We cannot achieve balance while DPS jobs have utilities

[Shurrikhan]

I disagree..
rDPS is the most important calculation method if you want to see if job performance

There is no single metric by which to account for the performance of any job that brings part of its contribution through buffing its raid. There is a single metric from which we can, over enough parses across varying contributions when those compositions are roughly balanced already, compare the performance of jobs who do not do so, but it isn't rDPS.

For every buff by which someone contributes varyingly towards the rest of the party, there are also varying degrees to which other jobs exploit that buff and therefore give the buffer their rDPS. Because of this, to know how much, say, a Dancer, can maximally bring, we need also to compare the raw damage of its potential dance partners under roughly the same conditions, or aDPS across a number of parses large enough to be less dependent on the specific composition.

While that number needed can then be quite large, exactly the same can be said of rDPS. But while rDPS necessarily and badly misrepresents the value of every job without a buff, it still doesn't correctly represent the value even of those who give more value via their buffs than they receive from others' buffs.

You use both. Both aDPS and rDPS, while also slightly mindful of the gap between aDPS and raw DPS of single-target buff recipients when considering the actual value of said buffers or their buffs best users in a composition with those buffers.

I would love to play job has higher rDPS than other jobs it is better for party and kill

Which, given that you seem to think rDPS just clearly and cleanly means "amount of value", was my point. People tend to whatever produces the best value for the amount of effort they're willing to put in, regardless of whether it has more or fewer support actions or a greater or lesser portion of its value wrapped up in said support.

If you look at fastest 5 clears for this savage you will notice something: All parties dps are supports!! why ?

Let's say we consider jobs that provide far more value by exploiting others' buffs than by giving out their own as "support" any time they give out at least one damage buff or carry at least one emergency external tool like a resurrection skill, such that there are only 3 non-support DPS in the whole game (SAM, VPR, BLM)...
...Even then, your claim is objectively incorrect.

Black Cat: MNK, SAM, BRD, PIC; SAM, VPR, DNC, PIC; MNK, SAM, DNC, PIC; NIN, SAM, DNC, PIC
B. Lovely: RPR, VPR, DNC, PIC; MNK, SAM, DNC, PIC; MNK, SAM, DNC, BLM; MNK, SAM, BRD, PIC; SAM, RPR, DNC, PIC

Across the top 9 for Brute Bomber, we see a SAM and/or VPR in each party, with an "all support" party (RPR, MNK, BRD, PIC) not appearing until place #10.
The only one to have an "all support" comp in the top 8 at time of writing --and at great stretch of the term "support"-- is M4 (again with MNK, RPR, BRD, PIC).
_____________________

No, really, go and look at the actual data. In doing so, you might even find a much larger takeaway worth noting: that PIC was used in 31 of 40 fastest runs across the game. Compare that to VPR's 22 of 40 despite being equally OP relative to its role (but that role being bigger) and similarly and new job... and you'll start to see what kinds of job push-outs are actually happening, such as PIC just being a straight better BLM 99% of the time. (VPR, btw, similarly pushes out NIN --and to a lesser extent RPR, DRG, and nearly MNK-- for simply sharing a gear class with NIN but being far superior to it.)

it clearly an issue that we see more pure jobs will suffer from utility jobs

No, they suffer when they generate less party value (especially, per party effort invested) -- i.e., when they're undertuned. But that's it. The mere existence of support jobs does not push out non-support jobs. Being undertuned pushes out jobs. In general. Regardless of portion of indirect contribution. Regardless of degree of support tools.

Requiring but capping "support" jobs helps nothing outside of increasing queue times for those who can only make themselves do work while waiting on a duty to pop. Which, to be clear, is generally bad for the rest of us.

Saying the truth is important,

first 10 clears for savage there is no BLM and few VPR and SAM
Other jobs appearing because there are utility jobs (supports) who do less 2% of selfish jobs which is crazy

If by "first" you mean "first to clear":

"No BLM": BLM parties were the third to clear M1S, fourth to clear M2S, first to clear M4S, lagging behind (swapping briefly to PIC, most likely) only on M3S (32nd clear).

"Few VPR and SAM": In the third clear, and in roughly half of first 20 clears thereafter of M1S. 6th clear, and in roughly half the top 20 thereafter of M2S. Second clear of M3S, third clear of M4S, etc. VPR, especially is slightly overrepresented even in early clears.
______________________

If by "first" you mean "nearest first place (in speed)":

"No BLM": At time of writing, there is a BLM in the top 7 of M1S, top 3 of M2S, top 6 of M3S, and top 5 of M4S. There are 7 BLMs overall, making up 7 of 9 non-PIC caster choices in the top 10s of each run (40 runs total).

"Few VPR and SAM": VPR is in 22 of 40 top-10s across M1S-M4S. SAM is in 14 of 40. In a random distribution assuming 2 melee DPS per raid, we'd have expected to see 13 of each. Both SAM and VPR are represented in the top runs far more often than more supportive Melee jobs like DRG (6) or NIN (5).

Seriously, look at the data before making claims on it.