Abstract
We examine the disruption of researchers with long-lived careers in Computer Science and Physics. Despite the epistemological differences between such disciplines, we consistently find that a researcher's most disruptive publication does not occur at random during their career, as it cannot be explained by a null model. Such publication is accompanied by a peak year in which researchers publish other work that exhibits a higher level of disruption than average. Through a series of linear models, we show that the disruption achieved by a researcher during their peak year is higher when it is preceded by a long period of focus and low productivity. These findings are in stark contrast with the dynamics of academic impact. In these dynamics, researchers are incentivized by the prevalent paradigms of scientific evaluation to pursue high productivity and incremental-less disruptive-work, as evidenced by extensive literature.