The acceleration of the universe, which is often attributed to "dark energy, " has posed one of the main challenges to fundamental physics. Galaxy clusters provide one of the most sensitive probes of dark energy because their abundance reflects the growth rate of large-scale structure and the expansion rate of the universe. Several large galaxy cluster surveys will soon provide tremendous statistical power to constrain the properties of dark energy; however, the constraining power of these surveys will be determined by how well systematic errors are controlled. Of these systematic errors, the dominant one comes from inferring cluster masses using observable signals of clusters, the so-called "observable--mass distribution." This thesis focuses on extracting dark energy information from forthcoming large galaxy cluster surveys, including how we maximize the cosmological information, how we control important systematics, and how precisely we need to calibrate theoretical models. We study how multi-wavelength follow-up observations can improve cluster mass calibration in optical surveys. We also investigate the impact of theoretical uncertainties in calibrating the spatial distributions of galaxy clusters on dark energy constraints. In addition, we explore how the formation history of galaxy clusters impacts the self-calibration of cluster mass. In addition, we use N-body simulations to develop a new statistical sample of cluster-size halos in order to further understand the observable--mass distribution. We study the completeness of subhalos in our cluster sample by comparing them with the satellite galaxies in the Sloan Digital Sky Survey. We also study how subhalo selections impact the inferred correlation between formation time and optical mass tracers, including cluster richness and velocity dispersion.