Job Balancing 6.08 (still needs work)

[Shurrikhan]

There’s a reason there’s such a ranged DPS shortage. No one wants to play that role. The ranged tax needs to either be removed, or reduced.

It hasn't been a "ranged" tax in a while. It's a "somehow far simpler than even the average FFXIV job" tax atop rez perhaps being mildly undervalued relative to TacticalSambaToir (though that balance is difficult to pin down, since rez has no value until things go wrong). Note the accordant differences between Bard, Dancer, and Machinist. The Physical Ranged role is no monolith despite their shared relative hypermobility.

[Aikaal]

If you look Phoinix or Hesperos 2 parses RDM is only around 250 DPS behind BLM right now and the gap is only getting closer once people are closer to bis gear and play better. If BLM would not get that massive 300+ DPS buff, it would already be behind RDM in 100% parse in Phoinix and Hesperos 2. Balancing selfish and buffer jobs is always kind of tricky because buffer jobs have so much more DPS potential after prog. Average DPS is always BLMs flavor because it is easier to have a good run with one job in a single run than a good run with 8 different people. To balance jobs around average DPS BLM would be out of the question for a speed kill meta. I think casters are currently pretty fine if SMN would be buffed somewhere between RDM and BLM right now.

To me, it looks like P3S is just a terrible fight for SMN/BLM and a really good fight for RDM. BLM is far ahead in any other fights. BLM is either top or second best on all fights but somehow is behind all Melees in P3S. It's fine this way. I still think RDM needs a minor buff and SMN needs a significant buff. Resolution is massive on dark flames.

[Kabooa]

It hasn't been a "ranged" tax in a while. It's a "somehow far simpler than even the average FFXIV job" tax atop rez perhaps being mildly undervalued relative to TacticalSambaToir (though that balance is difficult to pin down, since rez has no value until things go wrong). Note the accordant differences between Bard, Dancer, and Machinist. The Physical Ranged role is no monolith despite their shared relative hypermobility.

The primary wall I run into with Machinist is that all the buffs people want to put into it means I start destroying literally everyone but the max parses they refer to when pointing out how weak Machinist is, and frankly to me that feels wrong. Maybe it's stockholm syndrome, but I don't consider myself good at the rooty tooty run and shooty, just practiced, so even something that sounds mild like "Just removing the ranged tax" or "Increase to be the same as X melee" or something similar seems excessive, and grossly so.

Insert blah blah role system needs to be revamp blah blah shove the problem children into one role blah blah.

[Shurrikhan]

The primary wall I run into with Machinist is that all the buffs people want to put into it means I start destroying literally everyone but the max parses they refer to when pointing out how weak Machinist is, and frankly to me that feels wrong. Maybe it's stockholm syndrome, but I don't consider myself good at the rooty tooty run and shooty, just practiced, so even something that sounds mild like "Just removing the ranged tax" or "Increase to be the same as X melee" or something similar seems excessive, and grossly so.

Insert blah blah role system needs to be revamp blah blah shove the problem children into one role blah blah.

I think people just tend to forget that top-end balance, when some jobs are more difficult* than others, essentially obliges any less-skilled players towards playing whatever is easiest*, since there will never be a benefit for learning a more complex job as, even at the top 1%, they'll contribute equally.

* Though difficulty is idiosyncratic, hundreds of thousands of data points upon those idiosyncratic evaluations still creates... statistically relevant data. As such, yes, there are classes which are normally "easier". Those norms will not apply to every player (nor precisely to even a majority of players) but at the scopes which concern balancing decisions, they are wholly relevant.

I feel like a sensible general balancing paradigm, while perhaps difficult at times to calculate, shouldn't be that elusive:

• Jobs should be balanced for around the 90th percentile when including the more difficult to calculate in-practice rDPS effects of utility that isn't mere a damage bonus (e.g., Mantra, Shield Samba, or the rare ability to lose less party damage to doing a mechanic by having a hypermobile role take care of it). Prior to these calculations, jobs without "non rDPS" utility would appear to outperform those with it, while in practice they would be tightly balanced given a party of equal skill (percentile) to the contributor.
• Jobs favored below the 90th percentile should not also be favored above it.
• Ideally, jobs with utility value that is greatly increased at lower party performance (e.g., rez) should be "harder" to reach a comfortable skill floor on --preventing them from being locked-in at lower percentiles-- insofar as does not step excessively step on the toes of "pure" dps classes whose gameplay concerns must come solely from rotational complexity. (I.e., Ideally, SMN and RDM would have a noticeably larger % personal DPS gap between the 60th and 90th percentiles than BLM as to compensate for BLM's offering nothing else at lower skill levels.)

Now, to be clear, choosing to balance around a given percentile does not mean that you cannot make design decisions to further tighten balance at any/every given percentile, but it does set a threshold around what error we consider to be more or less a function of kit complexity vs. irrelevant/transient (soon to be passed through learning) player mistake. The higher a percentile we balance "around", the more the lower percentiles are obliged towards easier jobs and job diversity is reduced below that point; the lower a percentile is targeted, the more the mistakes made on theoretically stronger jobs are overcompensated for, thus reducing job diversity above that point (as the likes of Monk and BLM get buffed to the heavens to compensate for their being harder to learn).

(So long as the gaps at lower percentiles, too, aren't so huge as to make harder jobs seem a burden, I argue for a higher percentile because as we descend from the top 10% or so, aesthetic tends to have increasingly guided our job choice anyways.)

[Gruntler]

Now, to be clear, choosing to balance around a given percentile

is completely irrelevant to how this game is balanced and we really need to abandon this mentality. Jobs are not balanced specific to any percentile. FFLogs is not a balancing tool, never was, and never will be.

The data more likely to be used would be clear rates, relative difficulty to clear across multiple spectra, range of damage outputs across intended skill levels of player, and other metrics FFLogs either cannot measure with accuracy (because of rounding errors due to incomplete information) or does not have access to at all.

Because we only have damage output though, that's all we think about. But a game designer of a game where speedrunning is not something they design the game for would not design the game around what people do in speedruns, and while farms are not completely irrelevant, they're not nearly as relevant as the community would hope.

And if you were balancing around damage, best practices for data collection would require that you look at spreads, standard deviations, exclude outliers, and look at what the experience is across all players. And you certainly would not only look at clear data and the intendent selection bias that forces. How does looking at clears tell you how difficult it was for a job to GET to a clear?!

Anyone arguing for a single percentile is automatically wrong at the outset and doesn't grok data. And FFLogs might be the best data we have, but it's not nearly complete or accurate or valid enough for most of the arguments most players want to make in game balance conversations.

[Shurrikhan]

[targeting a threshold of performance] is completely irrelevant to how this game is balanced and we really need to abandon this mentality.

The data more likely to be used would be clear rates

Then our balance is already perfectly sufficient. Every job can clear with little difficulty essential to its kit. Done.

Because we only have damage output though, that's all we think about

It's sure as hell not all I'm thinking about, except in the sense that utility is only as valuable as the average improvement to clear times (including whether you clear at all) it produces. If it does not increase your chances of clearing and/or reduce clear times, it has done nothing.

The criteria in the post you're quoting already follows from consideration of spread, deviation, and outliers. But let's be clear here: an entire 10% of players is not going to be an outlier. As you go lower, moreover, the parses are that much less representative of their kits, as opposed to simply varied, and often short-lived, forms of mistake.

And if you were balancing around damage, best practices for data collection would require that you look at spreads, standard deviations, exclude outliers, and look at what the experience is across all players. And you certainly would not only look at clear data and the intendent selection bias that forces. How does looking at clears tell you how difficult it was for a job to GET to a clear?!

How does the phrase "to be clear" have anything to do with clears? That's literally the only the time that word appears in my post -- within that phrase. Where is this strawman coming from?

[CKNovel]

If we balance around the idea of "every job can clear" then there is no problem to buff ranged.
Because all jobs will be able to clear.

With that logic you could put BLM dead last and it wouldn't matter since "every job can clear".

It's true you don't balance a game around speedrun, but this logic applies to "single run" games.
XIV is a game where you repeat tasks and boss battles. Simply repeat boss battle until you can clear it.
That's why XIV deserves a bit more polish on this aspect, because it's designed to be repeated.

We only look at damage because Square Enix decided to strip ranged from anything else
They removed mana song, tp song, strong mitigation without replacing it with anything.
It's not that we choose to focus on damage, it's that SQEX forces us to only bring damage to the group.

Ranged DPS are support DPS with no support, therefore Ranged DPS are DPS.

[WiccaP]

The data more likely to be used would be clear rates, relative difficulty to clear across multiple spectra, range of damage outputs across intended skill levels of player, and other metrics FFLogs either cannot measure with accuracy (because of rounding errors due to incomplete information) or does not have access to at all.

Anyone arguing for a single percentile is automatically wrong at the outset and doesn't grok data. And FFLogs might be the best data we have, but it's not nearly complete or accurate or valid enough for most of the arguments most players want to make in game balance conversations.

While its true that FFlogs and the like can not be the only thing we measure balance with, they are the one of the only tools we as players have to utilize. Is it perfect? No. That being said, in the past if there was a problem with a jobs performance it would show in places like FFlogs. Granted there are a lot of people who dont always know how to read or analyze the data properly. I may very well be one of those people. But if there are a large number of people claiming that a job is underperforming and there is data on fflogs that appears to correlate with that complaint, it is not 100% absolute, but there is a good chance that that job is having performance issues.

[Picker]

I think reaper is balanced the way it is to act as an entry level job for endgame

[Gruntler]

Then our balance is already perfectly sufficient. Every job can clear with little difficulty essential to its kit. Done.

The ability of high-tier players to clear with everything does not imply it's easier for lesser skilled players to have equal proficiency in clearing with every job. Job balance matters more than at the absolute top end.

utility is only as valuable as the average improvement to clear times (including whether you clear at all) it produces. If it does not increase your chances of clearing and/or reduce clear times, it has done nothing.

Which is not something FFLogs has any capacity to measure, but in-game tools can actually provide input for. However, DPS also only matters for balance with how it affects your ability to clear or reduce clear times. Which gets to the more important point: The vector of balance is capacity to clear across different play levels.

[quote[The criteria in the post you're quoting already follows from consideration of spread, deviation, and outliers. But let's be clear here: an entire 10% of players is not going to be an outlier. As you go lower, moreover, the parses are that much less representative of their kits, as opposed to simply varied, and often short-lived, forms of mistake.[/quote]

But people don't consider spread, deviation or outliers. Case in point:

But let's be clear here: an entire 10% of players is not going to be an outlier.

An outlier is based upon deviation, and it is quite possible for 10% of players to be deviated too much to be counted as normalized data.

How does the phrase "to be clear" have anything to do with clears? That's literally the only the time that word appears in my post -- within that phrase. Where is this strawman coming from?

You're appealing to FFLogs data. FFLogs data ONLY includes clears. That introduces sample bias that excludes aspects of balance (including difficulty to attain clear) that are more valuable and relevant to job balance than speedruns. What you said or what phrases you used is irrelevant--you're appealing to FFLogs, and this is a criticism of that appeal with regards to data-validity.