AST's weakness in end game

[TatoRazzino]

Well, that's an extra 144 mp every 2 mins I suppose, assuming you didn't already spend it on something else, of course. I don't think it really stacks up to assize though

No, not 144. Adding 5s to LA gives you at least 707; 1014 if the refresh ticks are a little early.

[Ghishlain]

TLDR:

• On Living Liquid, AST use 17,692 MP when factoring in MP recovery tools
• On Living Liquid, WHM use 24,395 MP when factoring in MP recovery tools.
• Therefore AST MP efficiency > WHM MP efficiency.
• Please check my math / theorycraft, as it's a lot to write at once and I probably missed something on the way to my end result.
• Updated December 5th @ 5:30PM EST

=====

Assuming we're talking about equally geared WHM and AST at 14,500 MP ea. (both jobs have the same PIE formula) over the course of a Living Liquid Fight (11 minute fight as per FF Logs - some liberty taken):


Shroud and LA
Assume LA always gets two additional ticks via CO.

WHM MP Recovery = 6 * 707*5 = 21,210 MP
AST MP Recovery = 6 * 707*7 = 29,694 MP


Ewer
At 30 potency, I'm going to assume it's about 3/8 as effective as each Shroud / LA tick - which means approximately 265 MP / tick for a total of 1,325 MP. (I don't have the means to test this right now). No CO assumed since you can't always guarantee having Ewer + CO up together.

Assuming 20 draws in a 11 minute and no shuffles / RR of Ewer, you'll average 3 Ewers per fight.

AST MP Recovery = 3 * 1,325 = 3,975 MP


Assize
1,450 MP restored. 8 uses in a 11 minute fight. (More on free Medica savings later)

WHM MP Recover = 8 * 1,450 = 11,600 MP


Lightspeed
Assume you get to use Lightspeed on an Aspected Helios, Benefic II, and 3xBenefic. You get 5 Applications of it.

AST MP Saved = 5 * 0.25*(1,326 + 796 + 3*354) = 3,980 MP


The Fight itself
This is going to be a pickle to theorycraft for and something that should be taken with a grain of salt. The healing requirement of any fight will be dependent on the length of the fight, the mechanics of the fight, and the skill / gear level of the party doing the fight. With that in mind, I'm going with the following assumptions:

1. 11 Minute Fight = 660 seconds
2. 1,579,578 Healing Required - 710,810 healing for single target / 868,768 healing for raid damage (more on this below)
3. 1 Potency = 8.75 Healing (more on this below)

1,579,578 Healing Required
I'm using the White Mage HPS requirement for Living Liquid as per FFLogs HPS Metric. This means 2,389.3 * 660s = 1,579,578 healing required.

White Mage values are specifically used as White Mage should be assuming the bulk of the healing duty in WHM/SCH setup. We're also assuming the goal of this comparison is to determine if AST can fill the void WHM can fill. This means AST needs to be able to do what a WHM can do with the same support from the SCH. As soon as we start deviating from this, that means the AST will be "doing less" and the healing partner will be "doing more".

Raid damage itself is split 45% single target tank damage to 55% raid damage to meet the two values listed above as FF Logs has tank HPS requirement @ 1,047.5. 1,047.5 / 2,39.3 approximately 45% (some liberty taken)

1 Potency = 8.75 Healing
If we assume one Cure / Benefic = 3,500 healing, then 3,500 healing / 400 potency = 8.75 healing / potency.

AST - Raid Damage
868,768 Healing Required

868,768 / 7 = 124,109 healing required per person (not including MT - this value will be added to single target tank healing later).

Cascade does approximately 11K damage per application and there are five Cascades per fight. All five Cascades will be Disable'd, three will be CU'd with those Disables

(2 * 11,000 * 0.1) + (3 * 11,000 * 0.2) = 8,800 reduced via mitigation tools on Raid Buster

124,109 - 8,800 = 115,309 Healing Required per Person

Disable is a 10% reduction for 6s every minute for a [1.1*6 + 1*54] / 60 = 1.0% reduction over the course of the fight.

115,309 / 1.010 = 114,167 Healing Required after Mitigation

CU will occur five times in the fight since you're saving the final three CUs specifically for Cascade - assume 6 ticks per CU in total (waiting 3 seconds after first tick occurs)

5 * 6 * 150 * 8.75 = 39,375 healing via CU

114,167 - 39,375 = 74,792 healing via magic required per person

Synastry provides a 20% boost of healing for 20 seconds on a 90 second cooldown. [1.2*20 + 1*70] / 90 = 4.4% overall increase of healing during the course of the fight.

74,792 / 1.044 = 71,640 healing required for raid healing.

Assuming using Aspected Helios as its the most MP efficient (600 potency / person @ 1,326 MP)

71,640 / (600 * 8.75) = 14 Aspected Helios required

AST MP Spent for Raid Damage = 14 * 1,326 = 18,564 MP

AST - Single Target
710,810 Healing Required

Disable is a 10% reduction for 6s every minute for a [1.1*6 + 1*54] / 60 = 1.0% reduction over the course of the fight.

Bole is a 10% reduction for 30s every 180s for a [1.1*30 + 1*150] / 180 = 1.7% reduction over the course of the fight (assume no RR and one draw every six, like Ewer)

710,810 / 1.010 / 1.017 = 692,008 healing required after mitigation

Assume 14 Essential Dignity's over the course of the fight at 700 Potency each (tank @ 50% HP) = 14 * 700 * 8.75 = 85,750 healing

124,110 healing removed as collateral from tank healing.

692,008 - 85,750 - 124,109 = 482,149 healing required via magic

Synastry provides a 20% boost of healing and 40% increased healing to the tank for 20 seconds on a 90 second cooldown. [(1.2*1.4)*20 + 1*70] / 90 = 15.1% overall increase of healing during the course of the fight.

482,149 / 1.151 = 418,896 healing required for single target

Aspected Benefic
Requires 36.7 applications over the course of the fight to maintain 100% uptime.

36.7 applications = 220 ticks + 37 initial applications = 30,800 + 7,030 = 37,830 potency = 331,012 healing at the cost of 707MP * 37 = 26,159 MP

418,896 - 331,012 = 87,884 healing required to cover the rest. Assume 20% of that damage is spike damage that requires some more powerful healing overall, we'll get the following divide:

87,884 * 0.80 = 70,307 for Benefic = 21 Benefics required (21 * 400 * 8.75 = 73,500) and 21 * 354 = 7,434 MP spent.

87,884 * 0.20 = 17,577 for Benefic II = 4 Benefic IIs required (4 * 650 * 8.75 = 22,750) and 4 * 796 = 3,184 MP spent.

Total MP spent = 18,564 (Raid) + 26,159 (HoTs) + 7,434 (Benefic) + 3,184 (Benefic II) = 55,341 MP for raid healing


WHM - Raid Damage
868,768 Healing Required

868,768 / 7 = 124,110 healing required per person (not including MT - this value will be added to single target tank healing later).

Assuming Eye for an Eye is used to reduce overall AoE damage and assume only one proc per application (10% reduction @ 20 second uptime every 180s) for [1.1*20 + 1*160] / 180 = 1.1% Reduction

124,110 / 1.011 = 122,760 healing required after damage mitigation

Assuming 8 Assize is used specifically for the Medica effect, we reduce overall healing by 8 * 300 * 8.75 = 21,000.

Assuming 4 Asylum is used specifically for the HoT effect to the raid, we reduce overall healing by 4 * 100 * 8 * 8.75 = 28,000.

122,760 - 21,000 - 28,000 = 73,760 healing required via magic

Divine Seal provides a 30% boost of healing magic for 15 seconds on a 60 second cooldown. [1.3*15 + 1*45] / 60 = 7.5% overall increase of healing magic during the course of the fight.

73,760 / 1.075 = 68,614 healing required for raid healing per person.

Assuming we use 7 Medica II (7 * 700 * 8.75 = 42,875 healing) and 6 Cure III (6 * 550 * 8.75 = 28,875 healing) to handle the healing required (total healing = 71,750)

WHM MP Spent for Raid Damage = 7 * 1,503 (Medica II) + 6 * 1,680 (Cure III) = 10,521 + 10,080 = 20,601 MP

WHM - Single Target
710,810 Healing Required

Assuming Eye for an Eye is used to reduce overall AoE damage and assume only one proc per application (10% reduction @ 20 second uptime every 180s) for [1.1*20 + 1*160] / 180 = 1.1% Reduction

Assume Virus is used to help reduce overall single target physical damage on the MT once every two minutes (10 second uptime, 120s cooldown) for [1.1*10 + 1*110] / 120 = 0.8% damage reduction

710,810 / 1.008 / 1.011 = 697,496

Assuming 11 Tetragrammaton used for duration of fight, we reduce overall healing by 11 * 700 * 8.75 = 67,375

Assuming 4 Asylum is used specifically for the HoT effect on the MT, we reduce overall healing by 4 * 100 * 8 * 8.75 = 28,000.

124,110 healing removed as collateral from tank healing.

697,496 - 67,375 - 28,000 - 124,110 = 478,011 healing required via magic

Divine Seal provides a 30% boost of healing for 15 seconds on a 60 second cooldown. [1.3*15 + 1*45] / 60 = 7.5% overall increase of healing during the course of the fight.

478,011 / 1.075 = 444,661 healing required for single target

Regen
Requires 31.4 applications over the course of the fight to maintain 100% uptime.

31.4 applications = 220 ticks = 33,000 potency = 288,750 healing at the cost of 619 MP * 32 = 19,808 MP

444,661 - 288,750 = 155,911 healing magic required to cover the rest. Assume 20% of that damage is spike damage that requires some more powerful healing overall, we'll get the following divide:

155,911 * 0.80 = 124,729 for Cure = 36 Cures required (36 * 400 * 8.75 = 126,000) and 36 * 442 = 15,912 MP spent.

155,911 * 0.20 = 31,182 for Cure II = 6 Cure IIs required (6 * 650 * 8.75 = 34,125). 36 Cures means (36 * 0.15) = 5 Freecure procs over the course of the fight. 884 MP spent.

Total MP spent = 20,601 (Raid) + 19,808 (HoTs) + 15,912 (Cure) + 884 (Cure II) = 57,205 MP for raid healing


Final Calculations:
AST MP Spending = 55,341 MP (Healing) - 29,694 (LA + CO) - 3,975 (Ewer) - 3,980 (Lightspeed) = 17,692 MP

WHM MP Spending = 57,205 MP (Healing) - 21,210 (Shroud) - 11,600 (Assize) = 24,395 MP

Therefore AST MP efficiency > WHM MP efficiency.

=====

Any fact checking on my math / theorycraft is appreciated as it was a lot to think about and write about in one sitting. I feel I've covered most of the bases but can adjust if I hear some reasonable adjustments.

[TatoRazzino]

I'm sorry to say that, but your calculations don't make much sense. You're adding theorycraft to something that's situational and you didn't cover some basic things that AST brings to the table, like 10% AoE damage mitigation by using CU, which will also add a very powerful regen and doesn't cost any MP; also, you're ignoring Disable, which reduces damage by 10% and can stack with Virus/cover the damage when SCH's Virus is on cooldowm; you're also assuming that the AST will keep Aspected Benefic up 100% of the time, which is not necessary due to the initial burst healing/lower cooldowm on Essential Dignity, which reduces a lot of the healing burden on AST. And I didn't even mention the fact that any Bole used on the tank/party can reduce damage taken. It was a nice effort, but the playstyle you describe for AST is not that of an AST, it's that of a WHM. If you behave as WHM when playing AST, you'll always lose.
Even for WHM, which uses Cure III a lot in that fight. I don't think the metric you used is good for any conclusion about what you're trying to analyze. But again, nice effort.

[Ghishlain]

I'm sorry to say that, but your calculations don't make much sense. You're adding theorycraft to something that's situational and you didn't cover some basic things that AST brings to the table, like 10% AoE damage mitigation by using CU, which will also add a very powerful regen and doesn't cost any MP; also, you're ignoring Disable, which reduces damage by 10% and can stack with Virus/cover the damage when SCH's Virus is on cooldowm; you're also assuming that the AST will keep Aspected Benefic up 100% of the time, which is not necessary due to the initial burst healing/lower cooldowm on Essential Dignity, which reduces a lot of the healing burden on AST. And I didn't even mention the fact that any Bole used on the tank/party can reduce damage taken. It was a nice effort, but the playstyle you describe for AST is not that of an AST, it's that of a WHM. If you behave as WHM when playing AST, you'll always lose.
Even for WHM, which uses Cure III a lot in that fight. I don't think the metric you used is good for any conclusion about what you're trying to analyze. But again, nice effort.

Any suggestion on when to use CU and Disable? I could normalize Disable over the course of the fight though you'd obviously want to save it for Cascades if possible. The CU mitigation isn't exactly easy to push into the equation since it disables the AST for a period of time and CU is slightly weaker compared to Asylum (something I also forgot) if you only CU on one tick. I don't think you'd CU more than three ticks per use when I think about it. I can probably normalize Bole (3 draws on MT throughout the course of the fight).

Aspected Benefic is still the most MP efficient spell in the Diurnal's AST toolkit and should be maintained 100% for MP longevity's sake. You'll just need to heal less frequently on occasion because of the added spike it provides. If I don't 100% uptime Aspected Benefic, AST would end up consuming more MP.

Essential Dignity should be included and I'll add that when I get to editting it. Tetragrammaton will balance that but the favour is still towards AST in terms of instant heals.

I intentionally left out Cure III on my analysis though I can probably split it 50/50 just for completions sake tomorrow.

In terms of "behavior", the intent was to see if AST maintains better MP efficiency if it attempts to fulfill the niche WHM does. As soon as you shift your focus away from that, any discussion you have regarding MP efficiency becomes moot because you're pushing some responsibility off from the AST. If you're going to compare MP efficiency, battle conditions need to be exactly the same which unfortunately skews results and on top of that the fight dynamic will change depending on composition. It's also why I'm intentionally ignoring the 3% MND bonus AST may not have if you got AST/SCH composition and that somewhat balances out the lack of mitigation in my math.

Things to note since it was brought up in previous semantics:

-Even though LA + CO gives about as much MP as Assize's free Medica, you can Assize much more frequently than LA (4 Assize for every 3 LAs). which pushes Assize to be more efficient.
-Freecure should still push MP efficiency towards WHM favour. Even though AST spells cost less, AST needs to expend more MP for Benefic II to stay on equal potency footing as WHM.

I'll get to editting sometime tomorrow afternoon and see where the numbers stand then. I still feel WHM has better MP longevity though.

[EDIT] I'll probably end up pushing CU as an AoE heal and Aslyum as 50% AoE 50% Tank heal and use Tetra at 750 Potency (healing tank at 50% HP) when I get to the edit. Bole's will be normalized over the course of the fight (10% for 30s over the course of 3 minutes) and I'll probably add Virus to WHM's mitigation package for physical single target to contrast Disable's AoE damage reduction on Cascade (and normalize AoE damage to suit).

Don't discount so easily - even if we can't build it to be exact due to variables outside of our control, we can still build what I feel is a fair facsimile just to see the end result.

[TatoRazzino]

You can save it for Cascades, which reduces AoE damage for the group, which is what we require the most. ASTs can let CU tick for more than 3s on phase 2 when the party is stacked (and WHM usually uses Cure III there). You also have to consider Time Dilation affecting Aspected Benefic ticks, which will reduce the amount of casts necessary to keep it up 100% of the time. Another thing: it's smart to line up Lightspeet with LA to get 2 more GCDs during Lightspeed; you can normalize that in your equation as well. Synastry also allows you to heal two targets at the same time, which means that some of the individual healing you added to the tank numbers is not going to exist; I don't know how to calculate that, though.

A calculation mistake on the WHM part: you considered 600 potency for Medica II (probably from copying Aspected Helios' numbers), but Medica II is 700 potency, which will make WHM more efficient.

[Mutemutt]

No, not 144. Adding 5s to LA gives you at least 707; 1014 if the refresh ticks are a little early.

Woops, was looking at the wrong number.

But really, what I wanted to say is that, even then, there's a pretty noticeable gap in mp because the meta holds more value for the dps boosting cards rather than any that would help save or restore mp. Assize, on the other hand, always gives back mp. That one or 2 extra ticks just doesn't make up the gap. At least, that's what I feel is the reason.

[AzureFlare]

It's never a waste. NEVER. It could become a problem if you are running low and using every Ewer with Royal road
Of course, it could be better on another things but it will not be a waste.

Again, as I said before, it'll be inferior unsyncing CO and LA just for a 5 second Balance extension if you can get another round of DoTs out which you couldn't before. If you don't need the mana or have the globals to spend, sure, unsyncing will be in your benefit. Including normalised missrates and using the mana on just Malefic/Malefic 2, you'd need like 7425 raidDPS for it to be more mana-efficient than waiting for your first LA cast. In a burst session, sure, that's entirely possible, but I doubt that window lasts longer than 30 seconds which your regular expanded Balance already covers.

If you can fill all your globals without help of a BRD/MCH then of course it goes way deeper than that and using blanket LA will be fine if not better, but in fights like A2S or higher floors where bards/mchs lose DPS on giving mana thanks to large amounts of uptime, I have a hard time seeing that happen as opposed to mechanics-heavy, mana-slumberfest Thordan.

Theorycrafting

It is possible to fit 5 Lightspeed with Diurnal's boost, without it you'll need some spellspeed. Given, that does include Diurnal's speedboost. CU and Disable were mentioned, though I'd only count for Disable on critical moments as normal Virus would be close near a non-burst/buster Disable. Also didn't account for E4E on WHM to balance on that account, so CU would be the most noteworthy.
The 2 ticks from LA + CO are pretty much guaranteed as far as I know. It feels like the ticks are slightly faster than 3 seconds anyway, fast enough to guarantee 7 ticks in 20 seconds even if the first tick comes very late. (Edit: Just checked it. If the natural MP regen tick is 3 seconds, then LA definitely is faster. The first LA tick came after it and near the end, they synced up).
Honestly, we could discuss more points but I feel it becomes too contextual once we go deeper. Nicely done.

[Misutoraru]

i m fine with AST
but could we have less trolling RNG for card drawn please?
4 TP in a roll with shuffle is a nice troll when i m only fighting a 60 dungeon with all my teammate have their own ability to regen TP
or 4 Ewer in a roll, when other 3 member dont use MP
can we at least change shuffle wont draw the same card again?

[Alisane]

I don't think they have fewer tools to heal with, all 3 healers share the basic spells needed in this game to heal well.

I should have been more specific here. What I really meant was, because of the card abilities, you have less space for "extras". No Virus, no Eye for an Eye, no Stoneskin. WHM has an instant group heal, as does SCH, without having to use a cooldown. WHM gets two large instant heals, SCH gets however many as stacks they have, whereas AST gets one. Does this make or break healing? Definitely not. It is just one of the things that makes AST a little different than the others, which people newer to healing may struggle with. It also allows less room for big errors in the group because you can't blow three stacks of Lustrate on the tank, etc.

This doesn't mean that AST is bad, needs revision, etc. Classes are different and not everyone will excel with AST.

[TatoRazzino]

@AzureFlare @Ghishlain

What I think this math shows is that the gap between WHM and AST is minimum in the course of a fight. The extra MP AST spent can be recovered in 15 seconds and is diluted throughout the fight. That, of course, ignoring the fact that the described healing strategy is not ideal for AST (and I'm not even considering the co-healer) and also ignoring other variables that are hard to calculate (mitigation from CU and split single target healing for Synastry, for example). And also: WHM has a natural advantage in high AoE damage fights. The fact that the difference is low shows that the two jobs are pretty much balanced MP-wise.