Please decorrelate party support from party damage

[Shurrikhan]

What strawman? Every savage chart, no matter the type of party, shows the exact same patterns (BLM excepted).

Which has what to do with my statement that the point of balance should be to maximize the depth of competitive choices for Savage progression among decently skilled but imperfect parties (i.e., BLM would have a higher damage ceiling than RDM, which would have a higher damage than SMN, but not to the degree that someone especially bad at movement optimization would take BLM)?

If I'm saying it "should be X", you can be pretty certain it is not currently X. So how would my point be demonstrated in the existing charts?

Party damage, moreover, is a sum of its parts (not the sum merely of rDPS, no, because that is an incomplete metric, but absolutely the sum of it and aDPS for the given context, or simply averaged over any and all of them).
You cannot have damage disparity across compositions without also having disparity in potency-per-cycle and what portion of that potency may fall within raid buffs across the different jobs that may source those compositions.

I'm disagreeing with your point that somehow introduces differences between lower end and higher end in savage and raiding as I explained above once more.

...But I have never made that point. ???

I said only that making rDPS almost perfectly equal across all jobs is effectively to balance only for BiS speedrun parsers --since only they would be unaffected by the reduced reliability and ease of harder jobs-- while reducing breadth of choice for everyone else.

I've never advocated for different balancing paradigms based on what Week it is or what percentile one's at, nor anything of the sort. I have no idea how that would even work, and do not want anything like that. I simply pointed out who the actual beneficiary would be, outside of OTPs who happen to benefit from that imbalance, of making the likes of MCH perform as high as MNK, and it's not who most equal-rDPS-for-all advocates seem to think it is (since equal rDPS regardless of ease/reliability would damn near force less skilled players towards just a few jobs while maximizing choice only for the top 1%).

I also strongly disagree with anybody saying that rDPS shouldn't be equal across all DPS jobs, period.

Then we'll have to continue to disagree. While they should be closer, given the current context (much of the difficulty of melee having been reduced), equal rDPS for all jobs is position that benefits only the top 1% while screwing everyone else. So, no, I will never agree with that.

The target balancing point should be skilled but imperfect Savage raiders prior to farm/overgearing, not speedbarse runs.

Tl;dr: rDPS should not be equal so long as skill ceilings are highly unequal. Such just reduces choice for the majority of players. All jobs should have accessible floors and fairly high skill ceilings, with the same effort producing about the same performance among jobs that more or less equally click for a given player, rather than just giving additional performance for free to a few jobs (as would be the case if EW SMN or MCH had rDPS equal to a MNK, for instance).

Edit:
Heck, rDPS shouldn't be equal, regardless, by the mere fact that rDPS ignores the party synergy component of over a third of the roster (WHM, PLD, WAR, DRK, GNB, BLM, MCH, and SAM) -- and, no, homogenizing jobs to have an identical raid buff value is not a good way to patch up an insufficient metric. I don't know why people pretend rDPS tells the whole and complete story. It doesn't. It never has. aDPS tells the whole story for those without buffs, if of sample size sufficient to average out party composition, but for anyone else you need both aDPS and rDPS to get the full picture.

...Or actually use a potency map, including for different parties optimizing for comp raid windows when considering buffers, instead of just relying on fflogs bars to gesticulate at a mere part of the roster's contributions and insist that the product should all be equal despite their differences everywhere else.

[Valence]

Which has what to do with my statement that the point of balance should be to maximize the depth of competitive choices for Savage progression among decently skilled but imperfect parties (i.e., BLM would have a higher damage ceiling than RDM, which would have a higher damage than SMN, but not to the degree that someone especially bad at movement optimization would take BLM)?

I am not sure if I am expressing myself very badly, but i'll say it again: the patterns are similar no matter the skill levels of the party and show the same gaps of damage discrepancies between jobs in terms of balance, BLM excepted which slowly plummets to the bottom as you go down. I do not know how else to state it, but it proves on its own that uptime constraints are NOT a factor influencing skill levels a lot, at least not on the same level than balance discrepancies.

When it comes to support, I also said that even in lower skilled parties, bringing mitigation is a double edged sword, because lower skilled parties also use those a lot less optimally or at all (have you been in PF and checked how often people use their party mit besides healers?). On the other hand, higher skilled parties use mitigation very well, even though they rarely need it because the margin is already loose enough (unless maybe the last fight of a tier or ultimates, but then gear makes it a non issue past a few weeks anyway, and lower skilled parties don't go for week 1 clears anyway).

equal rDPS for all jobs is position that benefits only the top 1% while screwing everyone else. So, no, I will never agree with that.

I do not understand the meaning of that statement. This is why I brought up the chart patterns at various percentiles, since you brought up the top 1%. Those patterns are barely changing except for BLM.

I said only that making rDPS almost perfectly equal across all jobs is effectively to balance only for BiS speedrun parsers --since only they would be unaffected by the reduced reliability and ease of harder jobs-- while reducing breadth of choice for everyone else.

I think we're starting to run into circles, but as I said, an equal rDPS across the board also asks for similar accessibility levels for every job (and if possible similar ceilings as well), even though difficulty is subjective as hell, but it's not exactly rocket science that a job like SMN doesn't play on the same field than BLM.

But even without balanced floors and ceilings, having equal rDPS is a lot less obnoxious and toxic already than balancing over difficulty, which frankly, sucks the most out of all options, because it sucks the fun out of everything.

Tl;dr: rDPS should not be equal so long as skill ceilings are highly unequal. Such just reduces choice for the majority of players.

Yes, and how different is this from what we have already that advocates for skill balance? The most used jobs have always been the ones with the better damage output (including raid buffs), the only exception, again, being BLM. Ease of play only comes into the equation when the damage differences aren't enough to justify taking something better, unless we're talking about... BLM. It's always been like this since as far as I can remember.

And of course you'll always find players going for the underdogs as well.

All jobs should have accessible floors and fairly high skill ceilings, with the same effort producing about the same performance among jobs that more or less equally click for a given player, rather than just giving additional performance for free to a few jobs (as would be the case if EW SMN or MCH had rDPS equal to a MNK, for instance).

Then what are we arguing about? This is confusing?

Heck, rDPS shouldn't be equal, regardless, by the mere fact that rDPS ignores the party synergy component of over a third of the roster (WHM, PLD, WAR, DRK, GNB, BLM, MCH, and SAM)

There seems to be a mathematical fallacy here. Those jobs are affected by rDPS like any job. They just don't produce party wide raid buffs out of their kit, which is something rDPS also takes into account in the damage potential they bring to the party.

rDPS is a purely balancing tool between what jobs bring to a party overall. Obviously there will be finer details with a way lower weight in how some comps will skew rDPS a little (like fielding SAM/BLM with a DNC or BRD to inflate a little the latter's rDPS results), which is true though. But that's just not the same weight and I didn't want to enter into that kind of finer print before addressing the rougher outlines.

aDPS tells the whole story for those without buffs

No, that is a common misconception. aDPS is great to indicate when party buffs have been taken care of optimally, that's about it. Else it siphons a part of the damage that a job didn't bring on its own to the field out from another job.

Don't get me wrong, I do agree that all those metrics are flawed, but rDPS is the one that offers the best picture there is so far. It doesn't mean it has to be followed blindly, and as a stat freak I'd like to have better metrics, but that's what we have for now.

[Shurrikhan]

No, that is a common misconception. aDPS is great to indicate when party buffs have been taken care of optimally, that's about it. Else it siphons a part of the damage that a job didn't bring on its own to the field out from another job.

I never said to use it for considering the balance of every job. I said, very specifically and explicitly, that it tells the whole story as averaged across various party compositions of the jobs without any buffs.

Consider:
If you had a job that was in fact just a mobile totem and brought with it zero attacks, only raid buffs, rDPS would tell its whole story averaged across whatever various party compositions could be taken with it. That's because it'd have zero ability to exploit others buffs in/by itself. The difference between its aDPS and its nDPS would be zero, because it would have neither (zero buff-exploitation possible)

In the same way, a job without any buffs has zero buffing capacity and therefore no difference between its rDPS and nDPS, both trimming a valuable portion of their contributions (buff exploitation) while aDPS trims nothing.

So, yes, non-buffers are a lot easier to judge balance between (now that we're no longer pigeonholed towards Stormblood-style set comps). Everyone else requires one to look at both metrics, aDPS and rDPS.

Never, though, will rDPS be a decent metric by which to judge WHM, WAR, PLD, DRK, GNB, BLM, MCH, or SAM. rDPS is specifically designed to ignore non-buffers' party-synergetic contributions.

(If there's no buff exploitation, then the buffer brings zero rDPS over their nDPS. The exploiter is half the picture, and you cull half that partnership in insisting on fixating on rDPS alone.)

I do not understand the meaning of that statement. This is why I brought up the chart patterns at various percentiles, since you brought up the top 1%. Those patterns are barely changing except for BLM.

Fcs... It has nothing to do with what the rDPS at each percentile are right now. That would be like my responding to your "All jobs should have equal rDPS" by posting a picture of a current chart and saying "But they aren't." 'What is' is not the same thing as 'what should be'. We wouldn't be talking about this if what we wanted had already happened.

If you make it so every job has equal rDPS, not only do you overbuff every not buffer, making playing any buffing job a thankless risk, but you also make it so jobs with far less effort required to be optimized are essentially getting free value (which, yes, is already the case for SMN relative to RDM, etc., and is a problem, not something to be made yet more pervasive).

Again, I'd prefer to raise their skill ceilings instead, but in the meantime, no job should have greater throughput for the effort put in than the other jobs get. That applies as much in one job doing the same for less (see, SMN > RDM) as it does in one job doing more for the same (if all casters had the same --internal and contextual, together-- skill ceilings but BLM somehow still outperformed them).

Finally, as Shurrikhan said, SE does not balance jobs based on the top players.

I'm not perfectly sure where they balance it around, nor do I care, as the balancing philosophy itself (what criteria are rewarded and to what degree) is inconsistent in their application. Mine was a suggestion, not commentary on what is.

Again, I just think that both rDPS and aDPS need to be accounted for and that the balancing target should be Savage Progression by decently skilled but imperfect players. Because anywhere else, competitiveness of job choices rapidly stops being a bottleneck to breadth of job choice anyways.
_______
Aside / Not at Aco:
Perfect rDPS parity across all jobs would be a shitshow. It'd be less terrible if all jobs had roughly the same skill ceiling, but even then it'd slightly to noticeably overpower 8-10 jobs (since some jobs are especially strong exploiters even while also being buffers) over everyone else. I don't understand why some have trouble understanding that. It's a simple matter of what is or is not included in the metric and the disparities in the other aspects of jobs design.

[Aco505]

I'm not perfectly sure where they balance it around, nor do I care, as the balancing philosophy itself (what criteria are rewarded and to what degree) is inconsistent in their application. Mine was a suggestion, not commentary on what is.

No one really knows. Even now we use terms like rDPS, aDPS and nDPS when none of us has a clue of which metrics SE uses. They probably have their own, and their own criteria for balancing can sometimes seem to be a bit random.

But we use what we have at our disposal and what is clear is that they don't balance the DPS checks of the fights around 99 percentiles or the top playerbase. Otherwise, they wouldn't have nerfed the health values of P8S.

In any case, the point of my comment was to emphasize the fact that the enrage timers of Savage raids appear to be adjusted around a more average type of player.

[Valence]

Consider:
If you had a job that was in fact just a mobile totem and brought with it zero attacks, only raid buffs, rDPS would tell its whole story averaged across whatever various party compositions could be taken with it. That's because it'd have zero ability to exploit others buffs in/by itself. The difference between its aDPS and its nDPS would be zero, because it would have neither (zero buff-exploitation possible)

In the same way, a job without any buffs has zero buffing capacity and therefore no difference between its rDPS and nDPS, both trimming a valuable portion of their contributions (buff exploitation) while aDPS trims nothing.

This is a slippery slope at best, because the heaviest buffers in the game right now still get 80-90% of their total damage contribution from their own personal damage, which essentially puts them closer to fully selfish jobs than any 'rdps totem'. This in essence is what makes rDPS have immensely more weight when getting an overview/snapshot in the general trends of job balance. An aDPS chart on the other hand, will tell you very little on its own. Obviously, you can then use it to get into finer details once you're done with the former.