Interview: Technical Game Designer Chadd "Celestalon" Nervig talks Warlords of Draenor

I was lucky enough to head over to Blizzard Campus this week to talk to Technical Game Designer Chadd "Celestalon" Nervig. Chadd is a huge part of the class design team, key to a lot of the changes we saw in the recent Warlords of Draenor patch notes, which is just what we discussed. We were also joined by Senior Community Representatives Zarhym and Lore. You can also find a much-abbreviated summary on Wowhead.

Olivia: First up, is there anything you really wanted to clarify and get out there?

Celestalon: I've tweeted about pretty much everything. This was the first version of the patch notes, there have been more changes since then, those patch notes are about a week old or so?

Zarhym: Yeah it's like, tons of changes. [Rygarius] said he had a huge list of changes.

Celestalon: There's another five thousand words that aren't up there yet, which [Rygarius] is working on now.

There have been different amounts of patch notes released for different classes. Paladins have been complaining that they haven't got enough, rogues have been really happy that not much has changed. Is it safe to assume there's more to come? This is just step one?

There's definitely more coming. Like, for example, paladins had relatively few patch notes, and a lot of that is we were relatively happy with how things played out, at least for ret and prot, with the exception of a few things we can solve with tuning – changing numbers. So a lot of what you see in the patch notes now is what we call design changes, so the mechanics that we want to change so we can get to some design that we like.

There's still going to be a balance and tuning pass – several of those -- that come on later on, and we think ... I think it's fair to say that ret is not terribly strong on live now. I wouldn't say they're terrible or anything but they're a little on the weak side, and I think that the problems they have are not design problems, they're a little underpowered numbers-wise and we can solve that by tuning numbers.

This is alpha -- do we still have to expect six months of beta?

Celestalon: Blizzard doesn't necessarily call different versions of the game different things, like ... this alpha's going to last X duration, beta's going to last X duration. I guess it's fair to say it's more like how many people we're inviting to see it now?

Zarhym: Yeah, with Mists, obviously with the annual pass and stuff, a lot of that was getting people in as quickly as possible because that's what we said we'd do, but it doesn't necessarily make for the most conducive testing environment. So I think we're going to take it a little more slowly this time, as far as when we go into each phase. Not to say that we're just going to take our sweet time, just that we're going to steadily ramp it up, and that's partially why we decided to call it an alpha. And it has to do a lot with the new CASC system and wanting to test certain aspects of that so the initial test would not have that much content available or unlocked for people to try yet, because we're testing different things. So right now it's in take-home alpha, it's been in Dev alpha for a while so people shouldn't worry that ...

People shouldn't take this as where you are in content?

Yeah, don't read into it as a presentation of how far along in content we are.

One thing we hadn't heard much about were the Draenor Perks. What's the thinking behind them, are they to kind of replace level-up abilities or something else?

Kind of! We mentioned them very briefly in one slide at BlizzCon. The general point of Draenor perks are that it is cool to feel like you got something exciting when you leveled up, but it doesn't have to be a choice. Sometimes choice can be a bad thing. So Draenor perks are just a little perk you get when you level up. Not anything you have to worry about, plan around, anything that should stress you out. Just that hey cool, I leveled up, now my Frostbolt hits harder. Noticeably harder. Neat!

Is it the case that they're learned in a specific order? Or are they learned randomly?

No, every time you level up it randomly gives you one that you don't already know, so by the time you hit level 99 you've learned them all. But everyone's leveling experience is a little different and there's no choice involved, because if there's choice then that can feel like there's a wrong choice which creates a bad gameplay experience for some people. And that's not at all what we're looking for, we're looking for just a neat little perk that you get when you level up, not a character customization thing.

Another effect of the Draenor perks is that they'll change your rotation gradually as you level, so when you're going from 90-91, your rotation will be roughly the same, and as you pick up the perks, your rotation will alter until it's changed to where it'll stay at level 100.

And I assume, say you're a mage and you level as frost, you're also going to get the fire and the arcane ones?

Yep!

Follow-up question to that, is there a battleground bracket cutting those levels in half? Because obviously we've got 15-19, 20-24.. is it going to be 90-94 and 95-99? There was a question about the perks causing balance issues.

I don't think so, but I don't know the exact answer so we should just ask Holinka!

Note: Zarhym texted Brian Holinka and the battleground bracket will run from 90-99 with level 100 having its own bracket.

Obviously people are losing a lot of abilities, and are you going to go back through and look at the leveling experience and be like "Right now we need to redistribute all this stuff?"

Oh yes, absolutely. There's one really broad vague patch note line at the top of the ability pruning section that says like something to the effect of "we've re-tuned where you get things in leveling flow" and that's what that means. We aren't going to call out in the patch notes just for sheer volume what they would be, all the "this you learn at this level, this you learn at that level." But yes, we are going to re-tune to provide a much better experience.

Bashiok mentioned this a little while back, but is that going to include a look at low level battlegrounds, low level PvP stuff, like paladins get a lot of really strong abilities really early -- that sort of thing's going to be examined?

Yes, and also adjusting the damage and healing of various abilities. There's a change that's going to affect low level damage and healing fairly significantly. This is getting a bit techy, but if you look at a tooltip for a hypothetical melee ability, it'll say something like "Does 500 damage + 120% weapon damage." That 500 is what we call the base damage. We're getting rid of that; bringing it all the way down to 0, and increasing the weapon damage % (or attack power coefficient, or spell power coefficient, whatever the ability scales with) to account for that. Those base damage amounts created scaling irregularities, that were especially bad at low levels, such as in low level battlegrounds, where the base healing on some heals was most of your health! That creates some problems, and so removing that is one thing that's going to help solve that.

Being a bit more specific then, removal of cooldowns like Devotion Aura from tanks and not DPS -- is that coming for DPS?

Yes.

So all of the raid cooldowns that are being taken away from tanks are also being taken away on DPS and being put onto healers?

There's a section coming in the next version of the patch notes called "Stackable Raid Utility." The general philosophy here is that most of these things are survival cooldowns for the raid, at least most of these ones we're talking about. And that's a job that the healers signed up for. It's the healers' job to keep the raid and the party alive. It's cool for other people to be able to contribute to that, but their contributions have grown to the point where they're infringing too much on the healers' job. So what we want to do there is remove as much as we reasonably can to give that just to the healer, but not completely remove it all.

So let me explain that further. Right now the amount of stackable raid utility on DPS is not that balanced on live. What we want to do is balance that to a point that's lower than it currently is on average, but there are a couple of classes and specs that don't bring much at all if anything. We want to bring them up and give them something and the ones that give a whole lot, bring them down to that baseline that we establish. Tanks are one that we can bring really far down because there's no danger of you not bringing a tank. The prot paladin is not going to get sat from the raid because he doesn't bring devo aura, as long as we can make sure that none of the tanks bring devo aura and equivalent so we can safely go pretty low on that. But, we can't remove everything from all the DPS, and we think it creates a better situation if they bring a little bit so that's what we're going to go for.

So everyone brings a little bit, rather than some people bring a ton of utility and monks bring nothing?

Exactly! So something you'll see in the next version is Windwalkers. Their Stance of the Fierce Tiger provides them a slightly increased movement speed -- 10% -- that's going to become a 10-yard radius aura, so all the melee along with the monk move a little faster. Mages right now bring very little raid utility as far as a raid defensive cooldown. They're going to bring amplify magic as a raid cooldown, which increases all healing received by the raid by 20% for a short duration. But at the same time we're going to bring down the ones that bring a whole lot of raid utility. Retribution paladins for example. They bring Hand of Sacrifice which is a great tank cooldown, Hand of Freedom to break any snares, Hand of Protection for variety of uses, Sacred Shield if you talent into that, several things that ret paladins bring, and Devotion Aura. We're going to drop Devotion Aura from ret.

But they're keeping their hands apart from Hand of Salvation?

Yes. Hand of Salvation is going away, just as sort of a general ... threat has very little gameplay impact to it now. There is a little if you have really bad tanks, but we want to help those tanks improve rather than still keep this button around that you only use in 5% of LFR runs or something like that. We want to ensure that you don't need Hand of Salvation and then get rid of it.

Then rogues are keeping Smoke Bomb?

At a smaller percent.

Gateway is being nerfed?

Gateway is being nerfed. Healthstones are something that's getting set level with other healing cooldowns.

So Healthstones will share a cooldown with healing potions so you can't use both a healthstone and a healing potion in one combat?

Yes, if you bring a warlock you're basically getting free healing potions.

Are they being brought up to the same level as healthstones?

I don't know exact numbers off-hand, but they're being brought down.

So the healing percentage of healthstones and healing potions are slightly lower?

Yes. We doubled player health, so doubling player health meant that these percent heals effectively got doubled as well so we brought them down to account for that.

And health potions won't share a cooldown with things like Potion of Concentration?

Correct -- and all the ones that you currently use for DPS with the one minute cooldown.

And pre-potting will continue to exist for those?

Yes, and that's an interesting topic that I would love to touch on. We talked a lot about and I think we've hinted that we didn't really like pre-potting and were considering getting rid of it. Plan for now is to actually keep it and the reasoning is fairly subtle. If we got rid of pre-potting which we could do by making it not carry over a potion when combat started, or make it count as your one use. Right now you get effectively two potions per combat -- one pre-pot and one in combat. By getting rid of pre-potting you'd be getting rid of the second one which is the more interesting one because you get to choose where in the fight you put it. Do I cooldown for this phase with the potion, do I stack with Bloodlust or Heroism later in the fight. There's a lot of interesting gameplay to where do I put that one potion. If we got rid of pre-potting you'd always be using your potion on the pull.

Because I know people are going to ask, why do you think that if you removed pre-potting people wouldn't then go for that interesting one and would instead go for the pre-pot?

Because unless there's some specific fight mechanic, on the pull is the default ideal place to use any cooldown, because it means you're going to stack it with any other cooldowns that you've got. There may be a couple of exceptions but the vast majority of the time you'd pot on the pull and lose the interesting gameplay of where you'd use that second potion. It's definitely something where I understand why people wouldn't understand at first glance.

There are a lot of questions coming in on Resolve. A few people have come in saying that as a tank I like contributing DPS, and it's also something that comes up on meters and that I can measure. What was your thinking with resolve?

So the incarnation of Vengeance that we had in Mists worked very well in most regards. As far as its defensive value, we're very happy with how it worked out. As far as its offensive value, however, it went to a lot of standing in the fire for damage, we tried to fix that where we can by making like obvious pools of fire not give you Vengeance. But there's still like ... "I want to tank more than I really should," and there was this bad feedback loop of gameplay of "I want to do bad things, I want to do more damage," and it also meant that the damage the tank did was extremely dependent on the fight situation.

If this was not a fight that happened to hit the tank very hard, or you're in a different situation where you don't have a lot of Vengeance it meant the tank's damage was very very low. And conversely, on some important phase in a fight where the tank is absolutely getting trucked he's really high on the meters and you can cheese a few things. And suddenly you've got the tank beating the DPS which, either of those two extremes, all the DPS notice the times where the tank is beating them, and the tanks notice the times when the healer is beating them, and nobody ends up happy with that.

All that gameplay that we got out of Vengeance for defensive value was great. What we'd like to do instead is take that defensive gameplay, refine it a bit, and call it Resolve (renamed because a defensive-only ability being called "Vengeance" is a bit odd). Without Vengeance, tanks would be really low on damage, so we're going to bring tanks' damage up naturally; tanks will do good damage, regardless of whether they're taking no damage, or a ton of damage.

We're not going to tune them to beat DPS, that's DPS's job, to win the meters. But if you took how much damage a tank does on average over a whole raid tier we're fairly happy with that. We just want tanks to be consistently that much on every fight. So that means several tank damage abilities are going to go up in damage, passives are going to increase their damage, stats that were only previously defensive value for them like mastery are going to have offensive value, and we're just going to tune their damage just the same way we do tuning for DPS, just with a little less in there.

So they can kind of expect to be just below DPS? Are they going to be sitting at the top of healing meters?

Healing meters for tanks are always a little deceptive because it depends on whether or not your form of active mitigation counts as healing for meters. Like, a warrior blocks a ton of damage. A death knight takes damage and heals it back or puts a shield up and absorbs it. Both of the death knight things count as healing on healing meters, the warrior's don't. So tanks on healing meters is always really deceptive. Vengeance kind of played into that in that it had a sort of feedback loop, it was too strong in that if you got a lot of Vengeance you could tank a lot more than you really should which let you get a lot of Vengeance and tank more and so on, which made good tanks really hard to kill, so we had to burst them really hard in order to kill them. So as far as how much a tank heals, it's going to appear on the meters really varied, it's not something that we're going to balance.

So there's not concern about people being like "I do nothing that appears on meters as a tank?"

Right, very little of a warrior's defensive personal survivability shows on a meter, we're not concerned about that. We're not concerned at all about the balance of a warrior and a DK on meters.

Talking about raiding and utility and that sort of thing ... professions. Are you just thinking that all the combat specific things are going? So no extra sockets, no +DPS tinkers for engineers but you can still have your rocket boots and die every now and then?

Correct. For engineering in particular there are a few tinkers that, despite not being raw throughput, are very significant combat benefit. Our current plan, that may change, is to let Rocket Boots be something that engineers make and can sell and give to everybody ... so everybody can use Rocket Boots and get themselves killed sometimes!

And the invisibility belt? That would be the same idea?

Yeah, most likely, we're still deciding the specifics of that but the core goal is that profession is not a combat decision.

Things like the life spirits and water spirits from herbalism would be taken out as well?

I believe so. I'm not positive about that. I think that those sort of auxiliary benefits like that are things that we're still deciding on to make sure but if they became a significant impact like all healers have to go herbalism, then we would do something.

Yeah, there's the passive haste thing that you get with those as well?

That's definitely going away.

How are you going to make professions appealing now that raid bonuses are going away?

Celestalon: I believe that you guys are working on a blog about that?

Zarhym: Yes. Someone is.

Celestalon: Yeah, so blog coming up some time in the future about professions. It's going to have more info on that.

Lore: I think it's worth mentioning too that even without the combat benefit there's still a lot of gameplay in crafting for professions.

There are a ton of people who are really really excited about this rogue/druid combo point rework. Why now? Why not the other every single patch that it was asked for?

We just felt like now is the right time to try it. There's benefit and downsides to combo points and the system that they were, but we think we can get that same ... one of the differences that combo points provided was an innate downside to target swapping. That's probably the biggest obvious thing. We think we can do as much of that as we need through class abilities, and we probably need less of that than we had. And it creates, we hope, a more dynamic, interesting gameplay experience when you can do fancy things like oh I'm going to be DPSing this mob and switch over and stun that one and come back to DPSing this one, at a significant DPS loss to do that of course, but that's more added utility that we think they can use.

And you're going to be changing a ton of abilities too, presumably. Can rogues and particularly feral druids expect a big rework around this?

Not a lot. It's not something that we think needs a lot of tweaking. We think some of the impacts of that are we're going to have to see how they play out in alpha and beta, and I think the most obvious impact to that is how that impacts their target swapping and ramp-up time and if that causes any problems and what solution can we do those problems.

Why are agility classes the only ones getting more crit, and not strength or intellect ones?

Because that's how it was. The kit and theme of an agility class fits the sneaky, the precise, the well aimed attacks, whereas, talking in very broad terms here, the agility classes are good at precisely aiming their attacks, the strength classes are good at heavily hitting the target but not precise.

So it's thematic reasons.

It's thematic reasons and also that it's carrying over the gameplay that people are used to. And that's why they chose that class in a lot of cases, a class that didn't get crit chance from their primary attribute always ended up with less crit chance, so you're used to a natural level of how often you crit at various points in gearing so we wanted to keep that.

Feedback I'm getting is that strength classes feel like they're losing more than agility classes.

The problem there is maintaining the parity between strength, agility and intellect. As raw stats go, agility was a better stat than strength. Taking crit chance out of agility makes them equal, and that's what we're going for there. That was sort of a prevailing agility classes love agility; strength classes love secondary stats; and intellect was somewhere in between. We wanted to bring that to the same level so that we can better balance the value of primary stats versus secondaries.

Sorry, I thought the crit bonus was staying with agility classes?

Agility classes are getting a flat 10% additional amount of crit, but not crit from agility any more.

The last thing about agility classes, please tell us that more poison options are coming for rogues.

Yes, we removed several poisons from rogues so far. We realize that that means there are very few poison options, it's something that I do not have an answer to announce for yet, but it's something we're discussing. It's definitely on the radar.

Talking of removing things ... Why did you guys remove the healing wave/heal/holy light type heals?

They fit a niche that was only useful early on in an expansion before you got geared. For the most part, there are some exceptions to this, but talking broad strokes, they were the bad heal that you wanted to gear out of having to use. It never felt fun to cast Heal. You wanted to switch to Greater Heal and the like. On one hand it had this gameplay that you graduated out of once you passed some threshold, so they were button bloat in that regard. And the gameplay that they did provide we thought we could make in better tuning of the other two instead and still get the same gameplay which led to the notion that we've come to so far of just limiting to the simple single target heals, one high efficiency heal and one high throughput heal and you mix those based on how much mana regen you've got and how much healing the situation calls for.

One thing for a PvP shaman. You don't use healing rain a ton and you don't use chain heal a ton, you've taken away one of our four heals.

Right, we definitely recognize that. We're being very careful about this is the best that I can say, we recognize it's going to cause some fear, and it's definitely something given a lot of our attention to make sure that we get right.

Part of what's causing concern is some of the stuff coming out around the healing blog. People were worried about Cataclysm where mana was a huge deal, and you're also taking Hymn of Hope and Innervate, nerfing Mana Tide for the raid, etc. What's going on with mana?

That's definitely something that we understand the initial fear from. There were good parts of Cataclysm healing and there were bad parts. The goal we're trying to get to where we reference anything related to Cataclysm is those good parts of that.

We want mana to matter. But that doesn't have to mean that you're out of mana all the time. We totally understand why people hear Cataclysm healing and worry. We're taking steps to make sure that's not the case. We can make mana matter, and make the efficiency of a spell something you care about without making you out of mana all the time.

How do you plan to do that?

So, like Healing Wave, which we've renamed Greater Healing Wave to Healing Wave ...

People are not thanking you for that, by the way!

Yeah, we know, it's confusing. The problem is that Greater Healing Wave would be Greater than nothing! So anyway, Healing Wave on live, you could basically cast it while out of mana because it was so cheap. We're making sure you'll have enough mana regen to cast that in basically all situations. It's a little smaller than on live, but still most of a greater heal.

The way we're doing that is we're raising baseline mana regen a lot. Mana regen is going to be a lot flatter. It's going to be like you're in Throne of Thunder gear.

So that's your mana regen when you're naked. How does that impact mana regen through the expansion?

We have to make sure that the expensive heals like Flash Heal are much faster but also more expensive. The healing per second on that is much higher, so then the mana you're gaining by getting more spirit is being turned into Flash Heal capacity, and that's got a huge value.

You mentioned in the healing blog that you were making multi-target healing more expensive?

Yeah. On live, multi-target healing is very very strong. To the point where if there's one person injured and one who might get injured, it's probably worth it to cast the multi-target heal over the single target heal. We're going to push that tuning threshold out so that if there's one person injured you should single-target heal them. If there's two people injured you should probably single target heal them both. If three are injured, that's worth it to cast the multi-target.

There's going to be tuning differences on every spell, but the general ball-park is that it becomes equal at around 2.5. That's the general rule we're going, but that's going to vary a bit. It's a good benchmark. But then you have spells that break the mold a bit, like Rejuv, it's a single target heal but you blanket with it. There's a lot of complexity that goes into getting the individual heals tuned right. It's going to take a lot of iteration on alpha and beta but we're confident we can get it right.

So we're not going to have the situation where at the end of a tier or expansion everyone just spams multi-target heals because why not?

Right, because the amount of mana you're going to gain from spirit is less. What you do get is much more useful. A good way to think of it is you start with enough mana regen to spam Greater Heal, and a few Flash Heals, by the end of the expansion you get to do a lot of Flash Heals and fill in with Greater Heals.

Another cause for concern was terms like "triage."

Those terms have a lot of negative connotation. The goal that we're going for is making you care about the efficiency of a spell and making you care about overhealing. Overhealing used to be a bad thing that people didn't want to do, it meant you were wasting heals. We want to get back to overhealing feeling like something you want to be smart about. If someone doesn't need healing immediately you can use a slower heal on them, and that makes you a better player, that gives you a way to excel beyond quick reaction times. And it means choosing the right heal for the job becomes something you can do, and that created gameplay beyond spamming your biggest heal as fast as you can.

How is the melee-range balance going to be shifted with the removal of a lot of instant casts and cast-while moving?

Balancing ranged vs melee has been a long task. Part of that is very heavily dependent on the encounters, what movement they demand and how much damage they demand while moving. We have, in some cases, gotten to the point where the ability to cast while moving was strong enough that movement doesn't impact ranged DPS enough. We think it's OK if they do a significant amount of damage while moving, but not all their damage while moving, except for in specific examples where we tune for that. Hunters for example, we treat as melee for the purpose of their cast while moving, yet they can be at range -- that's a very strong perk of being a hunter.

But look at Siege, for example. In Siege, we didn't have damage perfectly tuned, but for the most part, if you ignore a couple of outliers, melee vs ranged was actually pretty well balanced. That's sort of a thing that needs to be reached through both class design and encounter design. Ranged doing appropriate damage while moving, and designers tuning encounters around that. We would like to tune both of those simultaneously to maintain a healthy balance between melee and ranged to get both less cast-while-moving and less need for cast-while-moving.

We also believe that when you're casting-while-moving less, you care about movement more and it becomes a gameplay element you want to consider, and it also means different spells become more valuable. For example, elemental shaman. They have had a very significant nerf in the cast-while-moving Lightning Bolt being taken away. However, that also means that you start caring about your other abilities that can be cast while moving for their ability to be cast while moving. So that creates some interesting gameplay. You now have a reason to cast unleashed weapon, you have a reason to use shocks other than just for Fulmination. Maybe if you know you're about to move, save your instant Lava Burst. That adds a little nuanced gameplay and that's cool.

And to allay the fears of the elemental shaman ... we're going to make sure you're all right. We know they're fearful, and it's not a case of we want to nerf you into the ground. This is a change because we think the gameplay is going to be better and we'll tune around that.

Would you make Flame Shock more powerful or..?

Probably Earth Shock, you can cast it with lower Fulmination, and that would be good in Warlords.

Are classes going to be getting more damage from instants?

Yeah, instants used to be good because they were instant. Not just because damage was applied immediately, but you could cast them on the move, and that's cool.

As far as instant cast and cast-while-moving, what classes are hardest hit?

Elemental shaman. Removing Lightning Bolt is a huge nerf, and we're aware of that. Fel Flame is going, but warlocks have a very strong cast while moving ability though Kil'jaeden's Cunning. We're still examining Kil'jaeden's Cunning, but warlocks have two instants in their rotation anyway, affliction has Corruption and Agony, and they have good interplay, and you can Soulburn: Soulswap as well if you want to put up all your DoTs. So because you can cast those at slightly offset points with Pandemic, maybe you want to line those up with when you have to move, or pool them for when you have to move. Kil'jaeden's Cunning might still be too strong, but we feel that other talents will still be far better like Archimonde's Darkness on stationary fights.

It's something we've seen before, where if someone has an ability that obsoletes the need to worry about something you stop remembering your other abilities that you have to deal with that thing. So elemental shaman may forget their other cast while moving spells, because they have Lightning Bolt.

Mages are keeping Scorch?

Yes, it's not high damage. The point is that it does a tiny bit of damage, but you can keep getting crits to maintain your rotation. Shadow and balance already didn't have a ton of cast while moving. A pattern you may start seeing is a few instant cast spells and a bunch of cast times in a rotation. That creates a nice setup where you have a mix of having to stand still and where you can cast while moving.

And DoT snapshotting's removal is going to help too.

Yes it means something like recasting your DoT at a specific point doesn't matter as much so you can recast it when you need to move instead of waiting for procs.

Do you feel like you're trying to even out CC across classes? Is that your goal? It seems like some have lost a lot, and some have lost barely none.

As a direct answer, somewhat! But not completely. It's OK if some classes have more CC, we just want to limit it so some aren't sky-high and some have nothing.

As a result of that are you going to be dropping a lot of CC breaks and immunities?

Not as much. We want to have less CC, so if we remove too many CC breaks we'll end up where we were, so we still need most of those CC breaks. A couple will go but the majority will stay.

Are we going to see many more ability removals?

The majority is what you've seen already. We might swap something. One thing we're talking about right now is Army of the Dead. It's something iconic to being a death knight, and it's something we're considering bringing back, it might mean we take something else instead, but maybe not. I would say that most of what you see is what we're planning to go with, but some might not go, and others might go in their place.

How have you worked out the balance between maintaining iconic abilities and ideas versus reducing ability bloat?

That's tough. What is iconic to a class is very subjective. But it's often something a lot of people can agree on. It's safe to say a lot of people would call Army of the Dead iconic. It's really subjective but it's what we're here to do. A lot of it relates to how visual a skill is, how long it's been in the game. A good example is Recklessness vs Skull Banner. We knew we wanted to remove the effect of Skull Banner, we wanted to reduce cooldown stacking and Skull Banner is effectively a cooldown the whole raid has. We wanted to remove the effect of Skull Banner, so did we want to merge Skull Banner into Recklessness, or make it so Skull Banner had a personal-only benefit? So Recklessness has been around since 1.0, I think, whereas Skull Banner is brand new. There's less attachment to that, if someone comes back who hasn't played for a while they won't miss Skull Banner because they may not have seen it. So how long an ability has been around is one significant factor.

Another is what we call the affordance of the ability, how much it represents and shows what it actually does in terms of naming and visuals. Skull Banner implies that it's more than the warrior that gets the benefit so that's another argument for Recklessness over Skull Banner. And there's also the gameplay of intervening to your banner, which we thought was good for Mocking Banner but not so good for Skull Banner. So all that pointed to adding the personal effect of Skull Banner to Recklessness.

Another case that wasn't so clear cut was paladins. Paladins had a lot of cooldowns. We wanted to get rid of some of those to reduce cooldown stacking. The solution we ended up with was keep one cooldown per spec from Guardian of Ancient Kings, Avenging Wrath and Divine Favor. And we would merge those into the one that we're keeping for each spec, with Divine Favor's effect for holy only, of course. So Guardian for Prot, Avenging Wrath for Ret, and Divine Favor for Holy. However, feedback to that was "I want my wings." The angel is more recent, but the wings are more iconic and have been there forever. Not an announcement, but we're currently strongly considering merging Divine Favor into Avenging Wrath, so Holy get wings, and when Prot pulls out the Guardian of Ancient Kings they might get their wings while it's out.

It'll make people happy -- we're not placating people, though. That's what we want to do. I'd say we've gotten it right in most places, and this initial round of feedback helps refine it further and we'll keep refining it. If people have feedback on how they love this ability and it felt awesome, that is very valuable feedback to us, let us know. There is still flexibility in what we're pruning and we might change it.