Dark Knight Megathread

[Shin96]

thats the problem, all jobs are braindead simple. The universal simplification and homogenisation was not neccessary in the jobs that suffered most from it.

It had to be done. Back in the HW era, SE had a clear target audience in mind and as such, the expansion was catered to them specifically. Dark Knight was born during that era and offered a more complex tank with rewarding gameplay for anyone that bothered.

Fast-forward, ShB and Endwalker are targeted at the "casual" demographic that doesn't give two flying shits about learning their job. Hence the huge emphasis on story-elements and less focus on gameplay mechanics. SE is smart enough to understand how complexity can get in the way of accessibility hence why they homogenize and simplify tanks as they bear responsibility for party wide failure.

Overall, if we had any sort of gatekeepers around tanking perhaps this wouldn't have happened, but we're not going to see 3.0 complexity ever again.

[Crater]

But I do think things went way too far in that regard, and I wouldn't mind dialing things back to a more refined version of the concept we had in HW.

I don't actually think it would be that hard to create a version of DRK that felt like HW DRK while still being in keeping with the general post-Shadowbringers design philosophy, if SE were actually of a mind to do it. The actual complexity of 3.x DRK is not really possible to bring back, but that's not really the main thing that DRK lost: The actual loss that you feel every time you try to play 5.x DRK is that its gameplay identity has been completely reversed from what it was supposed to be, and it's instead been shoved into a niche that other tanks already occupy (and are better at).

3.x DRK was an oGCD-focused equivalent to modern-day PLD: It was not a burst-damage tank; it was instead based on continuously doing damage across an entire rotation. While modern-day PLD achieves this by having three distinct 'phases' to their rotation, 3.x DRK achieved it by having oGCDs with irregular cooldowns that lined up differently from minute to minute, and rate-limiting how fast you could spend MP to deal damage.

If you sort of break down how and why 3.x DRK felt the way it did, it isn't impossible to imagine a post-5.x version of the same gameplay identity:

• If you had an excess of MP and wanted to turn it into damage, you had to do it through Souleater, giving it an effect 3 GCD cooldown. You could achieve the same thing today by adding an 8-10s cooldown to Edge/Flood, solving the two major problems of 1) 5.x DRK's burst windows feeling more spammy than 4.x DRK's "Dark Arts spam" ever did and 2) 5.x DRK not having anything to do outside of burst windows.
• Most of 3.x DRK's oGCDs were designed to not line up cleanly with 60s Trick windows, which meant that much of DRK's gameplay happened outside of party burst. Salted Earth was 45s, Blood Weapon and Blood Price were 40s, Reprisal's parry requirement gave it a randomized cooldown that averaged somewhere between 40 and 50 seconds, and Low Blow was 25s (with a randomized chance on parry to reset at essentially any moment). Only Plunge and Dark Passenger, at 30s each, were made to line up with Trick. Modern-day DRK could easily achieve this by rolling Salted Earth and Blood Weapon back to 45s and 40s, introducing Shadowbringer at 40s instead of 60s, and changing Plunge to 25s, not to mention any possible changes to Abyssal Drain, Delirium, or Living Shadow. (Obviously potencies would have to be adjusted to make sure it didn't come out too strong or too weak).
• Kind of a tag onto the last point, but: At 60 seconds, Carve and Spit was 3.x DRK's longest offensive cooldown. Everything else was shorter, and Plunge didn't have Charges, which (obviously) meant that DRK had things to do and buttons to press in between party buff windows. On 5.x DRK, once you realize that charges effectively give Plunge a 60s cooldown, Carve and Spit is actually tied for DRK's shortest cooldown, a complete reversal of how the class was supposed to work. Fixing this would go a long way to restoring DRK's gameplay identity.

I don't think that changing these aspects would run too contrary to the post-5.0 design philosophy, and they would go a long way to restoring the "feel" of 3.x DRK. Getting back the full range of gameplay nuance is probably not on the table, but there is an actual problem here that isn't just a case of lost complexity.

Ideally, with four tanks in the game, we would have:
1) A GCD-based, burst-damage tank [WAR✓]
2) A GCD-based, continuous-damage tank [PLD✓]
3) An oGCD-based, burst-damage tank [GNB✓]
4) An oGCD-based, continuous-damage tank [DRK✗]

Instead, DRK has been changed into something that's trying to wear the same pair of pants as GNB, and the results are, unsurprisingly, that DRK's gameplay feels unsatisfying and lacking in unique identity.

[Lyth]

I'm not sure how well the burst/sustain dichotomy will hold up in Endwalker with every tank operating on 60s burst windows. What made HW DRK different was not so much a lack of burst but the fact that you had to situationally respond based on whatever procs or resources were available at the time. The current problem that tanks face is that all tanks, including ones which are supposed to be based around resource-management, operate on fixed rotations and timers, just some are more blatantly obvious (PLD/GNB) than others (WAR/DRK). Your resource gains happen every time at fixed points on your combo. The end result is that every pull ends up playing out the same when things go correctly. I personally find that to be extremely boring.

What I would like to see is more variable resource gains on DRK. Even something simple like 'every time one of your barriers absorbs damage, you gain a small amount of resources' leads to variable resource totals, which can make a difference over a ten minute fight. I know that this will make the job less accessible for players who can't make decisions on the fly, but there's three other tanks for them to choose from if they want to cling to their spreadsheets. Really, all I want this game to do is start mixing up pulls so that they don't always happen the same way.

[Ignimortis]

Ideally, with four tanks in the game, we would have:
1) A GCD-based, burst-damage tank [WAR✓]
2) A GCD-based, continuous-damage tank [PLD✓]
3) An oGCD-based, burst-damage tank [GNB✓]
4) An oGCD-based, continuous-damage tank [DRK✗]

Instead, DRK has been changed into something that's trying to wear the same pair of pants as GNB, and the results are, unsurprisingly, that DRK's gameplay feels unsatisfying and lacking in unique identity.

Brilliant post overall, really sums up how I feel about DRK's identity and what I enjoyed about HW/SB DRK, but the quoted part bears repeating. The "oGCD sustained damage tank with lots of small buttons instead of a few burst big buttons" was literally DRK's niche DPS kit-wise before the rework, and the rework removed every trace of it. I really, really want that back.

[Crater]

I'm not sure how well the burst/sustain dichotomy will hold up in Endwalker with every tank operating on 60s burst windows.

PLD operates on a 60s cycle, but I wouldn't say it operates on a 60s burst cycle. Its damage peak is considerably lower than the other tanks, and its damage in between peaks is considerably higher than the other tanks. It isn't uniform - neither was 3.x DRK - but it is a lot more continuous and sustained than the others. And, looking at the changes it's receiving in Endwalker, Square-Enix is doubling down on that damage profile: PLD's main offensive upgrade is effectively a buff to the lowest-DPS point in its rotation, rather than giving it another big hit in the burst window as they've done for WAR, GNB, and DRK. I think that's absolutely a dichotomy that they're aware of and are actively trying to maintain; they're just (erroneously, in my opinion) doing a 1/3 split instead of the 2/2 that I think would be more healthy for the game.

I'm in full agreement that things would be much better if we had variability in resource generation back. I liked that your decisions could be very different from pull-to-pull based on relatively slight changes in how the pull goes, and even that you occasionally ran into slight differences even when you were essentially playing identically. I'd enjoy it if they not only brought that stuff back, but expanded on it and made it more central to DRK's kit - and I definitely feel you on "There are three other tanks for people who like the current level of complexity and predictability - those people don't need to have all four." (Same applies to healers.) But I'm not optimistic that SE would really lean into that kind of design any time in the foreseeable future, and if they did, I think it would have to be a gradual process that starts from something a little more baseline and fundamental.

There are some good reasons why that approach wouldn't work for the current tank designs: PLD is based around the predictability of some fairly long strings of GCD sequences and you wouldn't want to introduce chaos into that system (which I think fits the class fantasy of being a traditional, disclipined knight). WAR is intentionally designed to be incredibly straightforward, because that seems to overwhelmingly be what the Warrior playerbase wants (I think we all remember the reaction to having to manage gauge in early Stormblood). GNB doesn't have enough granularity in its resource system to really introduce any chaos in a way that isn't frustrating. And the current version of DRK is so married to the burst window that a little bit of resource variability wouldn't move the needle (or at least, wouldn't move the needle in an interesting way - it would just solidify the cursed "2/6 rotation" type stuff). DRK and GNB aren't fundamentally incompatible with a little bit of chaos/RNG, but neither of their current designs is well-suited to having it bolted on. A hypothetical rework of DRK that moved it more into a sustained-damage model, though, would open up some opportunity for SE to experiment with some of those mechanics in the future.

I don't think we can expect SE to completely leave their comfort zone on class design any time soon, especially considering how much the current success of the game must be encouraging an "If it ain't broke, don't fix it" attitude. But if EW rolls around and it does turn out that "DRK is broke, needs fixing", I think there's at least a small chance that they could be persuaded to change it into something with a more distinct gameplay identity that also still fits fairly neatly within the job design parameters that they've set for themselves.

[Kalaam]

Having some slight random/variable ressource generation could honestly breathe a lot of life into DRK. That's why I loved the parry procs, Blood Price etc. I honestly wish MP wasn't as standard 10k as it is now, and have actual numbers instead of a %.

[Shurrikhan]

Ideally, with four tanks in the game, we would have:
1) A GCD-based, burst-damage tank [WAR✓]
2) A GCD-based, continuous-damage tank [PLD✓]
3) An oGCD-based, burst-damage tank [GNB✓]
4) An oGCD-based, continuous-damage tank [DRK✗]

Instead, DRK has been changed into something that's trying to wear the same pair of pants as GNB, and the results are, unsurprisingly, that DRK's gameplay feels unsatisfying and lacking in unique identity.

In what XIV-related world, though, is a continuous-damage tank not simply a worse tank? The possibility of just barely getting in one extra CD cycle vs. not having the risk of barely losing out on one further CD cycle is rarely an acceptable balancing point in a game that's still largely ruled by raid (de)buff windows.

PLD doesn't hold up against WAR because it has "continuous damage". It holds up despite that, and only through some rather unique capacities that do far more to generate its identity that does the shape of its fflogs graphs.

[Crater]

In what XIV-related world, though, is a continuous-damage tank not simply a worse tank?

In the current world that we've been in for probably at least 3-4 years now, where encounter designers specifically go out of their way to give everyone nearly 100% uptime if they want it. We're a long, long way off from Heavensward fight design where bosses had so many jumps and forced disconnects that burst damage was worth twice your sustained damage.

Burst damage isn't better than sustained damage because it lines up with raid buffs; FFXIV class balancing (as of Shadowbringers) very, very obviously takes that into account - if it didn't, Paladin would be thousands of DPS behind every other tank on almost every encounter. But that's not what we see: The Tanks are all weighing in at approximately the same DPS, usually with less than a 500 DPS gap between the top and bottom, out of 14000+. Being a sustained versus burst class is very emphatically neither an advantage nor a disadvantage in modern-day FFXIV.

Rather, burst damage is "better" than sustained damage because of the (these days, largely hypothetical) utility that it brings: Burst can be used to kill adds faster, to push specific HP%-based phase thresholds, or to deal with any downtime or forced disconnections better. These are all things that were very common in previous expansions, and have been largely excised from modern FFXIV.

The situation that we exist in now is pretty simple: There are no particular objective advantages or disadvantages to being a burst versus a sustained damage class - not from a class design/balance perspective, and not from an encounter design perspective - but they provide very different playstyles, which different players may subjectively find to be significantly more or less fun than others. And with two clearly defined separations between tank playstyles, we should have a tank that exists in every combination, so that everyone has at least one tank with the playstyle that they find fun - but we don't. And that should be fixed.

[Shurrikhan]

Rather, burst damage is "better" than sustained damage because of the (these days, largely hypothetical) utility that it brings: Burst can be used to kill adds faster, to push specific HP%-based phase thresholds, or to deal with any downtime or forced disconnections better.

Which is still a capacity -- even if one we are never asked to sacrifice total/sustained damage (e.g., uses per fight) to exploit (for dps checks) -- that "continuous damage" (a euphemism simply for "less burst than most") jobs simply won't have.

"Continuous damage" does not, in itself, have a compensatory factor. Just as being able to do deal one's damage from range is, in itself, just objectively better than being only able to attack from up close, having burst is simply an allowance that not having burst does not have.

We can give "continuous damage" jobs other goodies to compensate, but the original lack of whatever (burst, range, mobility, etc.) will not be the basis for an job identity. It's what the job does to make up for that lack that makes a PLD what it is, what makes a GNB what it does, etc.

So what is it that you want DRK to have in exchange? To simply "not have X" is not going to be sufficient basis for DRK any more than it's been for WHM. It's how we fill that space differently that will give a well-designed job.

[Crater]

So what is it that you want DRK to have in exchange? To simply "not have X" is not going to be sufficient basis for DRK any more than it's been for WHM. It's how we fill that space differently that will give a well-designed job.

I think you're trying to reinvent the wheel where there's no reason to. The only thing that DRK needs "in exchange" for this sort of playstyle is "a lot of people find that sort of gameplay to be much more fun than the alternatives". That's it. No further justification is needed. PLD isn't some loser class that people only bring along when they absolutely need a second party mitigation skill or a tank that can briefly attack from range; a lot of people like its playstyle, and so it gets a ton of use. Neither would this version of DRK be.

We don't need to hold out for some galaxy-brain innovation or sit around wringing our hands about how it's hypothetically possible that a fight could come around that would be somewhat less advantageous to our class now that we find it fun to play at a base level. The aspects of the game that could conceivably turn this design into a disadvantage are things that the developers have intentionally minimized or outright removed from the game's design - if they ever decided to change their design philosophy and reintroduce those mechanics, then presumably that shift in design philosophy would also extend to class design, and the affected classes would be given something compensatory then. As of now, and as of the foreseeable future, there are no tangible disadvantages to an oGCD-focused, sustained-damage tank, and so we have no reason to oppose being given a class that fits that style.