
35: Teaching Statistics and Data Science through Sports
On this episode of “In the Interim…”, which is co-sponsored by the Journal of Statistics and Data Science Education, Dr. Scott Berry talks with Dr. Jim Albert, Professor Emeritus at Bowling Green State University, whose extensive work encompasses Bayesian statistics and computation, sports analytics, and decades of exemplary teaching. Dr. Albert shares insights on integrating sports into statistics education and discusses his transition from academic roots to consulting for the Houston Astros. This episode highlights the evolution of sports statistics—from manual data collection to sophisticated analytics—and critiques traditional metrics in favor of advanced systems. The dialogue explores career opportunities in sports statistics as well as the need for open research avenues in sports analytics, facilitating broader access and distribution of statistical insights.
Key Highlights
Use of sports to contextualize statistical concepts, providing practical illustrations over abstract textbook issues
Exposing misconceptions about randomness, streakiness, and “clutch ability” perpetuated by both public myths and sports simulations
Analytical evolution from traditional metrics like batting average to advanced assessments like OPS and on-base percentage
Regression-to-the-mean explained with sports scenarios and its analogous application in clinical trial progression
Challenges in adopting a unified approach to teaching statistics given students’ diverse cultural and sports familiarity
Barriers in publishing sports analytics research, prompting initiatives for accessible, open publications