Frostheim's work to compare DPS specs in raid content has always interested me, and for a while, I've been toying with the idea of doing a healers' version. However, healing's always been a lot tougher to analyze from meters than DPS. The whole point of DPS is to do as much as you can, but healing is more about doing as much as you can as efficiently and intelligently as you can. There's no point to topping the meters one minute into the fight if you're running OOM doing so. And then there's the minor point that it takes a lot of experience to parse healing meters accurately. I imagine most discipline priests have at least one horror story about a PUG raid leader trying to kick them for "low healing." Nevertheless, we shall do the best we can.
Frostheim uses Raidbots, which in turn pulls its data from World of Logs. World of Logs I was already familiar with, as my guild has used it to upload data after its raid nights, but I'm still new to Raidbots. For both that reason and my comparative inexperience trying to use these tools to generalize about a huge player population, I'll be blunt: This is going to be a much more tentative outing than you'd get from Frostheim. While acknowledging these limitations, I wanted to take at least a quick peek at how healers are faring in Dragon Soul less than two months into patch 4.3, with the promise that we'll revisit this topic in a few months with a much more in-depth look.
Fortunately for me, some trends are so obvious that even I can't screw them up.
- If your goal in life is to land on a top 100 parse and you're not a paladin, you'll have a much better shot through 10-mans.
- Really good druids have weathered the Wild Growth nerf OK. Average druids haven't.
- Shaman are still struggling on most fights, although they seem to do better on Spine and Madness of Deathwing.
- Heroic 25-man raiding guilds are getting a lot of use from their paladins.
- On any given fight, expect the healing done to run from paladin, druid, priest, to shaman. Priest and druid performance seems to vary the most.
Class balance is not the same thing as raid balance.
Blizzard's made this point on a number of occasions, but it bears repeating: Any given raid encounter -- or even raid tier -- is not necessarily an accurate indication of how "good" a class or spec is in PvE.
The best healing-related example I can probably give here is from Malygos back during Wrath of the Lich King. Malygos has a mechanic called Vortex that tosses the raid group into the air and does arcane damage to everyone for several seconds. It wasn't possible to use any spell with a cast time during this, which meant that holy priests and restoration druids were your best options for raid survival due to their access to instant HoTs. Paladins and shaman were often left dangling helplessly once they'd blown an instant cooldown.
This did not mean that paladins or shaman were bad healers; it meant that one of Malygos' mechanics had a punitive effect on healers without frequent access to instant-cast spells or HoTs. If every fight in Wrath had had a mechanic like Vortex, then paladins and shaman would have been less effective than priests and druids for the length of the expansion, simply because there was no raiding context where their weaknesses were not being hammered.
If you take nothing else away from this article, take this:
Class balance is not, never has been, and never will be raid balance.
Now that we've got that bothersome little caveat out of the way, let's take the above points one by one.
10-man healing seems a little more democratic than 25-man healing. This one's pretty apparent when you start looking at class representation between the two, largely because the top 100 parses for 25-man raiding (normal or heroic) are a sea of paladins. I'm not capable of analyzing how and where other specs are scaling from 10-man to 25-man raid content, but as far as druids are concerned, I suspect the recent Wild Growth nerf has something to do with it (more on this in a bit).
My guess would be that if your spec mostly has abilities that are capped at a limited number of players, you will have more issues competing in 25-mans than specs with more liberal caps. For example, a glyphed Wild Growth for druids hits six players, but the paladin's Holy Radiance hits everyone within a 10-yard range of the target, albeit with diminished effect on 6+ players. As many paladins will tell you, that does not mean that Holy Radiance actually did a damn bit of good to everyone it hit, but it does pad the meters -- and, in all likelihood, it's having a significant impact on paladin representation in the top 100 parses in 25-man raids.
I will grant it's just my theory that Wild Growth is at the bottom of this behavior, although it could just be a skill-related difference anyway. On a related note, I would love to see how many resto druids still have this spell glyphed and how many of them are doing 25-man raids. I've been wondering if the glyph's increased cooldown is exercising a negative effect on throughput for players who haven't trained themselves out of trying to hit the spell every 8 seconds, because even small portions of this in every fight would be enough to account for the disparity in throughput that I'm seeing. Anyone know how I could find this out?
Shaman are still struggling on just about every fight barring Spine and Madness of Deathwing. It's tough to find shaman in the top 100 parses anywhere except on these two fights. Otherwise, they seem to have the least throughput of the five healer specs when you account for the full raid, which as far as I'm aware has been an issue for a while. No matter how I juggle the populations, they're pretty consistently at the bottom.
My resto shaman is still level 80, so I'm not going to touch the issue of shaman problems with a 10-foot pole. I suggest bugging Joe Perez and seeing what he thinks.
Heroic 25-man raiding guilds are getting a lot of use from their paladins. Most guilds aren't raiding heroic content, period, but this is still relatively early for most hardcore guilds to be making extensive inroads into heroic Dragon Soul. However, of the guilds occupied with heroic raids in both 10-man and 25-man, they do seem to be relying heavily on holy paladins, who are even more heavily represented among the top 100 parses here than they are in normal (though this is far more true of 25-man).
On any given fight, expect the healing done to run from paladin, druid, priest, to shaman. This is pretty consistent regardless of the type of raid or size you examine, but priests and druids seem to be the most common wild cards with respect to performance. As I wrote above, the adjustment to the nerfed Wild Growth and its glyph could be accounting for the swings I'm seeing in resto druid healing, but I don't know what's going on with priests.
So what does this mean for us now?
I don't really know. I'm still uneasy comparing throughput as a means of how balanced the various healing specs are, but in an ideal world (one where blood death knights do not exist in Arena and nobody rolls on my frigging gear), every healer at an equivalent skill level would do the same amount of healing on a given fight. Obviously, this doesn't happen, but it's only a problem if low performance continually dogs a spec.
Once I get better at doing this and more comfortable with Raidbots, we'll return for a far more thorough look at the numbers and how the specs have fared in Cataclysm's final raid. I'm curious to see if particular healing specs scale better with gear than others and if druids begin to train themselves out of trying to hit Wild Growth too early (or just dump the glyph).
Shifting Perspectives helps you gear your bear druid, breaks down the facts about haste for trees, and then digs into the restoration mastery. You might also enjoy our look at the disappearance of the bear.