The Scouting algorithm is an evolutionary based search mechanism that is advanced from the realm of optimization to that of problem state space exploration. It relies on the concept of surprise, or the difference between estimated and actual phenomena at a given state space location, to find the major points of interest in the problem space. The estimate is generated from an internal abstraction of the state space based on previous samples of actual phenomena, thus, over time estimates will approach actual values in highly sampled regions and will prevent being trapped in local optima.