Multi-Stage Pairwise Elimination Contests

Rudi Stracke
SEW, University of St.Gallen

     Full text: PDF
     Last modified: January 21, 2011
     Presentation date: 03/12/2011 4:30 PM in NH 1130, Session B
     (View Schedule)

Abstract
This article presents a solution technique for pairwise elimination contests with heterogeneous players. Players differ in terms of their effort productivities, which are common-knowledge among participants of the contest. It is shown under which conditions a subgame perfect Nash equilibrium exists when a Tullock contest success function is used. Moreover, the equilibrium solution is derived analytically for the special case of a lottery CSF, and approximately for the remaining cases. A distinct feature of multi-stage pairwise elimination contests with heterogeneous agents is that continuation values in early stages become endogenous due to feedback effects across different branches of the game. Those feedback effects are analyzed in some detail, as well as several other properties of the equilibrium solution.