Bard procs have little impact on damage variability

[AcerolaParacletus]

Bard has procs which allow for the execution of additional damaging actions or weaponskills with enhanced potency. While the governing proc statistics are known, there has been little exploration into how much additional variability they impart into damage distributions.

My recent theorycrafting project shows how damage distributions can be computed to account for all sources of variability for Bard: procs, hit types, and random damage rolls. Methods for computing damage variability of each proc type (Pitch Perfect, Heartbreak Shot, Apex Arrow, and Refulgent Arrow) were developed and compared to damage variability when the most likely proc rates are assumed.

While there can be a substantial increase in damage variability for specific actions, the effect is negligible for a realistic rotation. For a full two-minute sequence, accounting for procs only led to the standard deviation of the DPS distribution increasing by 9.5%, compared to assuming most probable values. These findings suggest that damage variability due to procs is negligible compared to hit type variability and random damage rolls. This indicates that Bard's procs are more likely a means of making the job's rotation more engaging through proc management, rather than a major source of variability. For most cases of accounting for damage variability, assuming the most probable proc rates is sufficient.

Link to full write up: https://media.githubusercontent.com/...mage_model.pdf

[Carighan]

This is because they're not procs per se. Or I guess in the literal sense they are since they will be executing extra code based on a chance. But they're not in that they are quasi-reliable effects that happen expectably. An 80% proc chance is far too high to give relevant randomization, and I'll be honest if you told a new Bard the proc is reliable, they'd play for years without ever noticing. It's that inconsequential to actual gameplay.

Damn amazing analysis though, TYVM!

[Rithy255]

I think its good when procs don't have too much damage variance, we need more procs in general to spice up a lot of jobs.

[Valence]

This is because they're not procs per se. Or I guess in the literal sense they are since they will be executing extra code based on a chance. But they're not in that they are quasi-reliable effects that happen expectably. An 80% proc chance is far too high to give relevant randomization, and I'll be honest if you told a new Bard the proc is reliable, they'd play for years without ever noticing. It's that inconsequential to actual gameplay.

Damn amazing analysis though, TYVM!

Maybe so, but the proc rare for RA for example is pretty low on the other hand. Most BRD and DNC procs have a very high sample of low level casts (Burst Shot or Repertoire ticks, or the base DNC filler + feather use).
But anyway, it's definitely drowned beneath crit variance on your nukes. And that also boils down to variability on low sample of nukes with disproportionate potencies. That's what makes up the bulk of damage variability even within rng jobs. If we had actual nukes that do proc that were low sample like that and used within a 2min buff window, you'd have crazy swings in damage.

[Reginald_Cain]

It makes sense for procs not to cause wild damage variance. This game is designed to be super static nowadays and bard would suffer the same way samurai did when setsugekka wasn't guaranteed to crit. Having variance can be fun for lots of people but sadly it would lead to the job suffering in the long run in high end optimization.

[Carighan]

It makes sense for procs not to cause wild damage variance. This game is designed to be super static nowadays and bard would suffer the same way samurai did when setsugekka wasn't guaranteed to crit. Having variance can be fun for lots of people but sadly it would lead to the job suffering in the long run in high end optimization.

Yeah plus any bigger rework to the combat system would do better to focus on functional/rotational variance due to procs, not damage profile variance (which usually comes as a result of that, but hence focusing on it would not make much sense).

[Absurdity]

Probably intentional as well.
When bard procs were still based on Storm and Caustic Bite crits while also having the potential to get double Repertoire procs the variance was probably a lot higher...and I bet someone was complaining about it back then.

[Jeeqbit]

Obviously it's intentional. SE has long tried to keep jobs balanced (by their metrics, not players').

In fact, I would argue this is why they changed BRD procs to be a fixed proc rate as opposed to being influenced by crit, which they were in the past. That meant you could literally increase the proc rate by stacking crit. To a certain extent, you can still increase the procs on many jobs indirectly by stacking Skill/Spell Speed, because the faster GCD means more rolls at trying to trigger a proc over, say, a 10 minute duration, but in many cases this is deemed not as good as just stacking Crit, DH or Determination.

[Carighan]

There's sadly some design space lost in that.

For example Bards could have gotten a trait instead that made crit provide procs from DoTs instead of the normal crit functionality. That'd be a unique job design then, something to balance as a separate tuning dial, while also allowing the job to shift between base damage centric vs proc centric. Could it be balanced well? Maybe not, but since Crit already isn't balanced (which is why it is flat out the best stat for everyone by a wide margin) it would not really cause any "extra" damage.

But you poke at another problem that is bigger than bard crit variance: Secondary stats are neither balanced in FFXIV, nor even truly a "system". There's a very fixed, very flat and very limited solution, that is also shared by ~95% of classes. And the other 5% only vary in minute details. There is no job that wants to stack, say, Determination instead of Crit, or wants to maximize Skill Speed with no upper limit, or even better that these secondary stats even provide meaningful player-decided options instead of being just one more damage optimization with a fixed solution.

As a purely hypothetical example, what if healers had only these options for materiae:

* Piety
* A form of tenacity that decreases damage taken and increases healing-received from all sources, nothing else.
* Runspeed (this would not exist on tanks or DPS, just healers), which caps at ~10% extra at the end of an expansion's gear.
* Spellspeed but it decreases potency and mana cost equivalent to the increase in speed. Also increases animation speed linearly, so no messing with weaving.
* Bonus points if the opposite to spellspeed exists, spell "impact" or so. Increases cast time and mana cost and GCD and potency and animation time.
* Emergency healing (direct healing is increased the lower target's health is, like with Essential Dignity).

This'd start getting close to a setup where the player is actually free to decide what they personally want to optimize for, or even pick in the context of their groups or the encounters they're struggling with.

And for damage / tanking jobs, equivalent solutions could be found. So naturally no stat a damage job can pick can result in increased damage, not even indirectly via runspeed on melee classes. Because the moment one exists, that's the default.

[Absurdity]

Obviously it's intentional. SE has long tried to keep jobs balanced (by their metrics, not players').

In fact, I would argue this is why they changed BRD procs to be a fixed proc rate as opposed to being influenced by crit, which they were in the past. That meant you could literally increase the proc rate by stacking crit. To a certain extent, you can still increase the procs on many jobs indirectly by stacking Skill/Spell Speed, because the faster GCD means more rolls at trying to trigger a proc over, say, a 10 minute duration, but in many cases this is deemed not as good as just stacking Crit, DH or Determination.

Repertoire procs are seemingly on their own tick system, which is just server ticks but counted from the moment you cast a song, so every 3 seconds. That would mean Sks does exactly nothing for that part of the job.

You could technically get more Refulgent Arrow procs from the increased amount of Burst Shot casts and slightly higher DoT damage, but that's pretty much the only thing it does and clearly not worth it.
The only reason bard wants any skill speed is for better cooldown alignment.