Despite the extraordinary impact science has on our life, little is known about the patterns
that lead to scientific excellence. Is the success of a particular work predictable? Can future
individual performance be predicted? Are scientific discoveries more likely in certain countries
and/or institutions?
By combining the tools of statistical physics and information science, we aim to uncover the
predictability and the limits of predictability of scientific success. Our focus is on scientific
performance captured at increasing levels of complexity: individual publications, scientists'
careers, and research institutions. The final goal of our research is to provide theory and tools
of potential utility to policy making, from funding decisions to discovering scientific talent.