[5.4 PTR] SS Nerf and EF

Naturally, Avoiding damage is better than Healing it. You can avoid someone from going below 1 hp, but you cannot heal someone back up from below 1 hp.

However, the size of the Avoidance/Healing matters here. Sure, with massive absorbs/heals, the absorbs will win, since they can save a life. If I could glyph LoH so it would be a shield for half my HP rather than a heal for my full HP, I would!

However, SS is not that lifesaving absorb that saves you from instant death. It's more of an over time effect that keeps your overall damage somewhat lower.

Let's assume SS is between 3 and 4.5 seconds. It will affect at least every 3rd hit, and thus make strings of 3 attacks less dangerous. It might from time to time make strings of 2 attacks less dangerous, but since there will be lots of strings of 2 attacks not covered by SS, we can't count on it there. Ergo, we get at least one SS absorb for every 3 attacks.

I haven't studied into EF yet, but from the comments up here, EF seems to tick twice as often as SS does. That means that if SS ticks between 3 and 4.5 seconds, EF will tick between 1.5 and 2.25 seconds. Thus, it will tick after at least every 2nd hit. It wil therefor make strings of 3 attacks less dangerous, as there will always be a tick between the first and third attack. It might from time to time make strings of 2 attacks less dangerous, but since there will be lots of strings of 2 attacks without a tick in between, we can't count on it there.

Ergo, with SS between 3 and 4.5 seconds and EF between 1.5 and 2.25 seconds, they both apply to strings of 3 attacks. If EF ticks for at least as much as SS absorbs, then it doesn't matter too much that SS is an absorb, and EF is not.

If SS ticks less than every 3 seconds, it will apply to strings of 2 attacks, but EF will tick less than every 1.5 second, so it will tick in between every 2 attacks, so more haste has no effect.

If SS ticks between 4.5 and 6 seconds, it will apply to strings of 4 attacks, while EF ticks between 2.25 and 3 seconds will still tick in between every 3 attacks, so less haste makes it even better for EF.



I'm just talking about SS vs EF here. I'm not too worried about the loss of BoG, as even with 0 BoG, a 3 HoPo WoG heals for a truck on decent vengeance levels. However, without 4 piece T16, I'm not sure I want to drop SotR to keep EF up... It hadn't clicked in my mind yet that 4 piece T16 would remove this opportunity cost.

EF and SS do not scale with haste at the same rate. EF gets its first extra tick at 5% haste, then each additional tick is another 10%. SS's first haste tick is at 10%, and each additional is another 20%.

Theck did the original spreadsheet of that on his blog, and I asked if I could update it for Holy use. He said yes, so here's the spreadsheet simply copied from an excel file on my computer:

https://docs.google.com/spreadsheet/ccc ... CMWc#gid=0

Thels wrote:Ergo, with SS between 3 and 4.5 seconds and EF between 1.5 and 2.25 seconds, they both apply to strings of 3 attacks. If EF ticks for at least as much as SS absorbs, then it doesn't matter too much that SS is an absorb, and EF is not.

If SS ticks less than every 3 seconds, it will apply to strings of 2 attacks, but EF will tick less than every 1.5 second, so it will tick in between every 2 attacks, so more haste has no effect.

If SS ticks between 4.5 and 6 seconds, it will apply to strings of 4 attacks, while EF ticks between 2.25 and 3 seconds will still tick in between every 3 attacks, so less haste makes it even better for EF.

Unless I'm missing something I don't think this logic is fully correct (if we're just talking about boss melee here). Let's say a boss has a 1.5s swing timer, which is most common.
At 3s SS absorbs it covers every 2nd swing, the equivalent EF ticks cover every single hit.
At 4.5s SS it's every 3rd swing with EF covering every 2nd basically.
Seems to me EF indeed has more coverage than SS due to it ticking twice as often, which is one of its perks.

Also fwiw, I'm sure we all know there's barely any bosses where you just sit there and tank melees. In between all the melee hits there is spell damage/bleeds/aoe/DoTs etc peppered all over constantly damaging you too (which for me strengthens the EF argument as is lowers overheal a lot).
So while I do agree there are certain haste breakpoints where it won't help for melee purposes to get more haste in one, or drop some haste etc, in reality the more ticking the better due to the constant damage stream.

Jackinthegreen wrote:EF and SS do not scale with haste at the same rate. EF gets its first extra tick at 5% haste, then each additional tick is another 10%. SS's first haste tick is at 10%, and each additional is another 20%.

Theck did the original spreadsheet of that on his blog, and I asked if I could update it for Holy use. He said yes, so here's the spreadsheet simply copied from an excel file on my computer:

https://docs.google.com/spreadsheet/ccc ... CMWc#gid=0

Good to have, but essentially the relative scaling is the same, which makes sense, since you get 2 ticks for SS absorb, and that's what's happening here. 20% spell haste = 1 absorb or 2 EF ticks, much like how they function regularly. Good spreadsheet to have!

Just to add in on an item that hasn't been discussed: I'm very interested about the change to Sanctified Wrath. Been a big fan of DP, but SW buffs are going to make it a strong contender.

30 Second Duration buff for: 20% increased healing, increased HP generation that lasts an additional 12 seconds longer than HA, and you get to control WHEN you get the extra HP and Healing vs DP. So it will help for Physical and Magical damage. Disadvantage is the 3 min CD. So DP will be default, but might be a few fights to leverage this on. I plan on trying it out.

We also have flexibility with Devotion Aura for a more frequent personal magic CD or the raid CD.

I plan on utilizing EF when I get the 4 pc. It's off the GCD, provides a very strong HOT and you also can't ignore the initial heal. Sure absorbs are great, but having a much more powerful HOT that ticks more frequently is tantalizing. Bastion Stacks won't be an issue with haste

I'm going to miss you sacred shield, and I hate that sloot basically put up all the compelling arguments against it.

I'd like to just highlight his points about our health pools and how they're going to interact with the two talent choices.

We're going to have 1m+ health going into heroic modes here, assuming all appropriate raid buffs and a smart shaman to keep you up there (might not always happen but a fine assumption). Most of the bosses don't hit for more than 350k, and even in H25 I can't imagine bosses hitting for more than 500k. Testing hasn't happened on the later ones, which are generally speaking the more important ones imo thanks to tuning.

Lets say we do get hit for 500k per swing, there's the points already mentioned but if we're looking at the absolute most dangerous cases then we want to be aware of the tick frequency. We already talked about 1-2-3 etc attack strings but I'm not sure this is entirely the direction to go when looking at events that can kill us. our first job as a tank is to not die so looking at times we can die will be pretty valuable.

Since our haste values wont line up exactly with swing timers we're going to have an interesting situation where we get a SS refresh the exact moment before a really big swing comes in, after SS that brings it down to say 350k (starting at 500k) if you include various absorbes from other healers and you feel like rounding. The boss swing timer is significantly shorter than your SS refresh timer. You're next hit wont benefit from SS.

Looking at the other side of the fence with EF and a <2s tick frequency with our 50+% haste the dangerous cases are going to have a swing speed slower than 2s. Aoe situations wont matter just because of the way damage is coming in but single large hits will. Important window for this is swings coming in at 2-4 seconds. Lets say your EF tick is completely wasted right before a big hit comes in, you still get the larger EH buffer from the larger heal value due to EF ticking before the next hit.

No matter the situation the worst case of EF comes out better than the best case from SS. The only situation where the absorb is better is effects that would otherwise kill you. Our health pools are so large that absorbs are becoming largely just a more efficient way of healing by preventing overheal. To reiterate, with our incredibly large health pools I'm more skeptical of the argument of absorb better than shield. There's mechanically no difference when you're usually not in danger.

The two questions or concerns I have with this are the extremely dangerous situations where you are out of range of your healers or they are dead for whatever reason, does a new EF cast with lower vengeance/HoPo/BoG rewrite the hot? This is where I see sacred shield being the strongest and it's not that rare of an event in my experience, especially when getting particularly close to a kill.

Second question is how viable is it to run without the 4p. The loss of BoG stacks is unfortunate for emergency situations which is where I get concerned, but my 100-0 is rarely something that an extra few stacks of BoG would save. I realise it's strictly a HoPo loss but it's not a loss in control of your holy power just the total amount left if you're careful with it baring any prolonged downtime in generation thanks to either no mobs to hit like lei shen intermissions.

Currently on ptr you can overwrite a 3 HoPo 5 BoG EF with a much weaker 1 HoPo EF :/

I'm not totally convinced by the "we have massive hp so EF is stronger" argument, I'm expecting Blizz to tune up how hard the bosses hit and how nasty the tank debuffs are.
With the legendary meta and cloak it just feels like they have to ramp up tank dmg intake if they don't want us to be nigh on invincible.
Just to be clear I'll most likely be using EF on some fights, just don't think the "massive hp" argument holds up that well at least until we see the final raid tuning.

Slootbag wrote:

Thels wrote:Ergo, with SS between 3 and 4.5 seconds and EF between 1.5 and 2.25 seconds, they both apply to strings of 3 attacks. If EF ticks for at least as much as SS absorbs, then it doesn't matter too much that SS is an absorb, and EF is not.

If SS ticks less than every 3 seconds, it will apply to strings of 2 attacks, but EF will tick less than every 1.5 second, so it will tick in between every 2 attacks, so more haste has no effect.

If SS ticks between 4.5 and 6 seconds, it will apply to strings of 4 attacks, while EF ticks between 2.25 and 3 seconds will still tick in between every 3 attacks, so less haste makes it even better for EF.

Unless I'm missing something I don't think this logic is fully correct (if we're just talking about boss melee here). Let's say a boss has a 1.5s swing timer, which is most common.
At 3s SS absorbs it covers every 2nd swing, the equivalent EF ticks cover every single hit.
At 4.5s SS it's every 3rd swing with EF covering every 2nd basically.
Seems to me EF indeed has more coverage than SS due to it ticking twice as often, which is one of its perks.

I was looking where EF could save our lives and SS could not. Say a chain of 3 hits from the boss would kill you. For SS to prevent that, it needs to tick before the first hit, or in between the first and second hit, or in between the second and third hit. For EF to prevent that, it needs to tick in between the first and second hit, or in between the second and third hit. Ticking before the first hit will not help us at all, whereas SS would help us. The advantage of the absorb.

If both SS and EF would tick between every 3 and 4.5 seconds, we could always rely on SS breaking a chain of 3 hits, whereas we could not rely on EF to do the same, as it could tick before the first and then again after the third, which means we ate 3 attacks in a row.

However, since EF ticks more often, that is not a problem, as we can always rely on it ticking between the first and second hit, or between the second and third hit. So EF will always reduce a chain of 3 attacks as well.

Naturally, if they can assist on death prevention for attack strings of the same length, then EF's total output makes it the better choice.

Jackinthegreen wrote:EF and SS do not scale with haste at the same rate. EF gets its first extra tick at 5% haste, then each additional tick is another 10%. SS's first haste tick is at 10%, and each additional is another 20%.

Theck did the original spreadsheet of that on his blog, and I asked if I could update it for Holy use. He said yes, so here's the spreadsheet simply copied from an excel file on my computer:

https://docs.google.com/spreadsheet/ccc ... CMWc#gid=0

You're talking about effectiveness per cast here, which, while interesting, is not something tanks care about. We will always refresh the spell before it has fallen off, so it doesn't really matter if every cast gives one tick more or less. Only the frequency of ticks really matters.

Agamemnan wrote:Just to add in on an item that hasn't been discussed: I'm very interested about the change to Sanctified Wrath. Been a big fan of DP, but SW buffs are going to make it a strong contender.

30 Second Duration buff for: 20% increased healing, increased HP generation that lasts an additional 12 seconds longer than HA, and you get to control WHEN you get the extra HP and Healing vs DP. So it will help for Physical and Magical damage. Disadvantage is the 3 min CD. So DP will be default, but might be a few fights to leverage this on. I plan on trying it out.

Since when was DP the default, really? I think it was a bit of an arms up between DP and HA.

With the buff to SW, SW now provides as much HoPo output per cast as HA. HA has a lower cooldown, so HA still provides a higher HoPo output per minute, but lacks the 20% healing received from SW. Note that you need at least 20% Haste for SW to grant 100% uptime of SotR while active, so SW sucks for Challenge Modes, and probably Proving Grounds too.



Final thing about overhealing. Sure, EF is probably going to cause some overhealing, but that's only when we're at 100% health, so we'll be fine there regardless. SS might not cause us to overheal, but since we're bound to get a couple of hots on us, it would make our healers overheal instead, so that point is pretty moot.

the on thing that bugs me here Dicone is that 500k intake is pretty unrealistic. but it doesn't need that much. Now if I really understood what Slootbag was trying to explain I can pack this into an example.

with this high haste values your SoTR uptime is pretty high (about >60%). Your Tank meta gem proccs for ~65% uptime in 5.4 and gives you a reduce of 20 ON EVERYTHING. your DP CD can be reduced to 25s

let's take Slootbags example form above.

At 250k vengeance (typical 25 man vengeance, give or take a bit with the new Vengeance nerfs but harder hitting bosses), we get a 102k SS absorb every 4 seconds, or we can get a 164k EF tick every 2 seconds, which can also crit

250k vengeance average would mean the Boss melees pre-mitigation with about 694k (as vengeance averages out at about 36%)

let's round up to make it easier and say 700k meleehits pre-mitigation on a 1.5s swingtimer.

now our armor at current gear ToT HM gear reduces about 58% of that.

so even with NOTHING up and turning our back to the boss we get hit for only ~294k

SS: 294k-102k=192k hit
EF: 294k hit. healed up by 164k, being a relative to a 130k hit

although yes the SS absorb is fully used and the next hit won't have this 102k absorb. so the only real difference I see here is tickrate. AND the fact that you really have to take the damage with EF every single time. but EF ticks after every hit where as SS absorbs only every second.

Now from a healer perspective. And let’s say. Considering 5.4 gear you are at 1 Mio HP. This looks like this in 1.5s intervals:

294k = 29.4% of our health.
192k=19.2% of our health

SS:

Code: Select all

------SS----_______-----SS----______----SS----
100% -- 80.8% -- 51.4% -- 32.2% -- 2.8% -- dead

So we drop pretty linear. Leaves us with 4 hits we survive without any heals, CDs or nothing but SS doing it’s Job. Looking at those 294 it is almost a third of our HP pool. So we would survive 3 hits without heals, meaning our healers have 4.5 seconds to heal us. SS extends this by one hit so 6s

EF (heals 16.4%):

Code: Select all

100% -- 70.6% -> 87% -- 57.6% -> 74% -- 44.6% -> 61% -- 31.6% -> 48% -- 18.6 -> 35% -- 5.6% -> 22% -- dead

We get 6hits before we die! This could almost be compared in a doubled EH if you want do be abstract. But the real dead is that it gives our healers a full 9s to react.

With SoTR up this looks a bit different.

With 50% reduction through SoTR. The incoming hit would be only 147k

now:

SS: 147k-102k=45k
EF: 147k hit. healed up by 164k brings us back to full health. yews there is overheal. but we are forgetting something. overheal is only relevant if it costs ressources. overheal from "Free" heals isn't bad!

SoTR is atm at about 60% uptime. letting us reduce almost 2 out of 3 hits.

But applying our model with SotR reduce 2 out of 3 hits. And working with double SoTR giving us every 3rd swing without SoTR.
It looks like this:

147k = 14.7% of our HP
45k = 4.5% of our HP

SS:

Code: Select all

---------SOTR--------______---------SOTR---------_______------SOTR------
-------SS--________-----SS--_________----SS----______----SS----_____
100% -- 95.5% -- 80.8% -- 61.6% -- 46.9% -- 42.4% -- 13% -- 8.5% -- DEAD

Leaving us with 7 hits before we die. Giving our healers 10.5s reaction time. (behold the power of SOTR)

EF:

Code: Select all

---------SOTR-----------------------_________------- --------- SOTR ---------------___________------- --------- SOTR---------------_________________------------
100% -- 85.3% -> 100% -- 85.3% -> 100% -- 70.6% -> 87% -- 72.3% -> 88.7% -- 74% -> 90.4% -- 61% -> 77.4% -- 62.7% -> 79.1% -- 64.4% -> 80.8% -- 51.4% -> 67.8% --
SOTR--
53.1% -> 69.5%    AND SO ON.

Now, in 5.4 we also have about 65% uptime on the metaprocc reducing every damage by 20% also being active for about 2 out of 3 hits. Making EF even better. So I guess the logic is. As 1 EF tick heals more then the Boss swings us for with SOTR up it makes us almost selfsustaining. With 100% SOTR uptime we wouldn’t even need a healer fort hat kind of Boss (considering only melee swings)

If my math here is correct (and I am a Math N00b so It can be pretty much fail from the top to the end) this make EF not even better, but OP!

There also seem to be different breakpoints. The point of Damage a Boss has to do to absorb SS fully with one hit. As well as the point where he hits so hard that one EF tick isn’t stronger than the hit with SoTR up. (considering the Vengeance scaling of both abilities this would make an amazing blogpost I would love to read on sacred duty ^^)

Schroom wrote:

At 250k vengeance (typical 25 man vengeance, give or take a bit with the new Vengeance nerfs but harder hitting bosses), we get a 102k SS absorb every 4 seconds, or we can get a 164k EF tick every 2 seconds, which can also crit

250k vengeance average would mean the Boss melees pre-mitigation with about 694k (as vengeance averages out at about 36%)

My only beef here, as usual with a lot of models, is you can't fully replicate a lot of raiding environments. Of that 250k vengeance gain, rarely is it ever actually from just the boss melee alone. There's obviously a plethora of factors, including magic damage, that contribute to the vengeance gain, making SotR not always a good metric for comparison and analysis.

Regardless, the point is illustrated well!

og course. but with the tools at my disposal (which is a piece of paper, a pen and a calculator) I can't possibly consider all singel possibility. so I have to do it as simple as possible.

I also didn't consider the meta or other cooldowns. special attacks are usualy dealth with accordingly. it was more about getting an impression how powerfull EF really seems to be.

my only question is if this is really correct by numbers? it would make EF OP in my eyes. Nerf is showing on the Horizon.

It's really strong, IF you have the T16 4 piece bonus. Without it, the HoPo cost tunes it down a few notches.

Hey, new here but I wanted to add something to the EF topic, with 250.000 AP it will heal for 172k every tick (with 30% mastery, giving a total increase of 150% from BoG).

Another addition to EF is that when you use AW, if you pop it right before you cast an EF with 5 stacks of BoG it will do 209k per tick, and you can refresh you get a couple DP procs by the end of it with another 5 stack EF, so you have a total of 50 seconds of 209k per tick.

Additionally we have the direct part of EF, which will heal for 523k, averaging at 18k HPS, of course this has potential to overheal most of the time if you want the HoT component to be active unless you know a big hit is going to be coming within 3-4 seconds, since 3 ticks of EF will be the same as the initial heal itself

Ontop of EF there is also SoI healing which I don't have math on and the T16 2-piece bonus, which can be anywhere between 13000-30000k HPS, just looking at some ToT logs and the damage taken from tanks, however this is average damage taken, lining it up for big hit phases would result in more HPS.

I haven't done any testing on PTR raids myself since I don't do the hardcore raiding no more, but if this is true Prot paladins have some seriously insane self healing, not sure how well the other tanks are looking, I know prot warriors are looking to get some pretty good self heals as well.

Oh, and all parts of EH can also crit which can help but it's random and can result in a lot of overhealing since you can't play expecting a crit when you have less than 15% crit chance.

If I use EF. I think I will switch my weapon enchant to windsong as well.

EF benefits from all 3 Proccs.

and windsong is generally considered a DPS gain anyway at about 150k vengeance. (if I remember correctly)

Kiea wrote:Additionally we have the direct part of EF, which will heal for 523k, averaging at 18k HPS, of course this has potential to overheal most of the time if you want the HoT component to be active unless you know a big hit is going to be coming within 3-4 seconds, since 3 ticks of EF will be the same as the initial heal itself

This is actually a benefit of SS, not EF.

The direct part is WoG itself, which SS also has. With EF, we actually cycle it into our rotation, whereas with SS we still have it available to cast when need be.

Schroom wrote:my only question is if this is really correct by numbers? it would make EF OP in my eyes. Nerf is showing on the Horizon.

I really don't think it's that OP, there's just finally a competitive, if not better choice for SS which was always king until now, and that I think gives the illusion that it's OP. It's very strong sure, but I don't see it getting a nerf. SS was nerfed for a reason as was EF buffed.

Kiea wrote:Hey, new here but I wanted to add something to the EF topic, with 250.000 AP it will heal for 172k every tick (with 30% mastery, giving a total increase of 150% from BoG).

Hello! Just a note, BoG has a base 10%, so a mastery of 30% would actually see BoG at 40% per stack going to 200%, not 30%. That obviously increases that tick number even higher!

Schroom wrote:If I use EF. I think I will switch my weapon enchant to windsong as well.

EF benefits from all 3 Proccs.

and windsong is generally considered a DPS gain anyway at about 150k vengeance. (if I remember correctly)

I think the middle ground for DS vs WS was 150-170k. So anything beyond 170k for sure. But if using EF you get further benefit from Windsong so that number is indeed probably lower as a benefit breakpoint.

Just to throw out a few random points for your consideration that I haven't really seen covered:

SS scales with avoidance. EF does not. Avoidance levels aren't all that high, but they exist. Partial-absorbs don't consume the whole bubble, partial overheals are wasted. The magnitude disparity in practice is almost necessarily lower than it appears. I'm taking care not to take any stance on what wins and what loses, I'm just commenting on a few things that might have been glossed over to some degree with the mantra "absorbs are better than heals". It's not just an EH component, it's that utilization of absorbs is traditionally better.

I agree almost entirely with Sloot's line of thinking, though. TMI is a model and a metric - it is not the be all and end all of tank gearing - whether EF or SS wins in "TMI" will not necessarily match up with best practices for any given individual boss. There are damage patterns which given extra weight to throughput (EF) and others which will give extra weight to spike prevention (which is liable to favor SS).

The additional tradeoff is, of course, in not necessarily having a high-BoG WoG available when you need it, if you're running EF. That's a playskill mitigated tradeoff, though. The more aware you are of the situation you're in, the better you can prepare to make sure you always have high-BoG stacks when you might need them. Sloot might not notice it being a problem at all, but if you're not always playing at a very high level, EF over SS does introduce opportunities for unforced errors.

SS is gameable in a way that EF is not. If you happen to know your tick-time (you should) you can cast/refresh SS in such a way as to guarantee an absorb bubble right as boss-ability-of-pain comes off cooldown, ensuring that you have the protection up exactly when needed.

Up to you guys to decide how major/minor any of these are - I haven't played WoW in a couple of years, I may be a little out of touch.

Slootbag wrote:

Schroom wrote:my only question is if this is really correct by numbers? it would make EF OP in my eyes. Nerf is showing on the Horizon.

Kiea wrote:Hey, new here but I wanted to add something to the EF topic, with 250.000 AP it will heal for 172k every tick (with 30% mastery, giving a total increase of 150% from BoG).

Hello! Just a note, BoG has a base 10%, so a mastery of 30% would actually see BoG at 40% per stack going to 200%, not 30%. That obviously increases that tick number even higher!

Yea I know, I meant 30% with initial bonus + mastery, since that is what I have on my paladin atm. I guess most people don't have a lot more than that even at near BiS levels currently, gear in next patch would probably bring it higher.


Also, I can't say much about 25 man, the damage may be higher so it could be useful to have SS there to prevent just falling over, but in 10man you can on most fights keep yourself alive right now just with WoG for a good while, and I've tanked all but Ra-Den & Tortos on my paladin, only fight where I would really use SS over EF would be Ji-Kun and Horridon, maybe Lei Shen aswell since sometimes he just drops your health, but I only really noticed that in P1, and P3 I think EF would be king because you have a dot on you, so having EF negates it more or less.

Lastly, SS would probably be superior on tank switching fights where you stop taking damage, like Lei Shen, Ji Kun, Animus (3 tank). You can put SS on the other tank, where as EF would lose the 100% bonus for being on you, and the however high your mastery is, but atleast 150% in current gear levels.

Some assorted thoughts:

- The "EF is mostly overheal" argument glosses over a number of details. The first being that overhealing is tough to quantify properly. Most people look at a WoL report, see 70% overheal, and conclude that the source was 70% "wasted." That's not usually going to be an accurate assessment though. Something can be overhealing 70% of the time and still be a critical survivability gain.

To illustrate, consider: If you're at full health because you avoid 3 or 4 attacks in a row, EF ticks and SoI procs will register as 100% overheal. Sacred Shield won't be counted as overheal, but it will still be "wasted" in the same sense. During periods where you're not in danger, you'll rack up large excess healing from any source.

However, now consider that you're in the middle of a spike of 3-4 back-to-back unavoided attacks. You'll get a few SoI procs along with maybe two SS absorbs or 4-5 EF ticks. Neither SS absorb will be "overheal" because you're in the midst of the spike, and every point you can get counts. However, you could make the same argument about the SoI procs and EF ticks that occur during that time. Except the EF ticks are 3x-4x bigger than the SS absorbs are. So during that spike, the EF ticks may very well do a better job of saving you.

This should illustrate the mistake most people make when they analyze EF. Their first instinct is to say "well, it does so much overhealing, it must not be worth it." And that's entirely wrong. First, because long-term overhealing is no more relevant than long-term DTPS. And second, because Sacred Shield causes nearly the same type of overhealing, it just neatly hides it amongst the overhealing of your other healers.

What you care about is throughput/effectiveness during a spike period, when you're actually at danger. And there's no question that EF gives you more throughput during that spike period. The question then becomes, "at what cost?"

There are two downsides to EF that have kept us going back to Sacred Shield all expansion. The first is a point that Meloree brought up already: absorbs interact positively with avoidance. Avoiding an attack doesn't consume your SS bubble, but may cause EF ticks to be overhealing. This is a big concern when the heal and absorb being compared are relatively similar in size. It's less of a concern here because each EF tick is going to be almost 2x as big as each SS absorb and those ticks are happening twice as fast. Those two factors together more or less negate any serious "time-shifting" benefit that Sacred Shield's absorb has thanks to avoidance.

The second and more important reason is opportunity cost. Putting up an EF HoT costs holy power, which means sacrificing SotR uptime. So you're getting a HoT for smoothing out spikes, but by doing so you're opening yourself up to more spikes thanks to bigger and more frequent holes in your SotR coverage. I think this is still a pretty big concern, and was sufficient to suppress EF when you were trading a significant amount of SotR coverage for a weak HoT.

But again, a number of factors have lessened the penalty of doing so. The sheer magnitude of the EF HoT has grown to the size that it's a much stronger smoothing effect than it was in the past, making the gap between that HoT and SS absorbs significant. We also have a lot more haste now, and thus higher SotR uptime, so losing a little bit is less punishing.

And of course, the T16 4-piece is the biggest factor, as it removes most of the opportunity cost of using EF. You still temporarily sacrifice your ability to summon up a really big "oh-shit" heal, but that's significantly offset by the fact that you're getting a large, continuous stream of healing for 30 seconds and almost assuredly starting that 30-second period at full health thanks to the base heal. That means you're probably safe for the next 5-10 seconds, and after that time you'll have another BoG stack or two anyway.

Amusingly enough, TMI is probably a very good way to measure these two. Remember that TMI is exponentially weighted, such that the largest events are the most "valuable" and contribute the most to your score. So whichever talent does more to blunt those large spikes will tend to win, regardless of whether it causes massive amounts of overheal during safe periods. It will also take into account the increase in spike frequency that the holy power cost of EF incurs without the 4-piece. And of course, SimC will also accurately report the DPS cost (again, w/o 4T16).

I just got back from a week vacation, and haven't had time to run any of these sims. The semester starts this week, so I'll likely be pretty busy, but if I can find time in the next few days I'll try and test them in SimC just to see what the numbers say. My guess is that against a general melee barrage (i.e. a TMI standard boss), they'll be fairly close without the 4-piece, but EF will solidly win once 4-piece is equipped.

That's actually less of a guess and more of a hope: the goal of this rebalancing was almost certainly to make them both viable choices, and if they come out close w/o 4T16 then I'd call that rebalancing a success.

Pls add another one downside factor: with EF we lose one of the best our o-shit! buttons - pure WoG.

I tested to play with EF yesterday on PTR, doing the first 4 Bosses of SoO in Flex (10m)

I did not use the 4 piece for those testing purposes.

the first downside EF presented was the ramp up time. It takes time to get the first 5 BoG stacks in order to "enable" it.
so we are either without it's benefit for quite some time after pull, OR we use a non BoG enforced EF with our first 3 HoPo. much like "inq" when you play Ret. The downside to this is that you delay your very first SoTR, and a spike at the sart of the firght usually surprises your healers more then a spike mid-fight when every HoT and absorb and whatever is running anyway, from healer rotation.

so if you decide on delaying your first SoTR make sure to run at least DP directly after pull in order to smooth it out a little bit (legendary gem should procc as well on pull giving you a good amount of reduction here)
also this delays the 5th stack of BoG even more.

being scaled down to ilvl 510 (and after reforging accordingly to fit 7.5 hit and 15 exp) I lost a huge amount of haste. Keeping EF up 100% of the time was not possible.

It takes a huge amount more concentration for the EF micromanagement than SS.

unfortunately the Bosses did not hit that hard, in order to test it with big spikes. There was no real "big hit" attack on any of those bosses, so keeping SoTR ready for special attack X was not needed. But that is the Flex scaling coming closer to reality as release is closing in. I'm sure normal and HM will look different.

EF could be better tested on "Sha of pride" as it is a longer fight and tiny bit more damage for us tank. (and was the only Boss we did not down on our first attempt)

On tank-swap encounters like this it was pretty fun keeping up EF on my co-tank and an off heal to a DPS who dropped all of a sudden made him standing pretty safe with EF ticking on him and doing it's Job. (seeing this huge big green numbers, especially crits IS fun in my eyes )

I also told my healers that I'm testing EF and asked for their feedback.
In fact they told me that the rarely had to use single-target heals for us but that we mostly got some splash-heals and stuff like "beacon" and "Lifebloom".

Because of this we even switched down to 2 healers on Sha of pride (one being a Paladin switching to his H-Monk, fist weaving all the time, the other one your was our shaman)


So all in all I tend to agree. Both, SS and EF have their down- and upside making it a real choice depending on group setup / Bossencounter / and Tactics.

the 4 piece will increase the utility of EF. I'm just not sure if the loss in stats compared to offset (warforged)pieces is worth it. I might probably try to get both and switch to 4p for a fight I favor EF.

gnumme wrote:Pls add another one downside factor: with EF we lose one of the best our o-shit! buttons - pure WoG.

it was mentioned here

You still temporarily sacrifice your ability to summon up a really big "oh-shit" heal, but that's significantly offset by the fact that you're getting a large, continuous stream of healing for 30 seconds and almost assuredly starting that 30-second period at full health thanks to the base heal.

gnumme wrote:Pls add another one downside factor: with EF we lose one of the best our o-shit! buttons - pure WoG.

That's not entirely true. We lose the huge 5-BoG-stack emergency heal, but we can still self-EF in an emergency. And we'll build up stacks of BoG fairly quickly at high haste, especially if we take Divine Purpose. So it's really only the short period immediately after the EF that's dangerous, because we don't have the large emergency heal. However, if banking properly, we should have the holy power for SotR/EF.

Some numbers to mull over. I'm using the trunk build of Simcraft (v530-7, SimC build r17386), which has all of the 5.4 changes implemented through PTR build 17116. SimC v530-6 is the latest release, and it's missing some things (along with several bugfixes I made while generating the results in this post). I'm using the action priority list:

Code: Select all

actions=/auto_attack
actions+=/avenging_wrath
actions+=/holy_avenger,if=talent.holy_avenger.enabled
actions+=/divine_protection
actions+=/eternal_flame,if=talent.eternal_flame.enabled&dot.eternal_flame.remains<2&buff.bastion_of_glory.react>3
actions+=/shield_of_the_righteous,if=(holy_power>=5)|(buff.divine_purpose.react)|(incoming_damage_1500ms>=health.max*0.3)
actions+=/hammer_of_the_righteous,if=target.debuff.weakened_blows.down
actions+=/crusader_strike
actions+=/judgment,if=cooldown.crusader_strike.remains>=0.5
actions+=/avengers_shield,if=cooldown.crusader_strike.remains>=0.5
actions+=/sacred_shield,if=talent.sacred_shield.enabled&((target.dot.sacred_shield.remains<5)&(cooldown.crusader_strike.remains>=0.5))
actions+=/hammer_of_wrath,if=cooldown.crusader_strike.remains>=0.5
actions+=/execution_sentence,if=talent.execution_sentence.enabled&cooldown.crusader_strike.remains>=0.5
actions+=/lights_hammer,if=talent.lights_hammer.enabled&cooldown.crusader_strike.remains>=0.5
actions+=/holy_prism,if=talent.holy_prism.enabled&cooldown.crusader_strike.remains>=0.5
actions+=/holy_wrath,if=cooldown.crusader_strike.remains>=0.5
actions+=/consecration,if=(target.debuff.flying.down&!ticking)&(cooldown.crusader_strike.remains>=0.5)
actions+=/sacred_shield,if=talent.sacred_shield.enabled&cooldown.crusader_strike.remains>=0.5


which is just the SimC default with a line appended to maintain Eternal Flame, but only recasts if we have 4+ stacks of BoG. I'm using Slootbag's character as our test subject and pitting him against the T15H25 boss (T15H in 530-6). The only thing I change is which L45 talent he has selected (EF or SS).

5.3 Results, 4T15:
EF: 313602 dps, 106138 dtps, 105735 hps, 339.5 TMI
SS: 315417 dps, 55366 dtps, 99790 hps, 150.3 TMI

5.4 Results, 4T15:
EF: 257571 dps, 107743 dtps, 106385 hps, 3979.9 TMI
SS: 258832 dps, 74490 dtps, 46100 hps, 10950.4 TMI

Note that this is with his current gear, i.e. without 4T16 (but with both T15 set bonuses), so it's a simulation of what will be preferential for the first week or two of progression. The buff to EF and nerf to SS clearly shift the balance in favor of EF by a fairly large margin.

We can artificially disable T15 set bonuses and enable T16 set bonuses using the code:

Code: Select all

tier15_2pc_tank=0
tier15_4pc_tank=0
tier16_2pc_tank=1
tier16_4pc_tank=1


Doing that, we get:

5.4 Results, 2T16:
EF: 252145 dps, 113765 dtps, 111276 hps, 9623.6 TMI
SS: 253640 dps, 80465 dtps, 52372 hps, 16518.3 TMI

5.4 Results, 4T16:
EF: 258854 dps, 106027 dtps, 105690 hps, 4097.9 TMI
SS: 253784 dps, 80441 dtps, 52372 hps, 16475.9 TMI

Note that TMI goes up by disabling the T15 bonuses, moreso for the EF setup since it's getting a significant benefit from the 2-piece (~45% uptime).

It looks like EF is stronger even without the 4-piece, but the 4-piece clearly makes it a lot stronger while having no effect on SS. This action priority list doesn't include any line to simulate emergency WoG usage, so the comparison isn't entirely fair, but it's a start. We could try and include emergency WoG usage with a line like

Code: Select all

actions+=/word_of_glory,if=incoming_damage_4500ms>health.max*0.7

Which would fire off a WoG if you took over 70% of your health in damage in the last 4.5 seconds.

Here are the links to the html output in case anyone wants to pick through them with a fine-toothed comb:
https://sites.google.com/site/theckhd/s ... ef_53.html
https://sites.google.com/site/theckhd/s ... ss_53.html
https://sites.google.com/site/theckhd/s ... ef_54.html
https://sites.google.com/site/theckhd/s ... ss_54.html
https://sites.google.com/site/theckhd/s ... _2T16.html
https://sites.google.com/site/theckhd/s ... _2T16.html
https://sites.google.com/site/theckhd/s ... _4T16.html
https://sites.google.com/site/theckhd/s ... _4T16.html

as always Theck amazing work. so my hunch was correct, that EF would surpass SS even without 4p.

I guess that HEAVY spike you notice in sloot_ef_54_4T16.html at the start of the fight is what I wrote about, that you start of the fight without protection and without BoG stacks so you are basically unprotected and should use a CD like DP to prevent that spike. but after that initial spike you should be amazingly smooth.

it is similar after each taunt if you don't keep EF up during Offtanking (i prefer running it on the other tank) although at least you should start with 5 BoG, so you only have to time it right to have those 3 HoPo when taunting. which is what I meant with "it needs better micromanagement"

Do we actually still need 3 HoPo when we use the T16 bonus? My understanding was that if we have 3+ BoG, WoG (and thus EF) become free.

And of course, awesome work, Theck!