Real-time strategy games are of such high complexity that consideration of trying to brute force all actions and states is not only impractical, but impossible. Approximations, information abstractions, and models are, therefore, the necessity when creating game bots that play this genre of games. To create such bots, the detailed data is needed to base them on. This article introduces a universal algorithm that creates reusable simulation data of one attacking unit on a building and tests the feasibility of doing such a task. This paper concludes that capturing all relevant data in a sub-segment of real-time strategygames is feasible. Gathered data holds valuable information and can be reused in new research without the need of repeating the simulations.