The macroscopic properties of any material system are dictated by its atomic structure; however, the presence of structural and topological fluctuations dramatically alters the properties and performance of the material for a given application. Glass is unique in that it has a disordered atomic arrangement, meaning that the properties of glasses are based on statistical distributions rather than precisely known values. Up until this point, there have been no rigorous theories established to predict these statistical distributions. Glass properties have traditionally been represented as mean values, which do not fully represent the complexity of glass structure. This paper introduces a rigorous approach for quantifying fluctuations in glass structure which will enable scientists to improve their understanding of fundamental glass physics and chemistry. The model is first shown for arbitrary glassy systems to clarify the physical understanding and outline the general approach for calculating distributions of properties in disordered networks. The established framework is then applied to real glass-forming systems, specifically phosphates and silicates, where the microscopic structure, ability for atomic rearrangement, and thermodynamic properties are predicted and validated against experimental data. Results reveal that statistical mechanical modeling is an effective, computationally efficient approach to investigate structure-property relationships in disordered networks.