Dark Knight Megathread

[Lyth]

I'm glad that people once again are acknowledging that Dark Arts spam in Stormblood was terrible. The community as a whole fought to get rid of it. It was hands down the most poorly designed iteration of DRK, put together by someone who had no understanding about why the Heavensward design worked.

Jobs which involve resource management have to be designed from the ground up around resource generation and cost. If the resource generation is too high relative to the costs, or the costs are too high relative to the resource generation, it doesn't feel good. This is why both Stormblood and Shadowbringers felt 'off' for different reasons. The biggest issue at the moment is that the container size is wrong. 3000 MP increments in a 10000 MP container? Why? Does it not make sense to go with an integer number of multiples, like 1000 MP/2000 MP/2500 MP/5000 MP? I feel like none of these design decisions would have occurred if we were using two unique job gauges, instead of relying on MP.

I don't think that increasing MP costs on Edge/Flood is the solution. The next multiple up is 5000 MP. Could you imagine only being able to store 2 Edges at a time (3 if you include Dark Arts)? That doesn't sound very fun. If anything, it probably makes more sense to go down to 2500 MP or even 2000 MP, and then adjust the rate of MP generation accordingly such that you don't cap too quickly. More storage makes for more interesting burst possibilities under special conditions (i.e. tincture/fight-specific mechanics). Remember, you can't double weave Edge with itself, so extra capacity will usually get pushed outside of burst. And if you really want to reduce the number of double weaves in general, remember that a 1.5s GCD during burst naturally removes that consideration.

On the subject of DPS queues, on a populated DC they're not really noticeable. RPR feels like it has a bit more in common with GNB than it does DRK, though.

[DRKoftheAzure]

I'm glad that people once again are acknowledging that Dark Arts spam in Stormblood was terrible. The community as a whole fought to get rid of it. It was hands down the most poorly designed iteration of DRK, put together by someone who had no understanding about why the Heavensward design worked.

That's how I felt about Dark Knight going into Shadowbringers.

Jobs which involve resource management have to be designed from the ground up around resource generation and cost. If the resource generation is too high relative to the costs, or the costs are too high relative to the resource generation, it doesn't feel good. This is why both Stormblood and Shadowbringers felt 'off' for different reasons. The biggest issue at the moment is that the container size is wrong. 3000 MP increments in a 10000 MP container? Why? Does it not make sense to go with an integer number of multiples, like 1000 MP/2000 MP/2500 MP/5000 MP? I feel like none of these design decisions would have occurred if we were using two unique job gauges, instead of relying on MP.

Guzzling MP like a Black Mage is funny though, just wish MP restoration did feel bad though...

I don't think that increasing MP costs on Edge/Flood is the solution. The next multiple up is 5000 MP. Could you imagine only being able to store 2 Edges at a time (3 if you include Dark Arts)? That doesn't sound very fun. If anything, it probably makes more sense to go down to 2500 MP or even 2000 MP, and then adjust the rate of MP generation accordingly such that you don't cap too quickly. More storage makes for more interesting burst possibilities under special conditions (i.e. tincture/fight-specific mechanics). Remember, you can't double weave Edge with itself, so extra capacity will usually get pushed outside of burst. And if you really want to reduce the number of double weaves in general, remember that a 1.5s GCD during burst naturally removes that consideration.

Yeah, that is a good point, about double weaving, but again, MP guzzler like Black Mage would better fit Dark Knight, since it is a magic knight of some kind, and would be better suited to that type of playstyle, at least for burst.

On the subject of DPS queues, on a populated DC they're not really noticeable. RPR feels like it has a bit more in common with GNB than it does DRK, though.

I play on Crystal Data center. But I also really like how Reaper was handled for the most part, though there are issues that I have with it. Though I wouldn't go as far saying Reaper has more in common with Gunbreaker. As for Gunbreaker, Continuation skills for Double Down, Sonic Break, Fated Circle, and main combos, and I would probably be more enjoyable to play than it already is, just wish the potency of the current Continuation skills were 100 potency more.

[Shurrikhan]

3000 MP increments in a 10000 MP container? Why? Does it not make sense to go with an integer number of multiples, like 1000 MP/2000 MP/2500 MP/5000 MP? I feel like none of these design decisions would have occurred if we were using two unique job gauges, instead of relying on MP.

Because you generate MP both per-GCD and per-tick, though, it's not as if it matters. You used up 25% of your MP instead of 30% of your MP. Okay, but a fraction of a second later, you may have 23% MP left to generate instead of 28%. It's going to be a little messy regardless.

...At which point it just comes down to...

• the intended frequency of those skills while ensuring that design can still keep things like Stalwart Soul's and Delirium-spenders' MP generation in sync,
• the the nearness to 100% MP required to put all possible spenders into raid buffs, and
• the portion of offensive spenders cast only during raidbuffs.

Presently, we hold 3 charges and a third (remainder), so that we really only need 90% MP to unload a full burst, reducing perceived threat of overcapping while prepping burst, and from 100% MP we exit that burst with 800 or 1000 MP left depending on tick timing, ranging between allowing for a lost/wasted an active and tick MP gen each and being able to outright start from 9000 MP (in turn allowing us to pre-pop BW even without also using a barely-prepull TBN, in case the latter wouldn't pop).

Make that an integer with zero margin, and you've merely gone from one smooth number with a bit of helpful margin to another smooth number without any helpful margin.

And by swapping to the likes of 2500 or 2000, the spenders would no longer be fit to the multiples present from/due to our combos (both 2 and 3), since we have a two-step AoE combo but a three-step ST combo (hence the 600 MP gen per builder), which would likewise be messy in that you'd then be dealing with more ugly MP/gcd values. There's going to be a little OCD-bait regardless, but having three-and-a-third charges seems about the least of those.

Finally, do we even want to those "more interesting burst possibilities under special conditions", when those casts are just syphoned away from having any spenders or otherwise interesting actions outside of minute-marked bursts? And as you just noted, holding 4 or 5 charges of Edges then strains our bursts even further. RIP any ability to casts defensives during the 2-minutes.

__________________


Tl;dr: DRK's spenders being 30% of their MP --instead of a specific 1/3rd or 1/4th, etc.-- seems a non-issue. Changing it to an integer denominator would probably hurt more than it helps, and this all seems irrelevant unless we specifically want to increase DRK's apm (especially, over its lulls) anyways. Similarly, decreasing the cost and MP gen both to allow for even more charges of Edge to be banked for 2-minute bursts would likewise probably do more harm than good.