I study statistical aspects of head to head competition, which can include basketball, chess, cardinals electing the pope, pecking order among chickens, and dogs choosing between two kinds of food at a time. Through the advent of cameras and computer vision, many new avenues of research in head to head competition have been made available (although these advances probably won’t help cardinals elect a pope).
Video footage of sporting competitions can be programmatically analyzed to extract temporal information such as player position, player pose, and ball position. This data can be used to answer questions like, “how good is Steph Curry at shooting the basketball?”, “what is the probability a goal will be scored given player position?”, or “what does a good shot look like?”. (See below for embedded media)
Students this summer will work with basketball player tracking data and attempt to push the boundaries of the extraction and analysis of this data in R. Requirement: STAT231. Preference will be given to those with additional programming experience in any language.
For those interested in doing this work this summer, apply here. For all SURF opportunities, see here.
Shooting Quality
They asked me to write an article on the best shooters of the decade; here’s what I turned in.
— Kirk Goldsberry (@kirkgoldsberry) September 16, 2019
THE ABSOLUTE BEST SHOOTERS OF THE DECADE
by Kirk Goldsberry
Folks, it’s Steph Curry.
***
That wasn’t enough so, they made me change it:https://t.co/T1khqalBbh pic.twitter.com/HpLTGToAQh
Player Position
— Neil Johnson (@neilmjohnson) January 29, 2020
Player Pose
Play Success Probability