Paladin too overpowered, nerf their damage output!

[MauvaisOeil]

Treating the datas are showing the difference between PLD and war rarely exceed at 70+ percentile 5%, which is pretty low overall. I wouldn't call it a "high advantage to warriors", especially as high percentile logs are the lesser represented and can favor some RNG changing the resulting datas.

I guess everyone sees what he want to see by treating numbers.

[Lyth]

...

Someone tried to make this sort of claim before, but really, who gets to decide what isn't a "relevant difference"? Who gets to decide which percentiles we should consider, and which we should ignore? It's easy for a player to say that they only consider 70th percentile+ in job balance when they are barely in the 50th percentile themselves and the majority of the playerbase is less than one standard deviation away from the median.

Likewise, results on the lower end can't be discounted either. If you got 8th percentile on Neo Exdeath, it doesn't necessarily indicate that you don't know how to play your job. It provides information about what a first prog-to-kill result might be like, with high tank stance uptime, relatively fewer fight-specific optimisations, longer clear time, fewer loot drops, and perhaps a death somewhere in the mix. It's worth looking at everything, which is why I took the time to tabulate it all.

As far as dps differences are concerned, it's easy to say that we're "only" looking at 5% differences as compared to the 7-10% differences that existed between PLD and the other two tanks last tier. It makes it look nicer, sure, but we're looking at the same absolute differences, just with higher overall totals. The difference wasn't okay in Creator. The difference wasn't okay in 4.01, which is why WAR had all of their main resource losses removed in 4.05. Why should we be satisfied with it now?

I suppose the bigger issue here, and the one most relevant to this particular thread, is why would anyone complain about PLD's dps output as a WAR when when WAR is doing the highest dps, for the most part?

I'm going to come out here and make a bold prediction. If you keep everything pretty much the way it is now, except that you remove the MP costs associated with Shield Oath, Sword Oath, Grit, and Darkside, while making Sword Oath oGCD (i.e. give all three tanks the 4.05 WAR treatment), you'll see a much narrower dps margin between all three tanks. It wouldn't fix everything, but you'd at least be in a situation in which you could address all three tanks' concerns with equal weight.

They are also influenced by the number of players that have cleared on each job. Made this point after week 2, and even today it is still holding. 321 WARs for Neo, 1019 for Pally, and 818 for DRK.

The funny thing is, there are now 18184 WAR runs, 19497 DRK runs, and 29528 PLD runs. So while there are proportionally more WAR runs than there were before (from 14% up to 27%) and proportionally fewer DRK runs than there were before (from 37% down to 29%), the general trends haven't changed a whole lot over the past two weeks. DRK still does the lowest dps, and WAR the highest. It's just that people are finally catching on.

[MauvaisOeil]

Someone tried to make this sort of claim before, but really, who gets to decide what isn't a "relevant difference"? Who gets to decide which percentiles we should consider, and which we should ignore? It's easy for a player to say that they only consider 70th percentile+ in job balance when they are barely in the 50th percentile themselves and the majority of the playerbase is less than one standard deviation away from the median.

Likewise, results on the lower end can't be discounted either. If you got 8th percentile on Neo Exdeath, it doesn't necessarily indicate that you don't know how to play your job. It provides information about what a first prog-to-kill result might be like, with high tank stance uptime, relatively fewer fight-specific optimisations, longer clear time, fewer loot drops, and perhaps a death somewhere in the mix. It's worth looking at everything, which is why I took the time to tabulate it all.

As far as dps differences are concerned, it's easy to say that we're "only" looking at 5% differences as compared to the 7-10% differences that existed between PLD and the other two tanks last tier. It makes it look nicer, sure, but we're looking at the same absolute differences, just with higher overall totals. The difference wasn't okay in Creator. The difference wasn't okay in 4.01, which is why WAR had all of their main resource losses removed in 4.05. Why should we be satisfied with it now?

I suppose the bigger issue here, and the one most relevant to this particular thread, is why would anyone complain about PLD's dps output as a WAR when when WAR is doing the highest dps, for the most part?

I'm going to come out here and make a bold prediction. If you keep everything pretty much the way it is now, except that you remove the MP costs associated with Shield Oath, Sword Oath, Grit, and Darkside, while making Sword Oath oGCD (i.e. give all three tanks the 4.05 WAR treatment), you'll see a much narrower dps margin between all three tanks. It wouldn't fix everything, but you'd at least be in a situation in which you could address all three tanks' concerns with equal weight.

Because you are treating datas from actual fight parses like they were the potential of the job itself in a perfect environnement, while it isn't. Sample parses from tank depends from uptime of tanking stance and game mechanics. While the latter are the same for every tank, the first is not.

Technically the true potentiel of a job is mostly a matter of dummy DPS, the fight then comes and begin narrowing the edges, then comes team synergy, which is very clumsy in FFxiv as every job brings or brings not some tools that are effective over a school of damage, or an other.

That's why the lower percentiles aren't relevant, even if I'm pretty sure I would fit in them if I was even doing some savage, which I'm not. They can't be considered as a regular way to decipher the tank disparity in DPS, because they implies heavy factors where the warrior will have the stressless job of pushing damage mindlessly, while the other tank will have the job of taking most damage, maybe even only with shield oath.

That's why I don't understand why you want the whole overall parses to be considered as "the dps gap factoring warrior's proeminence", and Frankly I can hardly see it different from "The more we push that nonsense terriffic gap, the more we can hope for a buff in next patches that will push us over everyone".

We all know the paladin have the easiest DPS cycle for now, and frankly that's cool for them. So why do low percentile parses show paladins so low with warriors so high ? Why do higher percentile show them so close ? It's probably not because the skill ceiling of a paladin is higher, or the gap between their skillfloor and skillceiling.

Studying datas is one thing, but turning them to obtain a specific conclusion is quite mischievous. I'm trying to figure the why, how, and in what conditions have thoses been made. But frankly I'd prefer raw dummy parses with ilvl and no outside buffs. That's more speaking of the capacity of a job to obtain DPS, while adapting it to a specific fight is where the player skill will enter the game.

Considering the relevant difference between tanks, I guess a sample look at the difference between mid-tier (factoring medium debuff, none on the supports), and highest tier (Sam/mnk) shoud give a % of acceptance ? Let's be realistic, no dps either pure dps or tank will ever be balanced at the point we consider them equal, and the playerbase will allways choose what they think is faster or safer with a close to none loss, because it seems playing a game implies player the shortest amount of time actually playing.