The Audacity of Armor

There's a few blog posts out there that I intend to reply to, but I'm going to hold off for the moment until I have a better handle on my thoughts and feelings on the matter in question. So in the interim, I'm gonna post this response to a new topic that cropped up on maintankadin, and has taken protadin bloggers by storm.

Theck, the bringer of numbers, posted another of his patented analysis threads based on armor, and in particular, inspired by the Glyph of Indomitability. Some readers will remember that I recommended the Glyph to newer tanks in my post on Chill of the Throne. However, I made a caveat.

However, the glyph is absolutely useless on fights where the majority of the
damage is magical in nature.

Wrathy, Rhidach, and Honors have all posted their thoughts on the meaning of Theck's analysis with regards to dogmatic protection paladin theory. This post has kind of turned the ideas behind the predominant effective health formulas on its ear.

However, the conclusions were something that I already had a good idea of, and I'm fairly certain that any of the more cerebral tanks in the game already had a pretty good idea of too. What Theck's analysis has done is empirically prove what we've known all along. That currently slavish dedication to a single ideal of effective health is the wrong way of going about things. Variations in encounter design and damage sources creates variables that the old 11 armor=1 stam effective health equation simply does not accurately map.

In order to determine the best gearing philosophy, it has to be done on an encounter by encounter basis. It also requires an understanding of the limitations of each form of survivability, and their impact on encounter mechanics. A tank has several forms of survivability.

Flat damage mitigation: This is typically involves talents and cooldowns. This is advantageous because it works on all forms of damage. However, it cannot mitigate damage in its entirety, you will be hit through this, and some fight mechanics, which are predicated upon you taking damage, will still remain dangerous regardless of how much flat damage mitigation you have.

Avoidance: Dodge, Parry, and Miss. Each has certain advantages with regards to itemization, however, with the exception of miss, none of these work on non physical attacks. The benefits of avoidance is that it occasionally provides 100% mitigation from melee attacks, which are often the biggest source of incoming damage on most fights. This means that any debuffs associated with melee hits won't land either. However, this suffers from a lack of predictability, leading to healer panic attacks, and straight up ineffectiveness against many different mechanics.

Expertise: Expertise is usually considered a threat stat, and it certainly is, however, against many mobs, it works almost as well as dodge in regards to reducing incoming damage. Once again, Theck brought the numbers. It's kinda like the anti-dodge. Whereas dodge makes the mob attack and not hit you, expertise simply makes the mob not swing at you. However, this has all the limitations of dodge, and another big one. Against certain mobs, Gormak the Impaler and Patchwerk being among the more famous, they don't parry haste at all. In those encounters, expertise does absolutely nothing to improve survivability.

Armor: Armor is flat damage reduction that only works against physical damage. It has all the strengths and weaknesses of flat DR, but also becomes absolutely useless in the face of magic damage.

Stamina: Stamina works by directly increasing a tank health pool. It works equally against magic and physical damage, does not suffer diminishing returns, and scales blessing of kings. It sounds like a delicious panacea, and in some regards, it is. However, it is a statistic that is subject to "magic numbers". The point of stam that lets you go from being two shot to three shot is so much more important than any point since the one that let you stop being one shot by the boss. What the majority of the points in between do is simply make for smaller overkill numbers. It does help with healer strain, but different itemization theory can sometimes be a smarter way to go when you find yourself in the wasteland between those magic numbers. However, while the situations where more stam is actually detrimental are very few and far between, they do exist. Fights with mechanics that scale off health can create a harder healing load. The most prominent example is Anub'Arak's leeching swarm.

Resistances: Resistances are a much more difficult thing to decide upon. If stacked, the right resistances can almost completely mitigate some of the hardest hitting magic abilities in the game. However, it does nothing against anything else. It's also on very specialized gear which asks you to make severe sacrifices in exchange for that resistance.

One finds the optimal balance of stats for an encounter by looking at the mechanics of the fight. You have to look for what kind of sources of damage there are, which ones can be avoided, which can be mitigated by armor, which will require cooldown coordination. One of the most important things to affect your gearing decision is the choke point of the fight, the point at which you are at your most vulnerable. For example, Gormak the Impaler is the choke point for tanks from an itemization standpoint. You need to ensure you can survive the impale DoT melee combo, and while phase two and phase three contain a good amount of magic damage, armor is still a strong choice for itemization due to the fact that Gormak is the biggest threat, and all the magic mitigation in the world doesn't save you when gormak punches you in the chops.

I suppose that someone with far more time and inclination than myself could analyze the parses of those encounters, and determine the choke points in the fights. Then they could theoretically refine that information into an effective health formula that takes into account all forms of survivability weighted by usefulness. The coefficients would change based on each encounter, and give us a valid mathematical setup for the gear set.

However, that's beyond something I feel like doing, so until then, all I can suggest is that instead of mindlessly adhering to the EH formula, do your own research on the encounter, and make educated decisions. Don't be the tank shunning avoidance for Deathbringer.

The Chill of the Throne.

Well, a lot of bloggers have thrown in their two copper on the Chill of the Throne, the 20% nerf to dodge in the upcoming Icecrown Citadel raid instance. This has brought up concerns with tanking balance, due to the imbalanced distribution for avoidance between the tanking classes. So I decided to take a look at how this is going to affect each class.

Warriors: This will be a reduction in the amount of revenge procs they get, however, the existence of critical block means that if the bosses hit softer than the ones in Ulduar or ToC, then shield block will become a significant physical damage reduction cooldown.

Druids: Dodge is the only source of avoidance that druids have, however, because of their lower avoidance, they have been blessed with more stamina and armor than their fellow tanks. This causes the increase in damage intake to actually be lower than it is for the other tanking classes. This will also help reduce the overlap in Savage Defense usage, when combined with the increased crit and AP they will gain from the higher itemization.

Paladins: Paladins are getting hit pretty hard. The 20% dodge reduction makes it difficult for a paladin in a serious progression gear set to remain unhittable with just Holy Shield up. Normally, paladins were the ones with the most to gain from decreased boss melee damage, because they could count on block value as true mitigation. This is no longer the case. We will find ourselves relying on redoubt procs to fill out the last portion of the attack table. This also severely decreases the overall value of our T10 4 piece set bonus.

Death Knights: Death Knights will be hit the hardest by the changes. They lack any form of block, so they don't benefit from the decreased melee damage, as the decreased avoidance will cause them to take pretty much the same incoming DPS as they are now, it will just be less spiky. They also suffer from a threat nerf in the form of Rune Strike. Their threat will suffer more than a warrior's because rune strike is their primary threat tool, and they don't have block to proc it, so they'll lose a greater portion of their procs. They will, however, likely be the highest avoidance tank in ICC, due to their higher parry values.

What can healers expect? Not a significant amount of change in the healing style. Damage intake should remain reasonably close to the current model, however, the spikes will be less drastic. You'll still be spammy in your heals, but more often there will actually be a need for the healing, so the overhealing will drop significantly.

This effects overall stat weights. Decreased boss damage increases the value of stamina because it reduces the gap between the magic numbers that buy you another hit between heals. It also increases the value of block value, and by extension, strength for shield users, and agility for bears. Agility in particular jumps in value for druids, because it gives them more block rating, block value, and armor. Dodge and Parry become a little less valuable, because with lower overall avoidance, more avoidance becomes less valuable. However, the value of dodge versus parry remains the same. You should not be gemming for parry in icecrown, unless you've already got more than 1.88 times as much dodge rating.

One serious gemming consideration I'd suggest for any tanks out there right now is to replace any +dodge gems with +agility gems. The increased armor will make up for the loss in avoidance, with a healthy chunk of threat baked in. While bears benefit more, all tanks will benefit more from agility than dodge.

From an itemization standpoint, I don't think this has pushed us to the point where we should abandon all attempts at balance and throw solid majestic zircons in every slot. I do think, however, that dual stamina trinkets with on demand cooldowns should be the standard for the normal progression tanking gear set. A Scarab+Heart of Iron combo will serve you well. This has, oddly enough, made Fervor of the Frostborn slightly less useless. For undergeared tanks who want to attempt to push into Icecrown who don't have access to some of these trinkets, the Glyph of Indomitability is actually more powerful the lower your overall gear level is due to the inverse scaling of armor. For primarily physical damage encounters, the Glyph will yield more effective health than the Heart of Iron if you have less than 24k armor. However, the glyph is absolutely useless on fights where the majority of the damage is magical in nature.

All said, the key to the majority of these predictions is the assumption that Blizzard will tone down the melee damage of the bosses in Icecrown. The mechanics are the deciding factor. This makes all tanks even more vulnerable to special boss atacks that do not reset their swing timer, such as Impale or Fusion Punch. The risk of having an unavoided hit line up with those abilities in an unhealably short window has increased. This has been the major threat to tanks in most difficult encounters in recent memory. Sustained melee damage is only a threat to groups with undergeared tanks or healers, and this is going to punish them the most. Most of the guilds that are already preparing for Icecrown will have no trouble with the Chill, however, the guilds that are new to raiding will find themselves getting bounced out even faster than guilds in similar positions did in Ulduar.