Background: Methodological shortcomings of gambling studies relying on self-report or on data sets derived from gambling operators tend to result in biased conclusions. The aim of this study was to analyze online gambling behavior using a novel network database approach. Methods: From October 13 to October 26, 2014, telecommunications network data from a major telecommunications provider in Switzerland were analyzed. Netflows between mobile devices and a poker operator were quantified to measure the gambling duration and session number. Results: Time spent gambling during night and working hours was compared between devices with longest (red group), intermediate (orange group), and shortest gambling time (green group). Online gambling behavior differed depending on overall gambling time, F (2, 3,143). Night and working hours gambling was the highest in the red group (53%), compared to the orange (50.1%) and the green groups (41.5%). Post hoc analyses indicated significant differences between the orange and green groups (p < 0.05). No differences were observed between the red and orange groups (p = 0.850), and the red and green groups (p = 0.053). Conclusions: On mobile devices, distinct gambling patterns were observed depending on the overall gambling time. This methodology could also be used to investigate online gaming, social media use, and online pornography.