
Originally Posted by
Renathras
So I compiled some data for numbers. I find it honestly too suspect to be useful, but the Wayback Machine's crawls of FFXIV Census. The issue I have is I can't tell enough from the numbers if FFXIV's census is trying to capture mains or not. It seems to be more "Job at level 90 on a player that has completed the current expansion's MSQ", but that's a pretty low bar. It MAY be counting only one Job per player, but I'm not certain enough to be sure. It is still interesting that you can see some trends, like DPS getting a spike in SB and again in EW (especially Melee in Meleewalker), and that Tanks got a bit of a spike in ShB and Healers got a small one in EW; but it's hard to draw concrete conclusions from this since that could all be artifacts of the data (another Job being leveled alongside existing ones for someone that's a dual/tri/omni-main of a given role will inflate its percent unless the data only counts one per player). The FFXIV Census data also had an error for two years, though this only affected "all characters" not the endgame "active characters" count - it missed/counted as deleted characters who didn't have at least 1 mount and at least 1 minion...but that's mainly going to be less than level 20 characters, so it's pretty safe to exclude them from the count. But that said, the data DOES help to draw conclusions in the sense of telling us what arguments are likely NOT accurate. For example, if we expect a massive decline in Healers but the percent goes up, even if we factor part of that as people leveling multiple Healers and so counting a second, third, or even fourth time, if the decline was massive, we'd expect to see a hint of that in the data. We do not.
The Lucky Bancho numbers are harder to get via the Wayback Machine, and I can't read the original Japanese anyway to really understand the older datasets. Translations on Reddit and the Official Forums tend not to be made by the same posters or with the same topic/thread titles, and aren't a static address to have been archived by the WM's crawler, anyway. I wish they were, because I find that data's more reliable (judging Job count by "Person was logged out on it" is a bit dubious, but it is by "Has current patch Extreme weapon or better" is a good measure for "the person actively plays this Job...probably; or at least considers gearing it important enough to spend Tomes/Totems on doing so"), but it is what it is.
The FFXIV Census data, given those caveats, is still the best data I could get my hands on, and it shows Healers as holding approximately steady. Here's the Healer % vs All Jobs over time:
HEALER %
2017: 0.25047
2018: 0.215359
2019: 0.191224
2020: 0.172927
2022: 0.19941
2023: 0.205038
What we can see from this is that Healer Jobs at level cap with MSQ completion declined as a percentage of the whole from HW through ShB, but then started increasing into EW. PART OF THIS is for people maining multiple Healer Jobs, they now had another one to main, but that only really explains the people (like me) that level all of them. While that's likely a portion of the data, it doesn't tell the whole story. But the super short version is: "The Healer population declined in middle-to-late HW going into SB, then declined further going into ShB, before somewhat reversing and increasing to about the SB level in EW."