Why More Spirit is not the Answer to your Healing Problems

[4/8/13: Six months later and I see this post getting quite a lot of attention on various forums. While I’m glad of that, I do want to make sure people finding their way to it just now also see the follow-up posts in the Healing Theory sequence. The first is here, and the third is particularly relevant to this discussion. Those posts are further developments of the ideas that first came to me while writing this one. While I still stand by the main ideas of this post, it was more of a reaction to certain discussions at the launch of the expansion, whereas the later sequence is meant to be a more general exploration of healing principles and conclusions. ~H]

When I recently made a post on EJ about how healers are, as they always do at the beginning of an expansion, overly obsessing about Spirit, the first reply I got explained it in the most simple and accurate way possible: “Fundamentally I’m pretty sure most healers are very, very bad theorycrafters. There’s always been a lot of magical thinking, faulty logic and poor contextual analysis.”

Based on my experiences theorycrafting and writing for a healing class for a few years now, I can’t deny that in any fashion.  When I quoted it on twitter someone told me, “you don’t have to be a theorycrafter to be a good healer.”  Well, that’s partially true.  You don’t have to be an “active” theorycrafter to be good at any class.  You don’t have to be the one making the spreadsheet–after all, there’s only one of him.  But you have to have enough of an understanding of the numbers underlying the class to interpret what that person tells you and reflect it in your play.  This is well-accepted for DPS, but for healing, people have difficulty thinking the same way.  Just because your task varies more with context, you’re not going to be a good player by just going and doing whatever you feel like without regard to the same kinds of tools.

Mana Regen

I’m just going to pick out one piece of it today, one that arises regularly at the beginning of each expansion.  Everyone goes nuts for more mana regen.  Picking Spirit items and gemming Spirit is one thing, but talk about using Spirit flasks and Spirit food is very common.  In fact in most communities the common wisdom is that that makes more sense than using Int food/flasks.  Now, for the spoiler version: that’s wrong.  Use Int food and flasks.  But you can read below for more on both the right and wrong thought processes here.

Spirit is in no way a bad stat (for Druids at least–I can’t comment in detail on other classes but the qualitative results below will be the same).  But most people’s understanding of why they should choose it goes about this far:

“I’ve been running out of mana so far this expansion. I should get more Spirit than I already have.”

Literally–that’s it.  I’m not actually aware of a single post or bit of analysis, for Druids at least, that takes it further than this.  Now, even if we imagine for the moment that the second sentence follows from the first (and I will explain below that it emphatically does not), the first point is that this is a hopeless mode of analysis.  It has no numbers in it and no limit or conclusion.  It’s a blind guess at how to play WoW correctly.  On any DPS forum (well, on EJ), a post like that would get tossed from the thread.

 The Basic Tradeoff

Fundamentally what this is about is the strength of your heals vs. the total amount of mana you can spend over the course of a fight.  Which stat is exactly best for the strength of your heals may vary by class, but the basic framework has been made vary homogeneous–Int increases all heals by a standard amount, secondary stats tend to improve throughput by around half as much as Int, and Spirit regens the same amount for everyone (specifically, each Spi adds 0.564 MP5 during combat, on top of the 6000 base MP5 you have).  The only real class-based variance will come from the occasional ability that scales extremely strongly with Spirit (Mana Tide is perhaps the only remaining one, and even then only in a 25-man) (edit: Chris notes in the comments that the new Spirit-based Rapture may be another).

How do we evaluate how much benefit this extra mana gets us?  That’s the key question.  And the point I’m going to keep coming back to is that if all you say is “well I’m picking Spirit because I think need more mana” or “with poor gear early in the expansion, I need to build my regen,” you’re not answering it.  In fact, you’re not answering anything.  Spirit is a stat like any other–you have to measure what it gets you and compare to other stats.  We’re going to get into that below, but first, the most common stumbling block for people:

“I’m running out of mana. I have to get enough Spirit to get through the fight, and then I can worry about other stats”

If the only thing you get from this post is why this doesn’t make sense, you’ll be on your way to being a better healer.

Just a Few Numbers (nothing scary)

Every class has their “core” spells–the strong, efficient spells, usually with some kind of cooldown so they’re not they only spells you use.  For example, for Druids, it’s Wild Growth, Lifebloom, and Swiftmend.  Circle of Healing, Riptide, etc. are similar.  You want to use these spells a lot and you generally use them regularly through the whole fight.  Most importantly, they are very inelastic with respect to mana: you want to use them when you need them.  Your mana resources will determine how much is left after all the “core” spells you cast, the amount you can spend on “filler” spells.  Filler spells aren’t necessary bad–they’re key parts of your skillset (Rejuvenation, Divine Light, etc.), but it’s important to realize that spamming them heavily tends to result in inefficient healing.

Most importantly this informs how to think about Spirit.  What it buys you is more casts of filler spells over the course of the fight.  Let’s work an example: as a Druid, let’s say I used a Spirit flask, 1000 Spirit.  Over the course of, let’s say, 3 minutes of healing, that 1000 Spirit will get me just over 20000 mana.  Or, roughly enough to cast 2 Rejuvenations.  So that’s the first step.  In my head, I’ve relabeled Flask of Falling Leaves as “Flask of 2 Rejuvs.”  Suddenly it does not seem very compelling.  Depending on your class, you probably feel the same way about the “Flask of 1 extra Greater Heal” or whatever it may be.

What about an Int flask?  Well, in my pre-raid gear I have well under 10000 Int (remember that armor specialization and other % bonuses from talents should not be included here).  So 1000 Int increases all of my healing by around 10% (edit: Erdluf points out in comments that 5% is a better napkin estimate here, but as he says, the conclusions are unchanged).  Including those Wild Growths and other core spells that didn’t benefit one whit from added Spirit.  There is no world in which I’d take the Flask of 2 Rejuvs over that.

Most importantly, definitely not the “new expansion, undergeared” world.  That’s when you should be most concerned about heal strength, not regen.  When everyone else is going “I’m undergeared, I really need regen,” I want you to train yourself to think, “I’m undergeared, I really need to get my heals up to a decent size.”  Because you know what will happen?  Your core spells will start to work like they’re supposed to.  They’ll keep people up more easily and you won’t have to try to spam Rejuv or Greater Heal just to keep everything from falling apart.


If you’re running short of mana, you can either 1) devote your resources to getting more, 2) make your key heals stronger while using the inefficient ones less.  Now, which is likely to better is a question of math.  If Spirit were far, far stronger than it is now, option 1 might be clearly superior.  But it’s not, and in any case, if you haven’t tried to do any math, you really have nothing to contribute on which option is more likely to be more useful.

I think there’s one big thing causing people to be so biased towards option 1.  Option 2 ask you to change what you’re doing.  Not only to relearn your habits slightly to cut back on inefficient spells, but also in some psychological way, if you believe me that option 2 is correct, it makes running out mana “your fault.”  It’s much, much easier to believe that you ran of mana and the group wiped because you didn’t have enough mana (if only you’d had that extra 1000 Spirit, those two Rejuvs would have saved everything, right?).  But if I’m telling you that healing can be done successfully without stacking Spirit out the wazoo, and you wiped because you ran out of mana, you have to come to grips with that toughest of all realizations: you could have played better.

So if you can’t let go of the notion that your healing problems are your gear’s fault, there’s nothing I can do for you.  You’ll stack more regen and learn to get through the content (little secret: you still probably eventually succeeded due to further practice, not due to the one extra Greater Healing Wave you got to cast).  But if you’re interested in getting better, let this post break you away from the common wisdom surrounding all of this.  It’s not the size of your mana pool that matters.  It’s how you use it.

52 thoughts on “Why More Spirit is not the Answer to your Healing Problems

  1. Pingback: [Resto] Mists of Pandaria 5.0.x (Release) - Page 5 - Elitist Jerks

  2. “3 minute fight” – yes, there certainly are a lot of these short fights in this tier! if you’re going to tip the scales specifically against spirit, try harder

    “10k int” – yes, that’s how much int you would have in greens, healers are walking in with closer to 12k int in gemmed/enchanted gear, reducing the 10% number you tossed out to 7-8%

    Here is my premise: Intellect isn’t better than spirit, unless it allows you to cast fewer spells.

    If I need to cast 2 Surging Mists to heal my tank from 20%-100%, regardless of my intellect, it doesn’t matter how much intellect I have, because I’m still casting two spells and consuming the mana from both. Intellect did not help me heal more (total HP) or more efficiently (same mana consumed). In fact, in order for intellect to pay off, it needs to tip the scales such that there are multiple times where I can NOT cast a spell due to the bonus in healing.

    Spirit, on the other hand, is valuable all the time. Regardless of how efficiently I play or how many casts I cancel, spirit is always going to give me those two extra heals. Intellect requires much more “wait and let the HoT take care of it” and conscious spell management. You have to leave people not-topped-off because you need to save that heal in order for intellect to be worth it.

    In the healing game, I see two major stats: throughput and number of casts. Intellect does increase your throughput, which is good if you’re having a raw throughput problem, which is rare in today’s raiding game (since otherwise it would be holy paladins all the way down). Intellect can also help boost your number of casts, but not unless the amount of intellect you gained can materially cause you to cast fewer times. This is incredibly rare, and an almost non-existent occurrence in a new raid tier where everyone is inexperienced. In fact, I would imagine that improving one’s play (standing in less fire, timing heals better) will result in far more heals able to be cast than intellect ever could.

    You are making the classic DPS mistake of assuming that all healing is created equal. Sometimes you have to heal your tank twice, regardless of if your heals hit for 50k or 54k. You are going to have to spend the mana twice. Intellect does not help you here. Spirit does.

    • Saying that spirit helps in the big heal situation is rather disingenuous. To use a druid as an example, a single big heal (HT or Regrowth) costs as much mana as a flask of spirit would provide in approximately 2 minutes and 15 seconds. Sometimes you may need to cast two big heals in a row, but you’d get far better results in terms of mana by having made a handful of more efficient decisions earlier in the fight. Or maybe you don’t actually need two big heals, but could get away with a Swiftmend/HT, letting Efflorescence/HoTs tick them up the rest of the way. Not all healing is created equal, but due to wildly disparate spell costs, neither is all mana.

      If you’re having to spam inefficient heals it’s likely something went horribly wrong somewhere. Actually, in many cases the more efficient heals actually get a higher benefit from Int than the less efficient ones further rewarding better play. Speaking a druid, if you wanted to calculate out coefficient:mana ratio the big heals end up below all the HoTs and Swiftmend.

      The entire line of thinking and arguing is very tricky, because healing has the lowest gear dependency as performance approaches skill cap. It’s very likely that you could heal any non-gear check fight for the next few tiers (if not the whole expansion) with pre-raid gear with optimal play, assuming the tank stays geared unless there’s some massive mudflation tier to tier.

    • Arentios and Erdluf below made some good comments, but let me add:

      There are these little fallacies that crop up all the time arguing about healing that more or less allow one to argue for anything without any numbers or other analysis. “It just leads to more overhealing” is a common one. Yours here is related: “Making my heals stronger won’t make me have to cast less” (notice that that’s actually equivalent to “it’s all overhealing”). The problem is, at Erdluf put it, that’s based on feelings, not any anything quantitative, and that means it’s notoriously unreliable when dealing with things that can’t be seen with the naked eye.

      The benefit of having all heals a few % stronger probably never manifests as an obvious moment where someone would have died otherwise. Nor does the few points of extra mana. In both cases, the argument purely from trying to visualize when the benefit feels helpful is totally arbitrary. (And actually, this is part of what biases people towards regen–running out of mana is scary and very visible). It’s just like anything else–to find which stat is more useful on the margin, you have to quantify what each one is getting you.

      It’s obvious to you that doubling the size of all your heals would be a constant massive benefit, right? So why isn’t it obvious to you that increasing all your heals by 10% is also a benefit (albeit 1/10 as large)? You won’t necessarily _know_ those 1 in 10 times when you saved a cast when they happen. But they do, because you’re outputting more points of healing, and that’s why we need the numbers showing how much–so we can make decisions based on effects that aren’t obvious purely through intuition.

      • Hamlet, let’s take your example a step further. What if my heals were all multiplied by 100? Would it improve my healing done 100x? Would it reduce the number of casts to 1/100? The answer is no. It would increase my healing by maybe 50-60%, or at best 2x. We’ve seen this before: players with the Red Crystal on Ultraxion were not healing for 200% of their normal values. Your statement that 10% increased healing per spell = 10% more healing is false.

        As I said before, there exists a minimum number of spells that must be cast per encounter, regardless of intellect/spell power. I have to cast a certain number of HoTs, heals, buffs, etc to maintain my healing. If I need to cast 100 times per fight, then I need to have enough mana+regeneration to handle those 100 fixed casts. Intellect does not help me if I have to cast the spell anyway.

        I provided one example: the tank needs two heals, and one won’t do. Another: two people need healing, and they’re not stacked for an AoE heal — this situation is constantly present. Two heals are required. There are situations where increased healing per spell is not necessary, but an increase in the number of spells I can cast is necessary.

        To take the example to the extreme: would I rather have 10 heals for 10 HP each, or 1 heal for 150 HP? The single heal would clearly allow me to deal the most healing, but the multiple heals allow me greater flexibility in how to dole that healing out.

        As you start reducing the number of spells you’re capable of casting per encounter, you start limiting the types of spells and types of healing you’re able to do.

        Your statement of 10% spellpower has already been proven inaccurate, even generously you are looking at a 5% difference in healing effectiveness in the perfect scenario. My point is that we don’t heal in the perfect scenario, and there are many occasions where the net effect of that intellect will be 0%.

        If we’re gaining 1 spell per minute via spirit, that could be a 5% increase in the number of spells we cast per minute, or 2.5% at the VERY worst (and never occurring case). My argument is that intellect’s contribution comes in with a 5% benefit at max and 0% at worst. When you factor in the fact that in gems, spirit is stacked 2:1 against intellect, it becomes clear that spirit is more than capable of holding its own. In the worst case 1:1 situation, spirit and intellect are both within the same range of healing benefit.

        Your blanket statement against Spirit is not backed by the math involved. There is no math to explain how a 5% increase in healing done is superior to a 5% increase in the number of spells that can be cast, especially when you factor in fight mechanics.

        • You are pointing out a relevant factor–what percentage of healing is likely to be overheal. And you’re saying that in extreme cases (2x healing), everything is so inefficiently allocated that your average overhealing goes up. That’s fine, and you’re even right to say that some small portion of that extends to the 5% extra healing case–since it’s lumped onto the same targets already healing, it’s going to consist of _slightly_ more overhealing on average than what you’re doing so far. Ergo, 5% raw healing is not quite the same as 5% effective healing.

          There are few problems though, but to get to the point, here’s the biggest one. You’re making up numbers in order to make a handwavy argument. You say you’d rather have 5% more healing by way of extra regen than by way of extra Int. So? Everyone probably would–that’s the same total healing with some added flexibility from getting to spread it over more casts. The whole point of the post is that the comparison doesn’t shake out that way at all.

          You keep pointing out the “soft” factors that soften the benefit of adding healing power, but nowhere are you doing actual math to even find the basic estimate to start from. I did, and found that a point of Int adds a few times as much raw healing of a point of Spirit. Against that basic starting point, I’ll happily accept a comparatively imperceptible increase in overhealing.

          • Let’s look at your 3 minute example.

            1k intellect provides me with between 0-5% increased healing. Correct?

            1k spirit provides me with 20,000 mana (or close to it). Correct?

            On my first boss fight of Mogu’shan Vaults, I cast about 50 mana-costing healing spells in a 3-minute window. You can check my WoL for reference.

            The 20k mana would’ve allowed me to cast 3 more Soothing Mists, or 6% more spells. The intellect would’ve provided me with 0-5% more healing. The extra SM casts would’ve also netted me about 3 Chi points, which are worth 1% of my total mana per point (via Mana Tea) for a total of 9,000 extra mana, and an additional 1.5 Zen Spheres or a free Enveloping Mists.

            When you remember that int:spi is often a 1:2 ratio, it’s clear that spirit wins the numbers game.

            Is this example concrete enough for you?

        • Yes, that’s at least a quantitative argument I can deal with (replying here since this blog doesn’t like nesting quotes so far).

          Now correct me if something is wrong in the below, since I really don’t know much about Monks.

          Soothing Mists is your “small” heal, right? (I see it only has an 18% coefficient). So claiming you’re doing 6% more healing because you’re adding 3 tiny autoattack heals onto your 50 casts (which presumably consist largely of heals that are many times larger than Soothing Mists) is silly. If it’s anything like Nourish next to all the other Druid heals, that’s probably less than 1% added healing.

          • Soothing Mists was cast 17 times in that window, and ended up being 18% of my overall healing, or roughly 1% healing per cast. At that margin, you’re looking at ~3% healing from the extra SM casts, plus a 3-Chi heal and 9k mana from Mana Tea. And, actually, Soothing Mists is replicated at 50% strength to a second target via Jade Serpent Statue, so it could’ve been as high as 4.5% bonus healing (JSS’ SM casts were 8% of my healing during this window).

        • Agreeing with this reply, full on. People keep linking this blog post as some proof that Int>Spirit, in terms of flasks/food. I just don’t agree, as a rule. Diversity of spell selection and longevity is almost always better utilization than having stronger heals. I think Chase does a brilliant job illustrating the point, so I won’t rehash the argument.

          There are situational cases. As a holy paladin, I roll Intellect for heroic Garajal because I contribute DPS, and barring extremely unfavorable RNG, I have opportunity to fully regen. Plus, my mastery shield strength goes miles to mitigate voodoo doll damage, preventing the need to heal replicated damage. Damage prevented is of course the best regen model.

          Also willing to concede I don’t know resto druids, and Intellect may very well scale better for them. But for priest, monk and paladin, at least, I will argue Spirit>Int in almost all cases. A lot of the (very well progressed) paladins I know sacrifice socket bonuses to gem straight spirit. I don’t do this, but there is a very sound argument for doing so.

          • At Garajal there’s not even a question that you want to sacrifice regen for throughput in essentially any amount, due to the way that Spiritual Innervation works (not only is the buff based on how much raw healing you can pour out in 25s, but yes, you get all your mana back frequently anyway).

            Chase and I hashed this out in another forum a bit, but I basically told him I wanted to leave it until/if the time comes that we’ve both looked into it in proper detail. I felt that 1) I’d like to look into Monk numbers more thoroughly firstand (PTR may provide a chance for this) and 2) his arguments here were a bit fatuous–for example in the exchange you reference above he equates adding 3 Soothing Mists (a small heal) onto “about 50 heals” to be 6% added healing.

            Given the degree of response to this both (both positive and negative) I’ll probably do a followup, so I’ll try to leave detailed comments for then. Briefly, I’d want to more clearly separate out the strategic argument and the numerical one.
            1) On the numerical one I’d back it up to great depths w.r.t. Druids, but make clear that while my knowledge of other classes is incomplete, the basic heuristic holds unless people provide specific information about things that change it dramatically (e.g. Rapture).
            2) But I think even more of the focus would be on the strategic/logical discussion, where I want to get into why people find such comfort in staying as far as possible from the OOM bogeyman, while not attaching similar value to the importance of heal strength.

  3. Brilliant post, this has single handedly made me rethink my gearing process.

    As a Disc Priest Rapture does make thinks a little murkier- due to how it scales it is essentially a cheaper Power Word Shield. So, once every 12 seconds, when your PW:S pops you get 150% of spirit back as mana. As a result, the more spirit we have, the cheaper the shield will be. We’ll likely get to a point where the amount of Spirit we have increases to such as extent that Rapture gives more mana back than the initial cast of PW:S

    So, where does this leave us? Spirit ends up becoming a quasi throughput stat for Disc – the more you have, the more efficient it becomes to swap a standard heal out for the absorb of a PW:S.

    That aside I am going to have a rethink – at the moment I have been going for spirit at the cost of all other stats. A more balanced approach may well be worth a look.

    • Interesting. Since Rapture has a 12s cooldown, let’s first see the max return: 1.5 mana from one Spirit every 12s is 0.625 MP5 per Spirit, which does slightly more than double the effective regen from Spirit. So this is potentially quite significant.

      It’s tricky since PW:S is not necessarily an efficient spell to begin with, but it is reasonably comparable to a “large heal” in its mana cost and coefficient (compare Greater Heal). So with the Rapture reduction, it could easily start to become something you want to use on cooldown (well–12s Rapture cooldown, not 6s PWS cooldown).

      It’s also tricky since if you’re using it on cooldown, it’s essentially a “core” spell that scales with Int only, and the mana returns will be used to fuel _other_ filler spells. But since in that case, each point of Spirit is returning so much mana already, that it’s going to be very strong.

      The result here does seem to be that Spirit is about twice as strong for Disc as other healers–I’d have to imagine it’s the best secondary stat by a good margin. But again, all the gains still tend to go into your least efficient spells, so you have to be very careful to keep the strength of your core spells at a good level.

      • However, disc also benefits quite strongly from the intellect increase while semi-sidestepping the overhealing issue if you time your absorbs well and consider aegis. I haven’t done any math on this, but just reading this article and thinking about disc mechanics (and even holy mastery) makes me guess int would be a very appealing choice for priests.

        • Also, for priests at least, active mana regen skills are tied to int, so an int flask will be increasing the amount I get back from a shadowfiend.

          As a side note, this is part of why I really dislike the new fixed mana pool. As a disc priest you used to be able to run with less than 1k spirit in Cata and use active regeneration smartly and you’d be fine for mana, due partially to how a lot of our skills scaled off our total mana % (it also made being a blood elf awesome). Now, it feels a lot more bland when it comes to playing with gear and effective playstyle choices.

          • An int flasks will not increase the amount you get back from a shadowfiend. Shadowfiend (and Hymn) are both based on max mana, which as you point out yourself is no longer increased by int.

  4. Seeming that in gems we get 2 spirit to 1 int that seems the place for ‘regen’ to be added.

    I think the major point is to stop making crappy 1-1 trades of spirit for int (food/flask) or even 1.5-1 trades.

    For druids Hamlet do you feel that 1 int = 2 mastery based on acquiring mastery at the 2/1 ratio? Or is the conversion not that simple?

    • Yup, 2:1 Spirit from gems is much more reasonable. I hope this post makes people reevaluate how they made they tradeoff even there, but there it’s quite a bit more balanced. When you do want Spirit, that’s the place to get it (or by reforging away from weak secondary stats, such as excess haste for Druids).

      Int and mastery are both just flat healing increases for Resto now hat Harmony is so easy to keep up. Can go into this more on EJ thread or examine TreeCalcs for your setup, but they’re looking to be pretty similar in value right now.

  5. DPS spells tend to have (approximately) zero base damage. At 10k SP, 20% of Wrath’s damage is from base damage, and 80% is from SP. Adding 1k SP is ~8% boost.

    Heals tend to have a lot of base heal. At 10k SP, most heal numbers are about 50% base and 50% SP. For instance a Wild Growth tick is 990 + 9.2% of SP. For most heals, 1K of SP is only a 5% boost.

    Having said that, very good post. The bottom line argument still stands. 2 rejuvs every three minutes is probably less than 1% increased healing. 5% increased healing easily wins.

    Chase, your argument is based on feelings, not numbers. If the tank always has to be at 100%, there are times when a 54k heal will get you to 100% and a 50k heal won’t. If there are times when the tank gets hit before he reaches 100%, then having your first cast heal for 4k extra will improve his odds of surviving. Besides, if you’re already at the point where your tank is always immediately restored to 100% health, you clearly don’t need more spirit.

    • Very nice observation, I was always curious about how high was Int/SP scaling relative to base damage/healing but never cared enough to check the numbers.

  6. While Chase has somewhat a point (regen is not entirely assimilable as throughput since healing is not DPS), I think that it’s an error to assume that healing is about “having people topped off ASAP”. This is a bad habit carried from past tiers where mana was abundant and there was a high “burst” healing demand and also a mistake I noticed both in myself and some of my mates. People need to realize that there’s absolutely no point in topping off people when the next damage burst is coming in, say, 10+ seconds when you could use efficient heals to slowly top off the raid. Regarding tank healing, most of the time you don’t need to have the tanks at 100% HP because they can survive several autos without heals and if they are not topped off, overhealing is 0 and therefore, over time, Intellect *does* reduce the number of heals you need.

    Also, usually tanks don’t die because they didn’t get a heal in the last 1.37 seconds. It’s usually several seconds of high damage and/or low healing and in this respect Intellect increases tank survivability by more than Spirit (in the short run scenario, Spirit increases tank survivability by zero), by increasing the amount of healing relative to incoming damage. It’s true that in some situations Spirit allows you to use less efficient heals over more efficient ones, but 1) that tradeoff can be quantified, 2) if you need to resort to inefficient heals anyway, Int will heal for more and 3) more efficient healing (since you have more Int) in noncritical situations will also save up heals and mana and allow you to resort to less efficient heals more often. In the current raid philosophy and specially in 25-man, the best approach to modeling healing is not “the tank HP is a, my heals heal for b, so I’m casting ceiling(a/b) heals and therefore bigger heals will usually go to waste since they will not decrease that number”. Instead, a more realistic approach is “the tank will receive d damage during the danger period, so we are casting ceiling((d + a)/b) heals over the period” and usually you need to take into account that d + a is significantly greater than b and even a 7-8% difference matters when you are casting about ~10 heals over some period of time.

    @Chris, for Disc Priests, it’s still the same thing – only that regeneration doesnt equal a constant*spirit, but constant*spirit + (1.5 spirit)/(~13 seconds) = (constant + 1.5/(~13 seconds))*spirit = (another constant)*Spirit. Same analysis as Hamlet’s, somewhat different numbers.

  7. Something I struggled with back when you first raised this on twitter, and I began closely examining my play in heroics, is that you must consciously choose to integrate a throughput based approach to conservation. As people have pointed out, you cannot use Int to scale if you’re a compulsive chain caster (spirit won’t save you either, but Int will show a little worse). As I was told when I first learned the resto druid, you have to trust your hots (for other classes its trust your efficient heals).

    Wrath was a long time ago and nobody instagibs anymore, healing is now less about reacting instantly as about continuously refining your 10 second plan, knowing what you’re going to do with your next couple casts and how much time you can sit doing nothing or treading water with a small cheap heal that is mostly covered by regen.

    I had to go so far as to shift around my binds so that my inefficient spells require more effort to reach, making me default towards my efficient “core” choices. As a shaman that means riptide on cooldown (in 5mans I opt against the glyph) and if I feel the need to spam use healing wave.

    Shaman also need to remember that crit is a regen stat as well as a throughput stat, so if we really feel like we need more regen we should probably turn to crit instead of spirit. It’ll continue to boost throughput while providing some gains in regen.

    I’ve also learned to be comfortable at low mana, to learn to pace my casts so that while I’m waiting for tide to come off CD I can still keep people alive even if they might be sweating their 45% HP. This requires knowing the fight and being able to predict damage and triage appropriately, something it seems most healers I observe utterly cannot do. The model didn’t quite get there in Cata, but it appears to be firmly in place with MoP. The only time players are really expected to be at 100% is out of combat.

    • RE: Crit as a regen stat:

      The rate of gain of regen from crit is very weak per point, so it scales somewhat poorly. It’s non-negligible, but the value comes primarily from the throughput increase. I did some analysis on the “overall” value of Crit compared to int and spirit (which I talk about at more length in a different comment) here;

  8. This opened a bit my eyes on my bias about spirit. Thanks for bringing back a bit of sanity into my world, gave me something to think about for the next weeks.
    Altough I have a few things I’d like to add.
    The bottom line of what you’re saying is, in my opinion, that the Spirit>MP5 conversion is just not good enough, and this is what creates the big difference between Intellect vs Spirit for healers.
    On this regard I wonder why the differences in Intellect/Spirit that you can find in gems (usually intellect ones give half the amount of spirit, like 160 vs 320 or 80 vs 160) are not also present in Flasks and food. Why not 2000 spirit for flasks? Why not 600 spirit for food? I wonder…
    Altough even in such a situation it would be a “Flask of the 4 Rejuvs”, which wouldn’t make such a huge difference in the end I guess.

    Two more things. I still think that while stacking for spirit is wrong and you showed why, there should still be a generic minimum spirit threshold, as much as this value changes according to different variables from player to player, or are you saying a healer should go raiding with something like 3k Spirit buffed or even less?

    And I’d like to put a bit more emphasys on the differences between classes. For instance Shaman compared to Druid. If I’m in raid with 2 druid healers they will benefit from my spirit-boosted Mana Tide (from my manatide in general), but I won’t benefit from their Innervate. I think this potentially makes mana slightly less of an issue for Druids than it is for Shamans? And let’s not talk about Telluric Current because as useful as it is, its manareturns are very small and you can’t really afford to make use of it anytime you want.
    I agree that even in this case spirit is probably not the solution, but at the same time I think it’s probably slightly more valuable for Shamans than it is for Druids.

    I’d love to read other opinions on these issues.

    • We need to remember Spirit is a secondary stat. Its rather odd actually that we even have Spirit flasks. I don’t think Blizz ever wants to go to a world where Spirit is your top stat. They just want a world like Jay is outlining here, where you need to think about your spell use over the entire fight and if it is *really* necessary to top everyone off *right now*.

  9. Isn’t all of the math used in this discussion vastly over simplistic? Since healing works by negating a deficit in HP, changing the strength of any heal would affect spell use, and therefore mana use.

    How are you coming to your 1000 int = 10% healing conclusion? Are you finding the average benefit of intellect to each spell? Are you finding the benefits of each spell then assigning weights to those values based on their use? I really don’t see enough math by anyone ITT to validate these postulations.

    • I think more than focusing on trying to put an absolute “end” word in the challenge of Intellect Vs Spirit, which I think would be quite more complicated for the huge amount of involved variables, it’s more interesting to focus on intellect vs spirit for Flask and Food slots, which probably was the main intent of the original post.

      Seeing the 1:1 rate of Spirit vs Intellect in those slots and seeing how the Spirit>MP5 conversion is not particularly efficient, I think it’s quite likely to agree that in those slots, as a general rule, it’s probably going to be more efficient to get Intellect instead of Spirit, getting this one in other more efficient slots like Sockets and Reforge.

      I think the little math provided, as oversimplified as it may be, does well in showing this specific point, imho.
      It’s different from saying that Intellect >>> Spirit anytime, anywhere, for any class, in any fight, for any spell in any slot. Which may as well be the case but it would be way more complicated to test, analyze and prove.

    • I’ve felt what Jay has for a long time – DPS and even tanks really seem to accept numbers, but healers have resisted it strongly. It is still a numbers game – you can still reason mathematically about the healing game. The community has just resisted really hard analysis for so long – partially because, tbh, it hasn’t really mattered as much as reaction time since the first few weeks of 4.0 and before that, BC. Healers in Wrath and most of Cata didn’t have to be min-maxxed…they just had to be fast.

  10. Hi again,

    This whole discussion is very interesting, and I like the points made about;
    1) The fact that there has to be some kind of numerical analysis in order to talk about this
    2) That most of the numerical analysis is very simplistic, inherently faulty, and therefore must be taken with a huge mountain of salt
    3) That one can optimise here by focusing on gaining more spirit where it’s weighted more strongly in the item budget (a good example being in gemming where the item budget ratio is 1:2)

    An interesting point of note is that your analysis focuses on the **mana saving** and an increase in **intellect** would bring. My original idea a few months ago was to do the reverse; quantify the **throughput increase** that **spirit** is worth. The results of that experiment were mixed, but I have a measurement framework for it which is a measure of what Derevka calls “pseudo-throughput”, which I’ve labelled Healing Effective Power.

    The results of my analysis on the differences between Crit, Int, and Spirit for Shaman were that spirit is very powerful in the low-regen limit for simple direct healing rotations; so much so that it’s **by that measure** twice as powerful as Intellect. Crit was on approximately the same level as Intellect.

    While the numerical analysis is admittedly simplistic, I would suggest that in the limit that you cannot double (roughly) your heal strength by increasing your intellect, then spirit is certainly a powerful stat to consider.

    I openly welcome any comments on my work or requests for conversation on the subject, and I’ll keep an eye here for any developments. By no means do I think that your original analysis is **bad**, just that I think there is more to it.

    • I do actually think that throughput by increasing regen is the easiest way to look at things. Talking about how stronger heals reduce how much you have to cast is a helpful explanation for people who aren’t getting the concept of why don’t focus on pure regen. In practice they should be very equivalent of course, and the marginal healing done from added points of mana is a much easier metric to work with.

      I’m definitely interested in your wildly varying results here. I tried skimming your posts, but best I can tell, you outline the framework in the post but all the computation is done in a sim or something where you only show the output. Not that I’m likely to dig through the code in detail, but some sort of basic understanding of how it’s all adding up would help account for the differences.

      Here’s the thread where my Druid spreadsheet lives: http://elitistjerks.com/f73/t116144-resto_treecalcs/p9/
      But assuming you’re not going to go digging through the entire thing, here’s a basic outline of what I’m trying to do:
      1) basic info (work out average heal size, cast time, mana cost of all spells based on current gear/setup, and scaling factors for each individual spell with each stat),
      2) put together a breakdown that at least generally approximates real-world spell usage–it’s essentially a weighted average. Based on using the best spells roughly on cooldown, being able to estimate how often various procs proc, and then a not-very rigorous allocation of the filler spells based on what seems to be common use. This is a way of getting pretty decent consolidated stat weights. The concept is basically “if you use spells in this proportion, then on average you do X healing and burn Y mana per unit time.”
      3) In theory, the “effective throughput” of added regen, then, is how that healing output would change if you adjusted spell usage to cancel out a delta in MP5. It’s essentially a partial derivative (my college degree was physics too, incidentally). In practice, I think a slightly simplistic but quite sound approach is to assume the variation in spell usage takes place only in filler spells. This is what’s kind of described in the post–the super-strong cooldown-based spells you’re generally using as often as you can anyway, it’s really the spammable filler spells that show the impact of added mana.

      So the result is that it values added (or removed) MP5 by changing “nothing” casts into “Rejuv” casts or vice versa. In other words, Spirit is valued by how it translates into Rejuv (and I think this is actually generous, since there are filler spells besides Rejuv that are less efficient).

      The reason I spell this all out is that your approach based on “healing phrases” is totally different, and I want to see what implicit assumptions it makes, when I can really understand how it works, that are totally different from what I’m doing. I’ll admit I’m skeptical of a marginal Spirit value that’s as high as what you say, because of the similarities in mechanics across the classes. But I want to see if I can get a ballpark accounting of where the mana is going in your model to produce that result.

  11. Pingback: Disc & Spirit | Reality AFK: Priesting Downunder

  12. Pingback: … I’m not dead! I feel happy! | healiocentric

  13. Pingback: Disc Priest - Mists of Pandaria - Elitist Jerks

  14. Pingback: Raid Readiness in MoP: A Healer’s View | Heals n Heels

  15. My brother and I are both healers in a semi-casual guild. I Have been trying to get this idea of int>spirit when its a 1:1 trade-off.

    He seems to not understand the idea of triage, and keeps claiming that you always benefit from topping people off. For example, “if people are at 80% instead of 100% and take a 60% hit, they will be at 20% instead of 40%, leaving less room for them to make mistakes in the fight.” It is an interesting point, because most of the math in this article is based around your raid not taking extra, unforeseen damage. Is there any way to take this into account?

    Also, he keeps coming back to the point that in some situations you must use 2 heals on someone, and in these situations spirit would become more beneficial because that extra healing you gained from your intellect is essentially wasted.

    Does anyone have any suggestions as to how to explain to my brother why intellect is better, or is it just a matter of him not understanding what triage is all about?

  16. Pingback: The great food buff debate | Restokin

  17. I’ve been wrestling with the Int / Spi valuation ever since one of our guildies posted this thread. I realized it wasn’t ever something I was going to “feel” my way to a resolution on, so I figured I’d better jump right in for some maths.

    Currently, for priests, the calculations at lower gear levels (barely LFR ready) seem to favor spirit. I’m using two major sources: for spell coefficients, Valen’s priest calculator linked on EJ, and for mana regen Derevka’s in combat regen model.

    So, my current numbers are 8615 Int and 5414 Spi. (Too much time farming, not enough dungeon-ing, clearly.) For the Int calculations, I looked at our Prayer of Mending spell since I think that’s one of the only “on cooldown” spells we really use.

    A 1000 Int flask boosts my Prayer of Mending tick from 13,797 to 14,368. That’s an increase of about 4%. A 1000 Spi flask boosts my in-combat regen from 9057 mp5 to 9621 mp5. That’s an increase of about 6%. So assuming that I want more of both throughput and regen, I’m going to get more more mana regen as a percentage than throughput given equal boosts to the relevant stats.

    At some point I’m sure two things will change: 1, I’ll get more spirit on my gear relative to int (which will change the percentage calculations) and 2, I’ll start to feel like I have “enough” mana (which will change the “want” part of the valuation to favor throughput). But for now, I can’t point to the pro-spirit mentality as objectively wrong given these numbers.

    • You’re on the right line of thinking, but this sort of analysis (compare % MP5 to % tick size of one spell) isn’t going to work–it just ignores too many factors to give a result you can rely on for even a meaningful rough guideline.

      The first reason is something outlined in the post–an increase to available mana does not result in casting all your spells proportionally more often. The mana gain will accrue mostly into your filler spells, not into the strong cooldown spells that you can cast regularly even with less mana (this is assuming that you play correctly when you’re mana-limited, and cut filler spells rather than strong core spells).

      The second reason is probably much more important though. Look at the major mana resources you have over the course of a fight as Holy:
      1) Base mana: 300,000 (over a 5min fight, this is around 5000 MP5 worth of mana).
      2) Base regen: 6000 MP5.
      3) Spirit regen: ~3000 with the 5400 Spirit you have.
      4) Shadowfiend: around 72000 mana (is 8 attacks about right?) every 3 minutes: ~2000 MP5.

      This is a point I should probably have emphasized more in the OP: just how small a portion if your available mana resources are actually determined by Spirit.

      As a detail, I’m unsure of how the 1000 Int is only giving you 4% to PoM–maybe show me how you got those numbers? Based on the coefficient and base amount of PoM, 1000 Int on top of 8600 already should be over 5%.


      Like I said, you’re on the right process with trying to evaluate stats with basic estimates like this, but many hidden assumptions in the quick and dirty comparison you’re doing are strongly skewing things in favor of Spirit.

  18. Yeah, I don’t see this as an analysis that holds up in anything like a real healing environment. Similarly, though, I don’t really play Holy with the “strong spell on CD plus filler spells” model you outline. Aside from PoM, I can’t think of another spell that I would cast that way barring specific raid mechanics. Any additional mana I gain goes into what is (in aggregate) the bulk of my healing: PoH, CoH, Renew, Greater Heal, Flash Heal, Holy Word:X, Cascade. (According to a recent parse, about those spells make up about 49% of it; the remaining is split between 20% PoM, 20% mastery-based bonus healing, and 10% random other stuff… /sigh for Lightwell.)

    But if we go with the definition of “filler” spells as “inefficient spells you cast when the stuff hits the fan” then PoH or HW: Sanctuary might be a good gauge. Those are 4.6% (13,800) and 6.4% (19,200) of base mana, respectively. So if my +1000 spirit is giving me about 600 more mp5, I can cast another Sanctuary for every 2.66 minutes the fight goes on. (PoH is more forgiving at 2 mins.)

    I agree that that’s not particularly compelling. There’s no denying how small the spirit gains look compared to how much regen is available without it. But what surprised me when I did the calculations was how small the Int gains seemed as well — 600 mp5 isn’t a lot, but neither is 600 to 1000 more healing per spell.

    We could stop at “4% on everything” but it seemed to me that I should try to quantify that similarly to how you looked at spirit/mp5. So trying to translate that gain into tangible benefit, I essentially get a “free” Heal for every 24 Heals I cast. (I’d give you a Sanctuary comparison to try to keep apples to apples here, but the coefficients completely elude me.) It’s kind of amusing to think about it in terms of total raid member health: I’d have to cast 300 Heals to have a “free” alive person — and that’s assuming they’re as poorly geared as I am.

    So while I’m not going to rush out for one free HW:Sanctuary every 2.6 minutes, neither is the “get your 25th Heal free” pitch going to induce me to buy. ;)

    For the numbers, I got these from Valen’s priest calculator (and 4% isn’t far off from Erdluf’s estimated 5%): http://www.temerityofwindrunner.com/pri … -calc.html. The calculator does seem to estimate somewhere about half a percent lower than the (base healing + X%SP) tooltips, but I’m assuming I’m the ignorant one here and that those numbers are correct.

    The numbers hold up across most spells:
    PoH = 16214 to 16885 / 4.13%
    Heal = 24773 to 25797 / 4.13%
    Renew = 6263 to 6522 / 4.13%

    It ended up boiling down to fairly minimal increases either way, so looking at the benefit of a stat relative to its current effect, while quick and dirty, seemed like one way to distill these types of marginal gains.

    • Yes, that is the general metric to try to keep in mind (relative stat value). Unfortunately I can’t see the backend of that calc. I won’t get into that nitty gritty too much on that % heal increase, can revisit that if I ever do a real Priest spreadsheet. This 4-5% ballpark is fine. Note that Spirit already adds somewhat less than that to your total mana resources.

      I think you’re still discounting the variation between more and less efficient spells, for one. CoH is extremely efficient and on a 10s cooldown, basically the same spell as WG. I can run rough coefficient/cost comparison on the others when I have some time, but there’s certain to be wide variation; that’s just the general design in WoW. And the variation means you have a subset of spells to weigh towards in lower mana situations–essentially by definition, if increasing your mana resources increases your usage of all spells proportionally, then you’re not putting any effort into playing efficiently around the mana constraint.

      Your bigger point though is how it’s hard to “see” the benefit of an added throughput increase, and the thought process around all this is the more interesting discussion. It’s true that it’s hard to see–this kind of bogs down all healing discussions, but marginal benefits of any kind are not very visible to the naked eye. I want people to not only be informed by the numbers (which you’re already trying to do), but really logic out how they apply to your real-world scenarios. For example, 5% throughput is not merely a “every 20th heal is free.” It’s telling that you even view throughput by trying to mentally equate to saved mana. Throughput makes you survive situations you’d otherwise fail–this is literally the only reason that we healers gear up in the first place!

      Healing is not DPS. The goal is not to trade off short and long-term increases for the highest total over the course of a fight. It’s to never, over all the ups and downs during an encounter, have a spike of activity that overwhelms you and causes a person to die. Increasing the _maximum_ intensity you can handle–the function of throughput–is a critical way to progress. Increasing your resource of available healing over the course of the fight gives you more cushion in various ways, but can always be substituted by learning to hug more closely to the curve of what’s required, and play more efficiently.

      Regen is comforting and lets you throw out heals more more uniformly over the course of a fight, giving a large leeway even when not strictly needed. But it comes at the cost of limiting yourself. Remember, we’ve seen that even at very conservative estimates, Int adds roughly as much as Spirit even in the long term. So why deny yourself the ability to actually scale up the set of situations you can handle without collapsing, with only the requirement of having to put more effort into distributing your heals efficiently over the course of a fight? Int is more flexibility, choice, room for skill, and with it, the ability to be as effective as possible.

  19. Hello all,

    I’ve recently come back to Druid healing after playing two expansions as Feral, and I must say, I fell into Spirit Mania. On fights such as Feng, and even Stone Guardians (25m) I found myself lacking in mana efficiency. So, clearly the first thought was, “MORE SPIRIT!” After reading through countless threads, I’ve slowly wrapped my head around the idea of healing smarter rather than healing in surplus.

    Just wanted to leave a little post saying I appreciate all the debating and fact-finding that goes on in both this thread and on EJ. For those lazy people like me, you number crunchers are our guru’s. Lead on! :D

  20. There is one thing about throughput that most people dont calculate, or forget.

    The spellpower coefficients of healspells are low. So gaining int is not worth much. actually, all /most secondary stats are worth more in throughput than int in general. I am sure, that restoshamans who socket int over crit haste mastery, are bad shamans. I know that paladins that socket int over haste /mastery, are bad paladins.

    And The most simple thing in the world: as long as u dont constantly use the most expensive and highest hps spell u have, when needed, your best throughput stat is spirit.

  21. A very interesting post – from my experience it would apply to pretty much every healing class (though I only have serious experience on a priest, druid and shaman).
    I never understood people’s spirit craze and have always found that playing better (specifically, paying attention, avoiding pointless overheals etc) will net me much more mana than gearing for regen. But I think you hit the nail on the head when you said people don’t want to take responsibility, basically.

  22. Hello,

    I agree with this post and the idea that people feel more comfortable with the idea of knowing they can cast more heals.
    I think overhealing should be strongly considered, because in the example of having to cast 2 HT to top your tank probably most of the second cast would be overheal and the ticks of WG, LB, Reju, SM, etc.. could do the work, and having more INT will top the tank faster without overhealing with 1 more “free” cast from spirit.
    Hamlet: I have been trying to find some kind of guide on how much spirit is “recommended” based on ilvl (since now everything seems to be about that)… I currently use Mr.Robot as a help and with 174 ilvl I have aprox:
    16k INT
    7.5k Spirit
    23.4 SP
    23.60% Mastery
    12.5 haste cap

    I do believe casting efficiently is just another way to play and for sure is a better way to play… but I have no idea how much spirit should i get or what would be an appropriate range.. i read that around 6-6.5k but I am not sure, can you suggest a range considering your theory?

    Thank you!

  23. Pingback: [Resto] Mists of Pandaria 5.2 - Page 33 - Elitist Jerks

  24. Pingback: Healing Theory Part 3: Mana and other Resources | It's Dangerous to Go Alone

  25. Pingback: MoP In Review, Part 2: Mana Matters | healiocentric

  26. I know this is an old post but I thought I would comment to give it abit of an update. It’s now going into Aug 2014 past 5.4 MOP. I can tell you as a Resto Shaman I would never take Int over Spirit. The current content is to mobile and fast pace at least in LFR’s to prioritize int over spirit!. We are having to heal to many targets rapidly with burst heals needed constantly. Manna usage is insane and requires extreme rapid manna regen. By the time we try to get those big deep int boosted heals off that your talking about the raid would be dead. Often times the Druid hot’s, shaman chain heals with riptide, etc rule the day. This requires more high regen than it does high throughput. Yes there are some boss fights that seem to take for ever and in those cases I can see where an int boost would be nice but we now live in a erra where the trash mobs here are often stronger than some of the old world bosses. I would rather risk healing a little inefficient at times than totally running out of manna and people start to die including me.

Leave a Reply

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