Alan L. Morse

The Effects of Roster Turnover on Demand in the National Basketball Association

The purpose of this study was to examine the effects of roster turnover on demand in the National Basketball Association (NBA) over a five-year period (2000-2005) and compare these results to previous research on turnover in Major League Baseball (MLB). A censored regression equation was developed to examine the relationship between roster turnover and season attendance, while controlling for other potentially confounding variables in the model. The censored regression model was used to account for the capacity constraints by forecasting the level of demand beyond capacity using information from the uncensored observations. The regression model was found to be significant with a log-likelihood statistic of 110.446. Previous attendance, current winning percentage, previous winning percentage, number of all-star players, and team history were found to be significant predictors of attendance. However, the variables measuring the effects of roster turnover were not found to be significant. There were substantial differences in the effect of roster turnover on attendance in the NBA compared with MLB. In addition, these findings provide evidence for using censored regression when dealing with constrained variables. Sellouts in the NBA appear to have an effect on all of the variables in the demand model. Future research will need to be conducted to help sport managers understand the role of roster turnover in specific professional leagues and to better understand the importance of using a censored regression model.

An investigation of the relationship between television broadcasting and game attendance

This study investigated the relationship between whether a major conference NCAA Division I mens college basketball game was televised and game attendance during the 2003–2004 and 2004–2005 seasons. Multiple regression analysis was used to examine this relationship while controlling for 19 other potentially confounding variables. A significant regression model was created (F = 111.586, p < 0.001), explaining 77.6% of the variance in game attendance. The model estimated that mens basketball game attendance increased by 564 spectators when a game was broadcast on television, ceteris paribus, representing a 6.3% increase over the mean game attendance of 8892. These findings support practitioners broadcasting home mens basketball games on television, as television broadcasting was found to be significantly related to a substantial increase in game attendance in the population studied.

Constraint factors affecting non-attendance in collegiate volleyball

The focus of this study was to analyse constraint factors affecting non-attendance in sports. Few research studies have investigated the topic of non-attendance, and even fewer have sample representation of respondents who did not attend games. A survey was used to measure the impact of 12 constraint factors (41 total items) on non-attendance at womens college volleyball. The results of a logistical regression indicated that non-attendance of those who had never attended a match was predicted by the factors of financial cost, lack of knowledge, and no interest from others, while non-attendance of those who had previously attended was predicted by other sport entertainment and lack of success. These findings should be used to help marketers of typically lower-attended sports realise in development of strategies to increase attendance, more of a focus should be put towards their core product and enhancement of the event atmosphere, rather than on constraints.