Dragoons: A Rotation Reborn

[EasymodeX]

I overhauled my spreadsheet to make it more dynamic -- manually resetting data every time I tweaked a rotation was extremely annoying. Now I can plug in a rotation and it automatically calculates everything except the overlap buff times and clipped DOTs. GG.

With that, I plugged in a few rotations that basically said "who cares about clipping?" and "so what if DE drops because I'm lazy?". All based on a 2.4s GCD. Here are the results:

HT-PH-FR, IDC offset

HPF IDC ttt
HPF ttt IDC
HPF ttt ttt

t 64.8
D 7208.3
TP 1860
TPS 28.7
DPT 3.9
DPS 111.2

IDC, interleaved alt PH/HT + FR

IDC H ttt FP ttt H
IDC P ttt HF ttt P

t 62.4
D 7040
TP 1780
TPS 28.5
DPT 4.0
DPS 112.8

HT-PH, IDC offset (no FR)

HP IDC ttt
HP ttt IDC
HP ttt ttt

t 57.6
D 6483.4
TP 1620
TPS 28.1
DPT 4.0
DPS 112.6

IDC, interleaved alt PH/HT (no FR)

IDC H ttt P ttt H
IDC P ttt H ttt P

t 57.6
D 6527.4
TP 1620
TPS 28.1
DPT 4.0
DPS 113.3

KISS: IDC PH/HT flat (similar to what Ricauer posted, but fixed, because his filler sets were reversed :P)

HP IDC ttt
HP ttt ttt

t 38.4
D 4281.2
TP 1080
TPS 28.1
DPT 4.0
DPS 111.5

Based on the above, I'm going with:

1. FR isn't worth it
2. The IDC-anchored rotation is still the best
3. You barely lose any DPS (~1.2%) or DPT with the super simple rotation, so there seems little incentive to try so hard.

Edit: I accounted for the clipped DOTs where applicable. Incidentally, PH never clips in those rotations (unless you stack even more skill speed to some amount). CT clips in a few. That said, one thing to remember is that the DOT damage in this game isn't actually that much, and the ticks are spread across a long time. As a result, losing 1 or 2 ticks isn't a big deal, particularly when the DOTs are not "hyper-efficient" compared to direct damage attacks (e.g. the DOT abilities are not super awesome, so there's no need to "protect" them so hard from clipping).

Edit2: All references to "D" for damage mean potency, for anyone new to the thread.

Sidenote: None of the numbers above account for autoattacks. From prior assessments, autoattacks don't change much no matter the buff uptime, and at the end of the day they would slightly normalize any deltas in DPS between the above rotations.

Edit3: Triple checked all manual loop and DOT clipping adjustments -- had to tweak the HT PH FR rotation and "simple" rotation slightly. Net result is that I'm not bothering with anything more complicated than the simple rotation unless I start hitting major skill speed points (probably lower than 2.25s GCDs), after which it's worth re-evaluating.

[Ayvar]

snip

Let's see if we can resolve a few issues at once here and hopefully come to some sort of realistic and applicable solution.

Issue #1: Does it really matter if you clip your DoTs?

Issue #2: Is Fracture worth it?

Issue #3: Which of our available rotations produces more damage?

Issue #4: How does the inevitable increase in spell speed factor in?

Let's go back to our 3 competing rotations, we'll call them A, B and C

(A)

HFP IDC TTT
HFP TTT IDC
HFP TTT TTT

(B)

IDC H TTT P TTT H
IDC P TTT H TTT P

(C)

HP IDC TTT
HP TTT TTT

We had already calculated the p/s for our rotations on the previous page (7594 vs. 7525), however we ignored the issue of DoT clipping, but it seems now the discussion has moved to a point where we should factor this in. In rotation A the only DoT that clips is the initial Chaos Thrust (never Phlebotomize or Fracture), so instead of a 200 potency DoT we get a 180 potency, or 198 with buffs, so I've lost 22 potency from DoT clipping. However, going back through rotation A I realized the Disembowel buff is applied to the second Disembowel, so that should've been 266 and not 242, so with that slight fix rotation A comes out at 7596 total. With your rotation both the Phlebotomize and Chaos Thrust DoTs clip, so from your original potency number you've lost 20+22+22+22+22+22 (Phlebotomize DoT is now 110 and not 132) for the three clipped Chaos Thrusts (remember we had to come back around for a few extra moves) and 3 clipped Phlebotomizes, so your total potency is now 7395. Again using our 2.4 GCD that leaves mine essentially unchanged at 109.1 p/s and yours at 110 p/s, so the gap has shrunk from 3p/s to 1p/s.

Here is rotation C in potency:

HP IDC TTT
170-187/110-198-242-193/220-181-242-363

HP TTT TTT
205-205/110-181-242-363-181-242-363

+187 for the first HT when coming back to the start. The DoT clipping numbers have been included, the Phlebotomize clips but the Chaos Thrust doesn't. In total potency that's 4385 over 17 moves (40.8s) or 107.5 p/s.

So we have our final potency numbers when including the rise and fall of our HT and Disembowel buffs, and now DoT clipping of Phlebotomize, Fracture, and Chaos Thrust.

A: 109.1
B: 110
C: 107.5

Now we've answered Issue #1 and Issue #3. Your rotation came down by 2p/s due to DoT clipping and mine (assuming I hadn't screwed up the math on the second Disembowel) came down by 0.3p/s. This also helps us answer Issue #4. With Rotation B and C, since by default most of the DoTs clip, as spell speed increases the lost potency will get worse and worse. Including DoT clipping in your rotation robbed it of 2p/s, and if you end up clipping a second tick that'll come down again to 4p/s. Since rotation A only clips a single Chaos Thrust, it suffers dramatically less from increased spell speed.

Now Issue #2 seems to be a continued sticking point, and perhaps we'll just end up disagreeing on this. When actually playing the game, you never have 100% uptime due to movement. If including a rotation with Fracture only drops your 100% uptime dps very slightly (e.g. Rotation A compared to Rotation B, and this difference will close with added spell speed) then in fights with less than 100% uptime, the rotation with more DoTs will always pull ahead.

This has been great, albiet time consuming, to figure out. I'm glad we're getting some good numbers behind these rotations.

[HoolieWho]

Nice discussion.

(A)

HFP IDC TTT
HFP TTT IDC
HFP TTT TTT

(B)

IDC H TTT P TTT H
IDC P TTT H TTT P

(C)

HP IDC TTT
HP TTT TTT

I like rotation B for the simplicity of the alternating combo-triplet, single skills. Your priorities boil down to "H or P" and "IDC or TTT." When engaged in content where performance is important, I want a simple base priority system to free my brain to deal with encounter mechanics.

I hadn't considered working Fracture in while Blood for Blood is active. As of now Fracture is not on my hotbar. I have been choosing between the utility abilities Featherfoot and Mantra.

I seem to recall that SCH has a fairy ability that enhances skill speed. If so, how much of a boost is this and would it impact your rotations significantly?

[ManLet]

IDC, interleaved alt PH/HT (no FR)

IDC H ttt P ttt H
IDC P ttt H ttt P

t 57.6
D 6527.4
TP 1620
TPS 28.1
DPT 4.0
DPS 113.3

So this is the highest dps rotation possible? How come you don't get more damage from using HT first, or getting PH up faster?

[ManLet]

HT-PH-FR, IDC offset

HPF IDC ttt
HPF ttt IDC
HPF ttt ttt

t 64.8
D 7208.3
TP 1860
TPS 28.7
DPT 3.9
DPS 111.2

IDC, interleaved alt PH/HT + FR

IDC H ttt FP ttt H
IDC P ttt HF ttt P

t 62.4
D 7040
TP 1780
TPS 28.5
DPT 4.0
DPS 112.8

HT-PH, IDC offset (no FR)

HP IDC ttt
HP ttt IDC
HP ttt ttt

t 57.6
D 6483.4
TP 1620
TPS 28.1
DPT 4.0
DPS 112.6

IDC, interleaved alt PH/HT (no FR)

IDC H ttt P ttt H
IDC P ttt H ttt P

t 57.6
D 6527.4
TP 1620
TPS 28.1
DPT 4.0
DPS 113.3

KISS: IDC PH/HT flat (similar to what Ricauer posted, but fixed, because his filler sets were reversed :P)

HP IDC ttt
HP ttt ttt

t 38.4
D 4281.2
TP 1080
TPS 28.1
DPT 4.0
DPS 111.5

Is this accurate for a patchwerk fight?

[EasymodeX]

Is this accurate for a patchwerk fight?

Yes, generally. It doesn't take into account things like Internal Release being synchronized with anything in particular. But, the numbers should be very close.

I must ask. What dragoon jobb specifics is accualy making it worth go for dragoon over lancer? I ask this as I find the archer 20% dmg buff and crit buff wery usefull and use them over MAR utilitys any day.

Sup Osi.

Yeah, I think that thought has occurred to everyone. To that end I updated my model to include crit% with an assumed baseline of 25% (so crit% with SS up would be 35%) and the results are pretty entertaining. For most rotations the potency-per-second actually dropped by using SS, because the SS buff is actually pretty weak, the attack itself is weak (140, so you need to gain over 60 potency from the buff to match your filler to break even), and using it caused a loss of uptime of other buffs (like Heavy Thrust).

Even if we assume the same PPS for the "LNC" (+SS) and "DRG" rotations, take the next step:

A LNC with 450 STR has a potency coefficient of 126.0. A DRG with 470 has a coefficient of 130.6. That's roughly a 3.7% increase in damage from the STR. Ref: http://valk.dancing-mad.com/?page_id=179

The Raging Strikes buff is up for 20/180 = 11% of the time, and adds 20% damage. That's +2.222% average damage.


DRG STR bonus > Raging Strikes (although Raging Strikes is still good for burst damage situations)

Straight Shot is a wash at best (or a DPS loss at worst).

The net result is that the DPS of a DRG is going to be ahead 0-2% of the DPS of a LNC using straight shot ... without counting jumps. Jumps aren't the best DPS increasers in the world, but they do increase DPS.

The biggest question would be "how much random utility does a LNC bring" versus the damage / utility of jumps. Remember a LNC can bring rez, cure, prot, 50 defensive cross skills ... they all do nothing for DPS, but hey whatever right.

Edit: Jumps add in the general ballpark of 4% total damage.