Interestingly enough, I'm running a Poll elsewhere which involves a small sample of people voting for a "best" sensor model.
https://www.dpreview.com/forums/post/63761776
Reading the above, it occurs to me that a panel of judges is quite a small sample, statistically speaking.
Take, for example, five judges, three of whom award 8/10, one awards 7/10 and the other awards 9/10. 60% of those 5 judges said 8/10 but the margin of error is a whopping 43%, so if there were say 100,000 judges the chances of getting 8/10 from that crowd is somewhat variable ...
https://www.surveysystem.com/sscalc.htm
Have a play, perhaps taking the USA Supreme Court (13 judges) as an example ...