Categories
News and Events

Drafting Agent-Based Modeling Into Basketball Analytics

The CSS seminar speaker for Friday, February 9 will be Matthew Oldham, CSS PhD Student, Department of Computational and Data Sciences. The program will begin at 3:00 in the Center for Social Complexity Suite located on the 3rd floor of Research Hall. The talk will be followed by a Q&A session along with light refreshments.

This session will be live-streamed on the newly created CSS program YouTube channel: https://www.youtube.com/channel/UC7YCR-pBTZ_9865orDNVHNA

For announcements regarding this and future streams, please join the CSS/CDS student and alumni Facebook group: https://www.facebook.com/groups/257383120973297/

For a list of upcoming and previous seminars, please visit: https://cos.gmu.edu/cds/calendar/

We hope to see you on Friday, February 9.

Abstract: Sports analytics (SA) has experienced a meteoritic rise in recent years, with the trend forecast to continue. Modor Intelligence reports that the market was valued at USD 83.56 million in 2015, and is forecast to grow to USD 447.23 million by 2020. at the market was valued at USD 83.56 million in 2015, and is forecast to grow to USD 447.23 million by 2020.

The growth of sports analytics has raised a rich variety of research topics pertaining to basketball, including: how at the macro level the distribution of scoring activity is a mixture of random walk processes and power-law behavior (Gabel & Redner, 2012), and, at the individual level, the question of whether players develop hot-hands and how the player and their teammates react to its possible existence. While the erroneous belief regarding hot-hands was first identified by Gilovich, Vallone & Tversky (1985) it has remained an active field of research (Bar-Eli, Avugos, & Raab, 2006).

Agent-based modeling (ABM) has great potential to assist and inform those engaged in sports analytics but to date it has not been utilized. The advantage of ABM is that it allows researchers to assess, in a silicon laboratory, the micro-level interactions that give rise to verifiable macro outcomes. This is achieved through heterogeneous agents adapting and making decisions based on their environment, including considering spatial, temporal factors and interactions with other agents.

To support the use of ABM in sports analytics, I will present a 3-dimensional model of a basketball game, where the fundamentals of play including player and court positions, a shot clock, and shooting performance are all included. Additionally, player behavior in deciding whether to shoot, pass or dribble is partially predicated on assessing the length of a player’s shooting streak (designed to test the hot-hand effect), and the consideration they give to any streak, plus their franchise status, a feature identified in Burns (2004). The probabilistic nature of the model allows for insights into the dynamics of scoring actions following a random walk. The model captures extensive data which was used to calibrate and validate it against comparable statistics from the National Basketball Association (NBA).