by linspatz » 09 Dec 2017, 16:39
I think you are right, the percentages you have are the chance for an item to drop within a subgroup and not for the subgroups chance to drop that item.
For example if you look at quest scrolls as a subgroup the data I have seems to fit. Instead of having percentages treat the numbers as whole numbers. C-grade is 203, B grade 23, A grade 7 and S grade 1. Adding them all together gets 234. Dividing the whole number by the parts gets the percentages, ie, C grade is 86.75% chance to drop when it is a scrolls turn.
Now if you look at my data I had 51 C-grade, 5 b-grade and 2 A grade scrolls drop for a total of 58. If you take 58 times each of those percentages you get that I should have recieved 50.32 C-grade, 5.7 B-grade and 1.74 A-grade scrolls. It could be a coincidence but I think that must be how it fits.
This also works with my data if you group all the items in the 400,000 to 500,000 ID range as one group.
Equipment is rather tricky though. As far as I can tell for equipment it my be grouped by armor, weapon, accessory and then by grade. Maille Lizardman Pursuer have the same percentage for all B-grade and C-grade weapons and armor but the drops greatly varied. The variation is not made up by the fact that their are 24 armor of each grade and only 18 weapons for each grade. This is why I think they are in different subgroups.
Just my thoughts.