I started to wonder if the fact that I was using the count of dates had something to do with this problem, so I used other fields. Still the same problem. From what I can tell of the COUNT definition, it discards NULL values in the csv file. So for example, every assignment has a Client Name attached to it, but if it is not repossessed it does not have a storage location. So a COUNT operations on a record that opened and was closed without being repossessed would yield 1 for COUNT[(Client Name)] and 0 for COUNT[(Storage Location)]. Based on that, in theory, I can then determine the number of orders and the number of recoveries then graph them over time. Additionally, I can divide the count of Storage Location by count of Client Name and get the recovery rate.
Problem is, when I am graphing these numbers over time I am getting the exact same graph for Orders and Repos. On top of that, the recovery rate is remaining 100%.
Any help would be greatly appreciated.