The Beta forums are a strange place, every class telling everyone else "we won't get taken" while their neighbours look over at them jealously and say "but why can't we have ability ". In truth many of the classes with the fewest buffs were those with the strongest position in the Burning Crusade, those that least needed the improvement. As I honestly couldn't compare DPS classes easily I will focus on the Tanking classes, or rather on the Warrior
vs Paladins et al.

It can really be summarised by the comment from the US Warrior Forums:Why bring X class when Y, W, & Z class bring all of X class' abilities and so much more

Who needs warriors when Blessing of Might is superior to Battleshout
Who needs warriors when Leader of the Pack is superior to Rampage
Who needs warriors when Blood Poisoning is superior to Blood Frenzy (and rogues can misdirect)

you should see a trend here...
Ghostcrawler, as typical for her, concluded a rather nice response:You would really prefer to be brought to a raid because your dps sucks but everyone likes your buff? Doesn't that make you feel like a mascot or something?

How about: you get to come to a raid because you're a good player and do good dps?
Followed closely in another thread by:
We believe Titan's Grip would be too much of a dps increase without some kind of penalty. I couldn't tell you with any kind of certainty that the current one is the one that we'll ship with. The math works out (for us -- I realize not everyone agrees) but if it's too punitive and nobody can make the spec work, we'll change it.

I realize some people are nursing long grudges for mistakes we've made in the past (and to be fair, sometimes when they *think* we've made a mistake). Absent a convenient CoT Portal, we can't undo any of that, so we'll just try to maintain better relations in the future.

It's a BAD thing if someone logs into Lich King, plays for a few hours, is unimpressed with the changes to their class and decides not to go explore Northrend. That is not something we want to happen. On the other hand, it's also a bad thing if you are totally dominated or shown up by other classes. But when you think about it from our perspective, we really do want everyone to be excited about their class.
Simply, Blizzard is aiming for balance, perhaps not what we see as balance but they aim such that each class has a reason to be brought to a raid beyond simply being a buff bot, of being asked to stand outside and rebuff every 30minutes or an hour to maximise the raid. Every person should be valuable beyond simply their buffs, you should be allowed to play with a minimal spanning set of buffs and still achieve the goal of progression with any reasonable combination of classes.

Warriors vs Et al
So lets get to the real meat of this, what is there really for a Tank to do, what does every Tank need to do and how do we differentiate the Tanking classes such that there is flexibility, variability, but also the simple competence to get the job done.

Firstly lets define what in my mind a Tank needs to be capable of to actually perform their role as a Tank, this assumption changes a lot, so if you view it differently your conclusions will be different.

I define Tanking in a few distinct areas:

  • Physical Mitigation and Avoidance
    • Physical Mitigation: Block
    • Physical Mitigation: Armour
    • Avoidance: Dodge
    • Avoidance: Parry
    • Avoidance: Miss
  • Magical Mitigation and Avoidance
    • Magical Mitigation: Damage Reduction
    • Magical Avoidance: Miss
    • Magical Avoidance: Immunity
  • Threat
    • Constant Threat Generation
    • Spike Aggro Generation
    • Snap Aggro Generation
    • Multi-target Threat
  • Utility
    • Environmental
    • Anti-caster
    • Emergency Buttons
  • Niche
Many of these are as simple as they sound, however addressing them is important, so lets look at each section in turn and see if we can achieve a balance without a Homogenisation of the classes.

Physical Mitigation and Avoidance
This is the bread and butter of Tanking, being capable of taking unmitigated hits of 20,000 to 40,000 damage, and surviving several of them. This can be performed in a few ways, firstly scaling damage reduction in the form of avoidance, typically any of the 3 types (Dodge, Parry, Miss) can be stacked, Parry being the best overall tool for mitigation and threat generation but also the most costly to stack, and Miss only being stackable through Defense. Typically a Tank will aim for 50%-75% avoidance, this means that in general we aim not to take more than 2 hits back to back (12.5% chance of 3 hits in a row with 50% avoidance, ~2% with 75% avoidance), Blizzard typically assumes a Tank will have ~50% avoidance as a base level when doing their damage assignments (So a T6 Tank will be assumed to have more, but you shouldn't regularly be taking 3 hits in arow).

The second type of mitigation is non-scaling, that is block and static damage reduction (the latter is applied before armour so sees a reduced effect, where as Block is applied after armour so at higher levels is effectively more reduction). Non-scaling mitigation is typically far superior against fast light blows, and almost irrelevant against high damage scenarios. Taking an example a Paladin can tank Caverns of Time: The Battle for Mount Hyjal in block gear with ~1000 block, and consider only the Abominations to deal any damage at all (a Retribution Paladin could heal me through the Ghoul waves for example), in contrast the same Paladin would never wear this to tank Teron as the 10-25% avoidance gained is far more effective on the Healers than mitigating 10% or so of each incoming hit and risking more back to back hits.

Typically these two are balanced against each other and with armour, the goal being that the overall Effective Health of the Tank (basically a calculation to determine what your damage reduction in terms of static reductions is since avoidance relies on probability), and the avoidance levels achievable should be balanced across all classes. For Warriors and Paladins this should be similar due to sharing gear, Death Knights seem to have higher armour than Warriors and Paladins (strange but true) and rely on cooldowns to achieve the time to live (very low effective health due to no block), while Druids have high armour and health. Overall these mechanics can be easily balanced using a time to live calculation and working from that, each tank should ideally survive around the same time without heals against a boss.

Magical Mitigation and Avoidance
This area is relatively new, while Paladins and Druids have been asking for help in this region for a long time it has never been addressed. What this means is essentially ways to avoid taking magical damage (which ignores armour, if it didn't we wouldn't need this topic as it would be the same as physical). Warriors essentially have a 16% Magical damage reduction, coupled with the ability to reflect spells at the caster for damage. Death Knights gain personal magical damage reduction and a prismatic resistance aura, and a 75% reduction of magical damage up ~ 1/3rd of the time. Paladin come in with a 6% reduction.

As there are bosses which deal magical damage, and trash which deals extensive magical damage this is important, while in the past Reliquery of Souls was essentially only tankable by a Warrior Tank (well phase 2, it has been tanked by Paladins and Druids but only if they overgear the phase massively) due to being able to reflect the deaden onto the boss. This as a design principle is flawed, its not that Paladins and Druids can Tank the boss, but rather that its nearly impossible to do so in a progression attempt.

In this kind of region I feel that we need to address this, thus looking at Paladins there is already a solution in the Illidari Council fight, Blessing of Spell Warding, making this into the new system and toning it down to a useful yet viable buff:Hand of Spell Warding:
30s cooldown, lasts 5s, single target within 30 yards
The target becomes immune to spell damage and has all harmful debuffs removed from them.
For Druids something playing into their mitigation idea, perhaps something like:Runic Tatoos:
5 minute duration, 20 minute cooldown, self buff
Magical Damage is counted as Physical damage
Both of these are simple solutions, the first for Paladins fits the design and gives them some utility in being able to reduce the damage taken by others, this means that it can be used in other fights from and Off-Tank (or even Main Tank) position to help the raid. Its not as powerful as the Death Knight version due to a 1/6th uptime (1/2 that of the Death Knight Anti-Magic-Shell) however it does fully mitigate the damage rather than 75%.

The Druid version is the weakest in terms of instant mitigation, however provides a well armoured Tank with incredible magical damage reduction for long periods. The Druid version is also hindered by the long cooldown meaning it may not be up for each attempt guaranteed, these could easily be massaged however to something more fitting if its not viable.

These solutions look to fit with the role need, and don't really infringe on each other, its a different solution to the same problem, and gives each tanking class the fundamental method to deal with spells without stepping on each other's toes.

Threat is the second key area of Tanking, a dead Tank makes no threat, however a living Tank that isn't ahead of the DPS isn't a Tank. In general threat can be divided into a few key areas. In reverse order we have:

Multi-target Threat
This is the capability of a Tank to maintain decent aggro on 2-4 targets at once (holding at least from the Healers), with the current design most tanks are fairly equal in this area using either Cleaves or AoE spells to maintain aggro. Typically on this number of targets any Tank can allow for moderate AoE (whether AoE on 2 targets is effective is a different question :P).

Area of Effect Threat
This was the Paladin niche role, the ability to maintain and manage threat on 5+ targets, in the Beta Paladins retain Consecration so can handle this fight area, Warriors gained an 8 target (glyphed) Thunderclap, an unrestricted frontal Cone (Shockwave) and reactive damage making them nearly the equivalent of Paladins. Death Knights have access to AoE diseases, Death and Decay and various other abilities that hit multiple targets, their overall threat generation on AoE pulls should be similar to that of a Paladin, however with less room for error due to the longer cooldown on abilities (higher damage / time, longer cooldown).

The odd Tank out here is Druids, who maintain only 3-4 target abilities. Typically for this I would simply assume something akin to Shockwave would be appropriate, or a buffing of thorns. Perhaps aiming for something slightly more unique:Frenzied Rage:
Instant attack, 15s cooldown, 15 rage
You instantly attack all targets within range dealing {Damage} and causing {Damage} over 8s. This ability generates high threat.
Its a little like shockwave, however is a bleed effect (while Consecration and Death and Decay are static areas of effect this one moves), and hits all targets you could normally hit (Shockwave is ranged, and cone). Its far from a "wow thats totally different to everything before it" but hopefully it addresses some AoE utility, while maintaining some flavour.

Single Target Threat
Single target threat is what defines Tanking classes, we generate a lot of it. In short all Tanks need the ability to generate more threat than a DPS class performing a reasonably optimal rotation with a reasonable set of buffs (holding from a continual Heroism effect on a non-salvation equivalent Warlock is not reasonable, holding from a Warlock spamming shadow bolts with pauses to life tap is). It tends to come in 3 major forms, constant aggro, spike aggro, and snap aggro.

Constant aggro is generated by Warriors and Druids, their abilities on short cooldowns generate a fairly linear ramp in terms of threat, very easy to predict, however doesn't scale at the start of a fight (thus why Warriors typically ask DPS to wait for 5 Sunders). Constant aggro is the best in terms of boss fights typically as its easy to predict when you will get your threat ceiling increasing.

Spike aggro is typical of a Paladin, we front load damage (Mana vs Rage) in a spike, then are forced to wait 8s+ for the next cooldown with only Consecration and Seals on the boss. Spike aggro works very nicely for picking up phase changes and such like and is similar to snap aggro, however it is as its nature suggests spiky so can be harder for a DPS class to maintain an idea of the threat ceiling if they are not familiar with the Tank.

Snap aggro is the ability to generate threat over a very short period of time, typically enough to at least overcome the Healing threat generated over a boss transition. Paladins with spike aggro tend to work very well for this, however all classes can perform some kind of snap-aggro move once they are in a fight (and thus have rage).

The tanking classes need to be fairly similar in single target threat, as with mitigation and avoidance, to ensure that each has a fair shot at Tanking bosses, any obvious threat leader would become the default Tank for threat races, and barring any negatives associated with it would be the default Tank. Trading some mitigation for threat would be viable if fights are still handleable by the reduced mitigation.

Environmental is really a misnomer, what I mean by this is the ability to avoid environment wide debuffs such as stuns, disorientates, fears and silences. These kind of features massively reduce the capability of a Tank to Tank and generate threat. At current most Tanks are fairly equally affected by stuns and disorientates, however Silences, Mana burns and fears have a disproportionate effect on Paladins and Druids.

Addressing the Fear issues is easy, removing forebearance from the hand spells (and the hand of spell warding above) Paladins would gain a fear breaker similar to a Warriors, however consume a utility spell to do so making it slightly less flexible. Druids I can see two good paths to go, firstly would simply being to reduce the cooldown on their fear break (from 3 minutes to 60s / 180s), alternatively to something more like:
Feral Rage:
Passive Self Buff, 60s internal cooldown
When an effect would cause you to lose control of your character you instead are immune to control loss effects for 20s and gain 20% damage. You cannot be affected by this more than once every 60s.
Its a damage boost, making Druids the Tank of choice for moderate fear fights (60s or so between them, similar to Nightbane) due to the damage / threat boost. It is however not enough to cover a fight like Archimonde alone (3/4ths of fears could be caught using the trinket as well assuming a 30s fear rate), meaning its both a massive benefit, and a large but situational benefit in other places.

Hopefully again we have added the utility to fear that is lacking in several classes, but in a way that is both interesting and not overlapping.

Addressing Silences and Mana Burns, the best solution I have seen is simply a deep Protection talent that makes Protection Paladins immune to silence / mana burn resistant. This gives them the ability to Tank any of these fights as well as a rage based tank without losing the flavour. Fights can simply be altered if necessary to handle this without affecting one Tank in an arbitrary manner, you can still Rage / Mana burn in RoS pt 3 if thats appropriate, but in a fight like Kaz'rogal or Azgalor where the silence and burn effect is designed to force the Tank to use avoidance moves (no heals due to silence) and force movement / non-strict enrage timer (Kaz'rogal's mana burn) then having these affect the tank is just a bad design decision if we are aiming for balanced tanks.

Spell Interrupt, moderate cooldown.

Yup thats it, a Tank needs to be capable of dealing with fights like Princess Delrissa, or Hex Lord Malacrass without ensuring that the group is stacked with an interrupter. This doesn't have to be fast, something as slow as the Mage 24s, or as fast as the Shaman 6s versions could all be made to work, however when a Tank cannot guarantee that they can Tank a boss without assistance its a bad decision.

Utility is the added fun in a Tank, its what you do when not Tanking, its what you do when you have a spare cooldown. This is where Blizzard can have a lot of fun with the Tanks because utility does not have to be balanced so heavily. Typically this is things like Paladin bubbles and Hands, Warrior Intervene, Druid Fluid Hybrid.

The key with utility is to make it a) viable, b) appropriate. No Tank should be asked not to Tank due to their utility being more important than Main Tanking the boss, however the same holds true that no Tank should be kicked out the raid for another DPS / Healer when not Main Tanking the boss. Utility is the reason a Tank should be kept in a raid even on single target fights. Whether this is DPS, Healing, or something a Tank actually wants to do it should all be considered. I am not going to suggest anything specific in here, lets hear what others think on Tank utility.

A niche role should be the area a Tank excels in, and should be the preferred Tank for fights in that theme. Typically this is:

Death Knight: Heavy Magic Fights
Druid: Sustained High Damage Fights
Paladin: AoE Fights
Warrior: Utility Fights (no other Tank can match the raw number of abilities and options a Warrior has, intervene, safeguard, mocking blow etc).

A niche should never be a unique role, no Tank should be thrown out, however niche fights can and will exist, but should be balanced across the 4 classes such that there is a reason and benefit to every Tanking class, and the best solution for a Tanking team is actually one of each class rather than four of a single class.

Conclusions and Thoughts
I hope I have made some suggestions, something you can take away and think about, that Homogenisation isn't a bad thing if done correctly. It should be a drawing together of cause and effect rather than of abilities, giving flavour and balance should not be something that you throw in the air and pray for, but something you look at the role of the class and if it has the tools to perform that role.

No one would ever take a Warlock to a Fire / Shadow immune fight, nor a Mage to a Fire / Arcane / Frost immune boss, doing so breaks their class design, yet by having multiple schools these classes can be brought to single school immunity fights and still perform. Good design up front gives the instance and boss designers freedom to use the tools they have access too. Homogenisation isn't about being the same, its about making sure that your sonic screwdriver works as well as my toolkit.


Andrea said...

I am a prot warrior tank, and I am all in favor of homogenisation of tanks. I prefer tanking as a warrior, but no matter what flavor of tank, we should all be able to tank encounters successfully.

I raid 10-mans casually, and it really makes me mad that in TBC, some encounters are heavily in favor of some kind or other tank or CCer. ZA e.g. favors paladin tanks, and that's just not fair to the other tanking classes, on such fights as Dragonhawk.

I am looking forward to a more even playingfield in WotLK. However, AoE tanking will continue to be the one real niche of paladins, as long as I don't know how easily available the TC glyph will be. Note also that Thunderclap was significantly nerfed in the last beta build, so it's become harder again for warriors to AoE tank.


2ndNin said...

The fastest Bear runs I have seen have actually all been Bear + Fury OT rather than a Prot Paladin, so there doesn't seem to be that big a worry, the dragon hawk adds can easily be handled by a warrior or druid with a mage or suchlike to help (hey they bring cookies too :P). But yeah, Warrior AoE tanking got way better, even with thunderclap nerfed damage shield is still awesome.

Our AoE tanking isn't really that great, DKs look about on par in terms of numbers and Warriors got a huge boost, just poor Druids left sitting out :(.

Andrea said...

I think they're still tweaking a lot, warriors did get a few nerfs with the last build, they announced another nerf to Devastate, and I think bears will probably get their spot in the sun soon. One can hope. I really want all tanking classes to be strong.

As for Dragonhawk, I am just bitter. I am the poor sucker in the two prot warrior group who tanks the Dragonhawks, and thus our best on him is only 39%.