A Bayesian Equilibria in the Game of Effort Estimation in Agile Global Software Development

Abstract

This research examines leader–team interactions in agile software development within a global context, with a focus on effort estimation. Drawing on principal–agent theory, we analyse the interaction on the assumption that the Scrum Master guides the development team under imperfect information. We model the interaction as a sequential game with incomplete information. In the first stage, the Scrum Master allocates resources to the development team; in the second stage, the team exerts effort. Both parties are characterised by types that capture their knowledge and skills. As these types are private information, we derive the Bayesian Nash equilibrium to determine the equilibrium effort levels.