The subjectivity inherent to assessment of tornado damage has been discussed in the literature since the advent of the F-scale, but with little quantification and few actual examples. The statistical results of an interactive F-scale damage rating exercise are discussed here, illustrating the large range of ratings assigned to damage scenes and the strong judgmental steering of the process. This exercise was performed across a wide range of damage survey experience within an audience of meteorologists, engineers, storm spotters and chasers, and lay people – each representing groups whose input has been considered in NWS tornado ratings. These results, along with other issues documented in the literature, support creation of well-organized training programs on damage assessment for non-engineers involved with surveys, as well as formal establishment of a ready response team of experts from the meteorological and engineering communities who can be deployed quickly to major events (analogous to NTSB practices). Other methods are also suggested for mitigating the impact of F-scale subjectivity on tornado ratings.