This document summarizes a simple real-coded estimation of distribution algorithm (RECGA). The algorithm discretizes a real-valued population into intervals, models the discrete population using a greedy maximum probability model search, and generates a new population from the model. The population size scales sub-quadratically and the number of function evaluations scales sub-cubically with problem size for this simple RECGA. Next steps proposed include more experiments on scalability analysis, the relationship between discretization and performance, and developing theory on convergence for real-coded genetic algorithms.