An Agnostic Approach to Modeling Conflict: A Maximum Entropy Foundation for Contest-Success Functions
An agnostic researcher who wishes to assign ex ante winning probabilities to a conflict between two sides exerting effort should commit only to information she can defend. We adopt the principle of maximum entropy, justified by the Shore-Johnson axioms. In the low-knowledge limit, the unique solution induces the Hirshleifer difference-form contest-success function. The Tullock form arises when one imposes scale invariance in efforts.
Martin Castillo Quintana studies the relationship between violent conflict, policymaking, and politics, with a methodological emphasis on formal theory. He develops and applies game-theoretic models to understand the effects of different policies on the conflict behavior of violent actors. He also studies how violence influences politics and how politics shape policymaking. His current research delves into how organized criminal groups fight each other and their interactions with government and state authorities. He earned his Ph.D. in Politics from New York University in 2023 and holds a B.S. in Mathematical Engineering and an M.A. in Applied Economics from the University of Chile.