The history of efforts to reduce ‘human errors’ across workplaces and industries suggests that people (or their weaknesses) are seen as a problem to control [1, 3, 15, 16]. However, some have proposed that humans can be heroes as they can adapt and compensate for weaknesses within a system and direct it away from potential catastrophes . But the existence of heroes would suggest that villains (i.e. humans who cause a disaster) exist as well , and that it might well be the outcome that determines which human becomes which. The purpose of this chapter is to examine if complex socio-technical systems would allow for the existence of heroes and villains, as outcomes in such systems are usually thought to be the product of interactions rather than a single factor . The chapter will first examine if the properties of complex systems as suggested by Dekker et al.  would allow for heroes and villains to exist. These include: (a) synthesis and holism, (b) emergence, (c) foreseeability of probabilities, not certainties, (d) time-irreversibility and, (e) perpetual incompleteness and uncertainty of knowledge, before concluding with a discussion of the implications of the (non) existence of heroes and villains in complex systems for the way we conduct investigations when something goes wrong inside of those systems.