Very good point!! I was about to walk away free but you were like "hey, come back"..
Possible solution: I can create a frequency table of character lengths of branbucket inventory VS character length of sales list and plot a graph.
Now If we observe there are direct correlation between the two then we can say the available character lengths has direct impact of what get sold and hence ranking 4L as 7/14 will be inaccurate and unfair but if no reasonable strong correlation then my initial claim remains true and case will be closed.
Shall we begin? Which time frame of data set should we look at?
From March 2016 - March 2017?
(Job title: Unofficial BB Sales Reporter)