The methods for the study falls primarily into two broad areas, landscape generation and disease spread.

One part which does not comfortably fall within either group is the evaluation and determination of starting location. A potential problem when using randomly generated landscapes is variability, so a way to decrease this variability is desireable. Also a way to evaluate starting locations might be practically useful in determining which farms would result in larger outbreaks.

The way used in this study is to determine the severity of each farm in a landscape, this is determined by number of surrounding farms and the distance to these, and ranking them. Simulations are then ran for best, worst and median case starts.