Compete with Me? The Impact of Online Gamified Competition on Exercise Behavior.

Online gamified competition utilizes competition as a core gamification design element with affordances from wearables and mobile applications to track competitive activities and visualize information in an integrated way to shape users' exercise behaviors. However, a clear understanding of how online gamified competition cultivates exercise behaviors in different types of individuals is still lacking. We take into account the individual differences in exercise behaviors and categorize exercisers into three groups (active, moderate, and inactive) based on an adapted recency, frequency, and monetary value framework using key exercise behavior metrics. Theorizing online gamified competition as a means of social and temporal self-comparison, we examine the effect of performance feedback from two distinct modes of comparison (performance rankings and performance gap), and participants' relationships with their social comparison referents (i.e., rivalry intensity), on the exercise behaviors of different exerciser groups. Our results reveal that online gamified competition has differential effects on exercise behaviors across different exerciser groups. Specifically, we find that positive performance improvements are more motivational for active and moderate exercisers, while performance deterioration relative to historical exercise performance level is more discouraging for inactive exercisers. Performance rankings exhibit a more salient effect for moderate and inactive exercisers, and rivalry intensity has a stronger positive effect on active exercisers' exercise behavior. The strengthening effect of awareness affordances in mobile fitness apps is more notable with regard to the impact of rivalry intensity on moderate and inactive exercisers. We derive theoretical and practical implications of gamified information systems that use competition as a core design element for shaping the exercise behavior of individuals in different exerciser groups.