Life in Group 5 – A Resto Shaman Blog
A resto shaman perspective on raiding


Healing

June 26, 2013

Status Check: The Healing Game as of 5.3

More articles by »
Written by: Vixsin
Tags: , , , , ,
ToT_Lei-Shen-2

These past few months of progression, combined with a work schedule that refuses to ease up, have left me with a billion ideas about the game and a folder full of half-written, conceptualized, or hastily brain-dumped pseudo-posts that I haven’t had time to see to fruition. And amid my musings about the decline of 25-man raiding, why proc-based trinkets stink, and the rise of alts over offspecs, is one topic that I want to talk about more than anything else—the state of the healing game as of Tier 15 and Patch 5.3.

I want to talk about it not because I’m about to launch into a death-and-doom post about the state of the healing game, or about Resto Shaman (though I’d have no lack of material there), but rather because I think in these past two tiers, we’ve been watching healing themes evolve and morph from very class-specific issues into ones that are affecting healers (and healer balance) across the board. There are 4 specifically that I want to focus on: the contribution of cooldowns, the effects of burst damage, the efficacy of absorbs in mitigating raid effects, and the explosion of healer mana. Because, I think we oftentimes get so caught up in the numbers and with isolating and dissecting issues that we forget how interconnected it all really is.

 

Setting the Stage

When 5.3 was launched a mere 11 weeks after raid teams gained entrance to Throne of Thunder, there were 3 seemingly minor changes nested in those Escalation patch notes:

  • Tranquility now targets 12 raid members (up from 5) each time it heals when used in a 25-player instance.
  • Revival now has a healing cap of 15 raid members (up from 6) when used in a 25-player instance.
  • Divine Hymn now targets 12 raid members (up from 5) each time it heals when used in a 25-player instance.

Forums and guild discussions were rife with speculation about the size of these impacts on throughput, and when logs started pouring in, we saw just how substantial they were. While Monks saw a nice, albeit small, boost to their throughput, Druids skyrocketed up in 25s (arguably due to the changes to Tranq *and* mushrooms) and Holy Priests, who maintained a comfortable position before, widened the gap between first and second even further. It was a 1-2 punch, which when combined with the nerfs to Illuminated Healing and Atonement, completely reshuffled the healing ranks.

 

Contribution of Cooldowns

But there’s more to it than that, and if you focus strictly on meters and who’s on top versus who’s on bottom (which leads to the typical lamentations about class balance), you miss the larger picture here. And that is … in the cases above and including Shamans’ Healing Tide Totem and Disc’s Spirit Shell, these massive HPS cooldowns are now wielding more influence than they ever have before. This is important because it represents a definitive change in the healing game, the strains of which we saw with Spirit Shell in Tier 14, but which I think traces even further back, to Firelands in fact.

Recall, at that time, that it was Resto Druids who were on top of charts and defending the position that they “won” healing because their sole raid cooldown (Tranquility) showed up on meters, whereas raid utility tools like Barrier, Aura Mastery, and Spirit Link’s damage reduction component, did not. (Blue Source) At that point, Tranquility was in a class by itself as the only powerhouse AOE throughput CD. (Divine Hymn was comparatively weak, which is why it saw a buff via Heavenly Voice in 4.3). But it was the usefulness of that powerful, raid-saving Tranq that, I think, laid the groundwork for the distribution of throughput cooldowns to each healing class. Which was a good thing in theory (I sure as heck argued for it), but which we’re now seeing the negative effects of, post-5.3.

Firstly, we have a problem with proliferation—the sheer number of throughput cooldowns that are available as you increase in group size, not only from mainspec healers, but from hybrids as well, is boggling. It’s not something that you’d really notice in a 5-man dungeon or smaller team, where you’re thankful for the ability to save a dying group. Even in a 10-man environment, you might still be thankful for that parity between healing classes. But when you get into a 25-man environment, where every healer (plus your Enh and Ele Shaman, Spriests, and Druids) has a mega-cooldown at their fingertips, you have the option of playing every fight like Ra-Den Phase 2, where you simply chain cooldowns for 2.5 minutes of the encounter.

Secondly, we have a problem with what I can only term power creep, which is not only a result of the increase of healer throughput over time but also the escalation of incoming damage on raids. Power creep is something that we saw in Icecrown raids, and is something we’re seeing the effects of with Patch 5.3’s buffs to AOE CDs—it is the upward spiral of needing to make mechanics that challenge and strain healers mana, but also needing healers to be well equipped, in either raid size, with or without absorbs, to handle that damage. So you buff healers, then buff damage, then buff healers … up and up and up.

Which leads to the third issue at hand: valuation. As I mentioned previously, these healing cooldowns have moved from contributing in the neighborhood of ~10% total effective healing in Tier 14, to the point in Tier 15 where they’re sometimes contributing upwards of 20% of a healer’s total effective healing. But when I talk about valuation of a spell, I’m NOT talking about its measured output on WoL.

What I am talking about healers’ ability to affect change on the outcome of an encounter. Healing is a reactive game, a way to offset the damage mechanics of a fight and account for simple human error, and so “skill” as a healer is evaluated on the ability to make decisions that have a positive effect. But when the effect of a spell, or a single choice, is so significant that it eclipses the healer’s other spells and decisions, not only does it narrow their focus onto appropriate use of that one spell (as I mentioned above), but it erodes at one of the most basic and oft reinforced tenets of healing—your decisions matter.**

But in all this talk about dominant cooldowns, it’s important remember that the power of a cooldown and the healing it does are characteristics dependent on one thing: the amount of damage your raid is taking.

 

Burst, the Anti-Triage

Perhaps what I find most interesting about the power of mega-cooldowns is that they oftentimes have a positive correlation to the amount of burst damage in an encounter (especially if that burst is spaced out at regular intervals). Whether it’s Lighting Storm, Dire Call, Discharge, Quake Stomp, Rampage, Quills, Ventral Sacks, Interrupting Jolt, Fist Smash, Nuclear Inferno, Thunderstruck, or Unleashed Vita—burst damage was front and center in Throne of Thunder. (Durumu is, in fact, the only ToT encounter without a regular raid-wide burst damage mechanic).

And, with the rise of burst frequently comes the death of triage. (I say frequently because you can have both burst and triage in an encounter, just look at T11 bosses—it simply necessitates that the burst be less intense). Regrettably, Throne of Thunder was no exception to this rule. Triage’s death knell was sounded by ever-rising Spirit values on gear, and the final nail in the coffin was put in place by the incredibly powerful healer legendary meta gem, which provides a game-changing amount of mp5. In addition to raid-wide burst damage, we also saw the resurgence of heavy tank burst, which was front and center on fights like Primordious, Dark Animus, Ji Kun, and Ra-Den, where tank deaths were, and are, still are possible in a matter of a couple seconds.

Unfortunately, not all classes are equipped to handle burst damage, which is partly why pre-5.3, both Resto Druids and Resto Shaman struggled to remain competitive with other healers. Druids have long been masters of raw throughput, the constraint on which historically has been that their hots had fairly long durations, making overhealing the inevitable result of a shorter effective healing window. Likewise, Shamans’ long cast times, mastery, and spatial constraints make them reliant on a more protracted healing window in order to achieve comparable throughput. Thus, in a burst environment, these classes are at a disadvantage simple based on class design, and in a burst environment comprised of an entire tier, they become the low men on the totem pole for every single fight. (Which is what we saw until the mushroom changes in 5.3).

Relating this back to something I mentioned above–—the heavier the burst damage, the more throughput healers need to be able to do in a short window of time. One way to control that window is to make sure you have sufficient CDs to cover the unmitigated amount (which is what the preceding section talked about), while the other way to control that window is to reduce the amount of reactive healing that your team is responsible for. Which leads me to my next talking point …

(See? I told you this all connects!)

 

Efficacy of Absorbs in Preventing Damage

When you think about it, it shouldn’t really be much of a surprise that when damage comes in brief, predictable waves, the best way to counteract it is through the use of proactive mitigation. It was, in fact, the reason that Discs enjoyed Sunwell-esque dominance in Tier 10, and to a lesser extent in Tier 14. The scaling back of SS with 5.2 was, I think, successful in shifting the spell distributions of Discs slightly away from SS/PoH spam, ultimately rewarding greater contributions from PW: S and Atonement healing. But it was the combination of effects, from Pallies’ Illuminated Healing and Disc’s Divine Aegis, PW:S and Spirit Shell, that ruled the time between 5.2 and 5.3.

I feel inclined to mention here that … [dramatic pause for emphasis] … there is nothing wrong with Absorbs being a strong and compelling tool in raid content. I may gripe about how they snipe my heals and I may secretly plot to disrupt our absorb classes’ internet connections, but deep down, I appreciate the breathing room that they give our team when we’re working on progression content. Interestingly enough, as Ghostcrawler recently mentioned, absorbs have taken over that very particular and valuable niche in the healing world—smoothing incoming damage so that an entire raid isn’t spiking up and down from 100% to 20% HP every time your raid takes a hit.

But, part of the problem that came to a head in T15 had to do not with Discs’ mitigation (which is seemingly making its way back to the norms of Wrath), but rather with Paladins’ mitigation, and the fact that we don’t see the same sort of pros and cons with Illuminated Healing as we do with Disc Priests. Whereas Disc has front-loaded mitigation, with little in the way of pure healing, Paladins’ absorbs are actually based on their raw throughput, gaining power not only from intellect and crit, but also from mastery, making their toolkit strongest in those situations where absorbs are typically weak, namely, consistent incoming AOE. And while the 5.3 nerfs to Illuminated Healing did serve to reduce the size of the bubbles being applied by raid healing pallies, it still doesn’t alleviate the fact that the class is straddling the line between absorbs and throughput, without suffering the downsides typically associated with either.

But the other reason that absorbs were allowed to dominate in Throne leads us further down the breadcrumb trail … mana, for most classes, ceased to be a controlling factor.

 

Healer Mana

Tier 9 holds many memories for me, but none so significant as the day when I received my first Solace of the Fallen. It was only trumped by a day a month or so later, when my set was complete, and a second Solace of the Fallen was safely equipped in Slot 14. If you were a healer during this time, then you know about the heavens-parting-angels-singing effect that double Solace had on your mana pool. As a shaman, double Solace was the key not only to Haste-stacking, but also to my success in the burst healing world of ICC, where tanks and players could die in a couple GCDs. (Hey, there’s burst cropping up again). Double Solace assured that I would be able to spam the ever-loving-beejeebus out of my Chain Heal key and never go OOM, something that I hadn’t been able to do with wild abandon before.

If you recall, it was as a result of this boom time for healer mana and, accordingly, the incredible burst damage that was present in ICC encounters, that the concept of triage was introduced in Cata, in an effort to make healing decisions more meaningful and make mana a bona fide point of failure. And despite the fact that many of us were dragged kicking and screaming into that world, I think the implementation of triage was largely a success and healers, as a whole, started caring a lot more about their mana. There was something that felt smart about leaving a player at 80% HP and knowing that he’d be okay. But, even in Cata, triage was doomed to die at the hand stat inflation, so at the start of MoP healers saw yet another thematic change aimed at that same lofty goal of getting us to care about our sole resource—the capping of healer mana pools at 300k and the disassociation of Intellect from the regen formula.

Unfortunately, even with these constraints, overhealing has steadily been on the rise since T14. Whereas in the first tier of Cataclysm, I sometimes ended progression encounters with ~20% overhealing (no, I’m not even talking about Chimeron here), in the first tier of Mists my overhealing on progression was in the neighborhood of 30-45%. And when our team stepped into Throne of Thunder, that overhealing percent inched even higher, in the 40-60% range for any given encounter. Part of the problem is the proliferation of splash healing in dps and healing toolkits, but part of that is because with mana restrictions removed via gear and introduction of the healer legendary meta gem, healers are free to address challenges through pure brute force … otherwise known as spam healing … and their teammates are free to look at death logs and once again (as they did back in ICC) why 2 seconds elapsed with no direct healing on the person in question.

 

Déjà vu?

It wasn’t until I began writing this post and connecting the dots that I realized how similar this situation is to one we were in just several years ago. Sufficiently gorged on mana, gearing for secondary or even primary throughput stats, dealing with damage that comes in increasingly intense amounts, significantly aware of the strength of absorbs, and having the power of pure spam to counteract the mechanics we faced. We stood on the steps of Icecrown, facing a healing game governed more by stats and less by good decisions. (Oddly, it does seem like we’re hitting this milestone a little earlier than we did back then—an indication of some escalation, perhaps?)

What emerged from that previous experience was a new buzzword for a basic concept—triage. And we’ve been using that as the model of “good healing” ever since, (which, I think, should be considered a huge win on the designers’ part, because it’s not every day you sell the majority of the healing community on a concept of “how to heal properly”). Every time the healing game deviates from that model, we bandy about ideas about how to try and fix healing at that state where healers are trying to do more with less. But maybe we’ve become a little too focused on the triage model, seeing it as an enduring goal instead of what it really is—the first step of a healing cycle that moves towards inefficiency.

In that regard, I think it’s interesting to consider that none of the issues identified above would be completely absent in a triage state. Cooldowns could still be strong, healers would still be limited by absorbs, burst could still be widespread, and mana could still be toeing that line between sufficient and plentiful. So, maybe we need to add another issue to this list, because while the above healing issues are (I think) still important to discuss, we seem to be missing something hiding beneath their surface. Maybe what we should additionally be concerned about is the one thing which provides the constraints for that healing state we healers have grown to love and which we know has been on Dev’s radars since development began on Pandaria—our stats. I don’t think it’s any coincidence that we’re seeing the above issues now, given the rapid stat inflation we’ve seen in the past two tiers alone.

In which case, I might have been a bit precipitous in saying that a discussion about the state of healing should be disconnected from the numbers. Maybe this discussion should, in fact, be all about the numbers. And maybe, the item squish that was postponed in Pandaria, will need to become a reality in the very near future.

 

 

** For people who still insist on misunderstanding this argument over cooldowns as being a fixation on numbers … The numbers associated with cooldowns are inconsequential. It doesn’t matter how Healing Tide measures up to Tranquility, or Divine Hymn, or Aura Mastery or whatever other 3 min CD is out there. What matters is the size of that cooldown versus the size of your other available tools. Think of it like painting a room. You have a small brush that is no bigger than your hand, and you also have a roller brush, which is as wide as keyboard. Which are you going to use to paint the majority of the room? Obviously, you’ll choose the roller brush. Your friend might even pop by and comment about how much of the room you’ve painted in so little time, which makes you feel great. But suddenly, when you’re halfway done with the room, the roller brush breaks, leaving you to finish the room with the smaller brush. Imagine how futile that would feel. And now imagine that your friend walks in again and notices that you haven’t painted all that much more since he last stopped by. That would likely be a bit frustrating, especially since it took far more effort to paint a small area by hand than it took to paint a large swath with the roller. That is the essence of the incongruity that exists between cooldowns and the rest of the healer’s arsenal. It has nothing to do with numbers and everything to do with how much, after that exchange with your friend, you’re going to value that roller brush over the little brush. 






12 Comments


  1. Talarian

    As a Paladin Healer, all I can say is, “Yup.”

    I agree with this post in pretty much its entirety. I find the stat inflation point particularly salient. Back in Wrath they said stat inflation got away from them because they weren’t planning on heroic tiers of raid content in their original power curve. Lo and behold in Pandaria we find that they were just designing the ToT ilvl jump from T14 on the fly (as evidenced by their discussions about making gear palatable for Heroic T14 raiders, as well as making ToT actually difficult walking in. Makes me wonder if they goofed again on stat inflation.

    They made it such that caps aren’t likely to be hit for percentage-based stats (like haste and crit), but something tells me they goofed on the linearly-scaling stats such as spirit. Or perhaps they figured decoupling intellect from mana regeneration would have been sufficient and just accidentally made spirit more powerful that they should have to compensate?

    Honestly, I just think they’re making it up as they go along, and we’ll either see a nerf to spirit in 5.4, or some sort of Sunwell Radiance/ICC debuff that reduces the effects of spirit in the final raid.


    • Kevhore

      Instead of a 30% buff, item upgrades are driving the huge increase in survivability and output. I don’t think spirit is out of control, actually. The legendary meta makes it so most healers are actually avoiding spirit at this point. Maybe that means they should nerf spirit more so we have to stack it a bit. But until they break this meta, I can’t see an end to the spamhappy healing strategy.

      It just blows my mind to think we’ve gained about 100 ilevels this expansion so far – alongside buffs to the utility of secondary stats (not just the output of them) we are getting incredibly powerful too easily. For Shamans, both crit and haste (from RPPM) give us more regen now. Next tier it will be near-impossible to oom, which means throughput stacking and overhealing will be the most viable strategy.

      Personally, I don’t like healing like this. I agree with Vixsin, it takes much more skill to manage scarce CDs, mana, and healing targets rather than trying to maximize efficient output like a DPS.


    • I think you’re right that the balance for %-based stats feels appropriate. I’m at ~25% of my chosen secondary stat, which is right about where I was in previous expansions heading into the last tier of content. But, I’d suspect that part of the problem with Spirit, is that there wasn’t ever a clear/solid decision that it was going to be a *primary* stat for healers. So, it was budgeted at secondary stat scaling levels (eg: 320 spirit gems versus 160 intellect gems). Meaning it simply exploded once you started getting gear.

      For example … I was around 9k Spirit when I first stepped into MSV; I’m over 20k Spirit now, in addition to trinkets and the meta. The number of sockets I have in my gear went from around 6-8, to 25. If I socketed only Spirit, that would amount to a 2560 –> 8000 (delta: 5440) increase in spirit. That’s an insane amount of potential increase to a healer’s primary stats.

      Now, I’m no game designer, but I can’t think of a way to narrow that incredible difference in values, without adversely impacting people in lesser levels of gear. Because you need people just getting into the game to be able to perform at a reasonable level. Maybe they could have introduced some sort of Regen DR into the formula, but then that would penalize classes like shaman and holy priests who have much higher passive regen needs than classes like disc priests and monks.


  2. Ryan

    Very observant post, Vix! I’ve been thinking about these things for a while now, too. I’ve noticed that stat inflation this expansion feels a lot faster than it did in Cataclysm.

    One additional thing I’d like to mention is that how much absorbs are made more valuable by the current burst damage model and how much our burst healing abilities heal. For example, Iron Qon’s fist smash is easily healed by druid mushrooms, priest halo, monk uplift. These abilities heal for A LOT. When all of them use their abilities at the same time, they’re obviously competing for a piece of the healing pie. The damage is instantly healed, so they each get a small piece of the pie, and the absorbs that come afterward can take a much larger portion of the pie because they’re not sharing.

    I’m still planning on making that forum post. I’m glad to have such a good model to reference and cite in posting my ideas! Hopefully the dev’s will read it.

    P.S. Have you done any PTR yet? I keep catching myself saying WELCOME TO ICC. Tanks get killed in 2.0 seconds literally on some fights. : /


    • Yeah, tank damage on PTR has been … uh … slightly painful with every testing cycle thus far. So much so that it’s now a running joke whenever we’re testing a new fight. A tank will get squashed and someone will blandly say “Huh, well … tank damage feels a little low”. The one fight that I was really concerned about when testing ToT was Dark Animus, which on PTR, had our tanks, with cooldowns running, dying a mere 10seconds into an attempt. Thankfully, it was tuned back a bit on live, so I’ll keep my fingers crossed that the same thing will happen this time around. :-P


  3. I was looking for a post here while waiting for 5.4 most needed resto shaman changes. As always, great post. That reflects exactly how I feel about healing in MoP.

    – Too many cooldowns, that often heal for too much (at least in normal 10 man). If you add to that our lack of spread healing tool I always feel like my spells either heals for too little or too much.
    – The feeling that my actions don’t matter because the spells that demand the less action (healing totems + Healing Rain) account for the majority of my numbers.
    – The stat inflation : the difference between ilvl 502 and 522 is already so big that I’d prefer +hit 522 item than an adequate 502 item. That’s explain how out of control the stat inflation is this tier, and how raid damage has to increase to provide challenge to healers.
    – In addition to your “more CD > more burst damage > more CD”, I’d add another escalation : Blizzard made spirit and regen appealing for healers, and then he gave us plenty on toys to give us this regen (trinkets, legendary meta…), to the point that we don’t need spirit and regen anymore.

    I want triage healing back ! (A zone debuff reducing healing and absorb effects in SoO would be great !)
    Zahia´s last post ..How to fix resto shaman (spread) healing : the best of !


  4. shammypie

    you said something about cds well guess what you get 1 more, and we gave you the 25 tranq buff and to top it off we made healing rain bigger and heal for more in 25 man!!!


  5. I honestly don’t think I’ve got to experience a single triage fight at all this expansion. While this might be due to a lack of progress in T14, my guess would be that if your first three fights of the tier aren’t triage then the rest are unlikely to be either.

    Thing is, on those early fights I spent more time trying to optimize my overall spread healing than I did deciding who to heal – making good decisions about where to send my heals didn’t make a big enough difference to throughput or success rates.

    I really, *really* miss the T11 triage paradigm. It was engaging, it was challenging, it was never the same from fight to fight. I want it back!
    stoove´s last post ..Experimentalcraft


  6. […] Life In Group 5 has a nice analysis of how healing is coming along in this post: Status Check: The Healing Game as of 5.3. […]


  7. Shamanhands

    Squishing is necessary to maintain the triage model, I agree. I’ve been wrestling with the fix of only squishing spirit, making it a far more expensive stat. If the regen at the end of 5.4 matched the regen at the start of 5.2 (and all encounters were adjusted for dmg accordingly), we would be in a good world. Of course, mechanics like Resurgence would need to be adjusted…

    In the end, what kind of incoming damage fits a world where healers have early 5.2 levels of regen but the same throughput as now? I’m still trying to figure that part out.


  8. Tier 11 stands in my mind as the best healing tier we’ve had. I would love it if we could keep that level of mana-reliance across tiers.
    Coreus´s last post ..My first compulsory server merge


  9. Clearlylol

    Waiting for the 5.4 patch review from miss Vixsin.



Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

CommentLuv badge