Why isn't there a shortage of healers?

[Kemiko]

This is definitely an interesting bit of information - but it doesn't take into account the introduction of NIN or probably even the changes to WAR (I forget when in the years that happened). But, my experience suggests that Healers are not 3% more common than Tanks are. Though, I'm sure it's just my server (Faerie) as each server likely has a different role population.

For example, I have 131 Friends on my Friendlist. Out of all the people on that friendlist, only 9 of them are Tanks. 31 of them are Healers and the other 90 are DPS. These figures are what roles my friends main - not what they have geared/leveled to 50. Now that information I don't have memorized. Excluding my experiences in my LS (everyone in my FC are on my FL) as I don't even know/played with everyone in my LS's.

Though, I expect a slight role population change come the Expansion. Nothing major but, perhaps a slight increase in Tanks and Healers.

[Saseal]

People like me probably also help in messing up the stats in that chart.

Main character, I'm on WHM about 98% of the time and play that character the majority of the time (with exception to recent hand injury, in which I forewent WHM for a few months and only played BRD since it was less stress) - But then I have three other DPS alts with BRD played a good chunk of the time, as well (with some lalafel and hyur alt playage sprinkled in). A level 50 BRD, a 31 ACN (uh.....got caught up in story quests and FATEs in the area and didn't change, yet... but I'll be changing him to SMN/SCH to further mess up stats) and level 36 DRG. Thats not even mentioning 40-something SCH on my main character.

I'm wondering how they collected these stats - what parameters were required to record whether the character was identified as one job over the other - by account or by character, by 50s or by hours played, etc? As well, pretty much like someone else said, they have multiple/all DPS jobs 50'd, but does that count if some people don't play certain classes after hitting 50? I don't play my MNK job on my main enough for it to count (being the FC's primary healer).

It's probably collecting ALL data, but 50'd alts and multi-50'd jobs on a single character are probably skewing those stats (and I know I'm not the only person with multiple alts). Unless they're taking the highest percentage of play duration on one specific job per account... (which makes more sense to me, as that would signify your main or preferred job).

Also, yeah.... the lack of SCH might be skewing the healer data, also... but that could just be they didn't label it correctly and the data isn't wrong.

[Cynfael]

As long as you do not now what data SE is exactly collecting about our online-behavior, how can you say "easy to figure out"....

So we don't know what data they are collecting exactly implies we have absolutely no idea how they work with said data and everything is more or less technical guesses.

Exactly. Statistics in a relative vacuum can mean whatever the reader wants to interpret them to mean.

There's nothing to "figure out" because not all of the data is provided to us.

[Risvertasashi]

Honestly, I am a bit worried that they just took a lodestone-esque snapshot. Whatever people were on at the time the snapshot happened is what they were recorded as. Also note that, being the 2014 census, it was well after the changes to Warrior happened.

But, in the end, Square isn't sharing their methodology so that does limit the conclusions we can draw...

Well, here's to hoping that DRK gets the tank population up near the 25% mark.

[KisaiTenshi]

While what you suggested is one possible method to tabulating the results of the census, I don't think what it's how they calculated the results. On my server, there is a plague of what I believe to be RMT BLM bots just going in and out of instances - I'm pretty certain this is the same for many other servers as well (except maybe the eternally locked Balmung). Therefore, if they did this method of tabulating census, it would probably skew the results heavily towards showcasing these 24/7 BLMs.

The BLM bots are the same on Excalibur/Primal DC.

Like if I were running the game census, I'd just count the hours spent in each job to figure out which job is "main" to rank each job the player has, and then use that to tabulate the game-wide statistics.

At any rate, the data is from the middle of last year, so they could very well do something else this year.

[Spoekes]

For duties, wouldn't it be an easy way to just filter out what jobs have registered for duties and go by that result?
This would only work for combat classes obviously, but eliminate all "idle" factors and such, while providing a reliable figure of the jobs that people do duties on.

[Khalette]

Honestly, because the GCD. I always play tanks in other games, but the GCD makes it feel so bogged down -- especially in emergency situations. Being able to cast helps that boggy feeling so in this game I prefer casting classes.

[KisaiTenshi]

For duties, wouldn't it be an easy way to just filter out what jobs have registered for duties and go by that result?
This would only work for combat classes obviously, but eliminate all "idle" factors and such, while providing a reliable figure of the jobs that people do duties on.

That would work if we were only looking at Duties and not Mainstream storyline, sidequests and all the crafting stuff. Like it's very hard to really go "let's see how many players are X role on Y date and Z time" and then say that's an accurate census. Take a look at the Lodestone at any one day and see which Role you are set as. Even the forum avatars tend to shift around slowly. So clearly SE has a way of eyeballing the current stats at any moment in time, but they need something to filter out otherwise the bots and RMT will skew towards BLM, and long-time hardcore subscribers will have 50'd everything even if they have no intent of playing all of them.

Like just counting level 50 jobs wouldn't work. Because that doesn't mean the player actually plays it. With the Combat jobs, simply counting how many Relic weapons there are would probably give a more accurate number and then further classify that by which one(s) are maxed out. I find it hard to believe that everyone would have a level 130/135 weapon for every single combat job. So given the ilevel requirements are 90, just counting mainhand gear over ilevel 90 would work. With DoH/DoL, that's a bit more complicated, and there are Lucis main hand gear.

So I guess what I'm saying is "figuring out" what a player mains could probably determined from statistical data (eg what they queue as), but I somehow doubt there are that many people who "main" crafting, and any you see in the wild are likely peoples alts/mules or bots. So SE could just as easily be looking at only one character per account.

[Spoekes]

That would work if we were only looking at Duties.

Yeah I said that in my first sentence.
Would at least be a figure quite representiative concerning df queues.

[Sida]

Looking in the data given, what strikes me is how equal it looks. The tank classes are slightly more common on 14% or so but still all of them are very near 10%. That actually makes me think the data gathered is indeed, just taken every class every player has and lumped together. Since many long time players have maxed every class, that'd serve to give heavy even base for the result.And of those who provide the variation, that is, those who only play single dmg class, many have also tank class levened to some degree for bearable queue times, thus giving slight over presentation of tanks. But that's just me thinking such census is gathered lazily.