We exploit methods of sample-based stochastic optimization for the purpose of strategyproof dynamic, multi-unit auctions. There are no analytic characterizations of optimal policies for this domain and thus a heuristic ...