Yoshi-P Sanpo comments re: Bard

[Cetonis]

In the Yoshi-P sanpo the other day, he made a passing comment along the lines that they're getting a lot of requests to "fix Bard", but with a lack of concrete suggestions as to what/how.

Soooo, I figured I'd post a bunch of them Not that I expect SE to give the time of day to constructive forum feedback (just ask MCHs), as they only seem to react to riots, but hey might as well.


Note: This is not a flavor thread. If you want to talk about rangers and job splits and identity polls and etc., there are all too many threads about that already. Please search up one of those for such topics.

This is a thread to talk about raw performance, numerical issues, reasons why Bard is providing less damage than the other ranged, and perhaps less than SE expects it to in their internal simulations.

And, practical improvements that can be made in the scope of a mid-expac patch. Yes, no-interaction song buffs were an awful decision, and there are some other core design flaws not brought up here. Realistically though, we're probably stuck with them until 6.0 at best. We should try and limit the scope here to ideally raw numbers changes, or other things that might (?) be 'easy' enough to change in a patch.


Credibility: I've been playing and theorycrafting for Bard for a long time, and am one of the Balance discord Bard mentors. Naturally, "how best to buff Bard" has been a frequent topic in recent months, and I'll be listing some of my favorite ideas here.

----------

Underlying Issues

First, it's important to point out that Bard's problems aren't quite as simple as "numbers are low". Fight design is a big factor, and there are some QoL / anti-synergy issues that result in damage losses that maybe SE's internal math doesn't account for. Not all of these things are "must fix at all costs" stuff, but having this context will help talk about how different changes might play out.


(1) Bard is not as bursty as the other ranged, especially Dancer. This is fine - not every job has to have big spikes and valleys - but it leads to gaps when you add long downtimes to fights. The quality and quantity of the cooldowns that the other ranged get back during 30-90 second breaks is just better than what Bard gets out of these situations.

(2) Bard gets punished by new-target situations. Any time a boss comes back fresh, or a new boss appears entirely, Bard has to spend two GCDs getting their DoTs back up. During that time, they're not using Burst Shot, can't use RA, aren't getting procs, or DoT the damage itself, etc. It adds up to a pretty chunky penalty that other ranged don't have to absorb.

(3) The damage gap between Iron Jaws and Burst Shot, as well as weaker party buffs / the removal of crit synergy, effectively nerfed Iron Jaws. For those unfamiliar, a good Bard in SB could optimize Iron Jaws and Foe Requiem / Refresh timings to out-perform what was expected from the job on paper. SE took all of that away in ShB (and we're still mad), but it feels like they might be erring low on Bard out of fear that the dynamic still exists somehow. It doesn't.

(4) Empyreal Arrow became a nightmare in ShB. In Stormblood with the old Barrage, the typical plan was to settle for using 5 EAs per 80 seconds - leaving three GCDs of free flex time to move it around as needed to deal with the autoproc. Now, with the Barrage QoL change, EA left as-is became a high-potency / short-cooldown oGCD that is very costly to delay. Yet, it also bops you for -150p if you use it at the wrong point in an invisible 3 second window. It's bizarrely punishing given SE's job design direction, and it leaks damage all over the place, even for experienced Bards.

(5) Army's Muse can often do more harm than good. It ends up shifting the end of your WM either by 1.2s or 1.5s, which causes a traffic jam between the end of WM and the start of a GCD. This means either leaving WM early every time (moving some overall proc chances into the AP dead sea), clipping heavily, or at worst ghosting PP (up to 450 potency lost) if you hold out and guessed wrong. It's possible to play in a particular way to try and get around this, but few do and it really shouldn't be a thing in the first place.

(6) Battle Voice's value swings quite a bit based on fight length, i.e. three BVs in 7:30 is a lot more value than three BVs in 9:00. This is an problem for any 3m buff, but other jobs with 3m buffs have strong bursts and buffs on other timings that hedge the issue.

(7) This is speed / parse run specific so not too important, but because it's on a different timing cycle than other jobs, Bard's rDPS-provided pales in comparison to Dancer especially. While the shift to rDPS has had some good impacts on i.e. jobs being locked out of PF, there's still a dynamic of "who does the most under my party buffs", which greatly disfavors BRD among the ranged.

----------

Potential improvements

These are all simple number changes that either target a weakness that needs shoring up, or improve QoL in a way that should lead to more damage output in the average case. While you couldn't do all of these things given that there's only so much room to make buffs, any collection of them would be a well-aimed improvement.


(A) Increase the potency of Stormbite, Caustic Bite, and Iron Jaws by 50 (would need to be part of Bite Mastery II). The main goal here is to mitigate (2) a small bit, cutting 100 off of the new-target penalty. The Iron Jaws change would be needed in turn to keep it relevant, but doing so would help with regards to (3) and (7) anyway. And both parts of this would help lower-skill Bards by reducing the penalty for DoT drops, as well as overeager safety IJs.

(B) Change Raging Strikes to 20% for 15s. This is mainly aimed at (1), though not every downtime plays nice with the 80s cycle. Still, some do, and we can manipulate it at times. As for the duration nerf, there's really no need for Raging to be 20s at this point (we only have two key GCDs to fit into it). If it were 20% for 20s, aside from burning more off the "budget" to buff things, there'd be some collateral impacts we probably don't want.

(C) Change Battle Voice to a 2 minute cooldown or, if possible in a patch, give it a second charge. A charge would be a smaller buff (only ever +1 usage) but better for gameplay - it'd feel a little like Foe (3) but more predictable for the devs. Making it 2 minutes instead would just align Bard's fight-based variance with Dancer's a little more (6).

(D) If possible in a patch, give Empyreal Arrow a second charge. Even though numerically +1 EA per uptime doesn't sound like much, allowing flexible EA timings would help a lot (4), and it'd also add to Bard's downtime recovery (1).

(E) If charges are a no-go, make Empyreal Arrow higher potency but with a 20s cooldown. This would put EA at predictable points in the 80s cycle, lighten the cost of delay a little, limit it to only one per MB, and allow it to be kept far enough away from the end of WM to cause problems around the exit (4).

(F) If possible in a patch, give Bloodletter a second charge. Bards have wanted this since literally 2.0, and it'd mitigate some of the worst impacts in (4) while aiding downtime recovery a little. It would also bring MB very close in power to WM, creating some fight plan flexibility that does not exist now, though it's hard to evaluate what if any impact that could have numerically.

(G) Change the duration of Army's Muse 15 to seconds. This would solve the problems with current Muse (5) by making it gain 2.1s fairly consistently, moving the exit GCD up enough to be easily managed. Another small-on-paper buff that'd have a bigger impact in practice.

(H) Buff Sidewinder, by a lot. As our only ability that plays well with most party buffs, it would help with our synergy (7), as well as downtime compensation (1), if Sidewinder was a bigger deal. Make-number-bigger is kind of lame because there's so much room to do cool creative stuff with this skill, but I doubt such things would be patch-friendly.

(I) Cut Pitch Perfect's cooldown to 1 second. There's never been any great reason for it to be 3 seconds. All that ever does is exacerbate end-of-WM issues (4, 6) and lead to clipped GCDs or wasted procs. More QoL than anything, but like a lot of stuff here it'd tick up the damage output in practice, especially for newer Bards.

(J) Speed up the DoT application delay on Stormbite and Caustic Bite, as has been done for some other skills. If these applied their DoTs closer to instant, it'd not only help with Sidewinder QoL, it'd also slightly mitigate (2) a little more by making startup a tad faster. This is a small deal that's actually a small deal, but it'd be good to do.


Obviously, if all you want to do is to make numbers go up, it'd be easy enough to revert the DoT potency nerf, make songs self-apply, etc., but such things fail to address any sort of real weakness or gameplay issue. They're pretty low quality band-aids, and we should be looking for some nice HQ band-aids to use instead~ More HQ ideas welcome!

[Ramura_Sono]

In the Yoshi-P sanpo the other day, he made a passing comment along the lines that they're getting a lot of requests to "fix Bard", but with a lack of concrete suggestions as to what/how.

That is what having a Team of Battle/Class Designers is for. There is a multitude of feedback for them to parse and discern a concrete direction to take the job in in an attempt to fix it. "Players haven't given us concrete suggestions" is just a way for the Devs to pass the buck. It has to stop. Monk players have routinely mentioned they enjoy the speed, dislike being overly punished by things outside of their control, hate all the niche skills; Bards have continually mentioned a desire for more support in their kit, more musical inspiration to their design, etc. Every single job has a set of core design principles that players enjoy about the job. And a multitude of issues they have too. It's not on the players to give exact specific suggestions. Especially when such suggestions are routinely ignored or misunderstood.

Their propensity to implement even the most simple and basic ideas in the most convoluted and asinine way also doesn't particularly make people want to provide feedback and ideas.

[BloodRubyXII]

That is what having a Team of Battle/Class Designers is for. There is a multitude of feedback for them to parse and discern a concrete direction to take the job in in an attempt to fix it. "Players haven't given us concrete suggestions" is just a way for the Devs to pass the buck. It has to stop.

Agreed. You will never find a more passionate group of players than ones on MMORPGs, especially optimisers. The amount of work people have put into sites like Bole and Akh Morning, and the Balance Discord all of those guides, google docs, spreadsheets like holy mother of god. Even WoW's top players don't go that deep, and this isn't even a game where optimizing is on the same level or as important. So I know without even researching it any Bard suggestions put forward were probably extremely deep with examples to back them up. I know it's the JP forums they run to first for backup but they're even more brutal than we are so I know they've probably gone into it too. BRD is just entirely centered around that stupid WM/RS window with not much interesting things to do or get more of outside of that, is this really the vision for the job they wanted? Bard does have it's place. Honestly in early prog I like Bard because I am not as reliant on RDPS to have an impact, but obviously the more comfortable the encounter becomes, the stronger dancer gets as I'm no longer using the higher personal DPS (which isn't even that much higher) as a crutch. Every pull has different decision making and different instances of small bursts on Dancer which means they are capable of making Bard as interesting as it once was and simply don't, almost like it's their design for the job to be training wheels for learning how a cycle and priority system works - and it's just sad.

[EONX_]

One change I'd also like to see is having their song rotation last either 60s or 90s before restarting in WM. Currently it's a 80s rotation which aligns with very little in the game. Having them align more naturally at either 60s or 90s would definitely help the job a lot and make it more enjoyable to play since you have more windows to optimize.

[Eloah]

One change I'd also like to see is having their song rotation last either 60s or 90s before restarting in WM. Currently it's a 80s rotation which aligns with very little in the game. Having them align more naturally at either 60s or 90s would definitely help the job a lot and make it more enjoyable to play since you have more windows to optimize.

The issue with this is that the 80s rotation is mostly player made. Since based on the songs durations the rotation should be 90s. If they are able to adjust Armys Paean so that cutting it short isn't desired then that, in theory, would adjust the rotation up to 90s.

It would be nice if in 6.0 the songs got some kind of an upgrade.

[Akiudo]

The issue with this is that the 80s rotation is mostly player made. Since based on the songs durations the rotation should be 90s. If they are able to adjust Armys Paean so that cutting it short isn't desired then that, in theory, would adjust the rotation up to 90s.

adjusting armys paeon so cutting it short isn't desired would only mean you would cut mages ballad short instead. Only way to really make you "want" a 90 second rotation with the songs would be if the recast cooldown on them were 90s each, but that would

a)be a straight up nerf even with perfect gameplay
b)punish accidental cutting of other songs even harder than it allready does
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

from the things proposed here i will just say that while i like basically all the points you mentioned, the biggest by far for me would be
1/2. (tie)buffing sidewinder (its simply too weak for the cooldown, especially with EA as comparison)
1/2. (tie)stacks for EA and BL, its simply overly punishing right now aswell as more or less forces you to just spam BL during mages ballad which well, "doesn't feel nice" to say it nicely
3. pitch perfect cooldown of 1 second, its bad enough getting the skill out when the server tick happens at 1~ second left, really no need to make it worse by also having to watch out that you use it early enough so the cooldown is even ready.

btw, one point you didn't specifically mention about a second EA charge is the more flexible timing this would offer would actually allow us to keep an extra charge for wanderers instead of using it during armys paeon when its not needed, which would be great because that would actually be a small way to optimize a bit more, instead of simply "more/bigger" number

[Eloah]

adjusting armys paeon so cutting it short isn't desired would only mean you would cut mages ballad short instead. Only way to really make you "want" a 90 second rotation with the songs would be if the recast cooldown on them were 90s each, but that would
a)be a straight up nerf even with perfect gameplay
b)punish accidental cutting of other songs even harder than it allready does

How does adjusting things so that cutting Armys Paean isn't desired, translate into a desire in cutting Mages Ballad. The main reason for the cutting isnt the songs CD, it's the affect. Armys Paean has the least desired affect so it gets clipped. SE would need to adjust things so that players want their song to play out for the full duration. The 80s CD is just a way to ensure that a song is always available when cycling through the songs.

[Akiudo]

How does adjusting things so that cutting Armys Paean isn't desired, translate into a desire in cutting Mages Ballad. The main reason for the cutting isnt the songs CD, it's the affect. Armys Paean has the least desired affect so it gets clipped. SE would need to adjust things so that players want their song to play out for the full duration. The 80s CD is just a way to ensure that a song is always available when cycling through the songs.

you gave the reason yourself, the problem isn't the cooldown, its the effect. one song will allways be the weakest, unless all 3 are identical thats inevitable so if you make the effect of armys better so it isn't the worst song anymore something else (mages ballad in that case) will be the worst song instead and you would cut that.

the only real way to get away from that is to literally force you to use all three songs for their full duration, in that case they could aswell turn their cooldown up to 90s and be done with it, that would be a nerf though, buffing armys paeon so its better than mages would indeed be a buff, it would however simply shift the problem to the next weakest song.

[Vogue]

How does adjusting things so that cutting Armys Paean isn't desired, translate into a desire in cutting Mages Ballad. The main reason for the cutting isnt the songs CD, it's the affect. Armys Paean has the least desired affect so it gets clipped. SE would need to adjust things so that players want their song to play out for the full duration. The 80s CD is just a way to ensure that a song is always available when cycling through the songs.

I agree with this, having a tight cooldown rotation for the songs can be bad and stressful. The 80s gives a room to clip songs shorter (normally Army's Paean) if needed to avoid having to pay attention to the song timer while doing complex mechanics, or to lineup songs with party buffs burst times. If there was no extra time, then clipping any song would mean songless periods.

[Kabooa]

adjusting armys paeon so cutting it short isn't desired would only mean you would cut mages ballad short instead. Only way to really make you "want" a 90 second rotation with the songs would be if the recast cooldown on them were 90s each, but that would

a)be a straight up nerf even with perfect gameplay
b)punish accidental cutting of other songs even harder than it allready does


from the things proposed here i will just say that while i like basically all the points you mentioned, the biggest by far for me would be
1/2. (tie)buffing sidewinder (its simply too weak for the cooldown, especially with EA as comparison)
1/2. (tie)stacks for EA and BL, its simply overly punishing right now aswell as more or less forces you to just spam BL during mages ballad which well, "doesn't feel nice" to say it nicely
3. pitch perfect cooldown of 1 second, its bad enough getting the skill out when the server tick happens at 1~ second left, really no need to make it worse by also having to watch out that you use it early enough so the cooldown is even ready.

btw, one point you didn't specifically mention about a second EA charge is the more flexible timing this would offer would actually allow us to keep an extra charge for wanderers instead of using it during armys paeon when its not needed, which would be great because that would actually be a small way to optimize a bit more, instead of simply "more/bigger" number

I'll preface this with I've only leveled it and haven't had much interest in doing more with it.

It's really odd to me that the GCD upgrades are baseline instead of baked into Soldier's Paen. I'm generally not a fan of straight upgrades, but it seemed a missed opportunity to me.

I also mentioned this in another thread, but I'll throw the "Apex Arrow should be based on Songs" back in here. Rather than have Soldier's Paen being your song modulator, Apex Arrow does it instead. I'm sure some might disagree with me, but Apex Arrow doesn't feel good to me. It's a gauge i have very little control over, and the feast or famine associated with procs means I can't really use it when I want to either.

As far as this thread goes though, I'll give it a generic "thumbs up" because they know bard better and none of the changes seem unreasonable.