Spectators’ emotional responses in tweets during the Super Bowl 50 game

Highlights

•	328,000 real-time tweets posted by spectators during the Super Bowl 50 game were collected/analyzed.
•	Expressed positive emotions when their team scored, while expressed negatives when the opposite team scored.
•	Became habituated with subsequent scoring, resulted in fewer expressions of emotions.
•	Expressed a surge of emotions when a team scored soon after a score from the other team.
•	Simultaneous experience of positive and negative emotions may contribute to patronage and health.

Abstract

The author explored spectators’ emotional reactions manifested on social media. By using Twitter search application programming interface, 328,000 real-time tweets posted by fans of the Panthers and the Broncos during the Super Bowl 50 game were collected. The lexicon-based text mining approach (a big data analysis in social media analytics) was employed to classify tweets into five different emotions. The findings indicated that spectators expressed positive emotions when their team scored; conversely, they expressed negative emotions when the opposite team scored. Interestingly, spectators became habituated with each subsequent score from either of their preferred teams, which resulted in fewer expressions of emotions. However, when a team scored soon after the opposite team scored, fans expressed a surge of positive or negative emotions, accordingly. The results supported both the theories of affective disposition and opponent-process. Spectators’ simultaneous experience of positive and negative emotions may contribute to fans’ satisfaction, continued patronage, and mental health.

Keywords:

• Emotions
• Big data
• Automated text analysis
• Social media
• Spectator sports

Acknowledgments

The author appreciates the invaluable inputs of Chris Janiszewski, Semih Yilmaz, and Ilrang Lee on an earlier draft of this manuscript. The author thanks the Editor George Cunningham, the Associate Editor Daniel Lock, and the entire SMR review team for their outstanding guidance and support. Any remaining errors are the author’s responsibility.