Abstract:
Since the turn of the century, the use of analytics has become more prominent in the National Basketball Association (NBA). Teams are using data-driven approaches to construct rosters and develop strategies, which have proven useful in turning previously underwhelming teams into championship contenders. Today, all NBA teams have designated analytics departments that use advanced metrics to evaluate the success of their players and teams. Media coverage of the NBA reflects this change; more than ever player evaluations in the news are bolstered with evidence such as shooting percentage, adjusted plus minus, or other advanced metrics. This analytics overhaul in the news has made it easier than ever for NBA fans to interact with analytics and has caused many to take a “Moneyball” approach to consuming NBA basketball, which relies on using numerical data to evaluate players and teams.
However, NBA affiliates are resistant to the league’s rapid integration of analytics into team operations, claiming that analytics are destroying the integrity of the game by reducing players to numbers. This could be problematic for players, as it creates a larger gap between how players evaluate their own performances, versus how organizations evaluate them, which makes it more difficult for players to meet team expectations. This, coupled with other ethical concerns, is why some NBA affiliates argue that the reliance on analytics furthers the divide between players and management, which in turn is harmful to the players and threatens their livelihoods.
3
This thesis seeks to evaluate the discourse surrounding analytics in the NBA, specifically looking at reactions by NBA fans to media posts about basketball analytics. Gauging the reaction of an entire community to one specific issue is difficult when relying solely on literature review, so this project will make use of Twitter scraping and sentiment analysis to quantify the nature of the discourse surrounding NBA analytics. Twitter is a representative sample of the NBA community because fans, announcers, and other NBA affiliates can post thoughts, as well as interact with other posts. The use of sentiment analysis to quantify tweets has been performed before and is useful in getting the tone of a discourse efficiently. Once this evaluation is complete, this project will conclude with re-examining the current implementations of analytics through the lens of the NBA community’s discourse and provide suggestions on how the NBA could change their current implementations of analytics to better suit players, fans, and league operations personnel alike.