Loss chasing is considered a defining marker, and potential risk factor, for problematic gambling. However, there have been few efforts to identify loss chasing behaviors among actual gamblers. Loss chasing also is not well defined, and often measured along a single behavioral dimension. In this study, we propose a novel multidimensional concept of loss chasing grounded in three betting domains (bet size, betting odds, time between bets). Using actual sports gambling records from a major European operator, we calculated bettors' (N = 12,992) loss chasing during their first month of activity. We created binary high/low groups for each loss chasing dimension. Membership in various combinations of 'high' groups was diverse, and none of the raw loss chasing metrics were highly correlated with one another. We assessed each high group's (and sum of high group's) ability to predict mounting losses over the following months of activity. Only one of these groups (bet size), was positively predictive of mounting losses in bivariate analysis, and none of the loss chasing groups were significant in multivariate analysis. Results suggest that loss chasing is a multidimensional concept, but the potential value of a multidimensional loss chasing concept in terms of predicting gambling harm is unclear.
