Medical education literature regarding assessment is traditionally based on psychometric insights. In psychometry, validity and reliability are essential parameters of assessment. A literature movement originating from the Maastricht school, which is extremely influential in the Netherlands and beyond, argues that the clinical context is too complex for traditional standardized assessment methods, and proposes to optimize the use of expert opinions. In this commentary we address the risk of bias and noise in human judgement. Multiple assessors in a clinic may share the same bias regarding certain trainees or specific behaviors. We make a plea for the use of objective data, like results from knowledge tests and objective structured clinical exams, when assessing clinical trainees. The combination of both objective data and an aggregation of expert opinions may be most feasible in clinical assessment, both for learning and for entrustment of professional activities.