Over the last 10 years I was part of teams that participated in two large-scale multi-institutional forecasting tournaments sponsored by the Intelligence Advanced Research Project Activity (IARPA). The tournaments involved elicitation of probabilistic forecasts regarding geopolitical events from large groups of, non-professional, volunteers and the optimal (accuracy maximizing) aggregation of these forecasts. I will describe some of the projects seeking to answer practical questions about the size and composition of the groups, early prediction of forecasting skills and the benefits of teaming. I will use this presentation to highlight some of the differences, and the interplay, between basic (theory driven) and translational research.