^^Abstract^^
Galaxy formation involves a complex interplay of many nonlinear, poorly understood physical processes. This fact has limited progress in the field. Recent work has made significant steps forward by systematically surveying model parameter spaces to find models that best describe observed data. However, this work has been limited by inadequate treatment of the observational uncertainties. In this talk I will describe the results of a careful program aimed at quantifying the uncertainties in both observations and models, and using these results to carefully constrain a simple galaxy formation model from which quantitative predictions can be extracted. I will also describe the next steps in this program to add additional constraints, incorporate more physics into the model, and meet the computational challenges of surveying high-dimensional parameter spaces.