|Ascendance and AG|
|Level 75 Talents|
|T16 Set Bonuses|
G’day, mates! It looks like I’ve forgotten all about this place. I hear that Patch 5.4 is coming, is there anything new for Shaman healers?
Oh, check the patch notes, you say? Well all right then … let me just load up Rygarius’ blog, here, and may as well check in on the PTR Discussion forum … and …
Wow, that’s a lot of Resto Shaman outcry. Like, a lot. People are panicking harder than that little guy over there. ->
But how much of the panic is warranted? That’s what I’m here to help you with, dear reader. That, and digging holes.
So there is a lot of change coming up for Resto Shamans in 5.4, and a lot of controversy surrounding the most prominent of these changes. Healing Tide Totem has gone baseline, and is being given the “Tranq treatment”; a new Talent has taken its place that buffs our Healing Stream Totem; Conductivity has been completely redesigned. Healing Rain and Chain Heal were buffed to improve Resto Shaman throughput. The Glyphs of Riptide and Chaining were buffed to make them less punitive for Resto Shaman needing more spread-healing capacity.
But the biggest change is a technical fix that is being applied to reduce “input lag” – a delay between pressing your button and the ability firing – that has been plaguing 25-player raiders all throughout Throne of Thunder. This lag has been rather convincingly determined to originate from healing events, causing the developers to change the way most of the AoE raid-blanketing heals will work in 5.4. As part of this fix, Healing Rain will now heal only six targets each time it pulses, and will choose the most injured targets within its radius.
This will have a run-on effect on several of our abilities – Earthliving Weapon healing, Ancestral Vigor, Ascendance, and Ancestral Guidance, most notably, but also on our legendary cloak proc and any trinket effects that proc from healing events. These effects are legitimately hard to quantify, and like any set of people faced with uncertainty and change, the Resto Shaman community is uneasy and angry. Accusations are flying of effective 30% nerfs to abilities, of useless Talents, of doom and gloom, of woe and YOLO.
In this article I’m going to go through the most prominent of these changes and break down how they’ll affect our healing in raids – at least, to the best of my ability. As I’ve said already, the new behaviour of our spells in 5.4 can be quite complicated to model, so some of what I’ll be doing is just a rough approximation. As usual, I’ll lay out all my assumptions and try to make it clear where the model fails reality, so you can decide for yourself how concerned you ought to be, on a scale of 1 to Watergate 🙂
Bear with me, this is a long’n. If you need a TL;DR, see the Conclusions section!
In order to address systematically low Resto Shaman throughput in stacked healing situations, both Healing Rain and Chain Heal have been buffed substantially.
First, Chain Heal. This isn’t specifically a stacked healing change, but since Chain Heal is easier to use while stacked than it is while spread out, simply due to the ease of choosing targets you know are in a cluster, it will boost our stack healing quite a bit. Chain Heal no longer drops off in effectiveness with each jump. This is a 58% buff to Chain Heal’s raw healing potential.
Combined with last patch’s update to give Chain Heal the ability to bounce through full-health targets to reach a lower-health target otherwise out of its range, this will result in very powerful Chain Heals. So long as it effectively heals two targets, it’s better than casting a Healing Surge in terms of total healing. Each player gets less healing from the Chain Heal, but it is often better to heal more targets for less than to heal fewer targets for more, and especially in 10-player raids you may well be better off Chain Healing two targets rather than Healing Surging one.
Similarly, this buff to Chain Heal makes Glyph of Chaining far more palatable, especially since the Glyph has been buffed as well, reducing the incurred cooldown from 4 seconds to 2 seconds. If your unglyphed Chain Heal would have healed 2, but your buffed Chain Heal would heal 4, then Glyph of Chain is a mostly neutral change. For extreme-spread, heavy-movement fights like several of those coming up in the next raid tier, this will be valuable in that it allows you to target ranged as well as melee players, and greater flexibility of this very powerful spell will very likely result in increased throughput.
So what’s the problem? As my GM so eloquently put it on Twitter,
The combined change. All shaman smart heals got better. Result: Smart shaman less valuable.
— Moshne July 9, 2013
Now that every bounce of Chain Heal will heal for the same amount, you no longer need to worry about finding the best possible target to chain from. You no longer are rewarded for having had a Riptide up on a target who then took the lion’s share of the damage. You can Chain Heal off of any damaged, Riptide-buffed player and be reasonably assured that your Chain Heal will be almost as good as it could have been had you spent time choosing the right target. And the smaller the disparity between thoughtless and thoughtful healing, the more likely that thoughtless healing will pull ahead, because it is easier, more convenient, and reduces reaction time by eliminating decision-making.
This also has reduced the incentive to use your single-target heals unless there is literally only one player in need of healing. (Of course, use your own judgment here. You may prefer using single-target heals for players in dire circumstances, like Heroic Council of the Elders’ Frostbite target, or Heroic Primordius’ Volatile Pathogen target.)
Now, these are buffs. I know. But they are a degradation of the healing style. The more our budget shifts into smart healing, the less important our decisions become. Using our Tidal Waves intelligently, choosing proper targets, forecasting Riptide, and selecting the appropriate heal for the job – these are traits I’ve always loved about Shaman healing, even if I’ve often fallen into the “spam Chain Heal” mentality that plagued Hyjal and ICC raiders. They’re all less meaningful now, and while I recognise that our throughput will increase, and I do understand that the developers needed to do something to bring us parity and protect our raid spots, I’m just a little disappointed regardless.
Healing Rain itself was subject to a 30% nerf to its base healing and spellpower coefficients, but there’s a new addition to our Purification tooltip that doubles its healing for Restoration-specced Shaman only. As a result, we are seeing a 40% boost in the amount that Healing Rain will heal for in each tick (2*(1-0.3)=1.4). The radius was also increased 2 yards, which really won’t do too terribly much, but may be moderately handy for fights like Iron Juggernaut in SoO where not even the melee want to stack.
But the bigger news is the change in Healing Rain’s function – going from a soft-target-capped blanket AoE to a hard-target-capped smart heal. What is this really going to do for us?
Well, for fights where everyone is stacked up together in our soothing, beautiful blue circle and taking the same damage and receiving the same healing, this is a very, very mild nerf. In these situations, it’s much better to heal more targets for less, than to heal fewer targets for more. However, these situations are incredibly rare to the point where I’m happy to say they don’t exist – particularly because every other healer in your raid is spamming their smart AoE heals as well, and some players have damage reduction cooldowns active, so there’s always going to be some targets who need more healing than others. And since it’s smart, those targets will be prioritised, and your Healing Rain overhealing numbers are likely to go down.
For fights where players are taking highly disparate damage while stacked inside Healing Rain – imagine, for example, a Rampage that only targeted a fraction of your raid with each volley – this is a huge buff. Rather than healing 25 players for (6/25) of a Healing Rain tick, as you would have in 5.3, when only 8 of them need the healing and the other 17 are on full health – thus creating 68% overheal – you’ll now heal 6 of the 8 injured players for a full Healing Rain tick each, yielding 0% overheal. (Yes, I am assuming that the incoming damage is stronger than a Healing Rain tick here. This is definitely true of Megaera’s Rampage.)
For fights where only 6 or fewer targets can stand in your Healing Rain, this produces no change to the healing done.
I will note that this is a nerf to our Earthliving Weapon throughput. Healing Rain ticks have a 6% chance to trigger the Earthliving Weapon HoT on each player it heals. Before 5.4, in a 10-player raid and with 8 ticks of Healing Rain, you’re bestowing an average of 8*10*0.06 = 4.8 players with Earthliving Weapon per cast. Now, it’ll be 8*6*0.06 or 2.9 players with Earthliving Weapon. For a 25-player raid, you’re going from a pre-5.4 Earthliving rate of 12 players per cast down to that same 2.9 players per cast. (In situations where player health is below 35%, though, these numbers look stronger; all players in 10-player raids are likely to have the Earthliving HoT, and 14.4 of 25 in 25-player raids are likely to have it.)
How big of a deal is this? Well, keep in mind that you’ll be casting other spells that can proc Earthliving as well – Riptide and Chain Heal primarily – and they have higher chances to do so than Healing Rain does. So you’re likely to see far more Earthliving on the raid than this simple HR analysis would lead you to believe.
Furthermore, Earthliving Weapon is an extremely small portion of our throughput, and a completely uninteresting and non-interactive one, so losing a bit of throughput here, but gaining 40% throughput or more from Healing Rain, does not exactly concern me. Yes, the chance for Healing Rain to proc Earthliving could be increased to something more like 8-10% in order to reach parity with what we had before, but *shrugs* it’s really not anything to lose your head over.
This Healing Rain change has more profound implications for Ancestral Vigor and our major throughput cooldowns, but these each deserve their own sections.
Now we’re getting into the real meat of this controversy…
In order to fully explain the way that the Healing Rain target cap impacts our Ascendance and Ancestral Guidance cooldowns, I’m going to need to walk you through a brief primer on how the previous versions of this spell have worked.
Healing spells with no hard target cap – usually noted on the tooltip with language like, “effectiveness diminishes for each player beyond 6 within the area” – operated with what we like to call diminishing returns (DR). What this all really means is that, if HR’s tooltip claims it’ll heal for 20,000 each tick, this is only true if there are 6 or fewer players inside the spell’s radius. If there are, say, 10 players inside the radius, Healing Rain would heal them each for 60% of the value of the tick, or 12,000 healing. If all 25 players were inside the radius, then Healing Rain would heal them each for 6*20,000/25 = 4,800 healing. Note that in both cases, each “pulse” of Healing Rain heals for the same total amount – 120,000 healing.
At some point, pets became valid targets for this sort of healing. I honestly can’t remember when this occurred, but I think it was at the start of MoP. The problem here is that, originally, the pets would also count against the target total of the heal. So if there were 25 players and, say, 10 pets in the Healing Rain, each player and pet would be getting healed for 3,429 healing. We still have a per-“pulse” healing total of 120,000. However, 34,290 of that is completely wasted on pets that we do not care about and that pretty much never leave full health.
This was actually kind of nerfing our ground-based AoE healing (and everyone else’s) by a ridiculous extent, so in 5.1, I believe, pets no longer counted against the diminishing return limit of these types of heals. They were still being healed by Healing Rain, getting the benefit (minus Mastery, since as we previously mentioned, they don’t really leave full health in raids) of the same size of heal as players got. Using the same numbers as above, let me illustrate what this did.
25 players and 10 pets in the HR, which heals for 20,000 per tick according to the tooltip. Only 25 of these targets count against its DR limit, so all 25 players and all 10 pets receive healing of 4,800. This boosts the total healed by that “pulse” of HR up to 168,000 instead of 120,000 – or in other words, you were getting 40% additional healing out of your HR just because your raid was lucky enough to have 10 pet classes in it.
Now, again. That extra healing was being dealt to pets that we don’t really care about because they very rarely take damage. In 90% of cases, it was pure, unadulterated overheal, so nobody batted an eye.
However, it was still existent healing. And we have a button or two that turns all healing, even overheals, into redistributed raid healing – it’s called Ascendance (and sometimes, Ancestral Guidance).
Ascendance copies all the healing that we do – overhealing or not – and divides it evenly amongst all nearby targets. If you’re thinking “Hey, do pets not count against that, either?” then you’re starting to understand the problem. Pets do not count against that, at all. So that 168,000 healing that you just did with a Healing Rain pulse turns into 168,000/25 = 6,720 healing per player, instead of 4,800 we’d expect to see. That 6,720 healing gets dealt to each of the 25 players and to the 10 pets (but again, we really don’t care, so I’m not going to bother calculating this at all – it all goes into your considerable overhealing amount, and it doesn’t matter one whit). The point here is that players are receiving 40% more Healing Rain healing from Ascendance than they would have if those pets didn’t exist.
If you’re particularly bright-eyed this morning you’ll realise that the Ascendance boost of 40% is equal to the proportion of pets to players (10/25 = 0.4). So you’ll realise that the more pets there are, the more prominent this additional healing becomes.
Now we reach 5.2, and in 5.2, we fitted out our melee DPS with the Bad Juju trinket, which summons three Voodoo Gnomes each time it procs, and we gave our DPS Death Knights a 2-piece t15 set bonus that summons Fallen Zandalari pets. Not to mention there’s still 8 Ghouls from Army of the Dead, as well as Risen Ally and the enchantingly-named Gary the Gargoyle, Warlocks with Doomguards and regular pets and Wild Imp armies, Feral Spirits (from Enhancement Shaman and Feral Druid Symbiosis) and Fire/Earth Elementals, Mirror Images (from Mages and Balance Druid Symbiosis) and Water Elementals, Hunter pets and Stampede-summoned pets and Dire Beast and Snake Trap, Xuen, Shadowfiend or Mindbender and Void Tendrils or Psyfiend, and well, as you can see, things can get quite out of hand. A top Resto Shaman parse for 25H Megaera had 44 freaking pets active during their Ascendance, which for those of you keeping track, is 176% additional Healing Rain healing from Ascendance that really pushed that cooldown usage over the top.
It is worth noting that Healing Rain isn’t the only source of healing that contributes to Ascendance’s power, but it is definitely a powerful and prominent one. Your regular healing like Chain Heal and existing Riptides are not affected by the pets’ existence, so depending on how many non-DR casts you squeeze into your Ascendance uptime, you will be seeing an overall throughput boost that is much lower than the number I’ve calculated above. However, pets are also eligible targets for the Earthliving HoT, and the Earthliving HoT is copied by Ascendance, so we were also seeing extra throughput from pet Earthlivings. I’ve left this out of my analysis since it’s not as easy to model as the rest, and it’s relatively small potatoes compared to the spurious Healing Rain healing.
Fortunately, the developers have agreed that this bonus, “free”, utterly skillless healing is not what Ascendance is all about, and that it never should have been happening in the first place. I completely agree. Our healing budget is already tied up in cooldown usage. It is not fun to have that also vary wildly based upon the number of pets that happened to be spawned by trinket procs, and it’s also not fun to have to “compete” for ranks with players who can badger their raid into using their pet-summoning cooldowns during Ascendance for extra ridiculosity.
In 5.4, with Healing Rain being hard-target-capped, it simply will not extend additional healing to pets at all. So this extra (pets/players)% healing will disappear completely.
Let me make this clear: This is undoubtedly a nerf to the “potential maximum throughput” of these cooldowns. But that doesn’t matter, because come 5.4, we will not be living in a world where that “potential maximum throughput” is attainable. We will instead be living in a world where our Healing Rain is 40% stronger, and our Chain heal is 58% stronger, and Ascendance has stopped cheating. And those buffs will more than make up for the loss of the pet healing, unlike some naysayers have been worried about.
I can prove this to you with math!
Using a 4pc t16 + all off-tier items Warforged gear profile, I simulated the healing that would be done by our Ascendance if everything was still working under 5.3 rules (e.g. unlimited target HR, pets not counting against DR total, Chain Heal still has fall-off) but in 5.4 gear (so I removed Voodoo Gnomes or Fallen Zandalari, but still included many other pets). I used the maximum pet totals from the top-ranking 10H and 25H Megaera Resto Shamans because I figure, hey, they probably got that top rank because they were lucky with pet procs, or at least, luckier than the average Shaman. For reference, that was 8 non-5.3-specific pets for the 10H Megaera Resto Shaman, and 29 non-5.3-specific pets for the 25H Megaera Resto Shaman.
I then calculated the expected throughput of Ascendance using that same gearset and 5.4 rules (e.g. 40% buffed HR with no pet shenanigans, Chain Heal that doesn’t drop off between targets). In both cases I used an average raid HP of 40% (because why would we pop Ascendance if the raid wasn’t low-ish) to factor Mastery in to the heals that strike players. (Since pets do not leave full HP, they get no effect from Mastery, which means that in practice – and in my model – the pet contribution to Ascendance’s player-healing is lower than the proportion of pets to players.)
I used the same exact rotation for both simulations – pre-existing Riptides, pre-Unleashing and casting HR, then filling the dead space with Chain Heals off Riptided targets until HR needs to be cast again. I am again ignoring ELW because hard and buh. It would undoubtedly make a difference, but I do not believe it would make enough of one. I also did not include the healing that these cooldowns would have done to the pets in the area – sorry to be so repeaty, but it just does not matter at all how much we heal pets when they do not take damage, so I am only looking at the healing we care about, sorry to be so repeaty.
Similar rules apply for Ancestral Guidance, so I also calculated this one, although I’ll spare you the most intricate details, and just leave you with this one simple table:
|Ascendance and Ancestral Guidance in 5.4|
Remember, this is just an example situation, not necessarily a written-in-stone prediction of the exact throughput of your abilities in your raid. This is looking at pretty much a worst-case scenario of how many pets you’re losing access to in 5.4. If your raid tended to have fewer than 8 (in 10-player mode) or 29 (in 25-player mode) pets, then you’ll see a much larger swing from having HR and CH’s healing boosted. If your raid tended to have more than 8 or 29 pets, then you’ll see a much smaller swing, and at some point, it may actually become a nerf, but I am hoping that most players will experience a throughput increase from the 5.4 changes.
So let me just be clear:
The Patch 5.4 changes in no way make Ascendance or Ancestral Guidance “worthless” or “nerfed below live levels”.
You can certainly argue – or be disappointed by – the fact that the buff to these throughput cooldowns isn’t as strong as you’d hoped for from a 40% HR and 58% CH buff. I get that. I really do. But our cooldowns are bananas anyway, even without the cheaty pet bullshit, and the cheaty pet bullshit was a mistake that really needed to be rectified. These cooldowns still perform admirably when compared to other healers’ cooldowns – especially because they are predicated on us casting spells and therefore we’re putting out an equivalent amount of (or in the case of AG, 55.6% as much) healing on our own while the cooldown is active.
These cooldowns are already stronger than their live-server counterparts and will continue to perform incredibly well. Ancestral Guidance is not a “useless Talent”. It will still be strong in the same sorts of situations that it was strong in for tier 15 content – and these situations arise more frequently in SoO than they do in ToT, so we’ll see some real diversity in the Level 75 Talent selection like we’ve been pining for all along.
A much more valid concern is the affect that the Healing Rain change will have on our Ancestral Vigor uptime. This is incredibly difficult to evaluate because the value of the effective health buff is somewhat ephemeral in nature. It does nothing at all for a player who takes no damage, is very valuable for tanks, and is somewhere in between for all the rest. It definitely has a non-zero benefit, and it is definitely, definitely a good buff, and an excellent reason to bring a Resto Shaman to a fight, but nailing down exactly how much it’s worth – and how big a problem it is if it falls off – is much harder.
The first thing I want to note is that the size of the AV buff you do get from HR is going to increase with these changes, assuming a stacked raid. 10-player raiders who are healed by HR will get an AV buff that is more than twice as large as what they would have gotten in 5.3, and 25-player raiders will get an AV buff that is nearly 6 times as large. This largely counterbalances the effects of receiving HR ticks less often – so long as your buff is refreshed every second HR tick in 10s, or every 6th HR tick in 25s, you should be able to stack your AV to full at the same rate as occurred in 5.3.
That’s it for the math … for now. It’ll come back, don’t you worry. But I’d like to delve into some philosophical/theoretical implications here.
First of all, there are loosely two kinds of fights: Fights where everyone can stand in your Healing Rain at least some of the time, and fights where they can’t. For fights where they can’t – where players really must remain spread out to avoid death mechanics – the AV change will be unnoticeable. Your HR will be on the melee where it always has been, and so long as there are fewer than 12 targets in your HR, it’s extremely likely that AV will stack just as fast as it did pre-5.4, and won’t fall off those targets. Similarly, your ranged players will remain at the mercy of your direct heals in order to gain their AV buffs – just like they do now.
For fights where everyone can stand in your Healing Rain at least some of the time, we’ll start to see the effects of the target-cap change. Pre-5.4, each of the 25 players would receive a very small AV buff each time HR ticked, slooooowly stacking up to its maximum effectiveness over the course of around a minute or so. Post -5.4, each “pulse” of HR will only hit 6 people. This does mean there is a chance that certain players may never receive an AV buff, or that their buff will fall off because they do not get healed during the 15-second buff duration.
However, note that HR is a smart heal now. And it is a pretty sizeable chunk of healing, too. If you were the only healer in the pile, you could be pretty sure that the people who got healed by the first tick of the Healing Rain would still be on higher health than everyone else once the second tick occurred, and HR would kind of “round-robin” its way through the raid group. In this situation, each tick would apply a stack of AV to 6 new people, and you’d have the entire raid covered by the 5th HR tick, which is only 10 seconds. If this pattern kept going, you’d be able to keep AV up indefinitely on everyone.
This is unrealistic, because of course there are other healers in your raid. And those healers are going to be using their multi-target smart heals, splash healing, single-target smart heals, and party-limited direct heals, which is going to put this whole “round-robin” idea out of whack. The more smart healing and splash healing you have flying around your raid, the closer to pure randomness the HR targeting becomes. In this case, we can use statistical principles to determine the likelihood that AV will fall off of any player, using the simple formula of (1-(6/raid size))^(# of HR ticks in 15sec). This is calculating the likelihood that a given player won’t get healed each time HR ticks.
At the Haste levels I’ve used to do all my other computational work here, there are 10 ticks of HR in 15 seconds. (2 upfront ticks, because 2 casts, plus 8 ticks between the two casts.) This means that the likelihood of AV falling off in a 10-player stacked raid is 0.01%, and the likelihood of AV falling off of a player in a 25-player raid is 6.43%.
This assumes a pretty long ramp-up time, though, on distributing the buffs throughout the raid. In situations where you stack only briefly, and then spread out again, like the Malkorok fight, this ramp-up time of approximately 10 seconds (really, less, given that the melee are likely to already have AV on them) before the entire raid is affected by AV may be noticeable. And it may well be undesirable. But you are still very likely to get the whole raid covered within that 10 seconds, by at least some AV, and it’s also very likely that the AV amount will be larger than it would have been in 5.3 (10 seconds is 5 ticks of the old AV, but 1 tick of the 5.8-times-larger new AV, so you come out a bit ahead).
Now, the purely-random model is not exactly accurate either, especially in 10-player raids where smart healing is less of a factor and “round-robin” is far more likely to kick in. I don’t predict AV falling off of a 10-player stacked raid very often at all. Even in 25s, players who have not yet been healed by Healing Rain are probably going to be more likely to be healed by future ticks of it, so the purely-random model is probably overestimating the likelihood that AV will drop off.
Remember, there are other spells you can cast to stack Ancestral Vigor, too. Chain Heal, for example, still being smart, is still likely to target people who haven’t yet been healed up by Healing Rain. You could always put an Ancestral Vigor indicator on your raid frames and CH-target or Riptide the players who are in danger of losing it (or who have not yet gained it), if it was something you’re really worried about, but I honestly don’t believe it will be worth changing our behaviour in order to keep 100% uptime on everyone in a stacked raid situation. The game will probably take care of that for us.
I also want to point out that, with more of our healing coming from smart sources, it’s the players who take the least damage or have the most self-healing that are likely to be skipped over by AV. They’re the players who need it the least, so it’s kind of self-correcting. I’m not saying that there is no reason at all to be concerned – this is something we’ll have to watch to see how it plays out, but on Flex raiding tests on the PTR, my AV uptimes for occasional-stack fights look pretty similar to the way they look now on live servers.
There is not a whole lot to say here. Healing Tide Totem was made baseline for all Shaman specs so that it could be given the scaling-with-raid-size treatment that Tranquility, Divine Hymn, and Revival got earlier in the expansion. This couldn’t be done when HTT was a Talent, because then it would definitely pigeon-hole 25-player raiding Resto Shamans into that Talent.
I’m really not sure the scaling-up treatment was a good idea for the DPS specs’ HTT – given that Ancestral Guidance for DPS specs has always been very powerful when well-timed, and that 25-player raiding is already dominated by cooldown choreography that makes each individual healer’s skill less significant, I worry that the additional cooldown will drain a little more depth out of the healing pool, and let’s face it, in 25-player raids we’re already kind of paddling in the kiddie end. However, HTT would have felt like a pretty useless button for them to press without the scaling factor, so I guess I can see it.
25-player raiding Resto Shamans can expect to get around 7.2 million healing from their Healing Tide Totem (assuming the same gear and the same 40% average raid HP as the previous calculations). For 10-player Resto Shaman, you won’t see any change in your expected HTT healing from now, other than, of course, the increase in healing you get from increasing your gear’s item level. Again, this is not as high as some of the other classes’ raid cooldowns in terms of throughput, but remember, you get to keep casting/moving/not dying while it’s up, and they don’t, HA-ha.
With HTT going baseline, the developers had a gap to fill in the Level 75 Talent tier, and they filled it with Rushing Streams, a Talent that buffs our Healing Stream Totem. Rushing Streams increases the healing dealt by each tick of Healing Stream Totem by 15%, and also allows it to heal two targets each pulse instead of just one.
This is a fantastic buff for our spread healing throughput, and with this Talent we may well see HST at the top of our meters on certain heavy-movement/mostly-spread fights like the Iron Juggernaut, Siegecrafter Blackfuse, and Paragons of the Klaxxi encounters. Another excellent benefit of this Talent is that it spreads our Glyph of Healing Stream Totem buff to two people per tick instead of just one, which is an incalculably fantastic buff to be able to hand out. I can’t model how much healing-equivalent this glyph buff will provide, but see the table below to compare Rushing Stream’s throughput with our other options on this tier.
Ancestral Guidance I already covered above at length, so let’s move on, except to note that in order to make this a fairer comparison I have removed Mastery from my AG calculations (since I can’t assume the raid is always on 40% HP for HST or Conductivity).
Speaking of Conductivity, this talent also got reworked. Now, casting any of our targeted, cast-time heals (that’s HS, HW, GHW, and CH) will extend the duration of our active Healing Rain by 4 seconds, up to a maximum duration of 40 seconds. (This last bit is a little misleading – the game doesn’t count the time it takes you between casting the HR and casting your first HR-extending heal, so you can normally get 42 seconds out of each HR; and under temporary Haste effects the duration gets even weirder. But for the sake of simplicity and the fact that I hope they’re going to fix this, let’s just say 40 seconds for now.)
This Talent provides several benefits:
- Without Conductivity, Healing Rain has downtime equal to its cast time between each cast. With Conductivity, Healing Rain can achieve true 100% uptime. In the best-case scenario, this saves 7 HR spellcasts per 40 seconds.
- Instead of every 2nd Healing Rain having no Unleash Life boost, the entire Healing Rain healing pool will have it. This saves a number of UE spellcasts that can be replaced with higher-throughput spells such as Chain Heal.
- We also save mana to the tune of several HR spellcasts and several UE spellcasts, although if we are replacing these gaps in our “rotation” with Chain Heal, then we’ll have to account for that.
One interesting thing about this Talent is that it does not provide as many extra ticks of HR as you might have expected. At the 30% Haste levels that I’m using for my t16 gear set, I am receiving 78 ticks over 120 seconds (assuming perfect overlap between the end of the first HR and the beginning of the 2nd) plus 3 additional ticks from the up-front tick that occurs when you cast the HR. This gives a total of 81 ticks. However, in 120 seconds without Conductivity, you could have cast 10 Healing Rains, and each one would have ticked 7 times + the up-front tick, so you would get 80 ticks of HR this way. Those up-front ticks from normal HR casts are pretty much balanced out by the extra ticks you gain from having 100% uptime on the spell, so it’s almost a wash.
Consider, though, that the difference between half of your HRs being UE-buffed versus all of your HRs being UE-buffed is 15% throughput, so Conductivity wins on those grounds alone. When you factor in approximately 6 extra Chain Heals you can cast in the time freed up by cutting out 7 HR casts and 2 UE casts, Conductivity’s throughput boost actually comes out somewhat competitively with Rushing Streams’ increase…
But Rushing Streams is a lot easier to use to its full potential than Conductivity is. Any time your raid has to move, you’ll have to recast your Healing Rain anyway. And on fights where you simply can’t have a fully effective HR, Conductivity will suffer more than Rushing Streams will. And as you can see in the table below, Ancestral Guidance wins over both of these Talents for burst healing, so I really think that most Resto Shamans will be choosing between the set-it-and-forget-it Rushing Streams Talent and the use-it-or-lose-it Ancestral Guidance.
|Level 75 Talents in Patch 5.4|
|Talent||Healing per 2 min|
There is a significant mana savings with Conductivity, under this best-case scenario, equivalent to the cost of 7 HR and 2 UE minus 6 Chain Heals; this works out to 82.7k mana over the 2-minute period being measured, which is an mp5 of ~3400, equivalent to about 6k Spirit. This could tip the scales toward Conductivity especially for lower-geared players, but remember that the best-case scenario is really, really uncommon, and that you’ll on average receive far lesser benefit from this Talent than I’ve calculated here.
Sorry, Conductivity. I’m just not feeling that spark, you know?
(Remember, the above numbers do not include Mastery. You’re more likely to get a strong Mastery benefit out of Ancestral Guidance than out of Rushing Streams or Conductivity, since AG is a burst-healing, use-it-when-shit-hits-the-fan cooldown, and Rushing Streams and Conductivity are more about sustained healing, so AG will still win the pure throughput contest in the end.)
Our Tier 16 2-piece set bonus is pretty solid:
When Earth Shield heals a target, the target will receive 300% of the amount healed as an absorb.
Earth Shield’s healing is unlikely to be overheal anyway, but fortunately for us, overhealing is included in the shield amount, so that really doesn’t matter.
This healing appears as “Nature’s Barrier” in World of Logs, but gets assigned to the target who has the Earth Shield on them, not to you. In order to view the benefit you’re getting from the set bonus, then, you’ll have to look at your tanks’ Healing Done tab and tally up the Nature’s Barrier healing.
Based on my Flex raid testing logs, this set bonus would be a sizeable throughput increase if it were attributed to me. Earth Shield typically does around 2.3% of my healing, and since the Barrier includes overhealing, I see a slightly more than 3x increase in my Earth Shield throughput. Restandardising as if the Nature’s Barrier healing were attributed to me, it works out to around a 7% throughput increase. This is strong for a set bonus, even if it is boring.
I will note that Earth Shield does not proc until Nature’s Barrier falls off, so in some low-tank-damage situations you may see your Earth Shield proccing less frequently, and may lose a small amount of throughput from that. However, you gain much more than you lose, and also, by re-casting Earth Shield less frequently, you get to maintain your 20% incoming healing buff on that target while spending that extra time casting spells with higher throughput, such as Chain Heal, so it’s a net gain anyway.
The Tier 16 4-piece set bonus is a little more difficult to understand, and has been behaving erratically in previous PTR builds, so a lot of people are discounting it as “not worth” giving up two Warforged items.
So what is it?
Spiritwalker’s Grace will also summon a spiritual version of yourself that will mimic all cast time targeted healing spells for 15 sec.
And what does that mean?
<I won’t rant too much about how I hate this tactic of merging throughput with utility cooldowns. I did that earlier and I’m sure you can find it yourself if you care. :P>
Well, it’s not 100% additional throughput. The Spirit Champion you summon with this set bonus has his or her own version of your eligible spells (again, HW, GHW, HS, CH) that it will cast instantly upon you completing your spellcast. These versions have very low Spellpower coefficients compared to normal heals, and so feel somewhat weak. I spent way too much time deriving these equations, and mind you, I haven’t gotten any confirmation that they’re right, but they’re predicting my Spirit Champion’s healing to within 1% of the in-game observations, so I’m pretty happy with them:
- Chain Heal: 8669 to 9845 plus 3.6% of your Spellpower, per target (max 4)
- Greater Healing Wave: 17461 to 19670 plus 6.9% of your Spellpower
- Healing Surge: 14304 to 16227 plus 5.9% of your Spellpower
- Healing Wave: 9537 to 10815 plus 4.1% of your Spellpower
These low Spellpower coefficients are disheartening, in that the Spirit Champion’s healing will not improve significantly as our gear improves. However, it does mean that waiting for your trinkets to proc before popping your Spiritwalker’s Grace will only net you something like a 100-200 HPS increase. This means don’t bother – seriously. Use your cooldown when you need it, not when you get lucky with procs. (Granted, the two trinkets with Int procs have 115-second ICDs, so you’re pretty likely to have the procs up when you use Spiritwalker’s Grace if you sync them up even once, but again, it’s really not worth losing your autonomy over when to pop this CD for a mere 200 HPS.)
The Spirit Champion’s heals are boosted by Purification – so you can multiply those equations above by an additional 25% – but do not interact with any of our other passive or active Talents; you cannot increase your Champion’s healing with the Reinforce/Empower abilities of your Primal Elementals, nor will your Champion copy Echo of the Elements healing, Ancestral Awakening, Ancestral Guidance, or Ascendance.
The Spirit Champion will inherit our Critical Strike Chance and is affected by our Haste, in so much that the more spells we cast during Spiritwalker’s Grace, the more spells our Champion has to mimic. The Spirit Champion gains no benefit from our Mastery. I’ll also note that the Spirit Champion’s duration is extended by the Glyph of Spiritwalker’s Grace.
All of the Champion’s healing is smart healing, which means that your Champion will not necessarily target the same player you targeted. This makes the Champion’s mimicry of your single-target spells a little more useful – since you often don’t need to heal that target again if you just bombed them with something big – but it does pose some problems for Chain Heal. The Champion puts health deficit ahead of cluster detection when choosing a Chain Heal target, and as such you may not get all of the expected throughput out of the Champion if his or her target selection is poor.
The maximum throughput you can get out of your Spirit Champion in the gear that I’ve used for this assessment – without popping Heroism or Elemental Mastery since I’m stipulating you may save them for a more important cooldown – is 30k HPS during the 20-second uptime, or around 5k HPS sustained over the course of an encounter. This is obtained by still maintaining your Healing Rain and Riptide buffs, and filling the remaining time with Chain Heals, and hoping that your Spirit Champion will hit four targets each time. This burst healing is actually not bad – I’m estimating it’ll be around 8 – 10% of our stacked healing throughput when we’re not using cooldowns, and more like 4 – 5% if we are. More sources of on-demand burst do not hurt.
Note that this is more likely to occur when the raid is stacked and stationary than when the raid is spread out or moving. In other words, your Spirit Champion is less likely to be fully effective when you’re using Spiritwalker’s Grace to help you move or cope with a spread-out fight. This is a seriously counterintuitive result, but what do you expect when the game turns a movement cooldown into a throughput cooldown? *sigh* I will note, though, that your HPS tends to drop quite a bit in fights like this, so the throughput increase of your Spirit Champion relative to your own throughput may still be stronger than it would be during a stacked fight.
In comparison, upgrading to Warforged gear in two slots is approximately a 0.44% throughput increase thanks to 170-ish extra spellpower and 84 extra secondary stats. Even in a sustained-damage fight where there’s no clear opportunity for burst healing, the Spirit Champion is going to outweigh the miniscule improvement from upgrading to 2-piece t16 and 2 Warforged items. Chase this set bonus, because there are a lot of fights in SoO where you may want an additional burst cooldown, and besides, who doesn’t want a spiritual stunt double?
I am super sad, though, that you cannot activate Spiritwalker’s Grace while in Ghost Wolf form. A Spirit Champion puppy would have been amazing. When I am Queen of the World I will make this happen, I promise you all.
I’ll keep this short and sweet. No, really.
- The buffs to Healing Rain and Chain Heal are great for our throughput – we are doing very well in PTR raid tests – but IMO are a detriment to our healing playstyle.
- Don’t worry about Ascendance and Ancestral Guidance. They’re still very powerful. They’ve survived this Healing Reckoning.
- Ancestral Vigor probably won’t come off too much worse than it does on live now, assuming you also heal moderately intelligently.
- Forget Conductivity. Take Rushing Streams for heavy-movement, low-stacking fights, and Ancestral Guidance for fights with stack potential.
- Both the two- and four-piece tier 16 set bonuses are worth getting, no matter how much I might grumble that turning SWG into a throughput cooldown is bad for the class.
I know I’ve said a bunch of things in here that seem to contradict the community outcry. As always, I invite you to post your logs or math if you think I’ve made an error and am being too optimistic, and I’m happy to go through and explain any of the calculations underpinning these analyses in depth if anyone so desires. But for now, it’s back to the burrow with me, because I’m sick and I need some sleep 😦