Formalising Concurrent Test Interactions in Model-based Testing via Games

In preceding research, a strong correlation between model-based testing and game theory has been established and formalised. In this conversion, an assumption must be made on which transition in a game arena to take when tester and system propose conflicting actions. Several such assumptions exist, but always discard and ignore either of these actions. We investigate a conjecture from earlier literature concerning a novel concurrent test assumption taking into account both. We formalise the assumption and evaluate its efficacy. Based on our findings we conclude that the novel test assumption may allow for the accurate modeling of concurrent software systems, in addition to possibly providing an increase in overall testing efficiency. We identify multiple avenues for further research.