As we learned from Brad Pitt in Moneyball, data analytics has always been a huge part of the professional sports industry. Data gives coaches and managers clear insight into anything from determining player talent and ability, optimizing match-ups, and more.
With all the available data sources these days, however – Statcast, NBA Player Tracking, PITCHf/x, NFL Next Gen Stats, FIELDf/x, HITf/x, etc. – it can be nearly impossible to analyze all of the available data to get meaningful, actionable insights that can improve performance outcomes. There are a myriad of additional challenges to applying machine learning to sports analytics as well, including the lack of available data scientists and the long time horizons of traditional data science projects.
However, DataRobot has thrown a curveball to the sports analytics practice in the form of automated machine learning. Join Andrew Engel, Customer-Facing Data Scientist at DataRobot, as he discusses:
The potential for machine learning to help managers and coaches make sense of the wealth of available player and game data
The challenges associated with implementing machine learning initiatives and how automated machine learning helps sports analysts overcome them
An example case of using automated machine learning for catcher pitch framing in baseball
Andrew Engel Customer-Facing Data Scientist, DataRobot