High-resolution data from both ground- and space-based telescopes have enabled us to observe unprecedented details in stellar systems such as stellar multiples and star clusters. While it is well known that massive stars play a critical role in the evolution of these systems, their evolution remains highly approximated in most population synthesis codes. The evolution of massive stars is itself highly uncertain and model-dependent, with small differences in modeling assumptions leading to significant changes in their evolution. In this talk, I present results from the Method of Interpolation for Single Star Evolution (METISSE), which uses interpolation between sets of pre-computed stellar models to approximate evolution parameters for a population of stars. METISSE can readily make use of different stellar models, allowing the user to test the impact of various uncertainties in the massive stellar evolution. I discuss the importance of these uncertainties on the evolution and interaction of stars in binaries, and how they can impact compact binary mergers statistics. I thereby show that by using METISSE with population synthesis codes, we can better understand the contribution of massive stars in shaping the evolution of stellar systems.