Dragoons: A Rotation Reborn

[xyaie]

President Panda , making dps rotation affordable to all:

"You can keep your rotation if you like."

We'll see how that goes. Ill get she is gonna force us to change and put small non theorycrafting free companies out of business

[Ayvar]

You sure about that?

Assuming HT buff:

HT-ROT (3 targets): (170+150*3)/(70+120)= ~3.26 DPT, 310 DPG
CT combo (4 out of 10 CT ticks): (180+220+160*1.11+20*4)/(70+60+60)= ~3.46 DPT, 219.2 DPG
FT combo: ((150+200+300))*1.11/(70+60+60)= ~3.42 DPT, 216.7 DPG
CT combo (10 out of 10 CT ticks): (180+220+160*1.11+20*10)/(70+60+60)= ~4.09 DPT, 259.2 DPG
FT combo (Dis. debuff): ((150+200+300))*1.11/(70+60+60)= ~3.80 DPT, 240.5 DPG

It seems like all CT/FT combo variants are superior per TP (for 3 targets) and if you're aiming for damage/global for bursting (BfB, etc.) you may as well use 3-target DS (3 DPT, 480 DPG).

I think you're looking at how AE works from the wrong angle (also some of the numbers are a little off since if you're assuming HT buff is up then you need to apply it to HT-RoT). You might get slightly higher damage for your TP from single target abilities, but you're only applying that damage to a single mob. For example, take a fully buffed set of single target moves, and compare those to a buffed AE moveset.

HT must be refreshed if down, and DE must be applied to each mob, so you'd get this:
170+198+242+181+242+363

HT buffed HT-RoT gives a potency of:
170-165 followed by 187-165


We'll consider a pack of 3 mobs, each of which requires 1000 potency to kill. With Full Thrust the process would look like this:
HT-ID-Dis-TT-VT *switch* FT-ID-Dis-HT *switch* ID-Dis-TT-VT-FT *done*
170+198+242+181+242=1033 *mob 1 dies* 330+198+242+187=957 *mob 2 dies* 198+242+181+242+363=1226 *mob 3 dies*

Give or take some autoattack damage (such as on mob #2 which I cut at 957 cause an autoattack would make up the difference), it'll take somewhere around 12-14 GCDs to kill all 3.

With HT-RoT it looks like this (M1=mob 1):
HT@M1-RoT | HT@M2-RoT | HT@M3-RoT | HT@M1-RoT | HT@M2 | ID-Dis@M3
170-165 | 187-165 | 187-165 | 187-165 *mob 1 dies* | 187 *mob 2 dies* | 198-242 *mob 3 dies*

Most of the mobs were dead after 9 GCDs, with two extra used to kill off the final mob with single target damage.

You might get slightly more damage for your TP on a single target, but the overall group damage is much lower and so everything will take longer to kill.

[Swarna]

I think you're looking at how AE works from the wrong angle...

Look at what I was replying to. I was simply making the point that every variant of single-target chains yields higher damage per TP spent. I didn't make the case that it cuts down clearing time or that it boasted a higher damage per global (I actually showed that it was higher for HT-ROT with the stated DPG values). It's already very obvious that HT-ROT spam can kill a group of 3 in less globals than just single-targeting so your simulation of an example of events isn't contradicting anything I said.

[Ayvar]

Look at what I was replying to. I was simply making the point that every variant of single-target chains yields higher damage per TP spent. I didn't make the case that it cuts down clearing time or that it boasted a higher damage per global (I actually showed that it was higher for HT-ROT with the stated DPG values). It's already very obvious that HT-ROT spam can kill a group of 3 in less globals than just single-targeting so your simulation of an example of events isn't contradicting anything I said.

Except that you also calculated the damage for HT-RoT incorretly by applying damage buffs to all the other formula, but not to it. In addition, there's not much of a point to showing that single target abilities are 0.7 TP per second more efficient than AE abilities, but you'd still AE anyway because it's way better damage for almost the exact same TP cost. It's information that no one can use for anything.

[Swarna]

Except that you also calculated the damage for HT-RoT incorretly by applying damage buffs to all the other formula, but not to it.

Where did I do this? I said assuming you already have the HT buff active (meaning you already have used 1 unbuffed HT-ROT) the relative damage values should be accurate to eachother. The only additional damage buff I applied was the DE buff (x1.11 modifier) for the relevant attacks as stated (and Chaos Thrust's active portion right after a DE).

And I disagree on your latter points. I just did some more number crunching on the merits of using a mix of Doomspikes and single-target combos vs. just plainly HT-ROTing. I'm going to post it in a bit.

Edit: Also, you're brushing off a ~0.7 DPT advantage in favour of about ~50 DPG, but the same logic can be applied to forsaking a ~0.26 DPT advantage on HT-ROT in favour of ~170 extra DPG on Doomspiking.

[Swarna]

Back to the topic of determining whether or not HT-ROT spamming is superior to DS-ST alternation for 3 targets: after having some time to think about it in the gym I deduced that it's always going to be a better case to do the latter.

HT-ROT spamming has set damage per TP (DPT) and damage per GCD (DPG) rates. DS-ST's rates are variable based on the usage % of each. If you are a person already content with utilizing HT-ROT spam for 3-target pulls then that means you are content with either its DPT, DPG, or both.

I algebraically calculated (omitted for neatness) some values of Doomspike and some variants of ST chains that would yield the same DPT over time as only HT-ROTing. I decided to show only the results of using non-DE buffed FT combos (as it is typically the worst scenario) and of using CT combos on fresh targets and having the entire DoT tick (as it is typically the best scenario). Almost every other scenario of using single-target attacks lies in-between these two in results.

(We are assuming that the HT buff is active and as such all the globals we are talking about here is excluding the 2 you spend every ~20 seconds to refresh it by doing 1 HT-ROT.)

HT-ROT's DPT is ~3.263 as already established earlier.
DS (~37.5% usage rate)+ non-DE buffed FT combos (~62.5% usage rate) yields ~3.263 DPT. All this means is that if you spend 37.5% of your globals using Doomspike and the rest doing your full FT combo (1 part per global) you are doing the same DPT as if you had just been HT-ROTing.*
DS (~75.9%)+ fully-ticked CT combos (~24.1%) yields 3.263 DPT.

Now what is the difference? The DPG values. One method of damage dealing is outputting it faster than the other (and spending TP faster) at the same DPT rate.

HT-ROT's DPG is 310 as shown in my previous post.
Using the aforementioned usage rates:

DS + non-DE buffed FT combos: ~315.42 (480*0.375+((150+200+300)/3)*0.625)
DS + fully-ticked CT combos: ~426.79 (480*0.759+((180+220+160*1.11+200)/3)*0.241)

Since these are typically the best and worst ways to single-target, all other results will be in-between these DPG values. This means that if you alternate Doomspikes and single-target combos in a way that maintains the same damage/TP cost ratio as HT-ROTing, you will deal more damage in the same amount of time.

Naturally you might say "but my TP runs out faster" or "I like the pacing with which HT-ROT uses my TP". That's why we can also look at a different mixture of Doomspiking/single-targeting that matches the DPG of HT-ROTing and compare the DPT instead. In other words, if HT-ROT deals X damage in Y amount of globals, we can come up with a DS-ST mixture that also deals X damage in Y amount of globals and compare how much TP each method consumes.

After more calculations:

HT-ROT's DPG is 310 as we already know.
DS (~35.5% usage rate) + non-DE buffed FT combos (~64.5%) yields ~310 DPG (actually slightly more)
DS (~77.0%) + fully-ticked CT combos (~23.00%) yields ~310 DPG

Now if you calculate the DPT rate for maintaining 310 DPG with mixtures of DS and ST combos:

DS + non-DE buffed FT combos: ~3.272 DPT ((480/160)*0.355+((150+200+300)/190)*0.654)
DS + fully-ticked CT comboes: ~3.841 DPT ((480/160)*0.77+((180+220+160*1.11+200)/190)*0.23)

Which is higher than the 3.263 DPT rate of HT-ROT. What this means is that you can do the exact same damage per global as HT-ROT with DS and ST combos and spend less TP in doing so (from a small disparity in the worst scenario and up to a very high disparity in the best case scenario).

Also, if your usage rates lie in between the values I've shown for matching the DPT and DPG of HT-ROTing (i.e. you keep your Doomspike usage between the aprox. mentioned 35.5% and 37.5% mark when you are using non-DE buffed FT combos) you will have both higher DPG AND DPT than HT-ROT and this is how I can definitely say that DS-ST alternation is superior to HT-ROT.

You might argue that you could over-Doomspike by accident and do less DPT (waste all your TP) but you've still done more DPG. You can say if you don't Doomspike enough you do less DPG (don't deal as much damage in the same amount of time) but you are still using your TP more efficiently with higher DPT. The point being is that HT-ROT is never outright superior to DS-ST, while DS-ST can be superior in all ways (when you're within certain usage thresholds) and (if you're used to the dungeons you're doing) you can speed up your runs flexibly by deciding when you value a higher DPG (and Doomspike more often) or when you need to conserve TP for something (and Doomspike less often). Not to mention synchronizing your major buffs like Blood for Blood and front-ending all your Doomspikes into it (making up for it with ST combos after) benefits this type of rotation far more than it does the HT-ROT cycle. User error is only a valid excuse for people inexperienced with a dungeon or their lack of awareness of their party members' pace or something else. Playing Dragoon is all about planning your attacks in advance anyways so this shouldn't come as a surprise to any decent Dragoon.

Note: I didn't account for any additional damage you may do with Doomspike if you have DE debuffs applied as a result of single-targeting. Accounting for this would (very slightly) widen the gap further for Doomspiking/single-targeting so I didn't see a need to show it.

tl;dr version:

For packs of 3 mobs:

-HT-ROT and then switching between using Doomspikes and single-target combos intelligently is better than just HT-ROTing repeatedly.
-Either the damage/TP cost, damage/GCD, or both will be superior depending on how you use it and you can manually crank up or down the TP expenditure to adjust this and speed up your runs.
-On the other hand, HT-ROT is a static rotation with no ability to tweak the DPT or DPG to better optimize your runs and it is never going to be superior in both.

If you're inexperienced with the pace of dungeon runs it may be "safer" to HT-ROT. By the same logic, however, you can still TP-starve yourself with just HT-RoTing and it would have been "safer" to use a more single-target heavy rotation. Keeping a general rule of thumb of never letting yourself over-regen TP (tick over 1000) and dumping TP in a manner that ensures you start every pull at around 800-999 (also aiming to effectively use Invigorate every cooldown) should suffice in you using Doomspikes and single-target combos effectively.

*For less numbers inclined people: this is the same as saying that (very roughly) if you are doing 3 Doom spikes for every 2 full (6 globals) non-DE-buffed FT combos (outside of the mandatory HT-ROT, as always) you are using your TP with the same efficiency as just HT-ROTing over and over.

[Raminax]

Probably wrong thread to ask (though I'll admit it's lovely), but is anyone else having issues with Leg Sweep's stun not seeming to exist? Does anyone know what triggers it?

To be specific, at times when I perform Leg Sweep, I'll cause the damage and nothing else. I'm not getting any "Stun (Resisted!)" messages popping up. It just seems to act as if there actually isn't a stun component to the skill. It's quite rare that it happens. I usually try interrupting things such as 100-Tonze Swings if the tank is struggling to avoid them, and the stun of Leg Sweep not registering is... both painful and annoying. -_- Am I missing something here? Is there some separate accuracy check for the Stun? Or does it require you to hit from a particular angle, such as the Impulse Drive combo or Heavy Thrust?

[EasymodeX]

I usually try interrupting things such as 100-Tonze Swings if the tank is struggling to avoid them, and the stun of Leg Sweep not registering is...

There are two different things.

1. A stun can be resisted if you've hit diminishing returns. You're not seeing this; this is not what is happening.

2. Many boss attacks cannot be interrupted. The bosses are immune to stun -- not because of diminishing returns (Resisted!), but because they are coded to give a big F U ,,|,, to interrupt mechanics.


You're seeing #2. Run away from 100-Tonze Swing .

[Raminax]

That's not it, though. Just yesterday I was doing Aurum Vale several times and interrupted all Eyes of the Beholder without any issues on the Coincounter. The third run of the evening had a tank struggling to get away from the 100-Tonze Swings in time, so I stunned those for him. Then suddenly I failed interrupting the second Eye of the Beholder on the same fight, despite the fact that 100-Tonze Swings happen nowhere near as frequent that you'll run into diminishing returns. I reaaally don't remember getting any resist messages, the Stun effect just didn't seem to exist. :/ Though it may just be that my memory was faulty. Suppose I'll take screenies next time I go there of the combat log.

At least Spineshatter Dive ALWAYS stuns unless the enemy is outright immune to Stuns, but I don't want to rely on that for interrupts due to the long cooldown and animation.

[EasymodeX]

That's pretty odd then. Of course, I tend to not try to stun things like 100-Tonze Swings so maybe my memory is off :O.