How to Dragoon per Second: A DPS Paradigm

[GunksFoy]

Perhaps your latency is lower than mine or something. I'm at work now, but when I get home this evening I will definitely have to test this.

However, I will add that while this may be possible on a dummy, it's highly unlikely on a boss. From experience, I rarely have more than 1-2 seconds of HT left when I reapply and I am actually above the SS cap as DRG is not my main. That said, perhaps I'm doing something wrong.

The general consensus is that 2.4s, or more specifically 2.39s GCD is the optimal GCD for maximum dps and I agree, so I would aim for that regardless if it's *possible* to do with a slower GCD. I'm not sure how you would actually get any less SS without stacking way too much accuracy with the way gear is currently anyhow, but it's an interesting theory. I will definitely be trying it out when I get off work.

[Myon88]

Yes, I'm open to the argument that in actual fights against bosses, margins that small would be a liability. If you have to step away from the boss for a couple of seconds for a mechanic before reengaging - now your next HT (and possibly more) isn't covered by the previous HT.

My only issue with that line of reasoning is you can't quantify it easily, it's a really touchy feely argument similar to casters saying "More sks gives more chance that I'll be able to slidecast out of a mechanic". It's a potential DPS increase, but can it be proven with math that it outweighs biasing your stats toward sks instead of, say, det? That's not so simple.

Re: gearing, it's true we have a lot of sks on gear, especially sks that's bundled with gear with crit on it. Hypothetically, if it were favourable to lose sks, my hunch is we would do what the machinists do and opt for acc/det over crit/sks. I'm aware crit has a really good reputation among the community, but the reality is no stat is infallible. Would you take 9 crit, or 999 det? It's just a matter of the margins involved.

Anyway, I hope people will try it and let me know what I'm missing. I highly doubt such a lynchpin of HW DRG theorycrafting would be subject to a simple oversight like that.

[GunksFoy]

Anyway, I hope people will try it and let me know what I'm missing. I highly doubt such a lynchpin of HW DRG theorycrafting would be subject to a simple oversight like that.

I think this is the biggest evidence against your theory, however it is theorycrafting and not factcrafting after all.

While I agree it is not necessarily quantifiable, this logic doesn't necessarily apply to casters. Spell speed will always be a dps increase for casters because more casts in a given time frame is always a dps increase, and also allows for the potential of getting in that extra cast before having to move. Sure, crit could potentially increase dps more, but ss consistently makes you faster and therefore more potent, as well as the DoT scaling with SMN now. But that's all for another thread. Additionally, I don't think this necessarily applies to drg because we aren't stacking it, simply hitting a given threshold and then all but avoiding it.

You have a point about the idea of stacking acc/det instead, however the problem is that crit is so much more valuable than det. I don't believe there is enough Det in the game currently to compensate the crit and ss you would lose, as ss is still a dps increase and has a stat weight in the end.

[Myon88]

The stat weights for dragoons currently put a point of det at ~75% the value of a point of crit. There are issues with static stat weights but that too is another whole thread and I won't go into it here - let's assume they're reasonably accurate.

Crit wins as expected, but it's not as big a gulf as I would've thought. There is a very strong 'crit or die' mentality amongst the community, but the stat weights say it right there, 4 det is equal to 3 crit. If you offered me 5 det, or 3 crit, I would have to choose the 5 det based on the findings.

It seems within the realm of possibility that *if* - and this is a big if - some sks is extraneous, we could gain enough det (as well as det on the relic from all the accuracy we don't need there any more) to outweigh the loss of crit that was bundled with sks (on our legs, feet, neck, and wrists especially, since there we are choosing between crit/sks and acc/det).

As for casters, I won't deny sks is mathematically an excellent stat for them. That is a fact. It's just that the additional merits of potentially having less cancelled casts due to mechanics is a perk often touted that is above and beyond the base utility of the stat. For them skillspeed is a great stat even in a standstill scenario, but to say that it's an even better stat when you have to move? How much better exactly is something that needs to be quantified. Can you be certain that you had to cancel your cast because it was 3.3s instead of 3.28s? Maybe if you were bolder you might have gotten away with it anyway? My point is, it's anecdotal, it's not scientific either.

[GunksFoy]

That is a reasonable assumption. As I am a main Blm, I didn't investigate how much Det Drgs could gain by stacking Det/Acc instead of Crit/SS/Det. So the problem with stacking Det/Acc over Crit/SS is that the Det has to outweigh both Crit and SS as a whole, since accuracy has no value beyond the minimum. Basing this off static stat weights of course. I would like to see someone with a bit more experience with theorycrafting than myself weigh in as well.

As to casters, there are indeed merits beyond the "I can move easier" argument, however that is also a potential boon, even if it's situational. I agree you can't apply a value to it, but it is undeniably a benefit to stacking SpS and is worth mentioning. But I agree there is no measure to how much better.

[Shamox]

https://docs.google.com/document/d/1...FM8cdN5g/edit#

You'll find some usefull informations here.
Even if it's not up to date.

[muzzl]

I am not sure I understand the chart about blood for blood which shows that holding on to it for a cooldown is always a DPS loss. Is there an unstated assumption behind those numbers? Like maybe the idea is that the battle is going to end when you're under blood for blood so holding onto it for a GCD means that you are going to lose a GCD of it later on?

It seems to me that in many fights it would be possible to hold on to it for several GCDs while still getting the same amount of full b4b uses. If the chart is still accurate when you know you are not going to lose any b4b duration I would very much appreciate an explanation about why.

Thanks.

[Sai-Kon]

Continuing in the same notion that Myon88 started, I would like to go a bit further in the discussion of secondary stats for DRG and report an on-going experiment that I am currently conducting. My testing revolves around whether builds that favor heavy Skill Speed (more than the suggested) and builds that favor heavy DET (more than the current value of the BiS build) are a liability for DRG in a savage fight scenario or not. If yes (which basically confirms the stat weights priority by Dervy) then how big of a disadvantage are they?

According to the current stat weights for DRG, Skill Speed is the least valued secondary stat. Indeed, Skill Speed doesn't give a direct boost to stats but instead makes you hit faster and fit more spells in a given time. Moreover, more Skill Speed opens new possibilities such as the consistent 3.5 Geirskogul per minute that Thendiel has been vocal about in the past and a potential return to IDC4-TTT4-TTT4 combo before having to refresh Disembowel.

[Sai-Kon]

Before we go any further, I know that the first thing that comes to your mind when you hear high Skill Speed is obviously TP issues. I did sufficient testing and found that with Ninja's Goad (or MCH TP tower) and correct usage of my Invigorate I could DPS non-stop for around 5:30 minutes without running into TP issues with a 1000 Skill Speed build on a dummy. I could definitely push it a bit further if I didn't say to Ninja to stop giving me Goad given that you will almost never encounter a fight of more than 5:30 minutes non-stop DPS without downtime due to execution of mechanics. I have been experimenting with a high skill speed build (both high SkS and Crit to be precise) for a few weeks (between 930 and 1050 SkS and 1050-1300 CRIT) in an effort to see how it fares compared to the current BiS which revolves around Crit/Det. I will not report any precise numbers yet because, as I said, it is an on-going experiment but so far I have seen some great results DPS-wise.

[Sai-Kon]

Once I have sufficient data from which we can draw additional information and conclusions, I will post them here. I should mention that the testing has been conducted mainly in A9S and A10S for the time being.
Now that Savage lockout was removed I will try to get my hands on a specific build that revolves between 1200 and 1300 Skill Speed and test it out (ZTA3 and Z7V9 Ariyala builds).

Myon88's discussion about DET has piqued my interest to test a DET heavily oriented build and analyze it. More specifically, I am talking about the following Ariyala build: ZU7A, which prioritizes DET over CRIT and has low Skill Speed.

I need to clarify that these experiments are done for the sake of "science" and seeing how the stat weights that we have found based on theorycrafting translate into an actual savage fight such as 'The Creator'. They are not done to invalidate what we already know, but rather to add more data on the table and more food for thought for future theorycrafting.