Understanding equilibrium climate sensitivity (ECS, warming in response to a doubling of CO2) uncertainty is fundamental for making reliable climate projections. We leverage the Hector simple climate model in a probabilistic framework to explore how different ECS priors influence long-term (2081-2100) temperature anomalies. Specifically, we quantify how different ECS prior distributions broaden the uncertainty in temperature projections. This method demonstrates a computationally efficient probabilistic workflow that explores the effects of parameter priors on climate projections. Excluding process and paleoclimate evidence widens the resulting temperature projection uncertainty (1.12-3.03 ℃ and 1.09-3.33 ℃, respectively) while using priors that synthesize all lines of evidence leads to narrowed temperature projection uncertainty (1.24-2.89 ℃). Our study shows that excluding paleoclimate and process data in ECS priors broadens temperature projection uncertainty. In contrast, synthesized evidence provides a narrower and potentially more robust range of future temperature outcomes.