The role of promotions and other factors in attendance for minor league baseball – a study across all levels of play

Purpose The main purpose of this research was to determine what types of promotions increase Minor League Baseball attendance across all of the official leagues. The secondary purpose was to ascertain the role of other control variables such as win percentage, weather, days of the week, start time and city demographics. The research also includes a grouping of cities through k-means clustering to better understand what types of promotions work in what cities. Design/methodology/approach Data were gathered on all the Minor League Baseball teams and their respective cities. Regression models were run to test for the role of individual promotions (structured as dummy variables) and other controls. One model specification used city demographic variables, one used city fixed effects rather than city demographics and the final specification used k-means clustering to separate cities into distinct groups. Findings Promotions generally were found to increase attendance, although there were differences across levels of play. K-means clustering helped with the grouping of cities to ascertain which types of promotions were beneficial when comparing large metropolitan areas to high-income cities. Research limitations/implications Promotions were grouped into common categories, although some were difficult to classify (or were infrequent), so a miscellaneous promotions category was included to capture these promotions in Minor League Baseball. Practical implications The findings of this research are beneficial to those designing promotional schedules for individual teams. It also is beneficial to the leagues as the findings have implications as it relates to what fans desire to see when attending minor league games. Originality/value The originality in this work is the collection of all Minor League Baseball attendance, city information and promotional information across the different levels of play (AAA, AA, High-A and A). Using different model specifications and groupings, including k-means clustering to match similar cities, successful promotions were identified.