Brian Macdonald and Jun Yan
The 6th Connecticut Sports Analytics Symposium, formerly known as the UConn Sports Analytics Symposium, was held April 11–12 at Yale University, drawing about 150 registrants. Co-organized by the Yale Institute for Foundations of Data Science and the Connecticut Statistical Data Science Lab, the rebranded symposium expanded its reach beyond the University of Connecticut while maintaining its mission of engaging students from high school to graduate school in sports analytics and data science. With an emphasis on accessibility, the event offered a program designed to showcase the interdisciplinary effect of sports analytics while fostering collaboration between students, academia, and the sports industry.
Keynotes and Panel Discussion
The symposium opened with a one-on-one keynote conversation with Theo Epstein, senior adviser for Fenway Sports Group and three-time World Series champion baseball executive. Epstein shared his career path from Yale graduate to general manager of the Boston Red Sox and president of baseball operations for the Chicago Cubs, describing how data influenced him at each stage and how his approach to data-informed decision-making evolved over time.
The afternoon keynote was delivered by Lauren Poe, sports analytics engineer at ESPN, in a presentation titled “A Journey in Sports Analytics.” Poe described her career path, the nuances of multi-sport analytics roles at ESPN, and how sports data is used for both fan-facing and behind-the-scenes applications.
The final keynote featured Cade Massey, practice professor at the Wharton School, who gave a talk titled “Analyst as Difference-Maker: Lessons from Firefighting in the American West.” Massey used the example of wildland firefighting analytics to illustrate the gap between building great models and influencing organizational decisions, drawing lessons directly applicable to sports analytics.
The panel discussion, “Communicating Advanced Analytics in Real Time,” brought together Epstein; Brian Burke, sports data scientist at ESPN; and John Parolin, research producer for Monday Night Football at ESPN, with Seth Walder, ESPN sports analytics writer, as moderator. The conversation focused on strategies for making complex analytics accessible and actionable in live broadcasts and other real-time contexts, highlighting emerging trends and audience-specific tailoring.
Invited Sessions
The program featured two invited sessions. Sports Economics included Lee Kennedy-Shaffer of Yale University on evaluating rule changes using quasi-experimental designs; Hashan Peiris of Simon Fraser University on modeling player injuries as compound risks; Shane Sanders, Alivia Uribe, and Justin Ehrlich of Syracuse University on behavioral considerations in penalty-kick location optimization; and Toby Moskowitz of Yale University on disentangling skill and luck in tennis.
Advancements in Sports Analytics and Education included Ben Baumer of Smith College on definitions of fairness in sports analytics; Owen Fiore of the University of Connecticut on reaction time thresholds in track and field; Ron Yurko of Carnegie Mellon University on Bayesian hierarchical modeling of tackling ability in football; Elizabeth Upton of Williams College on hypergraph-based adjusted plus-minus in basketball; and Keith D’Amelio of Heretic Performance on factors shaping game demands in elite basketball.
Data Challenge and Poster Session
The 2025 data challenge centered on analyzing newly released Major League Baseball data on bat speed and swing length to study pitcher–batter interactions. In the high school/undergraduate division, Team Yonsei Blues (Juwon Lee, Jiyong Lee, and Yugyung Kim) from Yonsei University in Korea won with “Optimizing Batting Orders: Monte Carlo Game Simulation Based on Batter Swing Clustering.” In the graduate division, Reese Mullen from Boston University won with “The Effect of Injuries on Bat Speed.”
The poster session featured a wide range of topics spanning baseball, basketball, hockey, soccer, motorsports, combat sports, and beyond. The Student Poster Award winners were Eugene Han of Yale University (first place) for his work on leveraging alternative data for MMA betting markets, Jacqueline Wang of Yale University (second place) for her research on pseudo-Bayesian regularized adjusted plus-minus in hockey, and Abhi Nagireddygari of Bowdoin College (third place) for his squash match analysis using low-cost computer vision.
Hands-On Training Workshops
The symposium’s first day was devoted to hands-on workshops organized into parallel tracks for introductory, intermediate, and advanced audiences, with an additional track for educators. Introductory sessions included Introduction to R led by Lucy Liu, Introduction to Python led by Charitarth Chugh, and Data Visualization led by Rahul Manna. Intermediate sessions featured Web Scraping for Sports Analytics led by Melanie Desroches, Tennis Analytics led by Jaden Astle, and Basketball Analytics led by Addison McGhee. Advanced sessions included Causal Inference in Sports Analytics led by Shinpei Nakamura-Sakai, Player Tracking led by Quang Nguyen, and Leveraging Docker for Building Data Pipelines led by Zuri Hunter. Rachel Gidaro and Michael Shuckers led the educators track with How to Create a Module for SCORE. Most of the instructors were students, and the training materials are available online.
Committees and Sponsors
The symposium was led by co-chairs Brian Macdonald of Yale University and Jun Yan of the University of Connecticut, with support from the organizing, program, judging, workshop, local, fundraising, and web support committees and drawing members from academia, professional sports organizations, and industry. Emily Hau, associate director of the Yale Institute for Foundations of Data Science, co-chaired the local committee with Macdonald and played a central role in coordinating the onsite organization. Sponsors included Ceros Financial Services, Connecticut Sun, ESPN Analytics, The Red Sox Foundation, the Connecticut Statistical Data Science Lab, the National Science Foundation, the UConn Data Science Club, the UConn Department of Statistics, the UConn Joint Statistical Club, and the Yale Institute for Foundations of Data Science.
Looking Ahead to 2026
The 2026 Connecticut Sports Analytics Symposium will be held April 10–11. The data challenge will feature curling, a winter Olympic sport that combines strategy, precision, and teamwork. Statistics and data science instructors are encouraged to incorporate the challenge into student projects. Look for more information and updates on the CSAS website.
Leave a Reply