One of the strongest design parameters for contests is the prize structure, i.e., the number and level of prizes. In developing best practices, we are working to provide guidance to practitioners to optimize the use of prize funds. Optimal selection of prizes is a complex task. For tasks with diminishing returns to effort (the 100th hour of work improves the output less than the 1st hour), there is not a simple rule to follow, e.g., use a single grand prize, even in a basic contest model. In practice, contest environments have additional complexities: there maybe value in combining solutions from different contestants, contest outcomes may reveal information about the contestants for good or ill, contestants may be capital constrained and require financing before a final solution can be produced, etc. Our work aims to incorporate these features into prize design.
