Dark Knight Megathread

[Shurrikhan]

I honestly can't say that I've heard anyone refer to Shake it Off as being weak and useless outside of its very original implementation.

Shake is weird. It's honestly felt very weak compared to its competition in a few scenarios, like J-Wave in TEA, Terminal Relativity, and most of P4S P2. I know those are very specific scenarios, but they're at key points in difficult fights where the performance of those abilities is most important. Shake it Off is ridiculously powerful for single-hits, sure, but multi-hits always end up coming at the most important moments in the most important fights.

Well, that didn't take long.

Nevermind that a non-tank would have to otherwise have been killed by magic damage roughly 1 & 2/3s times over for HoL/Missionary to break even with Shake it Off... or that Veil's requires the PLD to have 50% more HP than the squishies to get the same barrier value onto them (and still it'd apply only two-thirds value to the co-tank and none to the PLD himself)...

And let's not conflate the number of hits with the amount of %HP that'd otherwise be dealt to a given player (or more usefully, a non-tank)? The number of hits is irrelevant; all that matters is how much damage the party takes.

[ItsUrBoi]

The number of hits is irrelevant; all that matters is how much damage the party takes.

Yes and no. I don't know what you actually mean but what your saying is not wrong but neither correct.
The number of hits is relevant because if there's more than 1 hit ( which happens only one, cuz its 1 hit, right? ) then % mitigation like Heart of light or Dark Missionary will mitigate successive hits too.
It also matters the amount of damage the party takes but when that damage is dealt across how much time / number of hits is also relevant.
HoL and DM are great for successive hits, while Shake it off and Divine Veil are good for single hits.

[Terhix]

The number of hits is relevant because if there's more than 1 hit ( which happens only one, cuz its 1 hit, right? ) then % mitigation like Heart of light or Dark Missionary will mitigate successive hits too.

That's completely irrelevant. If you take 5 hits that each does 10% of total HP of average DPS / Healer before mitigation, then with Shake it off party takes 35% of their total HP (0%, 5%, 10%, 10%, 10%) while with DM the party takes 45% of their total HP (9% * 5), which means Shake it off would have mitigated 3 times the damage despite doing nothing at all after the second hit, and the math would have been identical were it a single hit for 50% HP.

Edit: actually let me correct this, it's mostly irrelevant, I can see you could argue for mitigation over 15 seconds making healing easier even if the total damage is higher but less spiky towards the end.

[Claire_Pendragon]

Edit: actually let me correct this, it's mostly irrelevant, I can see you could argue for mitigation over 15 seconds making healing easier even if the total damage is higher but less spiky towards the end.

Agreed. When theres going to be prolonged dmg over a period of time, i always try to urge ppl to use shields AFTER taking some dmg, so it gives healers time to "catch up" rather than than playing catch up later, and maybe struggling.

[AziraSyuren]

I never said it was useless, just that its balance is very weird.

I’m also decently sure that most of the things I listed actually can kill you multiple times over in the span of 15 seconds. I know J-Wave (with enough damage up stacks, like 6 or so) and Terminal Relativity (the mechanic, not the raidwide) could, so Dark Missionary and Heart of Light were oftentimes a lot better.

But then you get things like Warder’s Wrath where Shake and Veil are suddenly the only things with any use? It’s not bad or weak by any means. It’s just weird. Like Lyth said, it’s HW-esque balance where it either vastly overperforms or underperforms, and often at critical moments. I’d actually lean towards Shake being too strong, but where it’s lacking, it needs to be as strong as it is to compare.

Just like most of WAR’s kit come to think of it.

Meanwhile, DW and HoL vary between “good enough” and “better than the alternatives in raw power,” which just about sums up DRK and GNB in general.

[Lyth]

Even with J wave, I think the break even point is going to be around 5 hits after you factor out other active effects. SiO just frontloads the mitigation into a single hit rather than spreading it out.

People make this same perceptual error when evaluating TBN as well.

[RyuDragnier]

Even with J wave, I think the break even point is going to be around 5 hits after you factor out other active effects. SiO just frontloads the mitigation into a single hit rather than spreading it out.

People make this same perceptual error when evaluating TBN as well.

Unfortunately frontloading the mitigation makes the move either incredibly strong (single hits) or completely weak (multi hits). Solutions to this can include extra mitigation when the effect ends or during its normal duration timer, or some prolonged healing effect during the duration timer.

[Shurrikhan]

Unfortunately frontloading the mitigation makes the move either incredibly strong (single hits) or completely weak (multi hits). Solutions to this can include extra mitigation when the effect ends or during its normal duration timer, or some prolonged healing effect during the duration timer.

But, again, it's not a single/multi-hit distinction. It never has been.

It's simply a matter of "A is better until X, and B is better after X." Anything other than pure homogeneity will have those same thresholds.

[Ryaduera]

The problem with shield based mitigation is not a matter of how many hits have been taken, it is that there is always a threshold where it is no longer better than percent based reduction. It doesn't matter if it's one hit, two hits, or 857864654877 hits, the math plays out that the numbers only matter for total damage taken in the time frame. As Shurrikhan said,

It's simply a matter of "A is better until X, and B is better after X."

X in this context is the damage threshold that makes shielding inferior. It is very important to note, however, that in a perfect composition, you have both shielding and damage reduction, so in the context of SiO there's usually another tank with you, and it's probably going to be DRK or GNB, so Heart of Light and Dark Missionary add to the strength of the shield. I bring this up because the relationship between shielding and damage reduction is a positive feedback loop. As damage reduction is applied on top of a shield, the threshold for damage taken for just mitigation to have been better goes up since shielding does not suffer from diminishing returns.

I've been MIA from the conversation but still playing actively, and was doing a lot more DRK for... Science. The problem with DRK isn't even its defensives, it's the reliance on healers. If, in dungeon pulls, you space out every TBN as well as you can (if a pull takes long enough using it every 15s can be hard if you didn't start with full MP, so its uptime can leave a lot to be desired) you technically reduce damage by more than the other tanks. The problem comes from the fact that, especially in AoE situations, once damage goes through, it stays there. You get 1 Abyssal Drain, that's it. Ive been doing Oblation+TBN, letting myself get sub-50%, AD, then TBN+more cooldowns, but after than second TBN, I just... start taking damage and can't do anything about the damage I took. Compared to the other tanks, PLD can regenerate off spells and HS, GNB has HoC and Aurora (Aurora is under-rated in pulls since you are almost guaranteed full value of every tick) and Warrior... well let's not talk about warrior healing. DRK just has to take damage and then sit there, dying slowly. It doesn't take more damage than the other tanks, it actually takes slightly less if TBN is used properly, but it takes 30 seconds after any pull for you to, start dying where the other tanks have ways of recovering. Defensively, this could be fixed as easily as reducing MP generation and potency of C&S and AD by half and reducing the cooldown to 30s, keeping all DPS the same while giving DRK a heal in pulls every 30s.

Bosses, however, would still pose the same problem since AD is not used on bosses. While playing DRK leveling it up again (New character on new data center for friends) I had several moments where parties did not end well and I was thinking "If I were any other tank, PLD, WAR, GNB, I could have carried that" and this was not a rare thought, this was a frequent utterance. The above proposed fix to AD only helps in dungeons, however I also think that it's perfectly fine to keep DRK the way it is (defensively, still needs a more engaging offensive kit) in raids and trials with the exception of LD purely because of the niche it fills. DRK is the perfect Off Tank. Shield an ally, Oblation 2 others, Dark Missionary (I still have problems with it being identical to Heart of Light, one of them should be changed for the same of keeping jobs unique but there is no denying the value of the skill) all while dumping damage and definitely capable of salvaging a bad run and taking the Main Tank role for swaps and recoveries from accidents. However, being the perfect Off Tank does not justify how terrible it feels to be the only tank that puts a demand on healers in dungeons. If the other 3 tanks are going to be mostly-self sufficient in 4 man content, DRK should be brought to the same standard. Warrior can fully heal itself 3-4 times over every 25s in a pull, let DRK fully heal itself once every 30s in a pull. It will still not heal as much as warrior, but will also be able to heal itself up from damage ever other TBN, and I think that will be good enough.

And TBN should give a stack of Dark Arts on break and on expiration. So many times I thought "Yeah this will break TBN" and it just didn't and I was like oops there goes 450 potency. I should not be punished for tanking too well.

[Shurrikhan]

I want to expand briefly on this dichotomy (between barriers and % mitigation -- or, front-loaded mitigation and steady mitigation, gear-scaled mitigation and content-scaled mitigation... however one may wish to name the two types).

Let us think of a fight as having periods of moderate and deep spikes of damage (each spike defined as or split at roughly the duration of a CD that'd reasonably respond to it) -- i.e., ones that fall short of the threshold by which content-scaled mitigation surpasses gear-scaled mitigation and ones that push past it. One might think then that because one works best over the highest densities of damage, it would therefore mitigate more across the whole fight, but that still leaves out two core factors:
First, a skill can have a higher maximum mitigation without having a higher total (such depends on the fight as a whole).
Second, as Ryaduera mentioned above, a typical party in any serious (and therefore 8-man) content has both types of mitigation, so small imbalances in which of two options (gear-scaled or content-scaled) would nullify slightly more damage over a whole fight are frequently overwhelmed by the available slight synergy in having both (being able to nullify more over both moderate and deep damage spikes, or stack both together without the anti-synergy of one source of GSmiti going partly to waste or of CSmiti being multiplicatively diminished by the first).
_______________

That said, damage taken isn't like Crit rates; it doesn't progressively increase (relative to tank HP) over the course of the expansion. As such, there is no more reason to frame this situation as "after a certain threshold, content-scaled mitigation is better" than "below a certain threshold, content-scaled mitigation is worse." It's not as if the situations in which the one with a higher floor and lower maximum will just be necessarily expunged over time.

The problem with DRK isn't even its defensives, it's the reliance on healers. If, in dungeon pulls, you space out every TBN as well as you can (if a pull takes long enough using it every 15s can be hard if you didn't start with full MP, so its uptime can leave a lot to be desired) you technically reduce damage by more than the other tanks. The problem comes from the fact that, especially in AoE situations, once damage goes through, it stays there. You get 1 Abyssal Drain, that's it.

While I'd still offer that Oblation should be stronger... agreed; though DRK has comparable mitigation, its total sustain (combined mitigation + healing, in practice -- not accounting for would-be healing where it'd be bottlenecked by... dying) is undertuned. And if we do not wish to give DRK absurdly more mitigation than other tanks, increasing its self-healing seems to be the way to go.

And TBN should give a stack of Dark Arts on break and on expiration. So many times I thought "Yeah this will break TBN" and it just didn't and I was like oops there goes 450 potency. I should not be punished for tanking too well.

I have mixed feelings on this point.

Though each still aims to maximize sustain (including recovery possible... from just not dying, where that'd otherwise bottleneck it, or more efficient use of healing, etc)...
I think of (1) gear-scaled mitigation, (2) meaningfully duration-limited gear-scaled mitigation, and (3) TBN thus:

1. No further constraints.
2. Wants to just barely meet the break-even threshold, as not to waste nullification (so long as the delay in optimizing any single event doesn't cost it overall sustain).
3. Same as above, but greatly increases the punishment for not meeting that threshold.

Admittedly, I kind of like that about TBN, and prefer that TBN has a bit of unique skill-gap element and consequent anti-synergy with other skills (not wanting to overlap it to such an extent that TBN won't be fully consumed), but the punishment is awfully steep for something that doesn't fundamentally change how the skill is meant to be deployed.

While I don't like the idea of removing that doubled punishment entirely --as I do find it unique, and a sufficient excuse to slightly overtune DRK's maximum sustain due to its likely in-practice losses-- I wouldn't mind seeing it reduced or TBN made just a tiny bit easier to pop (say, via a return to the Stormblood 12-second 20% TBN), especially if we're going to end up with another expansion of increasingly undertuned dungeons that progressively make TBN harder to consume, etc.