What if gamified software is fully proactive? Towards autonomy-related design principles

Computational agents are a type of software architectures designed to be autonomous and social, meaning that they can make decisions proactively, reacting also to stimuli from the environment. The use of such architectures is not common in the gamification field area, instead, gamified software has traditionally reactive characteristics, responding to user actions disregarding the possibility of proactive behavior. In this paper, we propose four formal principles for designing autonomous gamified systems, to ensure traceability of gamified outputs, internal consistency of gamification attempts, coherent agent-user interaction, and formal conditions to assess user actions from a rational perspective. We present our initial work on these general principles, highlighting our empirical future work.