The Theory and Practice of Computing the Bus-Factor

A team of computer science researchers has published a groundbreaking paper that finally puts the informal concept of the 'bus factor' on solid theoretical ground. The bus factor—a measure of how many key people a project can lose before stalling—has long been used anecdotally in software and project management. Sebastiano Piccolo, Pasquale De Meo, Giorgio Terracina, and Gianluigi Greco developed a formal, domain-agnostic framework that models any project as a bipartite graph connecting contributors to tasks. Within this model, they define the bus factor as a family of combinatorial optimization problems, specifically the 'Maximum Redundant Set' and 'Minimum Critical Set', and prove that computing these exactly is NP-hard, meaning it becomes computationally intractable for large projects.

Building on this theoretical foundation, the researchers propose a novel, normalized bus factor measure inspired by network robustness. Unlike previous ad-hoc methods, their measure tracks the largest connected set of tasks as contributors are progressively removed, capturing both loss of coverage and increasing project fragmentation. They also developed efficient linear-time approximation algorithms to make the computation practical. Through sensitivity analysis on controlled project structures, they demonstrate that their robustness-based measure behaves consistently with project management theory and provides a more stable, informative risk assessment than existing alternatives. This work transforms the bus factor from a vague heuristic into a rigorous, computable metric applicable across any collaborative domain, from open-source software to corporate R&D teams.