Proteolysis targeting chimeras (PROTACs) are bifunctional molecules that induce ubiquitination and subsequent degradation of proteins via formation of ternary complexes between an E3 ubiquitin ligase and a target protein. Rational design of PROTACs requires accurate knowledge of the native configuration of the PROTAC induced ternary complex. This study demonstrates that native and non-native ternary complex poses can be distinguished based on pose occupancy time in MD, where native poses exhibit longer occupancy times than non-native ones at both room and higher temperatures. Candidate poses are generated by MD sampling and pre-ranked by the classic MM/GBSA method. A specific heating scheme is then applied to induce ternary pose departure, generating an occupancy score and temperature score reflecting pose occupancy time and fraction. The scoring approach enables identification of the native pose in all the test systems. Beyond providing a relative rank of hypothetical poses of a given ternary system, the method could also provide empirical guidance to whether a given ternary pose is likely a native one or not. The success of the method is in part attributed to the dynamic nature of the pose departure analysis which accounts for solute entropic effects, typically neglected in the faster static pose scoring methods, while solute entropic contributions play a greater role in protein-protein interactions than in protein-ligand systems.