Legion Alpha: Resto Druid Mastery

The restoration druid’s mastery has a long history of changes. This is due in part to the fact that Resto druids are a really difficult concept to design, as we’re the only “HOT-based” healer, where our heals tick slowly over time.These HOTs are intended to be weaved together with a small number of direct heals we have access to (so that we can keep up with some amounts of burst damage). We are once again facing a change in the resto druid mastery bonus. The feedback on the Alpha forums from people doing mythic raiding has been overwhelmingly negative. This thread is going to try and contextualize this change, talk about where it falls short, and suggest what other design options might be available. Ultimately, we need to come up with possible solutions, given so many failed designs that have come before.

The short history of resto druid mastery bonuses

To understand why Resto druids have a new mastery (they don’t like) on Legion Alpha right now, we have to talk about all the failed Resto mastery designs. I inserted links to previous mastery discussions along with short summaries (but people are welcome to visit my old mastery threads to see the controversy & how our feelings about mastery changed over time).

  • Mastery #1: Our first mastery made HOTs heal more on low health targets. This mastery never hit the Live servers because it turned out to be numerically terrible, as I documented in early testing. This was similar to the shaman mastery (which shaman really dislike), but actually worse because each individual HOT tick heals for so little.
  • Mastery #2Our next mastery increased direct healing on people who had a HOT. I initially really loved the concept of this mastery, as the post I linked showed. However, it ended up being really problematic in raid healing, where it just didn’t work to constantly chase HOTs with direct heals. Due to the fact that each HOT was on a single person, the minute you had 20 people to heal, the mastery mostly fell apart because having to cast 2 heals on every person was too cumbersome to benefit in raids. This meant that 20-man raid healers didn’t enjoy healing very much during the short period of time where this mastery was live.
  • Mastery #3: Next comes the version of Harmony where casting a direct heal gives you a buff  that then increases your HOT healing done. I will quote directly what made this so appealing: “The best part of the new mastery is that it puts the buff on YOU, and not on your target.” This version of the mastery is one that people came to know & liked (after knowing what the alternatives were), and some version of this mastery stuck around the longest.

New Legion Resto Mastery Explained

The new resto druid mastery for Legion increases your healing done for each HOT you have on that particular target (which can stack up to the maximum number of HOTs the druid can stack). The current Legion mastery thus requires stacking multiple HOTs on the same target. Importantly different from Mastery #2, this increases all your healing done to that target, and allows for stacking the mastery buff multiple times. While the first HOT gets some benefit from the mastery, you need 2 or more HOTs on a target to get the full bonus. Note that this stacks on your target, and NOT on you, which is an important design consideration.

The good:

If you heal one target, it’s possible to stack lots of HOTs on them (lifebloom, 2x rejuvs, regrowth, wild growth). In 5-mans, you’ll have a decent number of HOTs spread around the group, so that most people will have 2 HOTs on them. Numerically, in small groups, the new mastery is the same or better in terms of total healing done today. This also has the potential to be a significant buff to our tank healing, something they want to be relevant in Legion.

The bad:

The new mastery in Legion is actually similar to the Mastery #2 described above. However, it does something slightly different, since the HOTs boost other HOTs in addition to boosting the direct heals on the target. For the same reason that HOT chasing with Mastery #2 was bad, the new mastery suffers the same fatal flaw. To do good healing, you have to cast a large number of spells on a small number of people, rather than healing the person who needs to be healed the most. This requires a lot of setup time, as the “buff” is specific to the person you are healing, and not to you as the healer. While someone else can jump right in, you have to anticipate who might be taking damage and then cast 2 to 3 heals on them to get the full mastery benefit. This is going to be somewhat tedious in raid dungeons, where you may end up devaluing Mastery to a great extent in 20-man raids (where only the tank will reliably get the full mastery bonus out of your heals).

Druids are designed as HOT healers, where we are slow and require ramp up time to reach our full healing strength. Rather than spamming true AOE heals, we weave multiple single-target HOTs on tons of people between AOE heals that have cooldowns. Stacking multiple HOTs on the same person is a slow process – especially if you are responsible for watching 20 people (at 20 people, HOT stacking becomes a potentially frustrating process).

Sigma suggests (on the alpha forums) that: I get the impression that people are significantly overestimating how many HoTs one has to have stacked to get reasonable value out of the mastery.” This is exactly the problem that makes the mastery feel psychologically bad, though. The answer is always going to be “more than 1 HOT”, and in that case, you are HOT-chasing like we did with mastery #2. People are always going to feel like the best strategy is to chase a HOT with another HOT so that you can maximize mastery healing. The mastery largely isn’t passive bonuses to your preferred healing style. It requires you to actively make decisions about whether or not you want the full benefit of your mastery or if you are okay only benefiting partially from it. Anyone interested in maximizing their mastery has to cast more heals than they might want on a particular target.

This is ultimately why Mastery #2 failed: It feels bad to have to chase your HOTs with more spells (either direct heals or HOTs). The mastery increases the feeling like you have to cast 2 heals for every person you heal (when everyone else around you casts 1 and moves on). This impact on our healing style in 20-man raids isn’t all that fun and makes it hard to keep up with other healers who don’t have to ‘waste GCDs’. So, for 5-mans, the new Legion mastery works fine, but in raids, it feels bad & taxing. Numerically on paper, the new mastery works out fine – but it feels psychologically wrong. The new mastery changes how we heal in ways that forces you to constantly think about the mastery, rather than constantly thinking about the best way to save someone’s life. In that way, you are investing a lot of time into a small number of people, which isn’t a viable 20-man strategy where your HOT investment in that smaller number of people is ultimately going to be stomped on by other people’s big AOE burst heals and result in overhealing.

Is there another option?

The most important point my trip down memory lane highlights is that the most popular druid mastery (the one we currently have today) works best because it places the buff on the healer. The most obvious solution would just be to go back to the old mastery if the new one won’t work. However, most of the time we were using Swiftmend to “prime” our mastery today, and the cooldown change for Swiftmend means we’d have to rely on a different strategy for getting our mastery buff (which would probably waste fewer GCDs than a HOT-chasing strategy). Overall, looking at the history of druids, it would be more rewarding and feel more natural if the mastery worked with our HOT spreading design, rather than requiring HOT stacking.

It might be possible to change the Legion mastery to put a stacking buff on the druid  that increases our healing done for each person who has a HOT on them. We’d obviously need a cap on any type of self-buff so that it didn’t spiral out of control (e.g., maybe 3?). That said, with the fact that druids always spread around our HOTs, that kind of mastery might feel too passive & boring. Then again, maybe a passive & “boring” mastery is better than returning to the days of mastery #2, where you chased your HOTs with other spells just for the purpose of gaining more mastery benefit.

Posted in Legion, Restoration Healing Trees, Written By Lissanna

10 comments on “Legion Alpha: Resto Druid Mastery
  1. RohanV says:

    I’ll approach the problem from a different angle. Pretend Mastery did not exist. How a healer heals is dictated by the spells she has, i.e. druids blanketing people with HoTs.

    Now, we introduce Mastery. Do we want Mastery to reinforce the basic healing style, or push the healer into a new style?

    If Mastery pushes the healer into a new style, the healer can still use the old style by ignoring Mastery and focusing on the other secondary stats. But if Mastery produces the highest output, then the healer is “forced” into the new style.

    It seems to me that the healers generally want Mastery to reinforce the basic playstyle, but Blizzard kind of wants Mastery to push a new style. I think a lot of the friction comes from that, especially when one class gets a reinforce basic style Mastery, but another class gets a new playstyle Mastery.

    Aside from that, as I was reading, I had a similar idea to yours, only have the buff be equal to the number of HoTs currently running, rather than players. That way you could set the cap fairly high, at 10-20 (not entirely sure how many HoTs you have in Legion), making it a bit of a challenge to fully cap out and maintain a full stack. Mastery could be non-linear with regards to stacks, such that you get the 80% of the benefit with 50% of the stacks.

    • Lissanna says:

      Most healers don’t actually have masteries that change how they heal. In the case of shaman, their heals simply heal for more on the people lowest in health. Those low health targets are always the people you want to heal.

      Most of the other mastery abilities simply just do more of what you already do. The Legion proximity based paladin mastery comes the closest, but this has to do with where you stand, rather than what heals you cast.

      None of the other healing masteries besides the druid one changes how you fundamentally heal. The Legion mastery changes the fundamental strategy of how you heal, encouraging more overhealing problems where you invest multiple GCDs into someone who is going to ultimately be topped off by someone else’s big direct heals. The HOT blanketing strategy (many HOTs on lots of people) means that you aren’t investing a lot of time on any one person. The mastery runs contradictory to HOT blanketing as a 20+ person AOE healing strategy and requires you to focus single-target healing on a smaller number of people.

      • RohanV says:

        Well, it’s been a while since I raided seriously, but my defining experience with Mastery was around 4.2 with Holy Paladins. We had 3 different builds which played differently, and had different strengths and weaknesses, depending on what stats you emphasized.

        See Zaroua on Holy Paladin Mastery and Holy Paladin Builds

        So I guess that period of time is my ideal of how a good Mastery is supposed to work.

  2. Bodhi Rana says:

    Recapping our Convo on Twitter…

    Ack! 2 GCD HoTs + 1 cast-time heal to properly heal someone who just got whumped on?

    Even going with the idea that a surprise whumping happened (which does happen) you have to make a decision (push out an “un-masteried” big heal [1 cast time] or a HoT and a “half-masteried” big heal [1 GCD + 1 cast time]) on how to keep this person from dying. Then you’ll want to make sure they’ve got 2 HoTs on them, then cast another heal to bring them back to the comfort zone.

    So you’re looking at 4 spells to triage an emergency (1 cast-time heal + 2 GCD HoTs + 1 cast-time heal or 1 GCD HoT + 1 cast-time heal + 1 GCD HoTs + 1 cast-time heal = 6 sec to fix them).

    The other option is to attempt to blanket the entire raid in HoTs to handle the pre-emptive triaging. “Oh, your raid is more than 15 people, or you plan on doing more than just blanketing rejuvs on everyone? Oh well, too bad. Hope the whumping lands on someone whose blanket rejuv isn’t expiring before your cast time lands on them.”

  3. Besheke says:

    Thank you Lissanna for hitting the nail on the head for me. I have been restoration druid now for almost 11 years. I actually have 2 active druids.

    I remember Mastery #2 and how frustrating that was. I have ridden the roller coaster thru the years.

    This recent past expansion has given me the option to play Balance as well and I have come to the conclusion that if this “new” mastery does go live I will not play Restoration, but change to Balance with a Guardian offspec.

    I want to have some enjoyment, and though I do enjoy healing a lot, I don’t feel that this new mastery is rewarding enough for the amount of work it could involve. At the gut level it just feels poor, and part of the gut feeling comes from the years I have played. I still have other concerns with the changes being done to the spec and how they will work as a whole. But I have always been decent enough to figure out tricks and get the job done.

    That being said I do like the idea you propose for mastery. With the amount of hots we can comfortably and realistically maintain maybe cap it at around 3-5 and tune around that? I have been doing a lot of reading and watching, but without actual hands on its hard to say.

    Maybe I am just getting inflexible as I get older, but I do have to say if something works well why fiddle with it? The track record so far has not been “fun”.

    Thank you,


  4. wrathsome says:

    That is a LOT of heals. Spamming HoTs like this would be over-whelming.
    Will the absence of Spirit and the flat mana regeneration going to be a problem?

  5. Berdache says:

    I suppose it comes down to what we are doing in the raid. If we are tank healing then the value of mastery goes up and it is worth stacking. If we are raid healing and can only expect to be able to cast a single hot on them, before another healer else gets in there, then its value goes down and we should potentially treat it as a dump secondary stat.

    I must admit I dont like the sound of it because, for raid healing, we are potentially casting two spells to make a third hit harder and the first two overheal more.

    I havent looked at Legion that much but does this new mastery invalidate several taleents and glyph choises? So we will always pick multiple rejuvs and never pick the regrowth glyph?

  6. stupidhero says:

    I wouldn’t consider the “Good Point” you mentioned to be that much of a good point, but more of neutral, as it creates a severe tuning issue:

    At 10% haste with Inner Peace, you can get a raid average of 0.9 HoTs per target.

    At 40% haste in a Cultivation build (assumed to be triggered when needed), you can reach a 2.7 HoT average.

    At 10% haste with cultivation and germination in focused target situations, you can easily reach a 4 to 5 HoT average.

    Currently mastery is tuned (rating conversion 400 to 1%) for a 1.33 HoT average, which means that we have 3.3 to 3.7 times the secondary stat budget in the third case, yielding about 70% higher performance to our spells. This isn’t much of an option for class balance overall, thus has to be accounted for via our single target spells inherently being tuned lower. In the end, we’re left without a way do deal with random damage spikes, and not neccesarily have much of an advantage in direct heals relative to where we currently stand on live. The current alpha values actually indicate this to be the case: unlike other classes, our direct heals did not get a coefficient increase over live (are now tagging behind 20-40%), and lb was nerfed with the recent buff to mastery.

    Furthermore, while the first two cases may not be problematic in the initial raid tier (fixed stat budget which can be tuned around), but it still means, that we double dip from haste, thus excessively scale throughout the expansion. It also creates this odd effect, that to maximize mastery benefit for raid healing, we actually prefer haste over mastery until we get sufficient haste to consistently beat the average. Small caveat here is, that such a playstyle would require sufficient mana, which is all but sure at this point. Though this also highlights another odd point with this mastery, as when you want to maximize efficiency (mana concerns), you’re inclined to spread HoTs first, instead of actively seeking out to stack them, which is incredible likely to just increase overheal, yet are forced to do to even achieve 100% performance.

    In any case, this shows, that the mastery is also rather problematic for talent design, as it puts strict preference on some talents over others.

    Regardless of the inherent scaling issues, it also cannot be tuned any different for precisely the reason you stated in “the bad”: the druids will feel like they’re forced to chase the HoTs, thus consistently beating the average (the 1.3 currently chosen in beta). In the long run, this means that the average would have to be chosen to be relatively close, if not equal, to the maximum number (this maximum increases with haste) of HoTs a druid could realistically attain by chasing HoTs, but then you just do not get much of a reward (10% over other classes), yet have to invest a significant amount of work, and still lag behind considerably most of the time. In this case, it just will reinforce the psychological impact, because wether it’s 0% performance at one stack (old #2), or something well below 100% at one stack (new mastery) isn’t percieved different by players (and the closer to 100% the “reward” for stacking is to 100% the closer to 0% the penalty will be – that’s simply how this stacking mechanic works).

  7. stupidhero says:

    Just to add something more, which I noticed after the recent Abundance/Cultivation swap:

    This mastery is heavily restrictive to talent design, as it inherently favors talents, which provide additional mastery stacks:

    Not only do those talent provide a benefit of their own (e.g. the reju strength HoT on Cultivation/AoE throughput on SB), or to be more precise, have to due to neccesarily having to be competitive even in a low mastery build, but also provide the mastery benefit, thus skewing heavily in favor of mastery in a high mastery build.

    Of course one could tune the talents in question for this, but then they’d be rather mediocre prior to reaching a given mastery level, i.e. you’d not pick a talent by your liking, but by what your gear tells you to do.

  8. John says:

    I just needed to comment to thank you for this post. I knew this “new” mastery felt like something I had used before, and this was exactly it! During the Cataclysm beta I very much remember playing with Mastery #2 and this new mastery just brings up all the painful memories of the feedback/iteration cycle. I hope the mastery will someday be something that benefits our spreading of dots, as that feels like a great nod to the resto druid.


Featured Blogs