How we determine balance and its flaws.

[Karshan]

You should specify that you're not using fflogs definition of rDPS.

rDPS on fflogs and the one that is often used nowadays when we say " x class rDPS is too low" is the whole contribution of a class, not the only the small buff part like before. With the new definition, everyone's rDPS (=complete final contribution) should be roughly the same.

[Vendalwind]

You should specify that you're not using fflogs definition of rDPS.

rDPS on fflogs and the one that is often used nowadays when we say " x class rDPS is too low" is the whole contribution of a class, not the only the small buff part like before. With the new definition, everyone's rDPS (=complete final contribution) should be roughly the same.

Your probably right. What is the term generally used to describe the buff portion solely right now if there is one?

[Kalise]

You are instantiating a limitation to knock it down. I spoke of that setup being generally the best, NOT of the the specific stormblood scenario singularly.

I'm pointing out that the "Optimal Scenario" that you envisioned, has existed and has not promoted the point about composition versatility that you claimed.

In StB we had your "Max DPS is 1 Selfish + 3 Support" optimal scenario. What did it lead to? It lead to a meta where people wanted just 4 Supports and non-meta jobs were shunned.

This is a very real aspect to balancing, cry as you might "But I only want to look at the numbers!" the fact of the matter is that balancing jobs goes beyond simply numbers.

I DONT CARE about the other support effects that aren't dmg because THIS POST WASNT ABOUT BALANCING those support effects. IT IS ABOUT BALANCE as an idea at numbers alone. You are so far off context here you are either intentionally derailing or incapable of understanding the scope I applied and limited my first post too. I specifically omitted class specific content because I'm not talking about anything class specific. I ONLY EVER HAVE BEEN TALKING ABOUT DAMAGE BUFFS and their effect on balance.

If you want to stop telling me I'm off context and actually read what you've been replying to, you'll note that I pointed out that your "Optimal Scenario" of 1 selfish and 3 support DPS was being critiziced by me as a poor goal.

As I mentioned, the ideal scenario is any 4 DPS jobs should be able to play together, irregardless of whether they are "Selfish" or "Support" (Barring things like duplicate classes) while being within a close enough DPS output so as to not matter.

This links directly back to meta and viability, because of the SITUATION WITH StB WHERE WE HAD A META THAT SHUNNED NON-META JOBS FROM BEING PLAYED IN CONTENT

Meaning that it was harder to do content on said non-meta jobs because people would refuse you a place in their party. I.e. It is not as viable to do the content on said jobs because you couldn't get into a party to do it.

You say that your way of balancing would achieve my ideal scenario where any combination of Support and Selfish DPS jobs can perform nigh equally well, whilst also saying that a situation that is exactly the same as what we had in StB is the "Optimal scenario" and disregard the fact that in StB there WAS issues with balance BECAUSE of the way the optimal party compositions ended up being.

...Thats my whole point. That its not common sense to these forums. Lots of places I've read someone saying something very close to 'pDps classes would be worthless if they didnt contribute more rDPS than utility classes'

The issue here, is that these sorts of statements aren't entirely related to rDPS related "Utility"

These statements are usually brought up in regards to non-DPS related utility.

I.e. If RDM and BLM did the same rDPS. Why would you ever bring a BLM when RDM has Vercure and Verraise?

You pull a bunch of numbers here out of thin air. Its meaningless ground those statements. There is no class in this game that has ever gotten anywhere close 6000 bonus rdps which is what your numbers loosely suggest. Where the heck are you coming up with numbers like 20700???

It was example numbers. To illustrate a point.

With the example numbers (Tanks dealing 10000 DPS each, Healers dealing 6000 DPS each, Selfish DPS dealing 15000 pDPS each and Support DPS dealing 13000 pDPS each) and Supports being a 10% buff to other players (Far in excess of current boosts purely for example)

20700 comes from the fact that the buffs to other jobs (10000 + 10000 + 6000 + 6000 + 15000 + 15000 + 15000)/10 = 7700

Plus their pDPS (13000)

Equals 20700 rDPS (Total DPS)

The further extrapolation from this (If you bothered to do the math, like you suggest other people do for you)

Is in this situation, if you pulled down Support pDPS to a point where 3 Selfish + 1 Supports is on par with 4 Selfish (You know, relating back to my previous point about all comp types being on par) then you'd have a case where any more than 1 Support is progressively worse.

Using example numbers, Support pDPS would have to go down to 7300 in order to achieve an rDPS of 15000 (The same as the 3 Selfish DPS in the party). However in doing so, it means that the rDPS of a Support in a 2 Support team would be 14230 and in a 3 Support team that would be 13460 and a 4 Support team would be 12690

Meaning that in this scenario, a 4 Support team would be dealing 70070 total DPS (All 8 players combined) compared to a 0 or 1 Support team that would be dealing 92000 total DPS (All 8 players combined). Approximately 25% less damage overall, which would be significant enough to make people shun those group compositions.

Of course, this is all exemplified using an inflated support boost figure, with more realistic figures things will be closer. It is however, an example of the difficulties of attempting to balance pure pDPS vs rDPS whilst trying to make compositions not boil down to a single optimal set up that becomes meta at the cost of others.

A Meta comp is a very specific definition of 'Most Effective Tactic Available'. Its not some Ideal, and it has nothing to do with viability. META is about the tip top best setup.

This is just laughably wrong.

The Most Effective Tactic Available, in video game (And specifically FFXIV) context is not the "Maximal DPS Tactic".

It is the Most Effective one. I.e. The one that more easily performs the best.

For example, look at the state of the game in Stormblood.

The Max DPS Tactic was 1 Selfish and 3 Support (Typically a BLM or a SAM being buffed to heck)

The Most Effective Tactic was 4 Support.

Why? Because it was significantly easier to perform well with 4 Support and thus would put out more consistently higher numbers than the "Max DPS" tactic would outside of paper.

Simply because, "Co-ordination" in FFXIV surmounts to pressing your CD's on CD.

[Gruntler]

Imagine pulling numbers out of your ass for stuff when we have access to actual data.

What a timeline.

Okay, so, while the dataset's small, it's actually looking like an interesting tier. Casters are absolutely dominating some fights, but on another, melee are absolutely wrecking house. Which job is the best is actually -fight dependant- in terms of raid dps contribution.

THIS, right here, is a desirable situation for casters and melee; a properly balanced mmo won't have one class consistantly on top on all fights, then another, then another, in lockstep order. You should expect to see variation where one class will be strong at one fight, but another will be strong at another. This makes class choice simultaneously meaningful, while also maximizing viability. Make it that there's one fight where ranged DPS destroy the house and you'll have dps balance pretty much nailed down perfectly--we're very close to, but not quite at, the ideal.

Another thing to consider is that in today's world, raid dps contribution isn't theoretical, it's now measured. We know how much contribution each job's buffs are, on a case by case basis. In past expansions, it was not measured, and so we literally guessed at it. Having access to this information has changed what the meta looks like because we no longer relegate buff strength to the theoretical; we know selfish DPS are strong and by exactly how much; we know how strong healer duos can be and we can analyse this a lot better than before.

So, we've gone past talking about theoretical raid dps contributions; we can actually directly compare buff strengths with data, not theorycraft.

[Vendalwind]

...

Physical data is still very hard to analyze. I've presented this exact same argument before with actual data last tier. And I calculated it specifically 4 different percentile gaps so it took a TON of work. Unfortunately i presented that information on the 14th page of a thread already in progress and although some people agreed with me the result was nearly exactly the same as Kalise's outcry where tangential arguments were created based on outlying data points. Once again as you can see above Kalise is inserting her own meaning into the original things i said and changing the core subject i created this thread to discuss. It's nice to see her use some math, but like i said in my post one scenario presented doesn't give the whole picture and Excel spreadsheets calculations are necessary to find the percent wise balance point that lends itself to proposed balanced team proportion of 1 selfish 3 utility.

So there is still absolutely value in theoretical or baseline analysis. First because baseline analysis allows you to develop a fair plan before looking at the variations in extensive datapoints, AND allows you to make judgements before and after utility and what makes classes unique is taken into consideration.

My post isn't trying to combat or change the meta. Its a critique on how we think about the meta. Currently many people see Class A is underperforming or overperforming and ask for a Buff or Nerf and usually its also tied to how players imagine that class should do in. For Example i can't remember the last time i met anyone who didn't think BLM should be the end all best RDS at 95th percentile and up. If you ask them why a whole can of worms, lack of mobility, rng, no raise, cast bars, memorization crops up and you can't target any core reasoning even tho in essence when all those reasons are stacked on top of eachother its because the class is difficult. But then in the same breath people say "The game shouldn't be balanced on difficulty" This is a problem. So we need to have a core reason for these statements.

And in the melee role, caster role, and ranged role it is still commonly accepted that NIN, DNC, and RDM should have the lowest RDPS for whatever reason. They should have the lowest Pdps for sure. but in actuality for the game to be balanced they should have the Highest AND the lowest RDPS (highest at 90th percentile up, lowest IMO percentile 50 ish and down) Those lines however are opinion and could be moved up or down the spectrum. But my point being I can't tell you how many times someone has told me that SAM should have the best RDPS in ANY percentile among the melee because it has no utility.

The only way to effectively have that conversation is to show people why that underlying thinking is wrong. And it's wrong because if selfish dps within each role are in fact the best it entirely disenfranchises all 7 utility roles. Those roles may still be viable tho of course. With a proper balance all 10 roles will be desirable, and with proper tweaks to the baseline a solution set can be found where a 1 selfish and 3 utility and at a distribution generally of 30% selfish and 70% utility thats the best match.

Which is exactly why i titled my thread like I did. Im not here to analyze the numbers as they stand nor make a statement of who should be on top where and why. I dont want people to just look at the numbers and say "muh this class needs to be stronger cause it has no utility" because its not that simple.

Im here cause i want people to think about the theory behind proper game balance and break the current accepted for a long time justification that selfish dps should just simply be the best. Not everyone thinks that way, but in my perusal of the forums over the years that trend is the most common.

[kajv95]

The only way to effectively have that conversation is to show people why that underlying thinking is wrong. And it's wrong because if selfish dps within each role are in fact the best it entirely disenfranchises all 7 utility roles. Those roles may still be viable tho of course. With a proper balance all 10 roles will be desirable, and with proper tweaks to the baseline a solution set can be found where a 1 selfish and 3 utility and at a distribution generally of 30% selfish and 70% utility thats the best match.

Basically, why have utility at all if it's not going to be useful in any way whatsoever. I've always agreed-defensive utility is overvalued by the devs and offensive utility is essentially non-existent if the jobs who don't have it perform better even after their adps is adjusted to rdps. And if we, as a community, are holding on to the saying that class balance isn't based on difficulty, then most of the points given as to why they should be are invalid.

[Vendalwind]

Basically, why have utility at all if it's not going to be useful in any way whatsoever. I've always agreed-defensive utility is overvalued by the devs and offensive utility is essentially non-existent if the jobs who don't have it perform better even after their adps is adjusted to rdps. And if we, as a community, are holding on to the saying that class balance isn't based on difficulty, then most of the points given as to why they should be are invalid.

Definately agree with your final statements. We definately do need diversity and utility tho. Im not a fan of homogenization either. Dmg at base spread tho should be evaluated before utility is of any form is considered, made healthy, and then adjusted to account for utility.

[kajv95]

The largest problem FFXIV is facing is the fact that we have so many party-wide buffs flying around that the jobs based around having those can't reasonably be powerful without outright deleting selfish dps. Ranged physical dps can't become strong because combined with other jobs with dps utility, they'll start outclassing the selfish. We can't balance the physical ranged around having better personal dps either because then atleast half the melees will feel invalidated due to them having to work more for uptime, and casters will feel shafted due to, you know, having cast times.

Pretty much the entire game would require a rebalance if we want "real" balance, so the party shackle bonus if the best we've got for now, quite frankly.

Though I would personally prefer living as a bowmage for the time being if that's what it takes. I like playing as a Bard because of the snazzy fukkin outfits, the bow and the "job fantasy" associated with it, atleast let me be powerful if you're going to take away the fantasy.

[Vendalwind]

The largest problem FFXIV is facing is the fact that we have so many party-wide buffs flying around that the jobs based around having those can't reasonably be powerful without outright deleting selfish dps. Ranged physical dps can't become strong because combined with other jobs with dps utility, they'll start outclassing the selfish. We can't balance the physical ranged around having better personal dps either because then atleast half the melees will feel invalidated due to them having to work more for uptime, and casters will feel shafted due to, you know, having cast times.

Pretty much the entire game would require a rebalance if we want "real" balance, so the party shackle bonus if the best we've got for now, quite frankly.

Though I would personally prefer living as a bowmage for the time being if that's what it takes. I like playing as a Bard because of the snazzy fukkin outfits, the bow and the "job fantasy" associated with it, atleast let me be powerful if you're going to take away the fantasy.

Very true! I feel tho that generally party wide buffs are better than specific buffs (such as the old terrible DRG BRD event) and they do make the game more interesting. That being said one thing I would LOVE is if those buffs were designed in such a way that they couldn't compound on top of eachother. I know that would likely be a homogenization, but it would be a good idea. It would limit the value of properly aligning things too in fights... making tactical unity less important, but i do think it would be healthy for The balance.

Alternatively (although it would require some serious recoding) all manuever could stay the same but be designed to keep dmges separate. I.E. if NIN trick attack is up and a Crit Boost from some source is up the game could be aware of whether the crit would have occured without the crit buff, and if so Trick augments the damage, if not that bonus crit dmg does NOT hit for the 10%. To do this code would have to be written something like follows during the damage calculation (very psuedo code):
New Var $BONUS DMG
If($dmg.type == 'buffed')
{
new var $Base dmg
New var $Buff damage
$Buff damage = $dmg/buff.type
$dmg = dmg- $Buff damage
$BONUS DMG += $Buff damage
}
Run that per each buff with some refining (would need to also be occuring i the portion of the function where buff is applied) and you could pool all singular additive effects from buffs into the bonus dmg so they don't compound.

I could make this better obviously. I do code work myself. this is very rough so i can go to sleep don't get mad at me if my sleepingess made this terrible I'm just showing it could be done.

[Eliadil]

The issue with damages buff is that they depend on the skill of your party in order to be decent.
Give Trick Attack to a pug group and more often than not the difference will not be that huge, on the opposite, give Trick attack to a decent and organized party and they'll squeeze the most out of it.

it's the same reason as to why peoples thought that dancer was bad at the start of ShB even with rdps. But as time went on and peoples became better and dancer was then one of the best physical ranged you could ask for when trying to do a speedkill.

That imo is why devs are kinda afraid to increase/decrease some party buffs. Because if you raise it in order to be better for the average green parser, then suddenly it becomes amazing and OP for the top-end group, which can be an issue because then it demands to rebalance new encounters around that in order for them to still be challenging... Which nullifies the whole point of the previous buffs.