This article examines assessment rubrics for the oral examination in the Italian upper secondary school final exam, focusing on the relationship between evaluative discretion, explicit criteria and standardisation. Drawing on a historical reconstruction of the exam from early local practices to the introduction of national grids, the study shows how oral assessment was long grounded in implicit criteria and broad judgements on students’ maturity, personality and cultural orientation. Since 2020, ministerial grids have attempted to reduce local variability by organising the oral examination around indicators, levels, descriptors and scores. The article critically analyses this shift, discussing whether such rubrics actually improve validity, reliability and transparency. It argues that, while common rubrics can make criteria more visible, their effectiveness depends on the quality of indicators and descriptors. When descriptors remain generic or mainly translate qualitative labels into scores, rubrics risk becoming a form of disguised grading rather than a tool for interpreting learning.