Dragoons: A Rotation Reborn

[Ayvar]

DoTs and How to Swing Them

One of the issues I have been attempting to resolve is whether or not there is any point to even using Phlebotomize. It is quite expensive TP-wise (this isn't much of an issue with Invigorate, but it does force you to keep half an eye on your TP bar when other stuff is going on), and how much dps does it really provide? While it is your single most potent ability that does not require a combo (170 initial + 150 DoT = 320 total potency from a single GCD), the problem is that the -10% piercing resistance debuff provided by Disembowel does not apply to DoT damage. Any direct attack used by a Dragoon (such as the initial hit of Phlebotomize) does count as piercing damage, but DoT damage is its own class. In addition, reducing piercing resistance is quite valuable. -10% piercing resistance doesn't increase your your abilities by a flat 10%, it increases the damage taken by 10% from piercing attacks. That means if you'd hit for 100 normally, the 15% from HT would bring that up to 115, which would then carry over into the 10% extra taken from Disembowel, bringing it up to 126 potency instead of 125 (115+10% and not 100+25%). That's because the damage you are doing has been increased by HT, but the damage the mob is taking has been adjusted by Disembowel.

This leads to my dilemma. Do we even want to use DoTs, or does using only pure piercing damage attacks (and potentially getting more Full Thrusts) provide more dps? I am going to attempt to answer this question theoretically using potency. I'll assume that potency is the damage you do, so if you have a 100 potency attack that has been adjusted by the 15% from HT, I'll calculate that out to be a 115 potency attack. This will not take into account the damage you are actually doing, but using it as a proxy for a paper napkin calculation should be fine. Lets look at our nice clean dragoon rotation. We'll ignore off-GCD buffs for this, but they could slightly adjust the math depending on how they are used.

HT-ID-Dis-CT-Ph-TT-VT-FT-HT-TT-VT-FT-Ph-ID-Dis-CT-HT-TT-VT-FT-Ph-TT-VT-FT

In terms of potency, this would come out as:

Heavy Thrust - 170
Phlebotomize - 170/150
Impulse Drive - 180
Disembowel - 220
Chaos Thrust - 200/300
True Thrust - 150
Vorpal Thrust - 200
Full Thrust - 330

170-180-220-200/300-170/150-150-200-330-170-150-200-300-170/150-180-220-200/300-170-150-200-330-170/150-150-200-330

However, we know these numbers are adjusted by the damage buffs. For example, while the first HT may have 170 potency, the second has a theoretical 215. This is because you have your 15% buff from the first HT= 195.5 potency, and then the piercing damage buff = 195.5+10% = 215. I'll always round down just to be safe in my numbers. So we have to adjust each potency number by the buffs that theoretically should be up when they are used. Take note this will almost never occur perfectly in a real fight, but we strive for perfection.

170-207-253-253/340-215/172-192-253-417-215-192-253-417-215/172-227-278-253/340-215-192-253-417-215/172-192-253-417 = 7360 potency

Note that the DoT portions of Ph and CT are only ever adjusted by the 15% from HT, while the initial damage portions are adjusted by both HT and Disembowel. This gives us a 24 move rotation with a ~2.4s GCD (assuming some spell speed) that takes ~57.6s to accomplish. We have to add in two extra moves since if our buffs are running by the end of the rotation, when we loop back around the initial HT, Impulse Drive, and Disembowel will be buffed by the final Disembowel from the end of the previous rotation, so we add +215+227+278 potency = 8080 potency over 27 moves (64.8s) = 124.7 potency per second (pps). Now lets create our theoretical rotation without Phlebotomize.

First we open with HT-ID-Dis-CT, this lets us have or 15% buff for longer than ID-Dis-CT-HT (1223 potency vs. 1117).

(1) HT-ID-Dis-CT

Now we have downtime, so let's TT-VT-FT until its time to refresh.

(2) HT-ID-Dis-CT-TT-VT-FT-TT-VT-FT

Here we run into the first problem with ditching our DoTs. We needed a gap filler between CT-TT to correctly cycle the HT buff. However, with this rotation you end up losing the buff on the second Vorpal Thrust, and Full Thrust and Heavy Thrust don't get buffed. However, if the piercing damage increase makes up for this then it's alright, so we continue. Refresh HT, and fill until ID-Dis-CT time.

(3) HT-ID-Dis-CT-TT-VT-FT-TT-VT-FT-HT-TT-VT-FT-ID-Dis-CT

Again we hit the same problem, the Disembowel and Chaos Thrust buff/DoT are falling off one move before you can refresh them. If we continue the rotation until it eventually wraps back to the beginning, we get this monstrosity:

HT-ID-Dis-CT-TT-VT-FT-TT-VT-FT-HT-TT-VT-FT-ID-Dis-CT-TT-VT-FT-HT-TT-VT-FT-TT-VT-FT-ID-Dis-CT-HT-TT-VT-FT-TT-VT-FT-TT-VT-FT

Compare to our initial rotation:

HT-ID-Dis-CT-Ph-TT-VT-FT-HT-TT-VT-FT-Ph-ID-Dis-CT-HT-TT-VT-FT-Ph-TT-VT-FT

We've successfully managed to make our lives 2x worse by swapping a 24 move rotation for a 40 move rotation. By this point I already begin to balk. I can get off the full 24 move rotation in most fights, but a 40 move rotation will never happen. But, for the min-maxers in us and the sake of argument, lets see if we would do more damage anyway by comparing the potency. Here's the monster rotation again:

HT-ID-Dis-CT-TT-VT-FT-TT-VT-FT-HT-TT-VT-FT-ID-Dis-CT-TT-VT-FT-HT-TT-VT-FT-TT-VT-FT-ID-Dis-CT-HT-TT-VT-FT-TT-VT-FT-TT-VT-FT

In potency, taking into account the rise and fall of both the 15% and -10% damage buffs, here's how it works out:

170-207-253-253/340-192-253-417-192-253-363-187-192-253-417-227-253-253/340-192-253-363-187-192-253-417-192-253-417-227-253-220/300-187-192-253-417-192-253-417-192-253-363 = 11453

Ugh. Also we have to add in an extra move for going back to the beginning since the initial HT will be buffed by the previous Disembowel now, equaling 187 potency. That's 41 moves, with a total of 11640 potency over 98.4s (again assuming ~2.4 GCD) = 118.3 pps. That means even if you were nuts enough to switch to a 40 move rotation over a 24 one, you'd be doing it for much less dps.

In short, Phlebotomize is worth it because it allows you create better buff uptime by helping you to cleanly cycle their timers, and so producing more damage in the long run. An extra 6 potency per second may not seem like much, but if you extend that over a 5 minute fight its an extra 1800 potency worth of damage, or the equivalent of about 7-8 extra abilities. Also who the hell wants to do a 40 move rotation with a bunch of boring thrust moves? Not me, that's for sure.

Ayvar

******************************************************************************************************************

Go Buffed or Go Home

I was playing around with a thought experiment, and figured it might be fun to tack it onto the end of this section since it is relevant to how useful DoTs are. Frequently you will not have a chance to perform the full dragoon rotation, and I was wondering what would be the best moveset that was only maybe 6-7 abilities long (for example the approximate uptime you might get on some adds on Garuda or a totem on Ifrit. What I wanted to know was if you only had a short time period with which to dps, should you bother applying DoTs and damage buffs, or just spam TT-VT-FT? Lets use 6 abilities as an example, since this is short enough to question the use of DoTs and damage buffs since DoTs wont last their full duration in a 6-ability moveset, and damage buffs potentially haven't been up long enough to show their full effect.

Let's start with a baseline and work from there:

TT-VT-FT-TT-VT-FT = 1360 potency

So anything that provides greater than 1360 potency will prove to be better than straight combo spam. Let's add damage buffs and see what happens. They obviously must be applied right at the start or there is no point.

HT-TT-VT-FT-TT-VT

So you took the time to flank the mob and cast Heavy Thrust, but this cost you the end of your combo since the mob died before you could finish. However, did you contribute any extra damage to make it go down faster? Let's convert that moveset to potency:

170-172-230-379-172-230 = 1353

Well that wasn't quite worth it. But a potential part of the problem was that you used TT-VT at the end and without Full Thrust its quite a low potency combo, perhaps there are two other abilities that would better fill those spots. Impulse Drive into Disembowel is a nice two move combo, and has the added effect of -10% piercing resistance.

HT-ID-Dis-TT-VT-FT
170-207-253-189-253-417 = 1489

So now we see that even for very short duration fights, getting your damage buffs up quickly is very important. We didn't combo into Chaos Thrust since the mob won't live long enough for the DoT to matter. Now that we've confirmed we want damage buffs, let's check DoTs. Fracture is pointless because it's initial potency is so low (100) and the DoT already doesn't last the full duration, so we won't even consider it. Phlebotomize on the other hand has quite good upfront potency. Over a 6 ability fight (15s) you'll get ~125 potency out of a 150 potency DoT (as the full duration is 18s).

Ph-ID-Dis-TT-VT-FT
170/125-180-220-165-220-330 : 1410

Well that didn't quite add up to our damage buff moveset. However, the effectiveness of DoTs increases with even just a short increase in fight duration. Lets compare two example 8-ability movesets, one that uses only damage buffs and another that uses Phlebotomize and Chaos Thrust. With 8 abilities, the Ph DoT would complete its duration and so tick for 150 potency, and Chaos Thrust would get about 3 ticks, so ~90 potency out of 300.

HT-ID-Dis-TT-VT-FT-ID-Dis
170-207-253-189-253-417-227-278 : 1994

Ph-HT-ID-Dis-CT-TT-VT-FT
170/150-170-207-253-253/103-189-253-417 : 2165

To summarize, if a mob is going to die too quickly for DoTs to tick then stick to getting only your damage buffs up. If it lives long enough to let most of Phlebotomize tick then get Ph up and apply CT.

Ayvar


******************************************************************************************************************

The Art of AoE

Dragoons are surprisingly good area of effect (AoE/AE) damage, which refers to damage applied to multiple enemies at once. There are three things that must be considered before you choose which of our AoE abilities to use: 1) How many mobs are there? 2) How much damage do you need to do? and 3) Do you need to have TP left afterwards? Each question will greatly narrow your choices depending on how you answer. If there are 2 or less mobs (or 2 mobs and a boss) it is highly unlikely you'll be AoEing at all. Our single target damage is quite efficient, so you need at least three mobs to warrant any consideration of whether you want to use your AoE abilities or not. Once you have enough mobs to warrant the use of area damage, the answer to questions #2 and 3 will determine which abilities you use.

Dragoons have two options (with some slight variations on both themes) for AE damage. The first is a circle AE combo using Heavy Thrust -> Ring of Thorns (RoT) spam, and the second is a single line AE ability called Doomspike (DS). The first option requires the constant recasting of Heavy Thrust, and the second option only requires you to refresh your HT buff when it is about to fall off. Non-combo RoTs can be used, and although they are very inefficient in terms of damage output for TP cost, there are some situations where they might be considered, which I will cover below. If you want a quick burst of AE damage (e.g. Garuda plumes or Turn 4) then DS can be your best option. If you want to put out some AE damage, but the AE phase will last long enough that you would drain your entire TP pool (including Invigorate), then HT->RoT is your best option. A good example of such a situation can be certain waves on Turn 4 since although you may want to get them down quickly, you still need to leave yourself with enough TP to continue single target dps for the rest of the fight.

Let's take a look at some of the math behind our options. The most important thing to consider is how much damage both options will do by the time you've drained your TP resources. Doomspike might seem nice, but if you blow all of your TP doing it then you'll be useless when it comes time to get back on the boss. TP regenerates at a rate of 60TP every ~3s (depending on server timing). You start with 1000TP +500TP from Invigorate, so let's see how each of your options performs:

Doomspike

You have two options here, you can open with HT-RoT-DS then Doomspike spam until it's HT refresh time, or you can just skip the RoT and go HT-DS spam, so we'll check both. HT costs 70TP, DS 160TP, and RoT 120TP.

HT-RoT-DS

First begin with 1000TP, so you'd begin as such:

HT (70) - RoT (120) - DS (160) - DS (160) - DS (160) - DS (160) - DS (160)

Add that up as TP:

70-190-350-510-670-830-990

So you've blown your first 1000. But we regen 60TP every three seconds, and since we've used 6 GCDs (we'll use a 2.4s GCD) worth of TP (the first HT doesn't count since we're full) then we don't finish with 10TP (1000-990TP), we actually finish with 250TP (2.4*6=14.4/3=4.8, round down to 4 = 4*60=240TP+10TP). Since we dropped the 0.8 remainder we are about to get another 60TP, but not just yet. Since we're at 250TP, we can get in another DS then HT since the HT buff is about to come off:

DS (160) - HT (70)
160-230

Now we have 250-230=20TP, but we're still not dry since we've bought another 2*2.4=4.8s of TP regen, or 1.6 regen cycles. If we add the 0.8 remainder from before that's 2.6 or 2*60=120TP, which is enough to RoT since we've just used HT.

RoT (120)
120

And we get another tick of TP regen (2.4s/3 = 0.8 + 0.6 remainder = 1.4 = 60TP+20TP. But 80TP isn't enough to continue AEing, so we truly are finally dry. Or are we? We still have Invigorate! so +500TP for 580TP total, and we continue. I'll skip the full timing/TP regen calculations and just give the final results from here on:

DS (160) - DS (160) - DS (160) - DS (160)
160-320-480-640

And we finish with 120TP left. We could wait a second or two for that final 60TP tick, but overall it would cost us dps. The final tally is:

HT-RoT-DS-DS-DS-DS-DS-DS-HT-RoT-DS-DS-DS-DS

Try it yourself on a target dummy, though it could vary slightly depending on server time, potentially allowing for an extra DS at the end. Converted to potency this is 170+195 (2nd HT is buffed) + 172*2 (2 RoTs) + 184*10 (10 DSs) = 2549 potency. But we need to factor in the number of mobs we're hitting, so the potency of both RoT and DS must be multiplied by mob number. We'll calculate for 3, 4, 5 and 6 mobs:

3 mobs = 170+195+344*3+1840*3 = 6917 in 14 GCDs = 14*2.4 = 33.6s or 6917/33.6 = 205 potency per second (pps)
4 mobs = 170+195+344*4+1840*4 = 9101 = 270 pps
5 mobs = 170+195+344*5+1840*5 = 11285 = 335 pps
6 mobs = 170+195+344*6+1840*6 = 13469 = 400 pps

HT-DS

Compare to the same rotation, but with DS instead of RoT:

HT-DS-DS-DS-DS-DS-DS-DS-HT-DS-DS-DS-DS-DS

By the end you finish with 40TP instead of 120TP.

3 mobs = 170+195+2208*3 = 6989 = 208 pps
4 mobs = 170+195+2208*4 = 9197 = 273 pps
5 mobs = 170+195+2208*5 = 11405 = 339 pps
6 mobs = 170+195+2208*6 = 13613 = 405 pps

Now we know that using HT->DS and skipping RoT is overall better damage, but not by an excessive margin.



Ring of Thorns

Now let's check RoT spam.

HT-RoT

HT (70) - RoT (120) - HT (70) - RoT (120) - HT (70) - RoT (120) - HT (70) - RoT (120) - HT (70) - RoT (120)
70-190-260-380-450-570-640-760-830-950

We've regenned 480TP, so there's still 530TP to use:

HT (70) - RoT (120) - HT (70) - RoT (120) - HT (70)
70-190-260-380-450

+500 from Invigorate = 550TP

RoT (120) - HT (70) - RoT (120) - HT (70) - RoT (120)
120-190-250-370-440

With 110 TP left over, but again we've regenned 480TP, so we still have 590TP:

HT (70) - RoT (120) - HT (70) - RoT (120) - HT (70) - RoT (120)
70-190-260-380-450-570

20TP left +240 regen = 260TP:

HT (70) - RoT (120)
70-190

70left + 120TP = 190TP:

HT (70) - RoT (120)
70-190

And we're done, so the final tally is 15xHTs (170+14*195 potency) and 15xRoTs (15*172 potency), so let's calculate for 3, 4, 5 and 6 mobs:

3 mobs = 170+2730+2580*3 = 10640 in 30 GCDs = 30*2.4s = 72s = 10640/72 = 147 pps
4 mobs = 170+2730+2580*4 = 13220 = 183 pps
5 mobs = 170+2730+2580*5 = 15800 = 219 pps
6 mobs = 170+2730+2580*6 = 18380 = 255 pps

Well we knew it'd be less damage, however the AE also lasted over twice as long. What we really want to know is how this damage compares to HT-DS if you waited for regen ticks. Let's check:

72s-33.6s = 38.4s of regen = 720TP, if we tack that extra TP onto our previous HT-DS sequence, we get:

HT-DS-DS-DS-DS-DS-DS-DS-HT-DS-DS-DS-DS-DS

plus

DS-DS-HT-DS-DS
160-320-390-550-710

3 mobs = 170+195*2+2944*3 = 9392 in 72s = 130 pps
4 mobs = 170+195*2+2944*4 = 12336 = 171 pps
5 mobs = 170+195*2+2944*5 = 15280 = 212 pps
6 mobs = 170+195*2+2944*6 = 18224 = 253 pps

Almost no AE phase should last this long, but it's interesting to see that as the number of mobs increases above 6, Doomspike will catch up in damage to HT-RoT, even if you have to sit there and wait for TP regen ticks, like so:

RoT - 7 mobs = 170+2730+2580*7 = 20960 = 291 pps
DS - 7 mobs = 170+195*2+2944*7 = 21168 = 294 pps

Additionally, remember to take the orientation of the mobs into account. Doomspike is a line AE and RoT is a circle, so if you can fit more mobs in your circle AE compared to your line AE, then it might be better to just use RoT instead of DS.

Here's a quick calculation for RoT spam over 33.6s to compare to the DS spam over 33.6s to show that the number of mobs you can fit into your AE can affect which one you choose to use:

HT-RoT over 33.6s, or 7xHT (170+195*6) and 7xRoT (7*172)

3 mobs = 1340 + 1204*3 = 4952 in 33.6s = 147 pps
4 mobs = 1340 + 1204*4 = 6156 = 183 pps
5 mobs = 1340 + 1204*5 = 7360 = 219 pps
6 mobs = 1340 + 1204*6 = 8564 = 254 pps
7 mobs = 1340 + 1204*7 = 9768 = 290 pps

For example, if you can only fit 4 mobs in your line AE, but you can get 7 mobs in your circle then that would be 273 pps from DS compared to 290 from RoT, meaning you should use RoT spam instead. This is also where you might consider using unbuffed RoTs. Though the damage of unbuffed RoT will outstrip that of buffed RoT over shorter durations, the TP cost to obtain this damage is worse that just Doomspiking (if added damage is what you're looking for). However, if there is a difference between the number of mobs you can fit in your circle vs. line AE then unbuffed RoT does much better than the regular HT-RoT spam. Here are the numbers for comparison:

33.6s = 14 GCDs, so two HTs, two buffed RoTs and 10 unbuffed RoTs = 170+195+172*2+115*10

3 mobs = 365+344*3+1150*3 = 4847 = 144 pps
4 mobs = 365+344*4+1150*4 = 6341 = 188 pps
5 mobs = 365+344*5+1150*5 = 7835 = 233 pps
6 mobs = 365+344*6+1150*6 = 9329 = 277 pps
7 mobs = 365+344*7+1150*7 = 10823 = 322 pps

So unbuffed RoT reduces the difference in number of mobs you have to fit into the RoT AE compared to Doomspike from 3 to 2. So, for example, now if you can only fit 4 mobs into a Doomspike line, and 6 mobs in RoT, then unbuffed RoT will now be better.

So over short duration AE phases Doomspike will do a lot more damage, but cost all of your TP after 30s. HT-RoT will do less damage over a short period, but will allow you to continue AEing for twice as long, and will do more damage than Doomspike over a longer period, but only if there are 6 or less mobs. If you want short duration AE damage, but are looking to reserve your TP then just use HT-RoT spam. Remember to switch HT targets while you AE, you want to make sure your HTs are always hitting the highest health mobs to ensure the damage is spread correctly. Also, the number of mobs you can fit into each AE (circle vs. line) will also help determine which one you decide to use.

Ayvar