Community Team: We’d like to start things off by asking a question that players often ask in regard to the very purpose of each class. In this case, we’re looking specifically at warriors, which have been stalwart base for comparisons since the start of World of Warcraft.

Q. Where do warriors fit into the larger scope of things currently and where do you see them going from this point forward?

A. Historically, warriors have always been one of the most dominant classes in World of Warcraft. In Molten Core and for raids afterwards, warriors were THE tank, no question. DPS warriors could also top the damage meters, and were a very potent PvP force. We think we allowed the warrior class to overshadow some other classes, which is probably to be expected given the iconic nature of the plate-wearing fighter in RPGs that long preceded World of Warcraft. We think they are in a fairer place now, in that there is room on the stage for other classes, yet they are still a very powerful and popular class to play. The warrior class has been a very tricky one to balance, largely due to the way rage converts into damage (which converts into rage, which converts into damage...), and we haven’t completely nailed that design just yet.

One of the things we want to do in the future is take a hard look at the Arms and Fury trees. There are several talents which just haven’t weathered the course of time well and pale in comparison to some of the newer Wrath of the Lich King talents. We’re happiest with the Protection tree -- we made a conscious effort to pare down that tree and remove a lot of mandatory talents in order to give the warrior more flexibility to take some more fun or utility-oriented talents. We need to make the same pass on the dps side of things. The reason we haven’t done so yet is that warrior dps is in a pretty good place and we don’t want to have to dilute the class across the board just to make some talents a little sexier. We will eventually do this though. We also need to make some decisions about the difference between Arms and Fury. Traditionally, Arms was the PvP tree and Fury was the PvE tree. We understand some players prefer that model, but we don’t like the way it cuts off such a big chunk of the class from players who might not have much interest in the PvP or PvE parts of the game. However, we would like to reinforce a little more the kits of Arms and Fury. Everyone (I hope) gets the difference between Frost and Fire mages. Arms is supposed to be about weapons and martial training and feel “soldierly.” Fury is supposed to be about screaming barbarians in woad. You get a sense of that, but it could be stronger. With the death knight, we allowed all three trees to more or less be able to tank. There is a desire among some players and designers to see Arms tank with a two-hander while Prot tanks with a shield. We’re still not sure that’s the direction we’ll go -- it’s a lot of re-design and will never work for say the druid or paladin classes.

Q. What is it that makes them unique compared to all other classes?

A. The big ones are stances and rage. Other than warriors, only bear druids use the rage mechanic, and that is pretty much just because that form is intended to mimic warriors. Rage is an unusual resource because it is infinite over the course of minutes, but can be very limiting over the course of seconds. While the basic mechanic of rage is interesting, it has caused us lots of balance problems over the course of World of Warcraft -- sometimes in the favour of the warrior and sometimes not. It’s probably time to give the mechanic another look.

Stances are intended to be a major battlefield decision for warriors, though we realise it doesn’t always pan out this way. You have access to different abilities in different stances, but pay a rage cost as well as sacrificing the potential to use other abilities. More on this below.

Warriors also have some unusual mechanics like say their ability to move quickly around a battlefield, to survive massive physical damage through plate armor and Defensive Stance, and game-changing abilities like Spell Reflect.

Community Team: Warriors have quite a few abilities that are contingent on certain circumstances like Overpower and Intervene.

Q. What is the reasoning behind this and do we have any plans to change that type of gameplay?

A. We like situational abilities. When specs don’t have situational abilities, it’s easy to fall into a very fixed rotation. We call this the metronome. Push button 1, 2, 3 on your keyboard over and over until the bad guy drops loot. We have made more of an effort in all the classes to have certain moments that require players to pay attention a little more and then reward them when they both cause those situations to happen and then execute on them.

Q. What would be the impact of changing those class mechanics?

A. I think if anything, abilities like this need to be more prominent. You should be less effective at your job if you ignore them, and ideally you’d also be less effective if you just macro’d them in. We like macros (obviously, or we wouldn’t have them in the game), but we like for them to simplify chains of things that you have to do often without making decisions in between point A and B. We don’t like it when playing your class becomes how clever your macro can be to the point at which you are pushing one button to play your class. That’s not playing an RPG -- that’s programming a robot.

Community Team: Stances have long been a debated aspect of warriors’ gameplay from the pluses and minuses each one offers to the restrictions they apply on what abilities are available for use.

Q. What is the overall purpose of stances and how are stances intended to be used?

A. The purpose of stances is for warriors to have to make decisions in combat. How badly do I want to Intercept now? Should I pay the cost of Spell Reflect? Ideally, we want warriors to switch stances in combat -- not every few seconds, but a few times over the course of a battle. Now we realise it’s going to be harder to enforce this in raid fights unless you have a battle with a lot of movement or other unusual circumstances.

We get a fair number of suggestions from players trying to basically slip the stance concept out of the warrior class: make it not take rage, or let them do more abilities per stance so they don’t need to switch stances so often. That’s not really what the warrior is all about though. You should care what stance you’re in and it should be a decision to change stance. Note that if you pay too high a price to change stances, that counts as there not being a decision though.

Q. Has there been any thought on moving away from restricting abilities based on the stance a player is in?

A. No. The design intent of warrior stances is that you change your toolbar when you go from one stance to another and that that decision isn’t a trivial one. Now, the third part aside from the rage cost and ability limitations is the penalties (such as 5% damage taken in Berserker). We cut those in half recently, and we’d eventually like to get rid of them altogether. We just don’t want to see Arms warriors in PvP in Defensive Stance 100% of the time. We have seen DKs stick with Frost Presence in PvP despite losing 15% damage, so I don’t think you can just argue “Oh, no warrior would EVER do that.”

Community Team: There has recently been a growing number of concerns with warrior damage, as a whole.

Q. What are our thoughts on the overall damage for warriors in each of the three specs?

A. Warrior damage was too high in Naxxramas and then a little low early in Ulduar. We think it’s in a pretty good place now and warriors will get a small damage buff in 3.2. Part of the concern here is we used to exempt warriors from the design philosophy that pure dps classes should do more damage than hybrid dps classes. We try to no longer play favourites here. Warrior damage should look like that of Feral druids, Enhancement shamans, Retribution paladins, and death knights. If their damage isn’t at that level, then it’s possible our numbers need some tweaking. However don’t always assume that you can’t possibly improve your gear or your button mashing either. Also remember that some fights just favour one class or spec over another. We’re totally cool with that, so long as it isn’t always the same exact class or spec that gets to shine.

Community Team: Warrior shouts have added some unique utility to the class in the past, but now they tend to be used very sparingly.

Q. What is the reasoning behind their short duration and do we have plans to improve the duration similar to the buffs other classes already provide?

A. The shouts are supposed to be buttons that warriors push in combat. They aren’t intended to be pre-fight buffs like Arcane Intellect or Prayer of Fortitude. We had a discussion about this recently and decided with glyphs and talents that the duration isn’t a problem. If you lack Booming Voice and the minor Battle Shout glyph, it might be more annoying.

Q. Demoralizing Shout tends to have a very minimal impact in most situations, are there plans to improve this ability?

A. I think by “most situations” you must mean “PvP.” Demo Shout has a massive benefit against raid bosses. It’s probably 20% less damage from a typical boss and literally like 50% against say Thorim’s Unbalancing Strike. However, removing 400 attack power from a Feral druid with 9000 attack power, or a Shadow priest who doesn’t care about attack power at all is of much more limited use. Monsters and players use pretty different combat formulae (which is one of the weird things about the old design of say Vindication). We would like Demo Shout to be more useful in PvP, at least against characters who rely on attack power.

Community Team: The rage mechanic as a whole is very unique, but sometimes leads to situations where players aren’t able to perform due to lack of resources. A prime example of this is when a tanking warrior’s gear is much higher than the content they are at, by taking less damage they get less rage which results in less threat and therefore cannot perform at a higher level.

Q. Are there any considerations in store for improving this mechanic and allowing more rage generation in these situations?

A. Yes. In 3.2 we changed Shield Specialization to provide a little rage on a dodge, parry, or block. This will help in say the 5-player dungeons or in the first few seconds of a raid boss fight. It does not solve the problem of the Prot warrior who is not being targeted (because they are there to pick up adds later in the fight or something). We want to solve that problem by letting Prot warriors generate more rage through doing damage. It could be in the future that we shift most of rage generation to damage done and have little or none in damage taken (and we would have to change a lot of other mechanics to make this work obviously).

Now, long-term we need a better solution to rage generation. Tying it to damage done is logical in the theoretical world of game design, but has problems in reality. When your gear sucks, you have rage problems. When you have great gear, you are no longer limited by rage. That’s just not a great model, and one of the reasons warriors are overly gear dependent.

Q. Where do we feel warriors fit into the current raid environment and where do we see them progressing in the future?

A. Obviously warriors were the traditional tanks and pretty much the only tank in much of World of Warcraft’s history. Warriors now share tanking responsibilities with three other classes, which can feel psychologically like a diminishment in role. In Ulduar, we think warrior tank balance is about where it should be -- death knights were a little ahead, paladins were a little behind, and druids were about even with warriors. We are making a few Prot changes to 3.2 to help in some of the areas where they fall short, such as damage done. Death knights are getting an adjustment, paladins are getting a buff, and druids might get anadjustment or stay as-is. There are plenty of guilds progressing through hard modes with warrior MTs on almost every fight, and we don’t see that changing in the Crusader’s Coliseum.
We’re happy with warrior dps in Ulduar. Whether you go Fury or Arms probably depends on whether you need Trauma or Rampage, and we know warriors in good guilds who flip between both specs. There is some evidence that Fury may overtake Arms dps once you get really good weapons. Dual-wield yet again shows its propensity to scale very well. Warriors will get a slight dps buff through Armored to the Teeth.

Community Team: Sticking with raiding content for a moment. Many tanking Warriors have felt the there is little value in both Strength and Block Value attributes.

Q. We have expressed an interest in improving Block Value for tanking warriors in the past; do we have any definite plans to update this?

A. Shield Block Value just isn’t a strong mitigation stat these days. However, the amount it would need to be increased is enormous in order to make a difference vs. bosses that can hit for 40K. The problem with improving shield Block Value by so much is that Prot warriors would be nigh invulnerable -- they literally might take no damage -- against large groups of adds, in easier content where opponents don’t hit that hard, and in PvP. The real problem is that the amount blocked doesn’t scale with the amount of the swing. We think block needs to be a percentage of damage blocked in order for the stat to do what we want. But the trade-off would mean that warriors (and paladins) couldn’t block every incoming hit, especially from large groups. Avoidance might also need to come down across the board, and many talents and abilities would need to be redesigned. This is a major change that isn’t the kind of thing we can crowbar into 3.2 with a clean conscience. It is almost certainly the future for the block stat.

Q. With strength on tanking gear currently providing a very minimal benefit and still using up a lot of stats on items, do we have plans to improve how this stat works for tanks?

A. Strength is good for dps and threat. It’s not a super mitigation stat (through block) but we also don’t know that it needs to be. We have made some big improvements to Prot warrior dps in Lich King, but too many players still view the primary role of the tank to stack avoidance and mitigation and then complain when their threat is low because they avoided all dps stats. Now, we do think the game of survival is more fun than the game of the threat management, but we also need to get players out of the mindset that it’s okay for tanks to ignore dps stats and just do trivial damage. They don’t need to top the charts, but their damage should be a meaningful component of damage done. We’re willing to change the way the game works to accomplish this goal.

Community Team: Let’s jump over to the topic of player-versus-player interaction. Discussions on the survivability of warriors in PvP have been ongoing.

Q. Do we have any plans to make warriors less reliant on healers in PvP conflicts?

A. We have taken small steps with Enraged Regeneration and increasing the healing on Bloodthirst. We don’t want the warrior to be great at healing as say a Shadow priest or death knight. On the other hand, we want healing to be a major part of the PvP experience. We’re okay with the occasional all-dps Arena team, but they need to be rare or a major chunk of the game just gets marginalised.
Not related to PvP, we do think warriors have too much downtime when levelling. Healing may not be the solution to that, but we think it needs a solution.

Q. With strength being the stat that provides the most benefit in dps scenarios, do we have plans to implement PvP gear like cloaks and rings that have strength instead of attack power?

A. Doing that just means the item isn’t of any interest to say leather or mail wearers, which means we have to create twice as many kinds of rings. The problem is that some classes value Strength and some value Attack Power. Things would work better if some valued Strength and some valued Agility, and Attack Power was a useful secondary stat to both. This has the added benefit of solving the whole problem where leather and mail look attractive to warriors. If leather had Agility on it and plate had Strength on it, then it’s pretty clear who is getting what item. Strength for rogues and Agility for warriors wouldn’t be junk stats, but they wouldn’t be as attractive as the other stat. Again, this is a big change. We wouldn’t just gut rogue dps by stripping Attack Power off all of their gear.

Follow up Questions:

Community Team: Some players feel that, while tanking, having to continually hit Heroic Strike to match your weapon’s swing can become somewhat tedious.

Q: Do we currently have any plans to change Heroic Strike to make it no longer necessary to hit that button on every weapon swing?

A: Yes, we really don’t like way Heroic Strike works. On-next-swing attacks just don’t feel right and for tanks especially you can convert every white attack into a Heroic Strike. Unfortunately, they do the job, which is why we haven’t changed it yet. It’s job is to let warriors bleed off excess rage by converting it to damage and / or threat. It’s possible to fix this in other ways, such as perhaps the hit consuming more rage the more you have Execute-style.

Community Team: It appears that many players who enjoy the Warrior class for its damage aspects continue to feel that, without best in-slot items, their class’s performance is very truncated.

Q: Is this an issue that we have seen in the Warrior class? If so, do we have any plans to accommodate those players who do not have best in-slot items, while still keeping those with the very best equipment from being too powerful?

A: This really just gets back to the way rage works, which is that damage leads to rage so you have to pick a point at which you balance warriors. High damage and high rage? Low damage and low rage? The way to fix it is to normalize rage even more so that you always get X rage per second regardless of gear. But once you always get X rage per second you essentially just have rogue energy. So, as with the previous question, we don’t like the way it is working and want to change it but we don’t have a perfect substitute in the can just yet.