Abstract

Some scientists are more likely to explore unfamiliar research topics while others tend to exploit existing ones, yet it is unclear how these choices connect to academic performance. This is evidenced by contradictory accounts of correlations between topic switch and performance in previous studies that often rely on single metrics and neglect confounding factors. To disentangle the complexity, we complement the switch frequency metric with switch distances, extensively control for confounding factors, and establish a link between past switch behaviors to future performance. We discover that `conservative adventurers' who switch topics frequently but do so to `close' domains measured by frequency and distance metrics, have notably better future performance than those who behave differently under these two metrics. The behavioral pattern of `conservative adventurers' is uncommon and seems to suggest an effective balance between exploration and exploitation. Moving beyond mere correlations, we ask whether `conservative adventuring' could be a strategy to adopt. While it is almost impossible to know whether any behavioral change in our scenario is an intentional strategy adjustment, it is plausible to assume that the rare ones who drastically change to `conservative adventurers' from the opposite behavior group shift their research strategies on purpose, given the observation that most scientists maintain consistent research agendas. These `drastic changers' experience considerable future improvements. Our findings are based on three datasets covering a total of 31,780,857 papers in biomedicine, physics, and chemistry, from 1800 to 2021, and shed light on understanding and planning of scientific careers, especially for junior scientists.