Abstract
This article presents an original model for assessing scientific productivity, research power ranking (RPR), which is based on the adaptation of the Elo-rating system to the context of scientific activity. Unlike traditional scientometric indicators, RPR accounts for the dynamics, multidimensionality, and relativity of research power. The model comprises three components—fundamental, applied, and commercial activity—each represented by a separate rating and updated on the basis of probabilistic “scientific games” analogous to chess matches. The scientific rating of each researcher is calculated as a weighted sum of the components, allowing the model to reflect not only their current position but also their career trajectory, including phase transitions, breakthroughs, and changes in scientific style. Numerical simulations were conducted for both the individual trajectories and population-level distributions of the researchers. Phase diagrams were constructed, and a typology of scientific styles was formulated. The results demonstrate that RPR can serve as a universal tool for objective assessment, strategic planning, and visualization of scientific reputation in both academic and applied environments.