Rationing Experiences
Dynamic one-sided matching mechanisms for amusement park scheduling, simulated with Disneyland data.
Final Paper for ECON 285: Matching and Market Design, Fall 2024
Abstract
We describe practical, lightweight algorithmic solutions for the problem of one-sided matching with dynamic ordinal preferences and multiple capacities. In particular, we focus on amusement park scheduling where guests have preferences over rides they wish to experience and rides have limited capacity in each time period. We explore sequential variants of the Random Serial Dictatorship (RSD) and Probabilistic Serial (PS) mechanisms, and we experimentally validate their efficiency using a simulation calibrated to ride capacity and wait time data from Disneyland.
Authors
- Lyle Goodyear
- Ivan-Aleksandar Mavrov