When scientists engage in Public Understanding of Science to communicate their research to lay audiences, a common suggestion is to structure their talk around storytelling. Thus, it is crucial to know the actual effectiveness of storytelling in science communication compared to other structures. For instance, a structure almost unexplored is the one of magic or illusionism. As storytelling, it has been evolving and improving over humanity's history to become ever more effective, granting magicians a prominent place in the entertainment and art industry. In the present work, we compared various storytelling structures and the structure of magic, through an agent-based computational model. The results open the questioning of story architectures; propose a new way to test ideas in science communication; and show that double-blind control studies are very much needed for further testing the structures of Public Understanding of Science and further developing agent-based models.
Keywords: Public Understanding of Science; agent-based model; magic; science talk; structure.