Exploring dynamics of sports fan behavior using social media big data - A case study of the 2019 National Basketball Association Finals. (April 2021)

• Record Type: Journal Article
• Authors: Gong, Xi; Wang, Yong
• Is Part Of: Applied geography. Volume 129(2021)
• Publication Date: 2021-04
• Subjects: Sports fan behavior -- Social media -- Spatial-temporal pattern -- Sentiment information -- Basking in reflected glory (BIRGing) -- Blasting; Geography -- Periodicals; Human geography -- Periodicals; Human ecology -- Periodicals; 910
• DOI: 10.1016/j.apgeog.2021.102438
• Languages: English
• ISSNs: 0143-6228

Abstract

Abstract: No one can deny the importance of fans in spectator sports. Sports fans' tweets are an excellent big data source for investigating large-scale spatial-temporal dynamics of fan behavior. Using tweets related to the 2019 National Basketball Association (NBA) Finals between the Toronto Raptors and the Golden State Warriors, we conducted exploratory analyses of fan behavior dynamics and validated the fan behavior theories. Several behavior patterns were identified. Peak tweeting frequencies happened near the end of each game, after key incidents, and during the daytime the following day. National identity could elevate Canadian viewers' enthusiasm for the games. States/provinces that are geographically close to the teams or have connections to key players had higher tweeting densities. Significant local hot spots of positive tweet percentages towards the Warriors were found near California, which clearly demonstrates the Warriors' fan base. Generally, a team's relative positive percentage of tweets increased after a win, decreased after a loss, and moderated by key incidents. The trends of fans' attitudes associated with the game results in Ontario were consistent with the theoretical fan behavior tendencies, including Basking in Reflected Glory (BIRGing) and blasting. The identified patterns and validated theories are transferable to other spectator sports. Highlights: Described fan behavior using spatial-temporal and sentiment information of tweets. More tweets near the end of games, after key incidents, and during the next noon. More tweets in regions that are close to the teams or linked to key players. Positive tweet percentages would fluctuate with game results and key incidents. Trends of attitude validated two fan behavior tendencies (BIRGing and blasting).