Question: If you have access to commercial optimization-via-simulation software, test how well it works as the variability of the simulation outputs increases. Use a simple model, such as Y = x2 + e, where e is a random variable with a N(O, 112) distribution, and for which the optimal solution is known (x = 0 for minimization, in this case). See how quickly, or whether, the software can find the true optimal solution as 112 increases. Next, try more complex models with more than one design variable.