insights for non-technical audiences, from coaches and players to fans
Prakash: AI is helping fans better understand the game by making advanced insights more accessible. For example, MLB recently released MLB Scout Insights in its app, which provides fans with near‑real‑time stats and commentary. More broadly, AI makes a huge difference by helping technical teams generate analytical insights and clearly explain those insights to non‑technical audiences, including coaches, players and fans.
Kakka: Our professor, Dusan Ramljak, is a big advocate of prescriptive analytics — using data to help you make good decisions. Researchers and data analysts have to be on the same wavelength with coaches so that it makes sense to them how to use the data in real life. Translating these nuances into terms that coaches and players understand — that’s really important.
The limitations of AI in sports analytics
Kakka: Sometimes solutions are overkill. Just because we have fancy deep learning or computer vision models doesn’t mean we need to use them. First we have to understand the problem by talking to people on the field to see what’s going on.
Sunkavalli: AI doesn’t know everything, just what it’s trained on. It’s just a recommendation. We need to use our own judgment and experience to make final decisions. While AI can understand the trends and patterns, it lacks human insight. The best results come from combining AI's analysis with the wisdom and adaptability that only humans bring to the table.
Prakash: Developing production‑grade AI systems is actually very expensive, from design and development to scaling, operationalizing, monitoring, and long‑term maintenance. Because of this, it’s important for organizations to clearly define their long‑term AI strategies before embedding AI into the foundational layers of their technology stack.
Hunsicker: AI is only as good as the humans who harness it. You still need that domain knowledge of the sport to be able to get the best out of the data. It helps to have a relationship with scouts, coaches and players to understand what they are looking for in the data and how they can use it.
This event was made possible through funding from the AMETEK Diversity and STEM Program Endowment.