Healing Theory, Part 7: Introduction to Active Mana Regeneration

Previous posts in this series can be found here

This is the first post in this series that’s explicitly about new mechanics in Warlords. As a general disclaimer, at the moment there is not yet a public alpha, so all we have to go on is information in patch notes and spell data. So I while I will be going into some numbers in this post, keep in mind that anything can change. The point will be more about how to understand active mana regen, and not as much, in particular, about comparing the relative strengths of each class’s new mechanic.

The Active Regen Spells

  • Druid: Innervate. The Druid casts Innervate (2 seconds), which lasts 8 seconds or until the Druid spends mana on a healing spell. If it runs for the full duration, it regenerates 5% mana.
  • Shaman: Telluric Currents. Lightning Bolt casts return 1.25% mana.
  • Monk: Crackling Jade Lightning returns 2% mana after a full-duration channel, which takes 4 seconds. Monks also still have Mana Tea.
  • Paladin: Divine Plea costs 3 Holy Power, and returns 7% mana.
  • Disc Priest: Penance, when used offensively, returns 1.1% mana per hit (and no longer Atones).
  • Holy Priest: When in Chastise, Smite and Holy Fire each return 0.75% mana. Notably, Chakra shifts have a 10s cooldown.

The idea behind these is pretty easy to see: give every healer an ability that lets them choose to regain some mana, at the cost of giving some opportunity to heal. This is a response by Blizzard to the problem that current mana-related abilities tend to involve little or no decisionmaking; you generally simply use them on cooldown. They were prime candidates for removal during the ability-culling process, and that is what happened at first. And while there was no huge problem with that (mana is still interesting due to the inherent choice in choosing how to spend it), abilities whose purpose is to regain mana are definitely a place to add something to healer gameplay.

General Analysis

Shaman: The simplest of the lot. Cast a Lightning Bolt at any time and get mana (but in Warlords, LB can no longer be cast while moving). During any moment where you don’t need to heal and don’t need to move, you get some mana.

I discuss Shaman first, because as the most basic mechanic, it’s a good template for general discussion of these new spells. One important goal is achieved easily: there’s no reason to ever be casting nothing. Remember that this was one of the purposes of the “small” heals that all classes had, that are being removed in Warlords because they never found a strong role. This iteration is much more appealing: when there’s no healing you need to spend mana on at the present moment, regaining mana is a more rewarding use of time than casting a tiny heal.

One concern (that will be easier to feel out once we have a game to play) is that this mana return is so freeform and unconstrained that it becomes uninteresting. Instead of the choice at any moment to cast a heal or do nothing (which is what we have now when mana-constrained), there’s the choice to cast a heal or cast LB. While you avoid the bad aesthetic of standing still casting nothing, it’s possible the resulting decisionmaking hasn’t actually changed much. Compare in your mind 1) the world where LB regens mana to 2) the world where LB does not regen mana, but passive regen is correspondingly higher and heals are slightly more expensive. They seem quite similar, except for the slight coarseness due to LB having a 2.5s cast and requiring standing still.

Monk: Very similar to Shaman. Marginally more demanding because the channel you must complete to get your mana is 4 seconds instead of 2.5. So a slightly more meaningful requirement to not cast a heal for a short time.

In addition, Monks still have Mana Tea, which also requires standing still and returns mana much more rapidly than CJL (around 8 times as much). Unless there is a danger of overfilling on mana, Mana Tea will have a higher priority than CJL. You check if you have stacks of MT, otherwise CJL. These two spells clash somewhat in terms of function, since it’s very hard to create an interesting choice between two things that both provide mana. Mana Tea itself walks an interesting line: it doesn’t incorporate the same choice as these new active regen mechanics, since you can’t choose how much you use it over the course of a whole encounter. But, you have to choose when to use it, so it does have some features of active regen. The result is this isn’t the greatest use of two different buttons (when not healing, you use Tea if you have stacks, otherwise CJL), but I think they felt the class didn’t have the true active regen experience with Tea alone.

Holy Priest: Again similar, but now you must shift into Chastise before getting your mana. This commits you to losing your Holy Word spell for 10 seconds (and has a secondary effect of, after you shift back to a healing Chakra, preventing a subsequent shift for 10 seconds). It is essentially the Telluric effect, but with a slight commitment to a healing penalty over a 10-second timeframe.

The element of commitment helps address the concern I mentioned above about Shaman. With the need to make a decision and a judgment about the upcoming 10 seconds, any moment not healing won’t necessarily have the reflexive response of starting a Smite cast. While I know that some Priests are unhappy with the Chakra effect in Mists, in this case it makes the decision about active mana regen somewhat more meaningful. Giving up your Holy Word for 10 seconds is not a big cost (in Sanctuary, it may not even give up a single cast), especially since the whole point is that you’ll likely be Smiting for those 10 seconds rather than healing. I hope it’s just enough so that you don’t do it in every spare second where you might not need to heal.

Druid: Going further down the spectrum of commitment, Druids must invest 2 seconds into an Innervate cast, and then not heal for 4 seconds to get a regen tick, and then not heal for another 4 seconds for the second regen tick. 0 healing for 6 seconds is the most that’s been asked of anyone so far, and quickly invited comparisons to the “5-second rule” from the days of yore.

One question this raises is whether we’re going to be encouraged to enter long extended periods of not healing. Not only does the Druid mechanic require 10-second increments to be used at maximum efficacy, it, like all of the ones discussed so far, can be chained indefinitely. It may actually gain extra efficiency (25% specifically) if you do so, if you can pre-cast a new Innervate to start right as the previous one ends, but I’m going to put this aside until we find out if it’s possible/intended. In either case, there is a potential incentive to take an extended break from healing whenever possible, and that’s why people think back (usually negatively) to vanilla strategies where one healer would sit out while others healed. I’m not too worried about this aspect of it, however.

Beyond the 10-second timeframe of the Druid and Holy mechanics, there’s no explicit incentive to cut off healing for a long time. You get the same amount of mana if you regen for 20s as you do if you regen for 10s, heal for a bit, and then regen for 10s again. So there’s a strong incentive to judge when you have a 10s period where you can stop healing, and I think that’s good gameplay (in fact the Shaman and Monk abilities may be better if they could incorporate a similar concept). But there’s no incentive to find a 20, 30, or 60 second period where you can stop healing. Combined with the fact that many classes have strong, efficient spells on 10-15 cooldowns, there will be every reason to come back and heal for a bit as a regen period ends. The 5-second rule caused your regen breaks to be more effective if you could break for as long as possible in one shot, but that is not true of the new abilities.

Note that all 4 mechanics discussed so far reflect the same 0.5%/sec regeneration rate for any time you devote to using them.

The final two break strongly from the Telluric concept, and are in fact the two I like the best so far.

Disc Priest: Every time you Penance, you have a choice of using it for healing or mana. This is simple and elegant. The choice is always there, and it’s built into your rotation. Among other features, it’s inherently limited–you can’t suddenly stop healing and devote yourself fully to regen. Instead, you can shift partly to regen, using one important spell for mana while continuing to heal with the rest. This should require a little more foresight in the mana management, to see in advance when you need to start regularly using Penances for mana to curb your spending.

The current value on Penance is interesting: 3.3% regen per each cast. Penance costs 1.26% with 5 stacks of Evangelism. 2.04% mana for 2 seconds of regen is a little more than twice what other classes get, presumably the premium for the fact that Penance has a cooldown. Without Evangelism, it’s down to 1.5%, still a 50% premium. So even with regular Archangel use, the premium is substantial. This is a pretty appealing tradeoff, since using a regen spell for 2 out of every 9 seconds is already quite a lot; it shouldn’t be seen as very limiting (it may not be expected that any class uses theirs more often than that in normal gameplay). Penance may also have a bit of a bonus for the fact that it’s the only of the abilities that doesn’t naturally scale with haste (assuming that Innervate does).


This unique approach makes you spend a secondary resource, rather than simply spending time, to get mana. There are a lot of math questions that surround this, and I’m going to try to keep it general on those points for this post, but overall I like this idea. It does the best job of breaking away from the “get mana back when you’re not healing” model that’s perhaps slightly similar to ordinary passive regen. Like the Priest one, it’s a decision that comes up regularly as you play, but not in the form of stopping your normal rotation. After looking at them on paper, this is my favorite of the lot, because, as I hope to at least introduce in the post and develop in later ones, it adds the most in the way of decisions.

Mathematical Interlude

This is a survey of some things I would keep in mind if you’re trying to start theorycrafting on active regen, or simply trying to think about how it will work out.

In current spell data, it looks like we’re going to have 160,000 mana at level 100 (this is from datamining in the available client), and that Spirit will give around 4 times the MP5 that it does now (so, relatively speaking, each point is 8 times as valuable). To put that in context, however, you need to have a better idea of how much Spirit you’re going to have on gear. With Spirit only available on a few slots, the breakdown of your total mana resources will look very different (I’m going to hold off on doing it until we have more information). For an early estimate, 400 Spirit in blues seems reasonable, which would translate into an equivalent of 3000 Spirit in current pre-squish numbers (not counting enchants and gems, or Spirit in slots that aren’t showing up in the data yet). This all goes to say that Spirit is looking to be a similar, if slightly smaller, piece of the pie chart in the previous post I just linked.

What about active regen? In current numbers, 0.5% mana/sec is equivalent to a bonus of 7500 MP5, or 13298 Spirit, anytime you’re not healing. So, against a backdrop of similar or lower Spirit from gear, getting that much of a bonus is going to be very significant. It would be as though, translated into current numbers, your Spirit was 5000 while healing and 18000 anytime you could break and cast a regen ability.

As I hinted above, one way to conceptualize this is that you have more regen, but spells cost more in practice. They have their real cost, plus the opportunity cost of not having the high regen for that amount of time. In other words, if spells didn’t change otherwise, long-cast spells, which are generally the more efficient ones, have their efficiency effectively reduced relative to short-cast spells, because you give up more active regen time in order to use them. This effect will be important to take into account when we start making spreadsheets to compare heal efficiency (which is the next thing I want to do; we should have info now that’s at least good enough to start seeing how it looks).

Finally, Paladins. Here’s a napkin calculation to get started. One Plea returns 7% max mana, which is the same as 35% base mana (due to Holy Insight). 3 Holy Shocks cost 28.5% (this was reduced in the most recent spell data from 15% apiece, possible to address this issue). Moreover, 3 Holy Lights cost 34.5%, still breaking even if you turn their HP into a Plea. 3 Flashes of Light are still much more expensive, at 69%.

It’s not exactly informative to say that you can get positive mana from Shock+DP, because you’re most likely to going to Shock on cooldown like you do now in any situation where healing or HP are relevant. You won’t be adding Shocks to your rotation just to spend them on DP. In the rare case where there’s nothing to heal for a long time and you’re also full on Holy Power, in theory 3 Shocks and a DP will have turned into 6.5% mana–not bad at all for 18 seconds, given that you’re doing some free healing to boot. What’s probably much more common though, is that you’re Shocking on CD and using HP spenders as needed to heal, but, as with other classes, you’re occasionally be in a situation where you don’t need to heal something (or generate immediate HP). But with this Divine Plea mechanic, you can use Holy Light during those times, almost in the same way that other classes use their active regen spells, knowing the full mana cost can be recouped anytime.


These abilities won’t cause a sea change in the healing game. Mana will factor into your spellcasting decisions generally as it does now. At minimum, however, they’re a minor improvement in that they improve the feel of what you do between heals. Even cementing the fact that there is something you do between heals is helpful for the ordinary pace of gameplay, now that they’re aiming to move healing more in the direction of efficiency-oriented decisions. Reinforcing that you’re intended to sometimes think about whether you need to even cast a heal is a good thing in this regard.

Even in the case of Shaman and Monk, where I suggested that the spells might not add a huge amount on the abstract and theoretical side of things, my gut sense is that they will play well. This will be much easier to comment on when we can actually play with them, but after a whole post looking at the question in the abstract, I don’t want to minimize the change to the core experience. Casting something and casting nothing are very different ways of getting back mana, and I definitely expect casting something to be the more fun way of doing it, and the one that feels more rewarding to utilize.

So I’m happy with how the system looks. I think it’s already likely to be an improvement to healing. The refinement I’d want to see is to add a small amount of tradeoff to the mechanics that don’t have any right now. The gameplay that defines active mana regen should be: I use my active mana regeneration because I judge that maximum healing output won’t be critical over the next few seconds. These spells have the potential to be more than a filler activity. They’re an opportunity for healers to be rewarded for their awareness and anticipation of healing demands.

4 thoughts on “Healing Theory, Part 7: Introduction to Active Mana Regeneration

  1. For holy priest, I like the decision, but making me hit 3 buttons to do it is just clunky.

    Also, if chakra still works the same, it has another effect than just taking away your Holy Word heal. It takes the bonus healing (+ to AoE or + to single target) away from all your other heals as well.

    It seems the intent is to effectively make a button that give mana but weakens the priest’s healing for a period of time after it is hit. Wouldn’t some sort of self-debuff achieve this in a cleaner way?

    • Chakra in 6.0 will no longer have bonus healing to AoE or single target. They’re removing those and making Chakra just give you the extra mechanics (different Holy Word spells, the Renew refresh, etc).

    • It takes away a side bonus (CoH cooldown reduction or Renew refresh), but yes, no more passive heal bonus. So not nearly as big a deal as it would be now.

      As to self-debuffs (like pre-5.4 Divine Plea):
      1) I think there’s an element of psychology. Casting heals with a -50% debuff on yourself feels really bad. Having fewer HP or having Penance on CD doesn’t; they should actually feel rewarding in a way because you used a resource productively. The Chakra one is a little worse in this regard, which is why it’s important that the penality is very minor.
      2) There’s a little more flexibility in swapping between casting heals and casting non-heals, than there is in simply being flat debuffed for a time. With Chakra, you can stop healing entirely and fully regen if it’s safe, or you can abandon the regen and cast heals (at near-full effectiveness) if it’s suddenly needed.

  2. Pingback: Healing Theory, Part 8: Three 6.0 Topics | It's Dangerous to Go Alone

Leave a Reply

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