Dark Knight under review: Feedback to the Developers and Community Team

[Lyth]

With the last live letter, it appears that DRK is finally being re-examined by the development team. So I've compiled a list of the core issues that need to be addressed.

Dark knight is under careful consideration at the moment, so please wait a bit longer.

Gameplay
DRK's main gameplay aesthetic is speed. This comes from a combination of a shorter average GCD length (due to Blood Weapon) and frequent oGCD use. In Heavensward, it was second only to NIN in terms of actions per minute.

Stormblood introduced a design decision that goes against this, namely, the move towards "Dark Arts spam". When the MP drain from Darkside was removed, DRK was left with a surplus of MP generation. The developer solution to this was to promote more Dark Arts usage to burn this off. The problem is that we have seen a shift away from short animation lock oGCDs like Dark Passenger and Low Blow, and been left with ones that have a longer animation lock, like Dark Arts and Plunge.

The reason why Dark Arts feels "clunky" is because it is relatively slow. In Heavensward, Dark Arts broke the steady flow of quick oGCDs in order to punctuate the wind-up for a big hit. Now that it is used between nearly every GCD, it makes an otherwise fast, smooth job feel slower and choppy.

This is also one of the reasons why people keep going on about fixing Dark Passenger for single target. It's not just about nostalgia (although DP is one of DRK's most iconic abilities in Final Fantasy). In a job designed around frequent oGCD usage, we vastly prefer shorter animation locks.

Another problem that arises from Dark Arts spam is that it forces unnecessary double-weaves. In Heavensward, it was much more straightforward to use Dark Arts on an oGCD ability. You could prepare the Dark Arts in advance, and then let it activate a couple of GCDs later. Now, because nearly every GCD uses up Dark Arts, if you want to use it on an oGCD, it needs to be a double weave. This is especially a problem for abilities like Carve and Spit, are designed to always be used with Dark Arts in single target.

It also creates a problem when you want to both use Dark Arts on an oGCD as well as on the GCD action that follows it, such as during a burst window.

I think there are two solutions here, in addition to shortening the animation time on Dark Arts. The first would be to just build the cost of using abilities like Carve and Spit into the action itself, and have Dark Arts specifically modify GCD actions only. The only problem with this is that it removes an element of player choice from whether you choose to 'boost' an oGCD or not.

The second option is to have Dark Arts expire only on the next GCD (giving 140 potency). This way, it would hit both the oGCD before it as well as the following GCD.

Recommendations:
-DRK relies on a lot of double weaves. Revise and shorten the animation locks on oGCDs, especially Dark Arts and Plunge.
-Consider making Dark Passenger a potency gain in single target, for variety.
-Make Dark Arts only expire on the next GCD, allowing you to activate both the oGCD before it as well as the GCD itself, rather than forcing you to sacrifice one for the other.

Situational Skills
For the tank job with the lowest number of actions left over after Stormblood's pruning, DRK has a very high proportion of situational abilities. We've talked about Dark Passenger already as an ability which, after the Stormblood nerf, is only dps efficient with multiple targets present. It seems a bit silly for an oGCD to be "for multiple targets only" when you have abilities like Circle of Scorn which are a dps gain in both.

Sole Survivor is another good example of this. The ability gives you a little under one Dark Arts use worth of MP every two minutes if you can find a dying mob or destructible object to use it on. In some encounters, you'll have things to use it on. In others, it just serves as celebratory fanfare at the end after going unused for the entire fight. There's really no reason why this shouldn't be a HP/MP drain over time, given that DRK has next to no self-healing. Perhaps this will help the skill actually live up to its name (i.e. surviving).

I have some mixed feelings towards Dark Mind. On one hand, in magic fights, it's really good when coupled with The Blackest Night. Neither cooldown is sufficient for the average tankbuster in isolation, but they combine to become very good. But it also creates a bit of a hole as far as physical tankbusters are concerned. I would personally prefer to see it be more versatile in terms of damage type, even if it came at the cost of a longer recast. Maybe have it absorb/reflect a portion of magic damage to keep it themed appropriately.

The biggest offender of all in this has been Stormblood's push to make DRK abilities stance dependent. DRK is probably the least favourable tank for stance dancing. It costs nearly 20% of your MP and a GCD to turn on Grit. Turning off Grit is pretty bad as well, as disabling buffs in this game tends to be sticky (as any pre-Stormsblood cleric stance user could tell you).

It seems bizarre, then, that so many parts of our kit are locked behind toggling stance. Grit off? No self-heals from Souleater, then. Grit on? You lose access to your only real burst dps tool, Blood Weapon. Blood Price is also in a weird place in which it either generates endless MP in big pulls, or next to none in single target.

Stance-based abilities made sense on WAR because switching was both oGCD and cost no resources. But even on WAR, we've seen a move towards making previously stance dependent like IR into stance independent. Stance constraints are not enjoyable. Being locked out of your burst cooldown is not enjoyable. It also goes against the job's aesthetic (Grit feels slow). Being locked out of any form of self-healing is not enjoyable. Neither of the other tanks is punished in this way for choosing a stance.

Lastly, Darkside occupies a strange place in DRK's ability set. It doesn't really fall off anymore outside of deaths. It really just serves as a way of manually toggling your MP regen on when there's no targetable mobs. Either turn it into a trait and have your MP turn on automatically when there are no targetable mobs, or remove the cost.

Recommendations:
-Convert Sole Survivor into a HP/MP drain over time. Consider making mob death reset the cooldown to keep the timing element.
-Make the Souleater lifesteal stance independent. If the amount of healing from this bothers you, just make the lifesteal into a Dark Arts property, not a Grit property.
-Make Blood Weapon and Blood Price stance-independent. Consider just reworking Blood Price into something different, because it seems like you're unable to tune it simultaneously for both single target and AoE. It's pretty much worthless in single target.
-Consider removing either the MP cost or the GCD cost on Grit.
-Consider either removing the MP toggle cost on Darkside, or make the ability into a trait.

Burst
DRK has historically tended towards sustained dps over burst. While these two damage styles may seem similar on a target dummy, they diverge in practical situations. For one, short burst works well with raid buffs. If a greater proportion of your damage is concentrated in the burst window, you get greater benefit under the buff. It also helps with short dps checks (burn down an add quickly) and tends to push ahead in lower uptime fights. It also allows for sloppier uptime as a whole.

The problem with this is that as the other tanks gain progressively more powerful burst windows, DRK falls behind. And it's not necessarily as apparent until you get off the target dummy and look at actual raid conditions.

The biggest culprit in this is Delirium. The ability is underwhelming on several levels, and reducing its recast doesn't really fix the problem. For one, it's the only offensive buff in the game with a resource cost attached. IR had a cost on release, but that was removed within the first few weeks. The 50 blood gauge cost on Delirium removes a large portion of its dps benefit.

It's also a finicky ability. In order to gain the full dps benefit, you need to drop your MP gauge low enough prior to use, or else you cap and waste MP. You need Blood Weapon active first, or else you lose the extension effect. You need at least 16-23 seconds of uninterrupted uptime in order to see any sort of difference over the base Blood Weapon. And as I pointed out before, concentrated burst is always better than one that is more spread out.

I think the solution is to rework Delirium so that it represents an significant improvement on top of Blood Weapon. Given that DRK's burst comes from Dark Arts usage, the easiest way to do this would be to make the Dark Arts potency gain increase under Delirium. I also think increasing the overall MP pool size relative to Dark Arts cost would help a lot in terms of preventing overcapping.

Recommendations:
-Rework Delirium. Remove the Blood cost. Consider merging Delirium and Blood Weapon (i.e. treat Delirium as an upgraded form of Blood Weapon). Consider having Delirium boost Dark Arts potency while active instead of just dumping an extra DA worth of MP into your pool.
-Increase the MP pool size relative to the cost of Dark Arts. If you are able to 'store' more Dark Arts, capping becomes less of an issue.

Balance and Identity
I've talked about the problems surrounding DRK's identity in Stormblood before.

Balance is controversial and it's difficult to get people to agree on the subject. The difficulty lies in allowing all jobs in a role to perform at a similar level while maintaining uniqueness. But a double standard exists here. Consider the following:

Letter from the Producer LIVE Part XXXIX
Warrior’s party utility will be increased by changing the effect of Shake It Off and adjusting its party defending capabilities to be more in balance with that of paladin and dark knight.

Letter from the Producer LIVE XLII
Q: While there were adjustments for all three tank jobs, we still think that paladin is the most efficient of the three. Will there be increases to warrior and dark knight’s support abilities and DPS?
A: Party supporting abilities are a part of the job’s identity, so we won’t make adjustments that will allow using actions with similar effect across all jobs.

What I would like is a clearer sense of the design philosophy. Party supporting abilities are either sacrosanct and part of a job's identity or they aren't. The problem with DRK at the moment from a balance perspective is that abilities that used to be unique to the job, like Reprisal and Plunge, have been shared for balance purposes. But the sharing has not been reciprocal. So we have been left with a lot of net deficiencies.

Plunge is a good example. In Heavensward, DRK was known for being a very mobile tank. In exchange, you would often end up using Plunge to negate things like knockback mechanics. When Plunge was copied over to WAR as Onslaught, it was improved on in every way. The recast allows it to be used virtually on demand. It generates bonus enmity if you need it. It has a significantly shorter animation lock. It has a longer range. When the other tanks received Fell Cleave equivalents, they were both designed to be significantly weaker so as to not eclipse it. Because of identity. But identity seems to apply only to certain jobs and not to others.

Reprisal is another good example. It created the sense of DRK being a counterattack-orientated tank. But while Reprisal was shared through the cross-role system (to give every tank access to party defense tools), the equivalents on other jobs were not also shared. It was not reciprocal.

The reason why abilities like Shadow Wall and Living Dead keep being brought up is similarly because they either have no unique identity, or because they bring unique penalties instead of benefits. The other two "30% mitigation" cooldowns bring something unique to the table. Sentinel brings it up to 40%. Vengeance reflects damage, lasts 5 seconds longer, and has a 60 second shorter recast. Shadow Wall offers just the base ability. It doesn't need to do anything powerful. It just needs to offer a unique job flavour. Like perhaps absorb MP or something.

Living Dead is similarly worth re-examining. Ultimate abilities should feel good to use. Hallowed Ground lets you shrug off damage. Holmgang lets you negate knockbacks and has a very short recast. Even if there's a penalty, it still offers upsides. Living Dead just feels like you're holding your healers back. People can work around it, but it doesn't feel good to use. There are lots of ways that this could be uniquely implemented. Perhaps let it prevent both damage and healing for 10 seconds. Either way, just get rid of the healing requirement. It gets progressively more obnoxious with every tier.

Recommendations:
-Shorten the animation lock on Plunge. Increase the range to at least parity with other gap closers. Consider adding in an enmity effect.
-Return Reprisal as a DRK job skill. Perhaps increase the recast to keep it in balance with the other party-wide defensives.
-Add an additional effect to Shadow Wall. It's very pretty, but very bland.
-Remove the healing penalty from Living Dead. Consider just having it negate both damage and healing instead, if you need to tie in a penalty.

Feedback and Communication
FFXIV is a big game. And it's getting progressively bigger. Each expansion brings more jobs, and more things to design and balance. This makes communicating with the player base and explaining design decisions all the more important. Large, keynote-speech styled feedback sessions every few months might have worked while this game was still in its infancy, but it's starting to show signs of strain. It's also very difficult for any one person, no matter how talented, to be an expert in all these things.

The Q+A sessions, while nice, have been too few and far between. This might be wishful thinking, but I want to see more of a dialogue between the developers and the playerbase. People are much more likely to get on board with design decisions if they understand why they happen and feel like they're being listened to. There have been a lot of design decisions that the playerbase has quite vocally questioned, not just regarding DRK, but tanks as a whole (VIT accessories, anyone?) This whole thing would feel better if there was a resident tank designer on the forums with whom we can hash things out with. I feel like this is nearly an industry standard at this point, or at least should be.

That's not to say that it isn't slowly starting to happen. Over in the PvP forums, for example, I see community reps like this popping in to acknowledge community feedback. It doesn't need to be anything complex; even a simple "Point taken, we'll pass it on" works wonders. Listening is an active process; if you don't acknowledge that you've heard it, we don't know if you have. We shouldn't have had to wait until a six-monthly Q+A session to force the issue.

When discussing the reasons why 1.0 fell short, Yoshi-p highlighted the importance of listening to the players. That's exactly what we're asking for here. We want you do more than just fix DRK as a job. We want you to show a commitment to listening to the playerbase, so this doesn't happen to another job, sometime down the line.

EDIT: Community Team Response

Greetings!

Thank you all for the extensive feedback and input offered on dark knight. Please rest assured that your thoughts have been gathered and forwarded to the dev. team, and we will continue to share your feedback as it comes.

It's also worth noting that we understand that as a community, it's important that you feel your voices are heard. While we cannot respond to every thread that we review and pass on, please know that we have community representatives (myself included!) who read through the official forums and report player feedback to the development team on a daily basis.

This means that oftentimes, your feedback is being actively shared and you might not know it. We'd love to be able to respond to every thread we read and share with the dev. team but the volume of feedback received and the nature of operating as a global community makes that unrealistic to achieve. I know that Kaposhipi has mentioned this before, but a lot of the work our teams do happens behind the scenes and we are often not at the liberty to discuss plans that are still in development.

With that being said, we will continue to try our best to keep the community up to date with the latest information and comment when we can. Please continue to share your feedback, thoughts and experiences with us, and we'll keep sharing it with the dev. team!