The localization of gambling establishments is a relevant topic in gambling research. In this paper, we analyze the spatial distribution of two types of gambling establishments—private and public—over the last 10 years in the municipality of Madrid (Spain). Using a spatial scan statistic, we identify the temporal dynamics of spatial clusters with high densities. The results reveal different spatial patterns regarding the locations of these two types of gambling establishments. While public gambling establishments do not exhibit spatial clustering, private gambling establishments show a growth in spatial clustering with dynamic behavior, seeking locations with specific sociodemographic characteristics. A machine learning tree-based algorithm is used to confirm that decisions on where to put new gambling establishments are based on targeting customers with a gambling profile.