Dear Dragoons. Your AoE is actually good.

[StouterTaru]

A drop from 40 dps to 16 dps (per mob) is a loss of 60% of your damage output from hitting the 0 TP wall, is what Keyln is saying.

And that's 40+ seconds into the pull that you hit that 0 TP. If your party is all doing AoE, the pack is all approaching dead at that point.

The ninja DoT goes to a similar argument; spreading out damage between mobs leads to more dps.

[JackFross]

You're ignoring the important part again. The GROUP potency.
If you insist on single-targeting, make sure you're swapping targets to the highest health mob at all times.

For some reason, I feel like you did not read my post.

>In a group of 6, your 0 TP, wait 2 ticks, hit Death Blossom 16 dps would be 16x6 or 96 dps, counting every hit of it.
>A Shadow Fang combo is 590. Over 5s this averages out to ~120 dps

I dunno about you. I spread my dots when I use them when multiple enemies are alive. Hitting Spinning Slash > Shadow Fang on one target, then switching to another to do the same thing becomes more dps than waiting 2 ticks for Death Blossom in pulls that are less than 8 mobs large. And I would eat my hat if cycling Shadow Fang from target to target is the optimal ST[AoE] rotation for Ninja, since I thought it up in 5 seconds. Hell, I'm sure you could simply add Mutilate and boost the overall potency (It adds 360 potency -> 950 -> but 3 gcds means ~7.5s -> 126.7, which makes it close to winning in 8-mob pulls).

The optimal ST[AoE] rotation for Dragoon is HT>IDC4>IDC4 with 4 gsk per minute. You spread your CT dots (and Disembowel debuffs) and drop Geirskogul on the targets. The reason Ninja is worse in AoE than Dragoon or Monk (Elixir Field, Rockbreaker built into the normal rotation) is because their base rotation has little in the way of AoE assistance. Not calling this an issue, just saying you need to actually read things I write if you want to try and say I didn't listen to your argument.

I'm not saying

PULL START - 6 MOBS - Let me start spreading Shadow Fangs!

I AM saying

PULL START - 6 MOBS - Let me use Death Blossom until my TP becomes unsustainable and start spreading Shadow Fangs.

And that's 40+ seconds into the pull that you hit that 0 TP. If your party is all doing AoE, the pack is all approaching dead at that point.

The ninja DoT goes to a similar argument; spreading out damage between mobs leads to more dps.

False information. A ninja using Huton (as every good Ninja should) will have a much shorter GCD than 2.5. Huton is a 15% attack speed increase, which means their gcd gets 15% shorter. Assuming 0 sks, as is ideal (though also not possible), a Ninja has a 2.12 gcd. Using this number, I made this table.



This includes cross-class Invigorate (+400 TP) and you can see that your TP will hit 0 ~34s into the pull. As I've mentioned in an earlier post (I sure do like to parrot myself!), the situations in which you will encounter a party capable of 30s kills on the trash in any relevant dungeons are approximately zero, outside of a pre-formed party with friends. And, in such a situation, you won't be bringing a Ninja, anyway.

However, that comment was me defending a correct statement against someone who drastically misinterpreted it to mean something wrong in their pointless smear campaign against someone who is trying to bring information to the thread.

(A case can surely be made for not using Huton in AoE seen here, since you can get more hits of the AoE off before your TP runs dry, and in this case you last longer than 40s straight up spamming it, but everything is still dying far slower than that, especially since you're a Ninja and every hit there is 100 potency per mob - Dragoon's single-target-aoe beats out Ninja spamming Death Blossom in mob pulls of 4 or fewer, and that's not gonna be breaking any records of add killing time.)

And, to reiterate, I'm not a Ninja expert and this is all very boring information to me that I really don't care about.


In addition, this is a topic about Dragoon. I'm applying what I've proven about Dragoon pages ago to Ninja. I don't play Ninja, so I really don't care to do math to figure out optimal rotations for this class. Hopefully you guys stop misrepresenting what others are saying now, at least. \o/

[FunkyBunch]

SNIP

I misunderstood what you were talking about. But I'm not sure you're understanding what I'm talking about either.
But math comparatively, even if you're doing 960 over 7.6s to one target, on that same target I've done ~617 potency with AoE.
Let's assume group of 3.
So every 6s (2 ticks) if you're doing 300 potency (death blossom) instead of shadow fang (590 potency including dot), you're losing out on my 617 potency not to mention the rest of the parties if he dies. Which means you've lost 327 potency when that mob dies vs if you had just waited for the two ticks. If your potency of 960 (which I think is a DPS number, but it's the number you used) then you've only added 43 potency which is negated if even one more party member is doing AoE.

So it really doesn't seem to me to be worth it at all. But I'm not a melee DPS person so I'm only running your numbers vs mine.

[Lilith_Merquise]

In short the way to think of it is that all of the mobs in a "Trash Pack" are one big boss monster that consist of multiple health bars that compile as on HP bar (like 2nd Statue Boss of Quarn HM). If you can hit An AoE and it hits multiple HP bars that adds up to 1 ultimate set of damage. So if AoE is hitting 6 targets, for a total of 960 potency for the collective HP, it is more productive to spam that then to just do 710 collective for a single portion of that collective HP Bar.

While you're doing a 1-2-3 (DRG as Example) (150 > 200 > 360) if you spam AoE you're doing (780 > 780 > 780) which is more productive.

While utilizing your TP regen abilities (most DPS classes have them if not MP based) and as long as people are adequately geared (if not you shouldn't be big pulling anyway) and all doing AoE for as long as possible things will either die, be very low to where you can easily finish off quick with single targets to not hit 0 TP, or reduced to 1-3 to where you can finishing off single target anyway. HP bars don't matter because the only things that have high enough HP bars to not AoE is bosses, but all trash is AoE fair game as long as it's in a pack of 3+

[StouterTaru]

False information. A ninja using Huton (as every good Ninja should) will have a much shorter GCD than 2.5. Huton is a 15% attack speed increase, which means their gcd gets 15% shorter. Assuming 0 sks, as is ideal (though also not possible), a Ninja has a 2.12 gcd. Using this number, I made this table.

The table assumes you just missed the 3 second TP tick when you used the first DB. Also, it leaves out the time of casting 2 Doton. I used 2.1 sec GCD and rounded up to get 40 sec.

[JackFross]

I read some things but I wont bother answer them, because in my opinion they're stupid.
Multidot, 60-1:20, Tp <-- my answer on these things still: lol still serious about that? boy
*vanish again*

It baffles me how obnoxious you are toward people providing proper information that you yourself have accepted and agreed with not two pages ago.

So it really doesn't seem to me to be worth it at all. But I'm not a melee DPS person so I'm only running your numbers vs mine.

Still completely ignoring or missing what I was saying. I give up.

The table assumes you just missed the 3 second TP tick when you used the first DB. Also, it leaves out the time of casting 2 Doton. I used 2.1 sec GCD and rounded up to get 40 sec.

Yeah, it's not exact. All that will probably bump it to 40s, but it's still very rare you're gonna run into a situation where an aoe pull lasts that long in a solo queue in df, so knowing how to handle an aoe situation when it stretches on longer is incredibly important.


This is all a silly discussion about optimizing dps in dungeons, where dps matters the least. I've discussed the topic to death and proven what I've been saying with numbers in this thread, so I'm just done trying to educate people on how to play this class when they clearly already have their (potentially sub-optimal) methods ingrained in their psyche. I feel like I'm saying the same damn thing again and again and having people counter with the same argument that doesn't even argue against what I said again and again. It's boring and rote. So, let me spell this out, in one sentence.

Using AoE skills (such as Ring of Thorns and Doom Spike) is absolutely optimal in any situation where you are facing four or more enemies at once, however, if you ever drive your TP to zero in order to maximize casts of those AoE skills, you will do considerably less damage to that pull before the group dies than if you cut early and shift into a Blood of the Dragon rotation, spreading Chaos Thrust and hitting Geirskoguls.

This statement can be expanded similarly to any other job, not including the casters, though some will slightly modify the "four" based on their AoE potential and how their AoE works (like MNK does AoE rotation in 2+ (pls no Arm of the Destroyer tho, thx), NIN should start AoE at 3+, etc). By AoE here, I'm referring to spending GCDs on AoE.

[Alistaire]

Using AoE skills (such as Ring of Thorns and Doom Spike) is absolutely optimal in any situation where you are facing four or more enemies at once, however, if you ever drive your TP to zero in order to maximize casts of those AoE skills, you will do considerably less damage to that pull before the group dies than if you cut early and shift into a Blood of the Dragon rotation, spreading Chaos Thrust and hitting Geirskoguls.

Saying "four or more" here makes this incorrect, again there's a # of enemies where there's at least that # where you're better off waiting for tp to aoe. Generally around 5+ though at 5 the difference is minor. You can't honestly think if you have 12 mobs your solo rotation over 2 hits makes up for hitting all 12 mobs...

[JackFross]

Saying "four or more" here makes this incorrect, again there's a # of enemies where there's at least that # where you're better off waiting for tp to aoe. Generally around 5+ though at 5 the difference is minor. You can't honestly think if you have 12 mobs your solo rotation over 2 hits makes up for hitting all 12 mobs...

In the case of Dragoon, I can expend 250 TP over 4 gcds (9.6s; net: 70 TP spent) to put out a combo that will run to completion for 1344 potency = ~140 pps, not including shared buffs. This can be dropped on subsequent mobs and never re-applied to the same one in a pull of sufficiently large (3+) size. In that same span of 9 seconds, I would get 180 TP back and then burn it on a 160 potency skill to hit all 12 for 160x12 = 1920 potency = ~213 pps. So yeah, in the instance where you run into 12 mobs, then sure, waiting on TP becomes a gain.

This threshold is only crossed at 9 mobs being hit with Doom Spike, and the threshold becomes even higher when you take into account the fact that using Heavy Thrust to buff the potency would *lessen* its overall effectiveness in a 0-TP situation, whereas if I incorporate a Heavy Thrust buff into the first combo (since it will most certainly have it), we can see that it'll be ~1545 potency = ~161 pps, which will beat out waiting on TP for Doom Spike up and through 9 mobs.

And, try as I might, I can't think of any situation where you fight 12 mobs at once where ALL of these conditions are simultaneously met:
1. Optimal strategy = kill all enemies
2. 12 mobs live long enough for your TP to drain to zero
3. 12 mobs actually exist in the same pull

So I don't know why you're randomly plucking 12 out of thin air like this.

Arboretum:
Pull 1 is 8 mobs; they spawn a Malboro that you pull to a pack of 6 Korpukkurs - a 7-mob pull.
Boss 1!
Pull 2 is usually 6 mobs. At absolute maximum, it would be 11, including the wall. If you're burning walls, you ignore these adds. Otherwise, that's a huge pull.
Pull 3 is completely irrelevant (also less than 12) because the optimal strategy is to burn down the wall and get to the boss, ignoring the adds.
Boss 2!
Pull 4 is 4 mobs.
Pull 5 is 6? I think. I forget exactly how many spawn here. I think it's 3 bears, 1 biloko, 2 little shits.
Pull 6 is 3 mobs.
Boss 3!

Pharos Hard shakes out similarly. Even the first pull is only 10 mobs, and the Corruption enemies die incredibly fast, so you'll never hit 0 before you're down to 6 enemies.
That's the biggest pull in the entire dungeon, unless you pull after the second boss and grab the 1+3+1+3 and then kill the furnace to get another +3 to get 11 mobs, still not 12, and there's no way you survived the pull to the furnace, due to AoEs from mobs slowing the dps burn and the sheer amount of health your tank will be losing due to the type of mobs that are on it.
The last pull in the dungeon is only 6? 7? Still not 12.

There MIGHT be a 12-mob pull in Fractal, but the mobs there die long before your TP runs dry unless the other dps in the party is wearing greens. I can't pinpoint it, but I think it's the second pull, *maybe*, and even that, I think, is just 8? 9? And the manikins all die in a couple shots.


But, yes. I agree with you that, in that hypothetical, has-never-been-realistic-since-geirskogul scenario, you should optimally burn your TP to 0 and then wait on refreshes and throw out Doom Spikes, even if they don't have a Heavy Thrust buff. Absolutely.

[Darwyn_Ulondarr]

Smart Person.

It seems that isn't a burden you share.

[Keyln]

I see what you're saying, now, but in the vast majority of circumstances in this game (like, for every 1 where this is false, there exist at least 8 where it is true), large groups of mobs either have similar enough health that (all things equal) they'll die on the same hit or they're spawned by a boss and thus obviously have less health than the boss mob, but similar health to each other.

I would say that it happens far more often than that. 3rd part of Arboretum comes to mind, as well as the latter half of Pharos Sirius.

Looking at damage per mob totally does work, but when trying to compare two different methods on their efficiency in killing a large group of enemies, it's much more beneficial and telling to compare the total amount of damage each rotation can dish out in a given period of time to the mob group.

The problem is that the mob group can vary in size and in composition. Damage done to the group as a whole can vary the number of enemies hit. Obviously hitting 6 enemies will show up as more damage than hitting 3 enemies, even though the potency of your attacks really haven't changed. And if the enemies vary greatly in HP, then the time it would take to kill the group would have to factor in the greater HP. You could look at it as 8000 HP, but the time to kill the group would be greatly skewed when the 1000 HP enemies die off midway through your attack spam. The reason why I prefer the potency per enemy model is that it automatically controls for the various group sizes and the varying composition of enemies without having really to adjust anything.

Unless you're running outdated content, there's no area in the game where a full spam of Doom Spike alongside a BLM using optimal AoE rotation will kill the mob before the dragoon hits zero TP, unless the BLM wastes his Raging Strikes on a triple flare, rather than saving it for the burst opener on the boss (or you're with a WAR/DRK who's doing high dps AS WELL AS YOU TWO and a healer doing the same). And, in the situation where your AoE is NOT killing mobs, looking at the damage it's dealing on a per-enemy basis doesn't tell you as much as looking at the total damage you did in those 10 gcds.

I would argue differently. Using the potency-per-enemy approach, we know how much damage we need to do to clear a group of enemies in a given amount of time. Using the equation AoE(T)=lmHP/DPSPE, where AoE(T) is the kill time using AoE, lmHP is the monster with the largest HP, and DPSPE is Damage per Second per Enemy, we can figure out how much damage each group needs. So, let's say that the majority of enemies has a HP of 30,000 (Anything about 30,000 we'll say we single target). We don't want fights that last over 30 seconds. So, thus, we plug the numbers into the equation, and we get 30=30,000/DPSPE. Using simple algebra to solve for DPSPE, we get DPSPE=30,000/30 or DPSPE = 1,000. So, this means that the group, as a whole, needs to do at least 1,000 DPSPE to clear a group, assuming similar enemies, in 30 seconds. The thing is that DPSPE for an individual can be measured by using the AoE attacks on a dummy, because DPSPE does not scale for the number of enemies. So, the damage you do on one enemy is the same as you would do on 12.

The reason I disagree on the "damage-per-mob" approach is just that it skews heavily toward single-target dps, unless you divide that by the amount of mobs present, in which case it becomes literally the exact same thing as just comparing total damage dealt, just multiplying AoE rather than dividing the single-target.

True, although I think you mean multiply, not divide. But, in either case, the two can be comparable if both are viewed as a function of time. Consider the two equations:

The equation for the kill time of a group of enemies can be described as follows: AoE(t)=lmHP/DPSPE, where AoE(t) is the amount of time needed to kill an enemy, lmHP is the enemy with the largest HP, and DPSPE is the damage done per second per enemy.

On the other hand, the formula for the time to kill a group of enemies using only single target abilities is as follows: ST(t)=tmHP/DPS where ST(t) is the time to kill a group using single target abilities, tmHP is the total monster HP for the entire group, and DPS is the damage done per second. Note that because single target only targets one enemy at a time, the PE part is unneccesary.

So, when AoE(t)>ST(t), then using AoE abilities would be more efficient. On the other hand, when AoE(t)<ST(t), then using Single Target abilities would be more efficient.

Basically, you can't at all compare HT>IDC4>IDC4 with 4 gsk/min to HT>RoT>DSx8 in a "per-mob" or "total-damage" potency examination unless you modify *both* of them to examine the situation more thoroughly. Saying DS is 160 potency per mob per gcd while the base rotation (H>IDC4>P>TTT4) is (not including buffs) ~280 potency per mob per gcd completely ignores the fact that, if there are two enemies, your 280 per mob is only doing 280 per gcd whereas my 160 per mob is doing 320 per gcd.

I'll admit that my DRG is only level 50, so I don't know enough about DRG to make any sort of comment specifically about DRG. All I can do is point up to the calculations above, and put both of them as a function of time, assuming a similar group.

But...

Using AoE skills (such as Ring of Thorns and Doom Spike) is absolutely optimal in any situation where you are facing four or more enemies at once, however, if you ever drive your TP to zero in order to maximize casts of those AoE skills, you will do considerably less damage to that pull before the group dies than if you cut early and shift into a Blood of the Dragon rotation, spreading Chaos Thrust and hitting Geirskoguls.

I agree with this conclusion.

Yeah, same reason I did it in my calculations earlier in the thread. Just as long as you are aware that the rotations don't last exactly that long, since discrete numbers don't operate nearly the same way as continuous ones. It's only really useful to give ratios of TP consumption and regeneration between different methods. It's very important to note that, while you regen 20 TP per second, you do NOT regen 70 TP in 3.5 seconds, you regen it in 6. I would actually think it'd be interesting to examine a 0 TP scenario and compare waiting for 2 ticks to pop a 120 TP skill (Death Blossom) vs using the TP as it comes back for your base rotation. Which would *actually* be stronger? Comparing pps here doesn't tell nearly the whole story.

Well, I did it that way to make the math a bit cleaner. I hate dirty math, and I prefer to do my math in a lab with industrial chemicals and...oh wait, that's something else. And as far as what's better, it would depend on if AoE(t)>ST(t) or not at TP starvation point.