Online Program

Return to main conference page
Thursday, May 30
Data Science Techologies
Practice and Applications
Data Science Applications E-Posters, II
Thu, May 30, 5:30 PM - 6:30 PM
Grand Ballroom Foyer

Forecasting NBA Fan Support using Time Series Analysis (306353)

*Victor Wilson, Cal Poly San Luis Obispo 

Keywords: NBA, basketball, fan support, popularity, bandwagon, sports, time series

Sports are a staple of American culture, and fans are always interested in statistics related to their favorite team, including popularity trends. Popularity trends can also be useful from a business standpoint, as it can provide advertisers with valuable data about which teams are notably popular at a given point in time. The support for a particular sports team ebbs and flows over time as well, varying with team success and time of year, such as an increase in popularity during playoffs and a sharp decrease in popularity immediately after. A recent example of this can be seen with the Golden State Warriors, who have seen a huge surge in popularity during the last 5 years, after winning several NBA Championships. Popularity for a certain team can be measured via search data available on Google Trends, which provides data on the relative popularity of terms that are searched for on Google. We focused on comparing the search popularity for all 30 NBA teams throughout time, and created a time series model to predict the yearly pattern of team support, correcting for team performance, playoffs, and other factors. This model is then used to discover the differences in support patterns between teams and potentially identify bandwagon fans, as well as whether some teams are more prone to attract bandwagon fans than others.