Antifragility as a design criterion for modelling dynamic systems

Abstract

AbstractHighly improbable events can have a substantial impact on complex socio‐economic systems and are frequently difficult to predict beforehand but easy to explain afterwards. Antifragile systems can withstand and benefit from this kind of outlier events, whereas merely robust systems cannot in any case. Yet the aim to design robust systems is almost as old as the system dynamics field itself. This research therefore aims to investigate the extent to which an antifragile system design criterion is more valuable than a robust one. By means of an extensive literature review, a simulation model was constructed, which is demonstrated to be antifragile. Comparing the antifragile and robust versions of the model shows that the former—as theorized—yields more favourable results in an environment with impactful outlier events. Implementing antifragility in systems involves the difficult task of changing policies (and, eventually, the mental models) of decision‐makers. Consequently, this research concludes that antifragility should not and cannot always be attained; its feasibility is to be assessed at the start of a system dynamics modelling project.