Healing Theory, Part 1: Principles

In my last post on healing, I outlined a rough mathematical argument to show how Spirit should not be valued as strongly as common wisdom indicated. The bulk of the criticisms centered on the fact that I was generalizing far too much about other classes without digging into their mechanics to the same extent I have with Druids. And while responding to this comprehensively would require a significant project of theorycrafting other classes more (which, as an aside, is still something I want to get into in MoP), I took something else away from that whole discussion. Which is that, theorycraft aside, there just isn’t much established foundation on the basic logic and philosophy of healer strategy. When talking about questions like regen and throughput, the math arguments tend to be sort of shallow because there isn’t even a mutual understanding of what we’re evaluating and how. So I’ll go as far as I can in this post, and continue to follow up as needed, about what should be going through our minds when we click the green boxes.

Healing Meters: 2 comments

I’m going to present two seemingly contradictory arguments here, each of which (hopefully) has no obvious flaws. Think about how you might reconcile them with each other while you read the rest of this piece, and going forward after that.

1) The only relevant healing “meter” is the one that counts how many people died during the encounter. With DPS, every added point of damage done affects how quickly the fight ends, which has an actual impact on your likelihood of victory. But with healing, your players took X total damage over the course of the fight, and your healers healed it all off. As long as nobody’s HP touched 0 between point A and point B, it matters very little exactly how you got there. Unlike the DPS meter with is continuous gradations of success, the healing team reaches a very binary result. At the end of the day it either succeeded or failed. In fact, the realization that the healing team can do almost nothing to change the total amount of healing they do over the course of a fight reveals the futility of healing meters–how can they reveal who’s contributing more to overall success when they merely show how good you are at taking little pieces of the same pie from others in your team?

2) People die when the healing output is momentarily overwhelmed and too much of an HP deficit (e.g. too much unhealed damage) accumulates at one point in time. The best way to avoid the chance of this is to heal off each piece of incoming damage quickly and fully. The more you can do this, the more you can keep the raid in a stable state where you’re ready to address the next bit of damage just as sharply. In general, the amount of effective points of healing a person does shows how often they’re the first to whack each particular mole, which shows how proficient they are in addressing incoming damage quickly. The more people you collect together who are individually good at doing this, the more responsive of a healing team you are likely to have on the whole, and the less often something will slip past all of them.

Healing is Different from DPS

Well, yes, it is. Usually this sentiment is merely meant to say, “the meters aren’t everything,” which I touched on just now. I want to dig more into a difference that permeates the actual playstyle, and I’ll do it slightly backwards by talking about what mistakes mean in each context.

For a DPS player, every deviation from perfection, no matter how small, is added together and tallied into the bottom line at the end of the fight. Any moment you didn’t execute to the razor’s edge can’t be recovered from–it’s damage lost even if you do the rest of the fight perfectly. Good DPS players keep these losses to a minumum even in the presence of complex encounter mechanics that they have to weave into their rotation.

For healers it’s quiet the opposite. Any error that doesn’t cause a death, not only can be recovered from, but almost certainly will be irrelevant a few seconds later. Once you square away any resulting instability from a spot of bad healing and get people to comfortable HP again, the relevance of the mistake is almost nil. At worst, you use more resources than necessary (mana or cooldowns), reducing your cushion for further imperfections later in the fight. Essentially your whole job is to play for some minutes straight without any single failure of attention so egregious that it allows a death (or minor error combined with perfect storm of bad luck that causes a death). Don’t get me wrong, on hard encounters it’s quite a lot of work to juggle all the things that require your attention without dropping one. But that’s what your game is–just do whatever you need to do in each passing moment to not let anything crash to the ground. Because if nothing crashes, at the end you still take a bow and nobody notices the difference.

Solo Healing

One thing I can’t stress enough is that you should try healing solo, in an environment that’s at least challenging enough that you have to think about your spell usage. Unless you’ve been playing for more than a few years, you might not ever have done this, though no fault of your own. The newly-instituted Challenge Modes are the first content in quite a while that’s meant to challenge a solo healer. The last time was the “heroic” 5-mans back in TBC and WLK (ironically, now the name of the easy 5-mans, but they were initially designed to be hard, before the advent of the LFD system). I suggest you do some Challenge Modes, regardless of whether your personal skill level has you achieving Bronze, Silver, or Gold scores.

The benefit of healing solo is not merely that you can’t rely on others to cover your gaps (also, it’s no longer ambiguous who’s at fault for healing errors), although that is good training. It’s that you can figure out how to address each incoming bit of damage on your own terms, using the most efficient tool that still is sufficiently safe. If you’ve been too long used to raid healing with its widespread obsession with meters, it might almost be a shock to realize that your goal isn’t to figure out how to slam every possible target up to 100% HP before the guy next to you does. And in fact, you’re going to have difficulty doing well in challenge modes until you make that shift, because healing that way is not only terribly inefficient, but tends not to prioritize the most important threats.

Put yourself in a situation where every spell you cast is put to the test of whether it’s really best one. Because if some other spell reduces the change of wipes you’ll start using it to stop having to reset the zone, or if some other spell uses less mana you’ll start using so you don’t waste time drinking, or if some spell isn’t really necessary at all you’ll DPS for a few seconds instead because it has a visible, measurable impact on your goal. And finally, if you need any more convincing, remember that a few challenge modes a week are by far the quickest way to get your VP up to cap.

Principles

I’m going to try to sum up in few basic guidelines to keep in mind. I think these do a better job of orienting yourself in the right state of mind than the common platitudes of maximizing effective healing, minimizing overhealing, etc.

1. Anticipate Damage. Much has changed in healing since vanilla raiding in 2005, but one thing has been constant: the hallmark of a strong healer is anticipating damage before it happens. You simply cannot transition from being a good healer to a great one if your attention is limited to the damage people have already taken. It goes from basic considerations like knowing instantly when somebody has aggro so you can start to react before the HP bar drops, to more complex tasks like knowing when certain boss abilities are coming off cooldown so you can save the healing spells that “match.” And beyond that, to having such a full understanding of an entire encounter that you know what damage patterns are likely to be coming for many seconds in advance.

2. Be Surgical in Spell Selection. This is a refinement of the usual “low overhealing” criterion. The underlying principle is to deal with each incoming damage event in way that preserves the maximum resources for future events. Resources are typically both mana and cooldowns. The goal of achieving the best possible matching of healing tools to incoming damage applies at many skill levels. Lower in the learning curve you might simply be looking for the lowest mana cost spell to address a given situation. But at more advanced levels, you might have a sense of which cooldowns are most valuable in the upcoming seconds (this requires Anticipating Damage) and weigh all of those factors together with mana efficiency and choosing the heal with the right “shape” for the damage pattern. Point is, finding the heal that minimizes overhealing is far too elementary of a guidelines for spell usage. The better underlying principle is to use a spell that solves the current problem while leaving maximum options available for subsequent events.

3. Consider the Entire Situation. Your life as a healer does not solely reside in the group HP bars and how you fill them up. Positioning, utility and personal survival are key. This sounds trite or obvious, but the point of including it here is not only that these things are just as important as your actual healing function, but that you have to integrate them into your healing strategy. You can’t plan for how you’re going to heal off a difficult boss ability unless you account for your own movement. You can’t make good use of support spells unless you know what’s going on across the raid (other than HP levels), and unless you have the GCD to cast them. And finally, a point that is woefully underserved by having only one sentence in an entire post on healing, you can’t heal anyone when you’re dead on the ground. Especially because healers are not tied to their bottom line the way DPS are, you can and should look for any opportunity to pay attention to these non-healing related goals.

Conclusion

All three of the above are in service of an underlying philosophy: an encounter is a puzzle and you have a limited set of tools and resources. Your goal is to, as closely as possible, map your available spells onto all of the events that happen over the course of the entire encounter in the way that most tightly fits. Doing it perfectly is impossible since you don’t know in advance exactly how things will play out (this is the fun part, of course). But you know a lot, and the ability to leverage what you know about an entire encounter into the best spell selection moment-by-moment is where all the intricacies of healing lie.

The point of this post was to outline how I think about healing. It doesn’t precisely address the discussion surrounding Int vs. Spirit post, but I hope it serves as a better starting point for discussion on a part of WoW that’s gone too long without more solid strategic development. The overarching set of guiding principles should not only pave the way for more detailed looks at issues like regen/throughput tradeoffs, but should preface more analysis of all healing classes that I hope to get into when 5.2 arrives on PTR.

8 thoughts on “Healing Theory, Part 1: Principles

  1. This was a Good Post.

    The major difference between healing solo and healing with a team is the idea of being a ‘role’ player. In a team healing environment, it’s common to coordinate cooldowns and healing mechanics to overcome specific obstacles (Anticipating Healing). One of the most common roles is the ‘AoE healing cooldown’ guy.

    If a druid is always tapped to use Tranquility to handle a big AoE attack, the meters are going to look quite favorable for the druid that simply pushed one (very effective) button. The counter example would be a mistweaver monk coordinating a Renewing Mist / Thunder Focus Tea / Uplift combo window up to 30 seconds in advance to handle incoming damage. Both might have done the same amount of healing, but one required significantly more effort and experience.

    Healing meters aren’t capable of determining between “easy” healing (obviously telegraphed AoE attacks with no healing competition) and “difficult” healing (heavy random damage to random targets at random times). Role players that specialize in AoE are going to appear quite powerful on the healing meters, which may or may not represent their skill level. Putting them into unfamiliar situations that are outside of their comfort zone is usually the best way to find out how skilled they actually are. Very difficult 5-man dungeons tend to provide this style of environment.

  2. This is an excellent post. I’ve yet to get into challenge modes myself but am really looking forward to them. I am one of those players (playing since vanilla, raiding since TBC) who grew up ‘healing alone’ and THEN learned to heal with someone.

    Viewing the encounter as a puzzle is a great way to express why I enjoy healing so much (still).

    So, in short, I agree with this. A++++ would read again.

  3. Excellent post, really outlines the issues that make theorycrafting healing far more subtle and complex than DPS. I personally look to healing theorycraft not as a puzzle with a single solution, but a way of understanding the exact shape of all the pieces I have so I can fit them together more quickly on the fly.

    The one thing about healing alone that I’d add is that not only does it mean you can’t blame anyone else for your failures, which is a very powerful learning environment but it also means you can see each individual element you contribute. Its very difficult to know just how good one way of solving a problem is versus another in a raid environment because you’ve got anywhere from 1 to 4 other healers sticking their oar in the water at the same time. In any solo healing environment you see exactly how far every spell moves someone’s health bar and you develop the intuitive feel necessary to make good decisions quickly under pressure in a raid setting.

    • Yeah, a lot of this is a lead in to figuring out how to more more healing theorycraft. The Resto spreadsheet is set up mostly like a DPS sheet, but it is used differently–more of the value lies in seeing properties of individual spells and how they compare to each other. It’s possible that that’s the limit of the hard math approach with healing, so then the question is figuring out what else to do. Because there are still a lot of questions–spell usage, regen/throughput, and then the really big issue which is class balance. These are still factual things that good theorycraft should be able to inform people about, but I think there has to be more of a solid underlying strategic picture first.

      And yes, that’s a good point about healing solo. You learn what all your spells actually do, so you can start to develop a sense of what you can expect out of each one when you cast it.

  4. I really like the encounters-as-puzzles analogy. I also like the analytical approach to what healers are really doing.

    Argument (1) is usually true, but not always. Sometimes it’s preferable to sacrifice one person to keep another alive, especially if the probable alternative is that both die.

    What a healer is really doing is minimizing P(wipe), not minimizing P(player deaths=0).

    Argument (2) is true on farm raids, but not progression. And on farm raids, I’d bet many raids bring too many healers. That leads to whack-a-mole contests, but isn’t worth as much theory-crafting IMHO.

    Granted, progression bosses have low-HPS phases too. But usually the progression healing challenges are new/botched mechanics and high-HPS phases.

  5. Pingback: Healing Theory, Part 2: A Tour of Your Character Sheet | It's Dangerous to Go Alone

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

  7. Pingback: Why More Spirit is not the Answer to your Healing Problems | It's Dangerous to Go Alone

Leave a Reply

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

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

Current day month ye@r *