This study compares Signals Passed At Danger (SPAD) severity assessments between train drivers and railway safety advisors. The classification and assessment of 58 SPADs caused by driver error in a Norwegian railway operating company in 2015 were analysed to identify differences between SPAD assessments conducted by train drivers and safety advisors. The results showed a normal distribution among the railway safety advisors, while the train drivers' assessments were skewed to the less severe. There was a significant delta between the assessments of the train drivers and the safety advisors. Thus, the analysis and discussion of these differences focus on the setting in which the assessments are made, the motivation for the assessments and the gap between the sharp and the blunt end of the organization. The differences between the assessments are thus interesting in a safety cultural perspective.