Title: Recording (Statistical) Errors in Baseball by Dr. Katy McKeough
Abstract: Understanding uncertainties from statistical and machine learning models is crucial in making informed, data-based decisions in sports analytics. Whether you are assigning value to a player or making recommendations to a coach, it is valuable to quantify and communicate variance around your predictions. This talk will review a way to estimate the uncertainty around a win probability model for Major League Baseball games and show how it can be used to make informed decisions for in-game strategy.
Date: Friday 12/8 1:10pm Zoom: https://williams.zoom.us/j/8294371163 (Stat442 class)
Speaker: Katy McKeough is a Senior Analyst within Research & Development of the Boston Red Sox. At the Red Sox, she works on statistical and machine learning models that help drive decision making for Major League player acquisition and in-game strategy. Katy has a Ph.D. in Statistics from Harvard University where she conducted applied research in both sports analytics and astrostatistics. She graduated from Carnegie Mellon with a B.S. in Physics and Statistics.