This is a discussion that I’ve been having a lot recently, in several venues. As people accumulate more Firelands gear, reaching block-cap becomes easier, and many more tanks are now able to do so. The question that keeps coming up is:
“Okay, now that I’ve reached block cap, I no longer want to keep pursuing mastery. But what do I want to stack?”
There doesn’t seem to be a consensus on the answer though. Some people have suggested that you should stack avoidance after block cap to reduce your damage intake, or to save healer mana. But we’ve already discussed that damage intake isn’t a meaningful metric, and that we’re usually not stressing healer mana in the first place.
I’ve been arguing that Stam/EH is still the way to go after reaching block cap. I’ve summarized my thoughts in a post on the EJ forums, but I thought that the topic might be of wider interest, so I’m reposting a chunk of it here for discussion.
There are three possible scenarios we could consider:
1) Able to be 2-shot (or less).
2) Able to take 3-5 melee hits without dying
3) Able to take 6+ melee hits without dying
Situation (1) favors EH pretty strongly. When you can be 2-shot, overall damage taken and healer mana conservation are the last things you’ll be thinking about. That’s a situation where you’re EH-deficient, so your first goal would be to increase your EH so that you can’t be 2-shot anymore, provided that’s possible. Adding EH means you buy your healers time, potentially giving them another boss swing timer to work with and allowing them to queue smaller, mana-efficient heals when you’re already at full health.
Brutallus is an extreme counter-example, but the environment has changed enough that he’s no longer very relevant. Decimation Blade isn’t a counter-example, because it’s not possible to increase your EH against that attack with Stamina. In some sense, they both represent the same scenario though – one where you can’t possibly increase EH enough to get “out of the hole,” and have no other choice but to stack avoidance. They’re the exception though, not the rule.
We don’t need to guess at this either, we have historical precedent to go by. Think about the last time a paladin tank was seriously in danger of being 2-shot: it was in Wrath. Healers were spamming big heals to deal with the enormous throughput. And what did we stack? EH. Because that was the only way to prevent getting globaled.
Situation (2), which is the category that almost all tanks in Firelands would fall into, is where some people think that avoidance becomes preferable to stamina. I believe that they’re wrong about that.
The usual argument is that it saves healer mana. In practice, I don’t think that tanks stress healer mana very much at all. I’ve yet to see a healer resort to Divine Light (DL) or Flash of Light (FoL) spam to keep up with tank throughput damage. Usually it’s a mixture of Holy Light and DL, modulated according to intake, and then FoL/DL/WoGs to recover from spikes. What saps healer mana, apart from avoidable damage (i.e. standing in fire, forgetting to use a cooldown) is reacting to those spikes by resorting to fast, inefficient heals to make sure the tank has enough health to survive the next melee.
Avoidance will reduce the frequency of spikes, of course, but the amount gained through trinkets, gems, and enchants isn’t enough to make a large difference. Certainly not enough to change how your healers play, or how they react to you taking damage. Stamina is still the most effective tool we have for that, because the EH gain is measurable and large enough that your healers will notice. It buy them time to consider their next cast, and lets them forgo more of the the fast inefficient heals for slower, more efficient heals. In that sense, I think Stamina is far more effective at saving healer mana than avoidance is, at least in the 3-4 hit situation. 10-20k health will make a far more noticeable difference in how they play than 2-3% avoidance would.
Some people also believe that EH is only useful in this scenario if it lets you reach a new “threshold,” meaning that you can now survive N+1 boss attacks instead of N. I think they’re wrong about that too.
Most of my deaths aren’t from being 2-shot. Nor are they due to the healer not being able to keep up with boss throughput while standing and spamming big heals. They happen when an encounter mechanic forces an interruption in the healing stream; for example, the healer has to move, and I don’t get enough direct healing to keep me alive through 3-5 damage events (boss attacks, DoTs, whatever). In that situation, the concept of “more EH than N attacks” is sort of muddy. Attack sizes vary, and you’ll be receiving a variable amount of healing from your non-moving healers, HoTs, and absorbs. So the net damage and healing you receive is variable enough that the waters between 3 hits and 4 hits are muddied – it’s entirely reasonable to die in a situation where you took “3.5 hits” worth of damage due to healing, absorbs, and variable hit sizes.
So adding 0.25 “hits” worth of EH isn’t meaningless, even if you’re going from 2-2.25 or 3-3.25 instead of 2.75-3. That said, the value of that extra 0.25 “hits” suffers from a sort of diminishing returns – once you can take 6 hits, you might be able to survive for 10+ seconds without heals. There are very few mechanics that will interrupt your healing stream for that long (though dead healers count!), which reduces the value of adding more EH. It’s still nonzero, but because the event is so rare it’s less effective.
Which brings us to Situation (3). This is where I think avoidance really comes into its own. Not because it’s a situation where avoidance really shines, but because it’s a situation where the diminishing returns on EH push it’s value down. If the boss can’t kill you in 6 or more melee swings (which may be ~10 seconds without healing, an absurdly long time), you’ve got little incentive to stretch that time window any longer, so why not beef up your avoidance?
Unfortunately, it’s also where the boss starts becoming trivial. If the boss can’t kill you in 10 seconds, then he doesn’t hit hard enough to test your survivability. And most tanks will start wearing DPS trinkets or putting on 2-piece Ret gear for lulz at that point. The relevant comparison would be a heroic Firelands tank wearing DPS gear for a troll heroic, or maybe a weak normal-mode T11 boss.
So in summary, the one situation where avoidance really does match or exceed stamina in survivability value is the trivial situation, where the boss isn’t dangerous on short time scales. That shouldn’t be surprising, in fact – the entire triage model was initially built around the concept that tank death would not occur due to spikes, but because healers ran out of mana over the course of an encounter. That scenario – low chance of spike death but noticeable time-averaged damage intake – is exactly situation 3, the weak-hitting boss that hits frequently. And that’s how you make avoidance valuable: by making time-averaged damage intake a relevant metric.
Unfortunately, that’s not how things played out. Meloree has pointed out in the past that healer mana regen scaled a lot faster than it should have, and that led to the current situation where tank throughput doesn’t end up being a consideration. Overall damage taken at the end of an encounter hasn’t mattered since early T11. And making it a factor, either by nerfing regen or modifying the boss damage and swing timer to smooth damage profiles, would run the risk of making things worse (for us, at least). In the scenario where tank throughput matters, but spike damage doesn’t, the tank with the most self-healing becomes more attractive. And that’s not us, it’s DK’s.
To be fair, I don’t think the system is bad the way it is. The choices we’re debating here might only affect the outcome of an encounter 1% of the time, and may even depend on the exact relationship between the tank and his or her healing team. But I do think it relegates avoidance to the role of “second class citizen,” and that we should recognize the limitations of avoidance that come out of this model.