Link to Table of Contents

In general, the playstyle of SMN in 4.0 revolves around job mechanics which ‘locks’ each other out. This, along with the slight unresponsiveness of pet abilities, makes the job appear to be clunky and punishing to play. In reality, the job flows a lot better than initially thought. This can be apparent with the establishment of a general workflow on how the job mechanics interact with one another. In this document, I will illustrate how the job mechanics interact with one another and attempt to form a system of priorities that might be useful as a guideline for general play/optimization.

1.1 General Workflow / Lockout System

The job mechanics can be split into 4 parts:

1) Aetherflow

2) DWT(Dreadwyrm Trance)

3) Bahamut

4) Rouse

1.1.1 Aetherflow

The core mechanic behind SMN is aetherflow. Without aetherflow, you are basically losing the ability to execute the majority of your other job mechanics, namely DWT and Bahamut. Therefore, it is extremely important to highlight that the no.1 priority of SMN is to attain the maximum number of aetherflow stacks. In order to achieve this, it is highly favourable to keep aetherflow on cooldown as soon as you can in most cases; the odd case being speed runs where you should have already pre-planned a kill timing. This is something that I feel some of the other rotations neglect. Although it may appear logical at first sight to fit in 3 festers and make full use of your first DWT, refreshing your aetherflow at ~20+ seconds opens up a huge possibility of you losing an aetherflow usage at the end of the fight. I will not go too much into the numbers, but it should be very clear that losing an aetherflow usage, which translate into a Tri-d, DWT and possibly even a bahamut is an overall potency loss.

So what locks out or prevent you from using aetherflow? DWT. With the new changes to SMN in 4.0, aetherflow stacks can only be gained after you have used up your dreadwyrm stacks.

1.1.2 DWT

With DWT having a duration of 16 seconds, in order to prevent ourselves from locking ourselves out from using aetherflow once it’s cooldown is up while still enjoying the full use of DWT, we have to use DWT at least 16 seconds ahead of aetherflow cooldown. Following this concept, a general rule of thumb is that you should be using your DWT when your aetherflow cooldown is at ~ 18-20 seconds.

So what locks out or prevent you from using DWT? Bahamut. In Bahamut trance, you are able to use aetherflow stacks but are unable to use DWT.

1.1.3 Bahamut

Here, we are further extending the interaction of ‘lockouts’ between DWT and aetherflow. With a duration of 20 seconds, you have to use Bahamut 20 seconds before your DWT lockout timing (i.e 20 seconds on Aetherflow CD). This brings us to our next rule of thumb, you should be using Bahamut at the very latest when aetherflow has 40 seconds left on its cooldown.

So what locks or prevents you from using Bahamut? Rouse. As Bahamut de-summons your pet, you should be trying to maximize the use of your pet abilities before using Bahamut trance.

1.1.4 Rouse

You should know the routine by now. The next rule of thumb is to use Rouse 20 seconds before Bahamut to get the maximum usage of Rouse before de-summoning your pet. This brings us to a full circle where you should be using Rouse at the very latest when your aetherflow has just finished its cooldown.

WARNING!

The aforementioned system of priorities is ‘general’ for a reason. It is meant as a rule of thumb in most generic situations or when you are unsure of what to do next. Depending on mechanics, downtime, existence of multiple adds, etc. the guidelines might not apply very well. The key here is to understand why the workflow works and adapt by doing minute adjustments in times of need.

Warning X2! - Common Misunderstanding about the flow chart

The flowchart is meant to show you the latest possible time to use Bahamut/DWT, etc. In essence, it is trying to inform you of the maximum amount of delay that you can afford before using Bahamut/DWT, of which if you fail to follow the guidelines, will result in the flow getting disrupted. This is important as it gives you a guideline to follow when fight downtime,deaths or other factors messes up your cooldown alignments.

On the other hand, if you want to deal optimal DPS, following the general rotation described in this section will give you a better idea of the optimal way to use your various job phases. This is different from the flow chart as it is showing you the particular order that you should execute the phases to ensure that you get the most out of your abilities.

You can think of it as having the rotation as your go-to guide. When things go south and not as planned, you refer to the flowchart instead to re-find your bearings and adjust for the fight.

1.2 Optimization in each job mechanic

Now that we have established a relatively simple workflow of the timings that we have to abide by in order to prevent ourselves from ‘locking’ out, we move on to the specific points to note when executing each individual job mechanic.

1.2.1 Aetherflow

For single targets (ST), you will want to spend your stacks on fester whenever possible in most situations. Some of the exceptions, which are not exhaustive, include using Painflare/Energy Drain to accelerate your entry into DWT for various reasons such as getting your first aetherflow CD as soon as possible to ensure that you get an additional stack by the end of the fight and blowing your burst before the boss dies.

For multiple targets (MT), Painflare is the obvious choice over fester. The usage of bane will depend on how many targets there are, and the cooldown of your Tri-D / availability of DWT. An example, if you have Tri-D up with 1 remaining stack of aetherflow and there are 3 enemies, it makes more sense for you to tri-d one enemy, manually apply dots on the 2nd enemy, use painflare and enter DWT, using the Tri-D reset on the last enemy. You lose 150 potency from manually casting dots but gain 600 potency from using painflare while making full use of your 2 Tri-Ds. This is far superior than simply using Tri-D on one enemy and baning onto 2 targets.

A good trick to master is that you are able to cast 1 Fester/Painflare every 3 gcds (e.g R2 (Fester) → R1/R3 → R2 (Fester). This is very helpful when you want to use your aetherflow stacks ASAP but not lose potency to clipping.

A minor point to note is to ensure that you are not overflowing on MP when using aetherflow. R3s should always be used to keep your mana levels below the cap.

1.2.1.1 Early Aetherflow vs Late Aetherflow

As this is a controversial topic with lots of debate that I have a particularly strong view on, I would like to take this opportunity to explain clearly why an early aetherflow is always better than a late aetherflow as a baseline, even if you are losing 200-300 potency on your opener.

When you look at this issue, it isn’t simply just about the risk of losing an additional aetherflow stack throughout the fight. When you take a look closer at what it affects, you realize that every second that you delay using aetherflow, you are delaying your bahamut trance timing (Assuming that you are using bahamut with aetherflow for additional WWs which usually more than makes up for the potency loss in the opener). This is important because most fights aren’t a dummy fight that gives you consecutive uptime on the boss for more than 2 - 3+ minutes.

At every point where there is a downtime due to the phase changes / mechanics, there is a check on whether you have sufficient time to execute your AF stacks, DWTs, Bahamut. An example is Susanoo, where the first phase ends at about 1min 30 seconds. With a late AF timing, it will not be possible to use bahamut prior to the first phase ending, but it is also a huge waste to use it on the sword. Now imagine another fight that has a downtime at 3min 30 sec. You will again be gimped if you used a late AF timing.

To make matters worse, outside of a speed kill where you are aiming for a specific kill time, in normal pugs/farm groups, how are you going to predict if a group is going to clear a fight at 8mins, or 8min 30sec or even 10mins?

When doing progression, let’s say that you have seen the first 2 minutes of the fight and realized that you could pull off your late AF / bahamut timing. However, as you progress further, you soon realize that you can’t pull off the 2nd one as the boss phases at 3mins 30/40 seconds. You will now have to revert back to an early AF and re-learn how your rotation works for the first 3mins.

Here is another example for progression that is equally applicable when farming raids. What if your team has gotten used to the mechanics and the gain in DPS pushes the original phase timing by ~ 10-20 seconds which prevents you from using your bahamut in a late AF opener? Are you going to tell them to stop DPSing like how MCHs screamed at their raid members to stop DPS-ing to get their Wildfire off in 3.0?

The point I am trying to bring across here is that there are multiple, countless ways where a late AF can screw you over that you cannot account for outside of a pre-planned speed kill. Is getting a 200-300 potency at the start of your opener worth the potential misery / troubles that you might get over the next 5 -10 mins of the fight?

Here is a logical analogy that stuck with me: Should I get my quick fix for ‘potency’ right now and lock out all my options in the future, or should I take it slow and look at my options from the present till the future before taking my ‘potency’ fix? If ‘potency’ was a currency in real life, it becomes pretty clear that you would prefer the latter. Please remember that you are not denying yourself out of the option to use a late AF if you have done your future planning and find it worthwhile to do so. There is no reason to throw all your eggs into one basket for a small bit of potency gain.

In summary, always use an early AF opener as your “baseline” (i.e Your first go-to opener in all fights) and then decide on what adjustments you want to make after you have done your “future” planning. Start early and plan for your future! :P

Note: I would like to emphasize here that using an early AF is a concept, not an opener. I am not encouraging or convincing you that my rotation is the most optimal. There could be other variants of early AF openers that are better or worse, but the concept remains the same.

Note 2: The potency gained from additional R3s might be a lot lower than initially expected, which further skews the debate towards an early AF. This is because an early AF opener will usually enable you to include additional festers and your deathflare into raid buffs such as Litany, Chain Stratagem, Hypercharge, etc. A late AF opener will be using 150 potency R3s while these raid buffs are up, which can be a bit underwhelming. Nonetheless, the main argument against a late AF opener is not about whether it is a potency gain or loss when compared to an early AF opener, but that it locks you out of future options which have a very high probability of making your life miserable as a SMN.

1.2.1.2 Bane vs Painflare

Early findings have shown Bane to be inferior to Painflare in the majority of cases. I will not get into specifics just yet as my friends on the summoner discord are in the midst of mathing it out.

I would just like to highlight conceptually why Bane is showing to be weaker than Painflare. The key reason is due to the lack of efficient AOE filler skills as a 4.0 SMN while our DOT application has become significant more efficient (3 dots to 2 dots, higher duration, Tri-D resets).

When we use Bane on multiple targets, our subsequent aoe filler skills are R1/R3 which are worth 150 potency. However, when we use Painflare on multiple targets, we have more options because the secondary targets are not afflicted with dots. Therefore, we are able to manually cast B3/M3 on each secondary target which is significantly more efficient than casting R1/R3s.

Here is an analogy to describe the situation (The numbers are not correct. Please view this as a concept):

Imagine the following in the context of Finance:

Bane = Option A

Painflare = Option B

Cash = Potency

Months = Number of DOT ticks on target. (DOT Uptime)

Filler skills = Additional sign ups after accepting the option.

Additional Signup in option A = casting R1/R3

“Special deal” in option B = casting DOTs (M3/B3) on secondary targets

You have 2 options:

A – You take $1001 cash over the next 7 months (The higher the months, the more cash you get)

B – You take $1000 cash right now

In this scenario, at first glance, Option A (Bane) edges out Option B (Painflare) at 7 months (DOT ticks). However, we have not included the additional sign ups(Filler skills) that happen during these 7 months.

While waiting for the next 7 months, here is what you can do depending on whether you take option A or B.

Each month, people who take option A are eligible for an additional signup that earns $100 (Casting R1/R3s) every month. Thus in the 1st month, they earn $100. The 2nd month, they earn the same amount($100) and so on.

Each month, for people who take option B, besides the additional signup offered in option A, they get their own “special deal” where they able to sign up for a plan where they get an additional $80 every month till the 7 months are over. I.E In the first month, they earn a total of $80. In the 2nd month, they earn ($80 + $80 = $160). In the 3rd month, they earn ($80 + $80 + $80 = $240) and so on and forth. (Casting DOTs on secondary targets)

Total amount of cash(Potency) held by A after 7 months = 1001 + 100 + 100 + 100 + 100 + 100 + 100 + 100 = $1701.

Total amount of cash(Potency) held by B after 7 months = 1000 + 80 + 160 + 240 + 320 + 400 + 480 = $2680

In essence, the “special deal” by option B provides a significant amount of cash over the additional signup offered by option A. Furthermore, that will only get worse as you wait for more months (“Special Deal” gets more and more attractive with more months). Also remember that option B can still use the additional signup offered by option A(Cast R1/R3s) if it is too short(Low amounts of DOT Ticks) to make use of the “Special Deal”.

What makes this worse is that the lower the number of months(DOT uptime on targets), the amount of cash that option A gives you will also drop (Bane loses effectiveness the lower the number of dot ticks).. But the higher the number of months, the “Special Deal” will offer option B a lot more cash than option A.

The actual numbers are not out yet, but using the aforementioned concept, and noting that DOTs are significantly more powerful than R1/R3s as filler skills (If they aren’t, why would we even cast dots on targets?), the amount of cash(Potency) earned by the “Special Deal” will heavily outweigh the cash you gain purely from choosing Option A over Option B.

Things look especially grim for Bane. But on the positive side, we are able to snapshot Contagion and other raid buffs if the situation allows for it. This may help out Bane to beat Painflare in highly niche situations.

Please look forward to the actual math when everything has been calculated and accounted for!

1.2.2 DWT

With the change to DWT not allowing you to use aetherflow stacks, the amount of optimization possible in DWT is very limited. The 2 main points of optimization here is the use of OGCDs (Tri-D & Shadowflare) to take advantage of the 10% damage increase and the problem of deciding whether you should clip or not to get 6x r3s + deathflare.

Using Tri-d in DWT nets you 2 benefits as both your dots as well as the 20 potency gain from ruination enjoy the 10% boost from DWT.

With the nerf to R3 potency, the potency gained from clipping your gcd to gain an extra R3 has reduced considerably.

Here is some very basic math to estimate the potency gains:

Assuming that 1 oGCD clip = ⅓ of a GCD (~0.75 sec cast time), a clip on average will cost you ⅓ of a ruin 1 / 2 cast ( 100 / 3 = 33.3 potency).

To calculate the gains of using a R3 over a R2 in DWT:

R3 in DWT = 1.1 * 150 = 165 potency

R2 in DWT = 1.1 * 100 = 110 potency

Potency loss from increased mana cost of R2 over R3 in DWT (every 5 R2 cast nets you 1 additional R3 cast) = 50 / 5 = 10 potency

Overall potency gain from using R3 over R2 in DWT = 165 - (110 - 10) = 65 potency

To achieve the 6th R3 in DWT and still use Tri-D in DWT, you have to clip at least 1 OGCD(using swiftcast). I.E R2 (DWT + Tri-D) → R3 x 5 → (Swiftcast) R3 → Deathflare

Overall potency gain from 6x R3 in DWT with clipping = 65 - 33.3 = 31.7

TLDR: You gain 31.7 potency from getting 6x R3 in DWT but it requires the expenditure of your swiftcast + clipping of your gcd.

However, there are further implications of doing this which makes this decision not so straightforward. Firstly, with how tight the dot rotation is ( 3 x Tri-D and 1 set of manually applied dots every 30 seconds), every clip results in your dots falling off for that ~ 0.75 seconds between your re-application of dots. Secondly, Swiftcast is very useful when using Bahamut too to use R4 procs.

1.2.3 Bahamut

For Bahamut, your goal is to trigger as many Wyrmwaves (WW) as possible. Each WW is triggered by an action that is targetable on an enemy. This includes normal GCDs(Ruins), aetherflow oGCDs (Fester, Painflare, Bane) and cross-class disables (Addle). It will not trigger on self-targeted oGCDs (Apoc, Surecast) nor AOE-target oGCDs (Shadowflare).

The optimization for Bahamut lies in having the right amount of oGCDs to weave and following a very specific method of weaving in the oGCDs to achieve the maximum number of WW.

The reason why it is not so straightforward is because WW has a recast time of 1.5 seconds, Akh Morn has it’s own internal delay and that Bahamut tends to move around and become unresponsive when it follows you while you are moving strafing sideways.

I have managed to find methods that consistently provides you with 11/12 WW as long as you fulfil the following conditions:

Have at least 5 targettable oGCDs up (3 x aetherflow stacks, 1x Tri-D & 1x Addle)
Remain stationary as much as possible
Decent Ping (I have heard that people are unable to get 11 off at ~ 190ms)

Low Latency(More risk but higher chance of 12 WWs):

R2 (Summon Bahamut as late into the GCD as possible) → R2 (Fester) → R2 (Akh Morn) → R2 (Swiftcast if you get a R4 proc) → R2 (Fester) → R2 (Tri-D) → R2 (Fester) → R2 (0.5 sec wait + Addle) → R2 (Akh Morn)

No Addle:

R2 (Summon Bahamut as late into the GCD as possible) → R2 (Fester) → R2 (Akh Morn) → R2 ( [Swiftcast if you get a R4 proc) → R2 (Fester) → R2 (0.5 sec wait + Tri-D) → R2 (Fester) → R2 → R2 (Akh Morn)

Disclaimer: Both of the aforementioned rotations are very ping reliant and risky to use as you might lose your 2nd Akh Morn to “Ghosting”. Use at your own peril!

High Latency(Less risk but low chance of 12 WWs):

R2 (Summon Bahamut + Fester) → R2 (Akh Morn) → R2 ( Swiftcast if you get a R4 proc) → R2 (Fester) → R2 (Tri-D) → R2 (Fester) → R2 (Addle) → R2(Akh Morn) → R2

Note 1: There is currently a “Ghosting” issue where Akh Morn or the last WW in Bahamut does no damage to the target if executed below ~2s left on the timer. The animation still appears and ACT registers the damage but it does not appear in the in-game combat log.

Note 2: It is possible to get 11/12 through other variations, just that it may not be as consistent. The main issue is that no WWs can be casted until 2 seconds have passed after an Akh Morn cast. Therefore, any oGCD used on the GCD right after Akh Morn is wasted. In addition, it is highly unlikely to get a non-ghosted WW cast after the 2nd Akh Morn.

I have made some breakthrough on getting 12 WWs. The timing required is incredibly precise as you have to squeeze in 1 extra WW between the interval of your Enkindle cast and the actual Akh Morn cast while ensuring that your Akh Morn does not “Ghost”.

Note 3: The best timings to use R4 procs is through swiftcast. It is not recommended to risk losing 1 WW proc just to expend your 2nd R4 proc to account for the possibility of receiving a 3rd proc (Over-riding your 2nd proc if you didn’t use it).

1.2.4 Rouse

With the new Enkindle CD reducing with each R4 proc, it is increasingly common for Enkindle to be on a pseudo 2min CD. Therefore, along with the general workflow we established earlier of using Rouse by the time that Aetherflow CDs, it makes sense that we should be using Rouse on cooldown ASAP. It is always a plus to have rouse up earlier to reduce the possibility of getting into a situation where your Rouse is at ~20 sec CD but your Enkindle is already up due to R4 procs.

In specific scenarios, you might want to delay rouse by a few seconds to fit in an additional Aerial Slash / Flaming Crush that is just about to have its CD below 20 seconds. Note that the workflow rule still has priority. If you are about to use Bahamut next then your Rouse must be spaced at least 20 seconds away from that moment.

1.3 Further Areas of Optimization

1.3.1 Contagion

This is an area that is up for strong debate. There are 3 different ways to utilize contagion at the moment. They are :

Contagion at the start of the pull, and then Contagion on CD or a slight delay to sync with Rouse/Enkindle

Contagion at the start of the pull, and then Contagion to sync with Bahamut and subsequent aetherflows

Contagion at 10+ seconds (right before Tri-D + Deathflare), and then contagion to sync with Bahamut and subsequent aetherflows

The difference between the 3 options lie in whether you should contagion at the start or 10+ sec into the pull and if you should save contagion for bahamut.

My personal view at the moment (without concrete math/numbers), is that it will be valid to contagion at the start of the pull if your group is running only SMN as a caster. Being able to snapshot the first Tri-D, full duration of Rouse, Enkindle, Shadowflare and possibly 3 or 4 aetherflow stacks will be a DPS gain for the SMN. Also, healers will be able to benefit more from the contagion at the start.

However, if you are running an additional BLM/RDM, with the buildup that these 2 casters require, it will be an overall net DPS gain to contagion at 10+ sec to line it up with their burst.

For the 2nd question, the debate is between the overall potency of 3 windows (I am being very vague here because of it depends on the rotation and how fickle contagion is).

1st Window (Used on CD with Rouse)

6-7 Roused AAs + Enkindle + (maybe even a DWT + Deathflare depending on the rotation used)

2nd Window (Used ~ 5-10 sec before Bahamut)

2-4 Roused AAs + Enkindle + (maybe a few R3s + Deathflare depending on rotation used) + 1 Akh Morn (for every 2 contagions) + ~ 2 Festers + ~5-6 WW

3rd Window (Used immediately before Bahamut/Aetherflow)

~8-9 WWs (for every 2 contagions) + 3 festers + Tri-D + ~1 or 2 Akh Morn (It’s really hard to get 2) (for every 2 contagions

I am currently opting for the 3rd window but the math really isn't concrete enough for me to justify one over another.

1.3.2 Clipping

Compared to 3.0, clipping in 4.0 is now a lot more punishing. The upgrade of potency for R1/R2 to 100 along with the nerf of R3 from 200 to 150 makes it a lot less attractive to clip GCDs just to save MP for R3s. Some useful math numbers to be aware of:

Potency gain from using R3 over R1 : (150 - 100) = 50

Potency gain from using R1 over R2 (5 R2 casts gives you 1 R3 cast): 50 / 5 = 10

Potency loss from clipping oGCD(assuming oGCD is ⅓ of a GCD cast): 100 / 3 = 33

Potency gain from clipping GCD for R3 = 150 - 90 - 33 = 27

Therefore in summary, you can view the potencies as such:

R1 potency = 100

R2 potency = 90

R3 potency = 150

Clip* = -33

Clipping 1 ogcd for R3** = +27

*Note 1: Clipping has an additional effect of messing up your dot uptime. With the tight rotation of 3 Tri-Ds + 1 manual set of DOTs every 2mins, every clip results in your dots falling off for an additional duration of time before reapplication.

**Note: This value is misleading as R3 is not an infinite resource. It is not favorable to clip your oGCDs to swap R2s into R3s unless you are overflowing on MP.

1.3.3 Snapshotting of buffs

Snapshotting of buffs are still very relevant for SMNs despite the loss of raging strikes + extension of dots via contagion. However, outside of the opener, due to the rigidity of the 3 Tri-D + 1 manual set of DOTs rotation, there is little room to adjust for the presence of raid buffs like TA every min.

Ruination, on the other hand, should be optimized as much as possible. The main trick here is to make sure that you use an instant cast skill at the end of it’s duration. (e.g Casting R2 instead of R1 when the ruination debuff has 1-2s left on the target). This nets you a 10 potency gain (20 from ruination - 10 from MP loss).

Syncing Ruination with DWT will also net you a slight potency gain. (0.1 * 20 (potency from ruination) * 6 (no. of ruin casts in ruination buff) ) = 12 potency

1.3.4 Mana Management

With the potency nerf for R3 from 200 to 150, the potency gains from being mana efficient has decreased. However, it is still a significant factor of optimization that will net you a lot of potency gains if managed well.

The introduction of Lucid Dreaming has made managing mana efficiently even tougher than before. The sheer amount of mana that it regenerates over time requires you to be very aggressive with your mana to ensure that you do not overflow. This, along with ranged classes having a MP refresh skill on a cooldown that they can pop without any repercussions, makes it really easy for SMNs to overflow on MP, especially in the early stages of the fight. Always use R3 > R1 whenever you have the opportunity to do in the first 1 - 2 mins of the fight. Beyond that, take into account the cooldown of Lucid Dreaming and any possible fight downtime that might regenerate your mana. The threshold of MP that you require will be dependent on the purpose of your raid. 4.0 SMNs will not require more than 2-3k MP to sustain indefinitely as we are casting far fewer DOTs than before but you should provide some form of allowance for resses to ensure that the raid goes smoothly.

Using R3s when raid buffs such as TA is up is always favourable, but it might be a bit harder to control the timings where you can store and burn your mp due to the aforementioned reasons. Never overflow on your MP just to align R3s with raid buffs.

Being mana efficient also involves minimizing the amount of R2s that you require to weave / move. Every R2 cast is a 10 potency loss so it should be avoided whenever possible.

1.3.5 Weaving of Pet abilities

While it is true that pet abilities can be clunky and unresponsive at times, I would like to highlight that in the majority of cases, the failure of pet abilities going off are often avoidable if certain precautions are taken.

Firstly, we should understand the constraints that we are working with when dealing with pet abilities. In ordinary circumstances where your pet is attacking the boss normally, you are only able to fit in 1 pet ability between each AA. What this means is that you are able to execute only 1 ability(Aerial Slash, Shockwave, Contagion, Enkindle, Devotion, etc.) per AA. You are only able to queue 1 of these abilities at any single point of time. I.e If i want to cast both Contagion and Enkindle, i will have to first queue contagion, wait for the pet AA to be executed, visually wait for Contagion to be used and go on cooldown before attempting to queue the next pet ability.

What this means is that you are not able to queue 2 abilities at once and you should always be conscious of this. A common mistake that people make is to attempt to simultaneously cast contagion and Enkindle at the same time. All pet abilities share the same oGCD queue. Although Enkindle and Devotion are oGCDs that are casted by the players, they are sharing the same queue as pure pet abilities such as Contagion/Aerial Slash.

In summary, treat all pet abilities including Enkindle and Devotion as unique pet oGCDs where you are only able to execute 1 oGCD at a time between AAs. Only begin to queue the next oGCD once you see that the previous oGCD has been casted.

1.4 Closing thoughts

In essence, 4.0 SMN revolves around having a general flow of one job mechanic locking and leading into another job mechanic. This flow is easily disrupted by variable downtimes/ kill times, deaths, missed Aetherflows. However, on a base level, it isn’t too hard to deal decent DPS as long as you roughly follow the flow by making use of the aforementioned system of priorities.

The real challenge and exciting part about playing SMN is the decision making required to alter this flow to optimize for any situation that we encounter.

Unlike other classes, SMN is given the liberty to choose exactly how they want to execute each job mechanic. Do you want to use Aetherflow early or late? Do you want to use your Aetherflow stacks on festers, or Painflares or bane? When do you want to use it? When do you want to execute Deathflare in your DWT? Should I use it early or late? When do you want to use contagion? When do you want to enter Bahamut trance?

There are limitless possibilities available to SMN, of which any one combination could be the optimal way for a particular unique encounter.

That is the beauty of SMN and I hope that everyone who decides to pick up SMN embrace the sheer amount of flexibility and room for optimization available to the class. Think out of the box and let your imagination run wild when trying to optimize for those numbers :P

Here is an example that showcases the rewards of throwing a rigid mindset out of the window and optimizing for a specific fight:

[4.0] FFXIV SMN SSS Dummy 2:21 [4653.09 DPS]

By using a F/PF/Ed opener with 0 R3s in the 1st DWT, I was able to achieve an incredibly early Aetherflow refresh which is mandatory for the 3rd DWT & Bahamut to be squeezed within a 2:21 window. This opener, when used under any other context in general encounters, would be unfathomable as it is vastly inferior to most openers in terms of initial potency in the opener.

Does this mean that SMN is no. 2 on the DPS ranking and overpowered? This is definitely not the case. How overpowered SMN can be depends on how well you, as a player, are able to abuse fight timings to your advantage.

All in all, have fun playing the class and optimizing each encounter. Smn is highly rewarding if you spend the time optimizing and squeezing every last bit out of every fight.

The next section will be a detailed breakdown of the rotation I use for general encounters.

2 General Rotation

Here, I will break down the rotation that I use into small segments to enable us to look at the micro perspectives of the optimization that goes into the core SMN rotation. As always, openers are dead but players aren’t, so please do be flexible and open to adjustments when the situation calls for it due to downtimes/phase changes/mechanics. I will be using Garuda as the main pet as the rotation for Ifrit will be somewhat identical but there are more things to actively look out for Garuda, namely due to contagion. I will not be covering the basic pet skills such as Shockwave/Aerial Slash, but they should be used on cooldown.

The key ideas revolving around this rotation is to incorporate all the information that I have mentioned earlier on in the document. There is a very strict timing requirement for particular areas of the rotation that will be mentioned in the breakdown below.

As a precursor, the rotation will utilize an early AF and enable you to have the maximum number of oGCDs up for bahamut. One of the main weaknesses in the opener is that you are forced to clip your “super-buffed” dots with a weaker one in order to align oGCDs for the later part of the rotation. I have taken this into consideration and determined that it should be an overall DPS gain through the potency gains that you will achieve later into the rotation. Do take note of this when analyzing/critiquing the rotation.

Note: I noticed a mistake with the inputs onto the rotation (additional rouse in 2nd DWT and have amended it.) Please refer to the updated link!

New: Part 1 Part 2 (The rotation was too long to fit into 1 sharable link).

Here is the video POV of the newly improved rotation! : Link

I will be amending the document to account for the changes.

FFLogs Link for Skill timeline (Note: Doesn’t show pet skills which are very important too)

2.1 Phase 1: Transition to 1st DWT

Note: Have Ifrit summoned initially.

Summon Garuda (@10-12 sec Pre-pull) + use Radiant Shield (When there is 2 sec left on cast time) → Diversion (Pre-pull) → Rouse(Pre-Pull) → R3 (Pre-pull + Obey on Garuda + Contagion when the pet starts casting its AA) --> R2 (Potion + Tri-D) → R2 (F + PF/Enkindle) → R2 ( Enkindle/PF + Swiftcast) → R2 (Fester + DWT)

In Phase 1, we are trying to fit in all our oGCDs(Rouse, Aetherflow skills, Pet skills, Potion) and get into our 1st DWT with as little clipping as possible while doing it as quickly as we can. Here are the points to note:

Ifrit’s Radiant Shield: We have Ifrit summoned out initially and hard-cast Garuda at ~10-12 sec before the pull. This enables the party to get ~ 10 seconds of Radiant Shield. This is possible because Radiant Shield shares the same skill slot as Shockwave and not Contagion. Therefore, we are still able to get Contagion off normally in our opener.

Rouse: Noting what we mentioned earlier about our general workflow and lockouts, we are trying our best here to cast Rouse ASAP and get into DWT ASAP to get the Aetherflow on cooldown.

The reason for using Rouse Pre-pull in this case is to enable us to get Rouse on cooldown ASAP. This has implications on your subsequent Rouse usage as the longer you delay your initial rouse, the lesser uptime of Rouse that you will get before entering Bahamut. This timing also fits very well within your GCD rotation subsequently (Rouse before DWT, Dot application,etc).

Clipping: Every additional GCD spent on weaving will have detrimental effects on the later parts of the rotation, especially as Contagion is used on pull. You will want to fit in as many oGCDs and high-power skills (R3, DF) into this contagion window. Note that we are still trying to avoid clipping as much as possible (Only clip in this phase being Pot + Tri) as each clip costs us 33 potency.

Garuda: There is a slight nuance here in controlling Garuda to cast an AA pre-pull as opposed to casting Contagion on pull. The rationale is that you are gaining a free pet AA by “pre-casting” it while queuing Contagion to cast right after. With the queued Contagion, the buffer time created by the use of Pot before casting Tri-D enables you to still snapshot Tri-D within contagion (contagion will be applied slightly before Tri-D does).

Furthermore, by casting contagion 0.75 sec into the fight, we will be able to snapshot higher-hitting skills into the contagion window. This will be clearer when we move on to the next phase.

Swiftcast: We are using a trick with Swiftcast that is only possible in 4.0 as R2 no longer consumes the buff. Here, we use swiftcast 2 GCDs before entering DWT as we have a spare oGCD slot. This enables us to cast our first R3 in DWT instantly to weave in SF + Tri-D. Credit to Aio Auroka from Tonberry for sharing the trick with me.

Painflare/Enkindle: The oGCD slot for Painflare / Enkindle can be interchangeably swapped. If you are having trouble getting Aerial Slash off on time, or if you are using a delayed contagion (Contagion after DWT), it might be better to use Enkindle early on.

If you prefer to align things up by using skills right on cooldown(no thinking involved), using enkindle on the later oGCD helps you to chain sync Aerial Slash + Contagion (2 Pet oGCDs in a row) for more consistent contagion timing as it will align everything perfectly assuming that you used Aerial Slash on CD the moment it’s up.

Enkindle/Devotion: It is possible to swap the oGCD slot of Enkindle(Phase 1) and Devotion(Phase 3), if your raid decides that an early devotion is ideal.

2.2 Phase 2: 1st DWT

R3 (Shadow Flare + Tri-D + Aerial Slash) → R2 (DF + Aetherflow)

In Phase 2, we execute our first DWT to fit in our oGCDs (Tri-D, SF) before ending it ASAP to quickly refresh our Aetherflow. This is the one of the most flexible phases for SMN as you are able to choose the number of filler R3s to cast to change your aetherflow timing. Here are the key points to note:

Aetherflow Timing: Earlier on, we have established how important aetherflow is to how our job functions and in practice, having aetherflow on cooldown as early as possible provides you with a lot more leeway and room for adjustment vmost fights/raids. Very rarely are we able to predict what the kill timing will be outside of a meticulously planned speed run.

Therefore, in most general cases, barring very scripted downtimes (like Lakshmi) or planned speed runs, it is always advisable to have aetherflow on cooldown. In theory, it may seem like a DPS/Potency loss, but in practice, the trade-off of losing a small amount of potency in comparison to a delayed bahamut/missed aetherflow stack is usually favourable to the latter.

Filler R3s: Do note that for my general rotation, 0 R3 fillers are used for the fastest aetherflow timing (good as a baseline) while still keeping the potency loss to a minimum due to buff alignments that are only possible with a fast DWT (Getting your first fester in the next phase (Phase 3) in contagion, and subsequently your 3rd Tri-D will be snapshotted in potion).

Aerial Slash: The Aerial Slash used here has to be precisely timed at this point or used ASAP or you will lose a usage of it prior to Phase 6: Bahamut Trance.

2.3 Phase 3: Aetherflow burn & the transition between 1st DWT and 2nd DWT

R2 (Fester) → R3 → R2 (Devotion + Fester) → R3 → R2 (Fester) → R3

In Phase 3, we want to use all our Aetherflow stacks quickly and enter our 2nd DWT ASAP to use our 3rd Tri-D. Here are the key points to note:

Early Tri-D: This is to provide 1 additional oGCD for bahamut and prevent the need to manually cast dots in bahamut which affects the amount of R4/WW procs you are able to get in Phase 6.

In addition, the early Tri-D will snapshot your potion(Infusion) if you used 0 R3 fillers in Phase 2.

IMPORTANT: The timing of the 3rd Tri-D has to be precisely executed at this point or earlier. If you end up using 1-2 more ruin spells accidentally while weaving in your Festers, it will negatively impact your Bahamut phase.

Festers: Using our Festers ASAP is ideal as raid buffs and potion(infusion) are still up at this point.

Devotion: Devotion is used here as we have free oGCD slots and also because most classes burst are around this timing. It is highly unlikely that we will lose a devotion cast by casting it at this point versus at the start of the pull because we will be in bahamut trance at the point of devotion’s cooldown (90s).

Enkindle/Devotion: It is possible to swap the oGCD slot of Enkindle(Phase 1) and Devotion(Phase 3), if your raid decides that an early devotion is ideal.

R3: R3s should be used whenever possible to burn mana in preparation for the usage of Lucid Dream and also to maximize the usage of raid buffs/infusion in this window.

2.4 Phase 4: 2nd DWT

R2 (DWT + Tri) → R3 x 4 → R3 (Aerial Slash) + R2 (DF + Lucid Dream)

In Phase 4, we are executing our standard DWT rotation without swiftcast. Here are the key points to note:

Tri-D: As mentioned earlier, if you executed Phase 1 - 3 correctly, this Tri-D will be snapshotted inside your potion(Infusion). It is also important to note when this Tri-D is casted as it will affect our Bahamut rotation in Phase 6.

Aerial Slash: Similar to Phase 2, the Aerial Slash used here must be precisely timed or used ASAP if not you will lose a usage of it prior to the Bahamut phase.

2.5 Phase 5: Rouse and Manual Dot reapplication

R3 X 4 → R2 (Rouse) → M3 → B3 → R3 x 3 → R2 (SF + Aerial Slash) → R2 (Aetherflow) → R1(Contagion) → R2 (Summon Bahamut [As late as you possibly can without clipping into your next GCD] )

In Phase 5, we are trying to burn our MP while lucid dream is up while making sure to use Rouse on cooldown ASAP. Here are the key points to note:

Rouse: Using Rouse ASAP ensures that we get the maximum usage of Rouse(20 sec) before entering bahamut.

Dot Reapplication: Your dot reapplication has to be as precise as possible by starting to reapply M3 only when the dot debuff is at 3 seconds or lower. This impacts the execution of your Bahamut rotation later on, and it can be devastating if you were to miss a fester while executing the rotation because your dots fell off before the intended timing as you reapplied them too early in this phase.

Aerial Slash: If you had executed Aerial Slash as intended in Phase 2 & 4, Aerial Slash should be coming off cooldown right before you cast R2 (Aetherflow). Ensure that you cast Aerial Slash before Contagion as long as it does not delay your Contagion timing.

Contagion: Contagion will be used right before Bahamut in this rotation. There are other feasible options of choosing when to use Contagion but we will not be going through it here as it has been covered earlier on in the document. You can find the particular section here.

In addition, always ensure your contagion goes off both visually and by looking at the hotbar before you cast Summon Bahamut. If you realize that your contagion is not going off, delay going into Bahamut till the pet casts the skill.

Summon Bahamut and Delayed Contagion: It is integral that you cast Summon Bahamut as late as possible in your GCD window to maximize the time that you have in Phase 6.

In the worst case scenario where your contagion goes off very late (It should be more consistent because you used aerial slash prior to this), consider clipping your gcd to wait for Contagion to go off and summon Bahamut immediately after. It is very important to minimize the gap between your Contagion debuff applying + Summon Bahamut in order to snapshot our oGCDs inside the Contagion window in Phase 6.

2.6 Phase 6: Bahamut Trance

Low Latency(More risk):

R2 (Fester) → R2 (Akh Morn) → R2 (Swiftcast if you get a R4 proc) → R2 (Fester) → R2 (Tri-D) → R2 (Fester) → R2 (0.5 sec wait + Addle) → R2 (Akh Morn)

No Addle:

R2 (Fester) → R2 (Akh Morn) → R2 ( [Swiftcast if you get a R4 proc) → R2 (0.5 sec wait + Fester) → R2 (0.5 sec wait + Tri-D) → R2 (0.5 sec wait + Fester) → R2 (Akh Morn) → R2

High Latency(Less risk):

Note: Ignore the last R2(Summon Bahamut) from Phase 5 if using this rotation

R2 (Summon Bahamut + Fester) → R2 (Akh Morn) → R2 ( Swiftcast if you get a R4 proc) → R2 (Fester) → R2 (Tri-D) → R2 (Fester) → R2 (Addle) → R2(Akh Morn) → R2

In Phase 6, the strategy here is to get Bahamut to use as many Wrym Waves(WW) as possible.

If you would like to know more about the nuances and optimization for Bahamut, please refer to Section 1.2.3 for more details.

Here are the key points to note:

Low Latency Bahamut Rotation: If you have sufficiently low latency(<120 ms), this is the most consistent method that I have found to get 11 WWs, with an increased chance of getting 12 WWs with all hits registering (No ghosting of Akh Morn/WW).

Disclaimer: This bahamut rotation is very tight and ping-reliant. You run the risk of losing your 2nd Akh Morn to “ghosting (Damage not registering despite ACT/FFlogs showing otherwise)”. In cases of high latency or when you are forced into heavy movement, it would be preferable for you to use the High Latency rotation instead. It has a much lower chance of getting 12 WW procs but is a lot safer to use.

High Latency Bahamut Rotation: The key difference between this and the low latency rotation is that Akh Morn is used 1 GCD earlier. Note that you might have to clip your GCD by a small amount while waiting for the 2nd Akh Morn to come off cooldown. To minimize this, use your first Akh Morn as soon as you can without clipping.

No Addle Bahamut Rotation: If you are using Addle for mitigation purposes, you should still be able to get 11 WWs in this rotation. You run the risk of not being able to fit your 3rd Fester in the contagion window but this is necessary if your raid requires Addle for mitigation.

Contagion Window: If Contagion was applied correctly and Tri-D was used correctly in Phase 4, you will be able to fit in all 3 Festers + Tri-D in Contagion.

R4 Proc: When you get a R4 proc, you will be able to use swiftcast(Replace the 0.5 sec wait) at any point in time to use it while not affecting the rotation. Any subsequent procs runs the risk of losing 1 WW. In general, I would advise against risking a WW proc just for the possibility of losing a R4 proc (Over-riding of procs) during this phase.

Early DoT re-application in Phase 5: If you had an early re-application of your dots in Phase 5 you run a risk of missing the 2nd fester due to your dots dropping earlier than expected. If you had delayed the Tri-D in Phase 4 due to a mistake in Phase 3, your Tri-D won’t be up at the point where you are expected to cast it in the rotation. In both of these cases, consider using addle instead of Fester/Tri-D first to cover that GCD, and use Fester/Tri-D in place of the addle later on.

2.7 Phase 7: Ruin Spam + Rouse Phase

R1 → R1 (Aerial Slash) → R2 → R1 X4 → R2 (Rouse + Enkindle)

In Phase 7, we are basically spamming ruins while waiting for our Rouse to come off cooldown. Here are the key points to note:

Aerial Slash: Similar to Phase 2 & 4, use Aerial Slash on cooldown to ensure that you don’t lose out an Aerial Slash cast before your next Bahamut cast.

R2: Used to snapshot the Ruination debuff and to weave in oGCDs.

R3: We are not using R3s here to conserve MP for R3s inside the next contagion window found in the later phases.

2.8 Phase 8: 3rd DWT

R2 (DWT + Tri-D) → R3 x 4 → R3 (Aerial Slash) → R2 (SF + DF)

In Phase 8, the main idea is to get into DWT and refresh our DoTs after using rouse. Here are the key points to note:

Shadowflare: By fitting in Shadowflare at the end of DWT, we are able to gain the 10% Damage buff from DWT while carrying Shadowflare forward to the contagion window in the subsequent phase.

Clipping: Unless you have swiftcast up, do not clip your gcds in this phase. If you do clip your gcds, you won't be able to execute the last R2( SF+ DF) as your DWT timer will run out before the DF cast.

Enkindle: If it wasn't up in the phase before this, you will have to weave it in this DWT. Remember the earlier point about clipping, if you do not have swiftcast, weave in Enkindle with a R2.
Aerial Slash: Similar to Phase 2, 4 & 7, use Aerial Slash on cooldown to ensure that you don’t lose out an Aerial Slash cast before your next Bahamut cast.

2.9 Phase 9: Aetherflow Burn (Transition between 3rd and 4th DWT)

R2 (Aetherflow + Contagion) → R2 (Fester) → R3 → R2 (Fester) → R3 → R2 (Shockwave + Fester)

In Phase 9, this phase is very similar to Phase 3 where you are expected to burn all your aetherflow stacks on 3x Festers and get to your 4th DWT as soon as you can. The only difference here is that there is a time limit on your execution which is dependent on your DOT timers.. The key points to note here are as follows:

Timing: The timing required here is as strict as the one in Phase 3. You have to use 3x festers before your dots drop off to get DWT for the trid reset. If you are unable to meet this timing(dots fell off) , consider doing a painflare to catch up.

Contagion: Note that if you used enkindle right before the SF + DF in the previous phase, it might delay your contagion timing. The optimal situation is to cast contagion before using your first fester, but if contagion is delayed, please do not delay your festers to the point where you won't be able to get 3x festers before your dots drop off. The threshold here is about 10-12 sec left on your dots before your first fester cast.

R3/R4s: As this is a contagion window, you should try to cast as many R3/R4s as you can while weaving in your festers. Feel free to use Swiftcast if you have it up.

2.10 Repeat Phase: Cycle repeats from Phase 4 - Phase 8

R2 (DWT + Tri-D) → R3 x 5 → R2 (DF + Lucid Dreaming) → Phase 5 to Phase 8

The rotation repeats from this point onwards from Phase 4 - Phase 8. Here is the only point to note:

DWT: If you executed your Phase 8 correctly, your Tri-D will be snapshotted in your contagion just before it falls off on the boss.

3 Videos and POVs

3.1 General Rotation

Please refer to the Section 2: General Rotation for the video and a detailed breakdown of the general rotation.

Here are PoVs of the Primal fights to get you pumping. None of the PoVs are even close to being 90% perfect. But it gives you an idea that there is a potential to get way higher numbers. The pursuit for the perfect rotation / run is never-ending :)

Here is my FFLogs profile if you would like to analyze my runs (Unfortunately it’s very hard to sieve out crazy AST runs (Pug ASTs just tend to give cards to the highest DPS in the party, so bear with me if there are a few odd runs with high balance uptimes) : Eirene Snow's FFLogs

3.1.1 Primals

3.1.1.1 Susano

Note: I was having some issues with pug tanks pulling immediately without countdown timers, so I opted not to use the pre-cast Radiant Shield for the 2 POVs mentionied below.

FFLogs - Susano [4033.1 DPS] - 100th Percentile without cheese at the time of posting!

Youtube - Susano EX SMN POV [4033.1 DPS] [No ST cards]

This was a reasonably good run considering there were no ST cards from the AST. There were some small minor mistakes such as not fitting in Deathflare into Contagion, and also 2 mechanical misplay where I got hit by churn in the 2nd phase when it should be avoided and got hit by levinbolt twice.

A minor optimization tip: By manually casting dots on Susano when it first comes out in phase 2, I am able to use the Tri-D reset from my DWT in the Bahamut phase that follows after.

FFLogs - Susano [3857.3 DPS] [No AST]

Youtube - Susano EX SMN POV [3857.3 DPS] [No AST]

Here is another decent run where the comp is not ideal for SMNs (No AST + SAM/RDM/BRD/SMN). Hope this provides an example that it is possible to hit good numbers without AST / heavy raid utility!

3.1.1.2 Lakshmi

FFLogs [4024 DPS] - 100th Percentile without cheese at the time of posting! (Was beaten by my own parse found below)

Youtube - Lakshmi EX SMN POV [4024 DPS]

A relatively decent run other than the messed up opener because I fat-fingered R2. There are also a few small mistakes such as 2nd Tri-D not being used early to snapshot my potion. And I couldn't get my aetherflow off before the ultimate because we cleared the adds too quickly (Bahamut FTW!).

In terms of optimization, please note that this fight is very un-relenting on aetherflow timings! You have to use it ASAP to safeguard yourself from losing aetherflow stacks. In most cases, if you are doing it right, you will be able to get your aetherflow stack before the ultimate.

Also, the method used here for the adds phase is currently the best way to utilize Bahamut that I have managed to find. If you try to squeeze Bahamut out before the adds phase, you will not have aetherflow up for it. This method of using Bahamut enables you to be tremendously more useful in bursting down low-hp adds.

I also managed to get another freakishly almost perfect run but unfortunately it wasn’t streamed. Sadly, my AST decided to troll me by giving me a ST balance at the last 20-30 seconds of the fight so it isn’t completely objective either. However, it was ~ 3.95k prior to that card. Here is an image of the ACT parser at the end of the fight: Lakshmi EX SMN [4049.64 DPS] on ACT

Edit: After I uploaded onto FFLogs, it appears that there was an abnormally high balance uptime for that run. It’s not that objective but nevertheless, a very clean run. FFLogs [4130 DPS]

I think SMNs could be in a very strong spot DPS-wise when optimized well!

4 Changelog & Contact Details

I will be updating the document with more content as I learn more about the ways of the SMN. There are still loads to learn about the new 4.0 SMN and we will develop as a community once we have more people coming up with new ideas and concrete math on the ideal way to play SMN. Hope you have enjoyed this document as much as I have writing it!

You can contact/support me on my various channels if you have any questions:

Youtube Twitter FFLogs Reddit

V1.0: 30th June 2017

Creation of the original document

V1.1: 13th July 2017

Completed the Rotation section with improvements to the readability of the detailed breakdown of phases; improved the formatting of the guide and included a changelog to enable readers to keep track of updates!

V1.2: 22th July 2017

Amended a minor mistake in the FFXIV Rotation link and re-updated the link.

V1.3: 6th August 2017

Corrected an error in Phase 7 and Phase 8 where there was an extra DWT indicated in the rotation.