5.2 Interview with Greg Street
As some of you have heard through the grapevine, we here at Totemspot were lucky enough to score a Q&A interview with Greg Street (aka Ghostcrawler). I should point out that due to the general nature of this site there will be some Shaman specific questions, but also that a number of them were fairly philosophical in nature. I'd also like to thank (in no particular order) TwintopTahoe
for helping out with the questions.
How frequently will Thunderforged items drop in 10- & 25-person raids?
I don’t think we’re going to announce that just yet. We understand players will eventually figure it out, but we like for there to be some mystery left in the game. Furthermore, most players (heck, most humans) have this weird relationship with random numbers. While logically many players know that random numbers are random, in their gut they still want them to follow a distribution. If I didn’t get a Thunderforged item in the last three weeks, then I am guaranteed to get one next time, right? Right? As I’ll discuss more down below, it’s really easy to turn a bonus into a penalty if the bonus isn’t there when you want it.
How do you decide on including player references (items, NPCs, etc.) in new content? Do you always plan on having a few each time?
We handle it pretty ad-hoc, and honestly I think this is another one of those areas where it’s better to remain a bit of a mystery. Parting the curtain too far and exposing a deterministic system might sound like we’re inviting players to try and game said system.
Do you think players are putting too much emphasis on tools like RaidBots and SimulationCraft when posting about relative class performance?
They are all great tools for evaluating individual performance. It totally makes sense that as e.g. an Enhancement Shaman you might dig into your raid performance and ask yourself questions about whether you used Maelstrom less effectively than you could have on that one add or whatever. Similarly, we get geeked out about simulations like many players, and we think they’re useful for asking questions about stat weights, gear choices or rotational changes.
Both tools are much worse as measurements of class balance. They weren’t designed to be used that way and don’t work particularly well for it. There are a lot of reasons for that (and this feels like it’s going to be a long answer already) but here are two big problems: Simulations are very dependent on the knowledge and dedication of the volunteer working on that particular sim. Warlocks may appear to do higher DPS than Mages because (as an example – not picking on anyone here) the Mage author is top-notch and the Warlock dude is a hack, and there is no easy way to determine this unless we spend a lot of time debugging the sims, which is not time we have readily available.
Similarly, raid parses don’t tell you how much DPS someone can do theoretically; they tell you how much DPS someone did. If only uninformed, stubborn or bad players play Frost Mages while anyone remotely informed plays Arcane, it makes sense that Arcane will appear to do much higher DPS than Frost simply because of the sampling bias. Skill affects those numbers a great deal, but is hard to isolate. Looking at all parses means you’re looking at every player who logs and not just the most skilled and best geared, but unfortunately looking at the top 100 also measures who is very lucky (or exhibiting weird behavior to win meters or perhaps even outright lying). I don’t mean to dismiss the power of logs, because you can get a lot of useful information out of that data, but it takes a great deal of effort. Statistical analysis is non-trivial and takes more time and expertise than many players have. Our players are smart absolutely, but I’m saying as a former ecologist that statistics on large data sets are no simple matter. Tweeting the developers a link to RaidBots is about the least useful argument someone can make. (We know the URL already, guys.) We ask for evidence of some kind because that gives us more to go on than just simple opinion (“I think you should buff Hunters because my DPS is low”), but really it takes a great deal of evidence.
And the things is, players don’t need to worry about it.
Here’s the deal: Blizzard develops the game and players play it. We appreciate feedback, like any smart developer does, but we don’t really want the relationship to go much beyond that. We’re not asking the player community to balance the game for us, and that’s one of the reasons we don’t publicize our target numbers. In general, players should try to maximize their performance (to the extent they’re interested in doing so) and worry less about whether we are meeting our own internal targets for things. Focus on killing the boss, not winning meters. There is seemingly an unhealthy obsession with not getting benched, but really that phenomenon is typically meaningful only for the most cutting-edge hardcore guilds, and those guys are also the first to recognize the benefits of bringing certain characters beyond just their DPS on meters.
Now, that said, nearly everyone here working on WoW systems is still a pretty hardcore player, so we understand the geek value of theorycrafting. Our providing really nitpicky details on the PTR forums for instance is a way to give something back to that community so that you all can spend more effort on the more challenging aspects of developing models, and not reverse-engineering trinket proc chances. On the flip side, given the community freak-out level to even modest balance tweaks, I’m not sure if it does more good than harm on the whole. Despite our caveats that we’re really trying to feed information to theorycrafters, some players get frustrated with numbers continually going up or especially down during our testing process. We haven’t come up with a good strategy for handling that yet, but would prefer not to return to the black-box, no-communication design of old.
What sort of healing strengths/weaknesses do you often look at that aren't given enough consideration by theorycrafters?
For DPS, theorycrafting is nearly always focused on the Patchwerk/target dummy scenario, because it’s much easier to do so. But nobody really cares if they win target dummies. They care about the actual encounters. On the other hand, we’re also not looking for every spec to be identical on every boss, and as I said before, we’re not particularly interested in publishing our targets for each boss. So we’re not really leaving the theorycrafters much room to look at anything but a target dummy. They (and more importantly the players who rely on their hard work) should just remember that encounter mechanics vary enormously. One of the most frustrating things for me personally is to see when players average out the DPS of every boss in an attempt to come up with a meaningful measurement of DPS for each spec. But bosses aren’t even designed equally. Nobody cares (if they are focused on progression that is) if their DPS is highest on an easy boss–they care about performance where it matters.
For healers it is even more complicated. Raw healing per second is pretty useless as a statistic in all aspects, except that it’s easy to measure. The Paladin I used to raid BWL with would always top meters, but his number one target was himself. (Jerk.) Conserving mana should be a big component of healing, but it competes with winning meters and when you’re focused on the latter, it forces everyone to use their hymns and totems to make you look awesome, instead of saving them for when they’re best for the group. Absorbs and HoTs are great at heal sniping because they tend to heal a target before conventional heals. Power Word: Barrier and Divine Hymn are theoretically both healer cooldowns, but the Priest gets full credit for using the hymn, but no credit for using the damage reduction. Using a heal to save a tank who is about to die is much more valuable than healing the Mage who was at 90% health, but meters tend to reward both the same value.
Healers also have a really weird relationship with crit. Yes it’s undependable, but at some point “undependable” became synonymous with “garbage.” Guess what: critical heals still increase your healing. Your HPS will be higher with 20% crit than 0% crit. Haste also tends to get too much credit for increasing HPS even though it also causes you to burn through mana faster.Honestly, I wouldn’t even bother looking at stack ranks of HPS in any system. Use the simulations to try and answer questions about when it makes sense to use an AoE heal instead of multiple single-target heals, or how much you should game your trinket procs, or when it’s the best time to use your mana return. Figuring out who is awesome and who is weak at healing is hard to answer even with objective data.
How hard is it to balance the “bring the player not the class” ideal against making every class ability a carbon copy of another’s? (e.g. raid & single-target healer cooldowns)
It is challenging, but to be honest, players focus on it much more than we do. “Bring the player” was our philosophy in Wrath of the Lich King coming off of the Sunwell raid. Long-term players will recall that the way you created raid groups back then was to stack the DPS specs in individual parties with the specs who buffed them. If you had, say, an extra Rogue in your guild, but you didn’t have room in your stacked melee group with the Enhancement Shaman and Warrior (providing totems and Battle Shout), there was no room for that player. When we first added DKs to the game, raiders told us they couldn’t bring them because there was no room in their rigid raid compositions. That is when we decided that all buffs should be raid-wide and spread among multiple specs, and that no spec should have a single ability so powerful that it always guaranteed them a raid spot (e.g. Replenishment, battle rez). Building a raid comp isn’t supposed to be the fun of raiding—and really it only affects the leaders and officers –killing dragons is. As long as you aren’t overly rewarded for over-stacking the highest DPS spec and ignoring everyone else, raid comp shouldn’t be a big focus. That’s ultimately the spirit of “bring the player.” We have 34 specs in the game and they are all fairly well represented in 10- and 25-player raiding. That was our goal, and we feel we’ve accomplished it.
(My answer would be different in terms of PvP representation, where we still struggle with whether our design should be to get as many specs as possible into PvP, or whether we should focus on providing a very fun environment for PvP-focused players even if that means far fewer viable specs.)
Some people love no-fly zones. Some people hate them. Some of us who hate them don't hate being unable to fly so much as the jarring, arbitrary feeling of flying mounts just suddenly not working for no discernible in-game reason. Have you considered, instead of just turning flight off in some zones, adding in-game discouragements to flight that would feel like an immersive part of the world rather than a rules change?
We tried a few modifications to the flying model, such as a mechanic where you start flying slow but sped up the longer you kept moving. The intention was that flight was used for long-distance travel, but not to avoid quest mobs or to jump from the auction house to a mailbox. It never felt quite right, though, and caused some of its own problems. Sometimes the right call is just the obtrusive one: You can’t fly here because it wouldn’t be as much fun if you did. The Thunder Isle offers some (flimsy) story for why you can’t fly, but the real reason is we think it just makes the world too safe and too convenient. I said on Twitter once a quote that I kind of liked, which is that the journey is an epic part (perhaps even the most epic part) of any fantasy story. Yes, you can wear down that analogy as you can any analogy, but I still like the gist of it.
Philosophically, one of the interesting things about communicating on the forums or on Twitter is how frequently I am pressured to improve convenience for players. I get why convenience makes life easier for players, but it doesn’t always make things more interesting or fun. Part of what designers are charged with is erecting barriers that players are supposed to try and overcome–that’s where the challenge comes in. We’re not trying to make a soul-crushingly frustrating game experience, but we also don’t want to make it so convenient and safe that there is no feeling of accomplishment.
Is acquiring Sunsong Ranch an experiment with player housing?
No. We just didn’t like the fiction that for all your hard work, the place still belonged to Yoon. The reason we phase it (which is similar to just plopping it in an instance) is that sometimes gameplay works better when you’re alone. We don’t want to have an entirely instanced game, but we’ll do it when it makes sense.
To do player housing right, we think there is an expectation for a lot of customization and a way to show off your house to other players. Sunsong Ranch doesn’t do either. Our biggest design challenge for doing player housing someday (aside from the production challenges of creating enough artistic content) is what role the house would serve. We like bustling cities and don’t want every player to just sit in their solo house all the time. In that respect, the farm did a good job of giving players a nice reward (typically Cooking or other profession materials) without replacing the cities or shrines as your gameplay hub.
Understanding that time spent on things like the Warlock class quest line is time that can't be spent on more universal content, do you anticipate more of this sort of thing coming up? Or will class-specific stories remain pretty rare?
They will probably always be rare. Our choice for a long time was basically to never do class-specific content, or to do it only when we could do something for every class (which effectively meant the same as never doing them). We’ll do more projects like the Warlock green fire experiment when they make sense and when we have time, but I wouldn’t, for example, start expecting patch 5.3 to have the Shaman quest line and patch 5.4 to have the Warrior quest line. We’re not trying to design class quests to a quota.
How well do you think the “mobile vs. stationary” DPS niche worked out? In particular, Mages feel heavily penalized during movement but amazing when stationary vs. Hunters, and Elemental Shaman, who can deal most of their damage on the move, feel a little behind the curve when stationary. Is it a case of making sure that there is a mix of mobile and stationary encounters, or do you think that these niches should be reduced or eliminated?
Movement is a very tough design challenge for us. As a designer, I think we have allowed too much damage on the move. Excessive damage while moving causes a lot of gameplay problems, including it being challenging to counter casters or healers in PvP, and eroding one of melee’s big niches in PvE. On the other hand, players love to move while doing damage. It makes the game feel more responsive, is more intuitive and is more consistent with other games, especially shooters where you can move almost continually.
(Note that I am specifically not including as a reason to let players damage while moving because it makes the game easier for them. As I said above, we like having moving as a challenge that players have to overcome and, when we let everyone do nearly 100% of their stationary DPS while on the run, that takes one tool away from encounter design.)
We do use movement as a way of distinguishing specs, especially within the same class. Arcane isn’t great in movement fights, but Fire is much better. Players can choose to change specs to maximize their DPS, or those who just prefer Fire all the time can have their moment to shine. Shadow isn’t great at moving, but is really good on multi-DoT fights. As you elude to, movement is one of the vectors along which we can promote differentiation.
What’s your opinion on specific-gearing tricks like haste caps/breakpoint? Would you like to see these removed so newer players don’t have to worry about them, or do you think players are focusing too much on something that doesn’t have an actual impact?
One of our design philosophies at Blizzard is simple to learn, difficult to master. Tricks like this fall into that category when handled well. A player who is focused on Raid Finder never needs to give another thought to haste breakpoints and they’ll be able to defeat every boss. Someone who is interested in more difficult content can choose to worry about it and in return be rewarded with a small performance increase. As long as the benefit is small, we’re okay with that.
Again, this is one of those situations where the community overall has matured beyond the degree with which it is generally healthy. If you’re in a guild that generally kills a few heroic raid bosses at most, then you could probably improve your DPS by 20-30% or more just by hitting the right buttons at the right time and worrying less about arcana like soft caps.
What sort of impacts have you seen as a result of the capping of healer mana? Are the same classes struggling with regen?
We’re pretty happy with the model overall. It lets healers focus on Intellect for throughput and Spirit for regen (rather than Intellect also indirectly providing longevity). Mana generally matters for all of the healers, though in 5.1 we think Disc priests are able to game Rapture too much by incorporating it with temporary buffs such as Mana Tide Totem. We think things look better in 5.2.
With Cataclysm, “triage” became a healing buzzword. But it’s been largely absent from Mists of Pandaria, replaced with burst damage and thus burst healing. Was this intentional?
As long as there is a risk that healers can run themselves out of mana, then triage matters. Most healers (except possibly Disc) can run themselves out of mana, but generally only during group healing. We have enough personal and group cooldowns now that a few healers can keep a tank alive through almost anything (provided they are paying attention to encounter mechanics) so I’m not sure we’ll ever be at a place again where a healer can run OOM spamming heals on a tank. But you can absolutely run OOM by casting indiscriminate group heals when the damage coming down isn’t that intense (i.e. you are using the wrong spell for the job). Overall, we think healing is more interesting when there is a cadence to the encounter, rather than it being, “spam your best heal all the time!” Those burst windows provide that.
A healer trinket, a la Deathbringer’s Will, was mentioned in a Q&A a year or so ago. Any chance we’ll see that crop up soon?
There isn’t one in 5.2. Maybe next tier? The limiting factor here is animation time. Very few creature models have the full suite of animations that a player can do, and it’s not fair for designers to throw our artists under the bus by using their art in ways they didn’t originally anticipate. We still do it (the transform part, not the under-the-bus part!), but we don’t want to overdo it.
How successful do you think the current Legendary chain is compared with previous versions like Dragonwrath or Fangs of the Father?
The current legendary chain isn’t really designed the same way. It’s a long-term goal that any raiding player (even LFR raiders) can expect to finish. There are some really cool story moments, and Wrathion is a fun character for us to get to use, and we think the rewards have been cool. I think it’s a fair complaint that sometimes the grind has been a bit too in-your-face, especially with the 6k-Valor step in patch 5.1. Lesson learned.
Do you prefer the “small bonus for all who want it” method over the “rarer but more distinctive weapon” system, or could both be used in future content?
I wouldn’t assume that more traditional and/or specialized legendaries are a thing of the past. The two problems we want to avoid from past attempts are creating the expectation that e.g. every rogue will eventually earn Fangs of the Father, or that everyone takes turns, and that eventually it will be time for e.g. the Single-Minded Fury Warriors to get “their” legendary. In that respect, the more random-based legendaries of classic and Burning Crusade felt better. They were more prestigious for being rare than something everyone eventually grinded their way through.
Will there be recipes that use raid crafting materials for Enchanting & Jewelcrafting in the future?
We haven’t ruled it out. It’s usually a case of coming up with cool ideas or not, rather than a philosophical bias against particular professions. I don’t personally think we’ve hit the sweet spot with raid crafting in a while. Usually it feels too mandatory for raiders to get those recipes, or else they out-gear the profession recipes by the time they have access to them.
5.2 includes a number of new trinkets with procs that are more varied than simple stat additions. Is this signaling a trend towards wanting to make the trinket slots less uniform? Is it to be expected that trinkets will start to have much more variance in value between classes/specs than other items?
Trinkets are one of the few items we can provide where players might really have to think about how they work (set bonuses are the other). A bracer is a bracer, but a trinket might be better or worse depending on how you play your class. We’re okay with there being some class or spec variance, as long as e.g. the Elemental Shaman feels like she has to stick with a trinket from a previous raid because the new ones are so terrible for her spec.
The Resto Druid mastery currently amounts to completely uniform percentage added to all heals in virtually all situations. Has there been any thought given to trying to differentiate it more from a completely flat Int-like benefit?
As far as the value of mastery goes, this is true. But it also serves to encourage mixing your heal types. It’s not the most interesting mastery, for sure, and that’s the sort of thing we’ll consider for the future, but for now it’s working and not causing any problems. Better to be bland and functional than interesting and broken.
In general, healing masteries are really passive (and we just made Disc’s mastery more so). Healers already have a lot of reasons (HPS, HPM, AoE) to choose one spell over another, so buffs to just some of them are a little dangerous. If we buff Lava Burst too much for Elemental, it doesn’t really break their rotation. If we buff HoTs too much for Resto Druids, then you might stack mastery and cast nothing but Rejuv.
The new Wild Mushrooms have had very few changes or discussion since their first introduction on PTR. Does this mean you feel this new approach addresses the concerns players had about them?
The 5.2 implementation is a redesign, so we believe it will take players some time to get used to it. As I mentioned in the previous question, healers have a lot more leeway in which buttons to push than DPS, so it can take longer to really figure out the right place for a new mechanic to shine. We found that to be the case for Spirit Shell, where Priests had to experiment for several months (on average) to really get a feel for the right way to use it. If we need to iterate more on healing mushroom in 5.2 or beyond, we’re totally open to that.
Whatever happened to Water Jet?
We still like the idea and it may happen at some point in the future, but we really try not to add rotation-changing mechanics in the middle of an expansion if we don’t have to. What we were really trying to solve here was a quality-of-life problem using the targeting reticule on Freeze, and we didn’t feel like that qualified as a critical change. Totemspot, like many of our fansites, serve a relatively hardcore component of our community, and we love you guys for it. You just have to remember that you’re in the minority. Think about friends of yours who might have taken a break from the game or who don’t religiously follow patch note updates (or GC’s twitter feed!). I know of so many players personally who said things like “Yeah, I hear the new Mage is cool, but I just don’t have the time or energy to try and relearn it.” All of these big changes do have a cost, even if they are very good changes. When we ask players why they quit playing, up there with financial and free-time concerns we always see “I can’t keep up with all the class changes.” We have to balance improving things for our core players with the risk that too much change can turn players off of the game.
Do you have any long-term plans for getting Frost out from under pretty attainable soft caps on both crit and haste?
Haste, yes. Crit, no. It’s intentional that Frost has a low soft cap on crit, and that they switch to haste (or potentially mastery) after that. The Glyph of Icy Veins helps with that, but there will pretty much always be some instants in your rotation, leading to some haste soft capping, but it is potentially too easy to reach that with the number of temporary haste effects we’ve given out lately. We don’t really like that that Glyph becomes (unintuitively) “mandatory” once you’re at a certain gear level, but it’s a necessary evil for now.
We know the forums are only one of your many feedback pathways. Do you find that the level-90 talents are more popular in general than they appear to be on the forums? How do you feel about them at this point?
The level-90 Mage talent tier is actually one of the better balanced tiers of talents in terms of usage, power and diversity. In that respect, they are “good” talents. Despite that, many players don’t think they feel good and that opinion still carries a lot of weight. One of our design tenants is to “make it a bonus,” and this is a classic example of failing at that. Sometimes lacking the bonus feels like a penalty. The level-90 talents are strong, and they fill their function, but you don’t feel better when they’re up–you feel lousy when they’re not. There’s some really fun potential gameplay with min/maxing them, but we acknowledge that it takes a pretty dedicated Mage to do so.
The biggest problem with these talents is that they’re trying to do too much. They’re very much tied to mana, but Frost and Fire don’t really care about mana. We’ve strongly considered changing them, but have decided to hold off for now. It would be a huge change to Mage gameplay, usage, power, etc. One might even need to be completely replaced (I’m not sure how you can take the mana-management gameplay out of a talent whose core premise is removing the cooldown on Evocation). Since they’re still solid talents, all get a lot of use, and are a distinct, meaningful choice as is, we’re going to leave them for now, but plan to change them in the future.
Priests have 3 tiers (T3, T5, and T6) of talents that directly impact their mechanics or add new abilities, many more than other classes. Given how these changes must account for both two diverse healing specs and a difficult-to-tune DPS spec, has any consideration been given to making some of the talent effects baseline and retrofitting some of these bonuses?
Baseline is probably not the way we’d go. I agree that the Priest has been one of the more challenging trees to design. We have been considering changing the way talent trees work in some cases so that every spec doesn’t have the same options. In that world, Holy and Disc might choose from among 3 healing talents in some tiers, while Shadow could choose from among 3 DPS talents. We would do this when it makes sense (such as having Solace and Insanity be separate talents, not a bundled talent). This should not be interpreted as our being unhappy with the new talent model and willing to throw it out for the next expansion. It’s a safe bet that this basic talent model is going to last for a while.
With regard to Shadow Apparitions, is there a particular reason the developers chose to attempt to improve SA AI/pathing, rather than bypassing the pathing mechanic entirely? Has there been any consideration given to changing SAs into something that attacks at range, similar to the Nibelung proc?
So here’s the deal with Shadowy Apparitions. We want them to be a different way to cause damage, and that way will have some limitations. Yes, they would be more reliable to the Priest if they stopped and channeled a Mind Flay or flew to the target. They would be even more reliable, if the Shadow Priest just did extra damage and there was no apparition at all. However, we like that different mechanics have strengths and weaknesses. Now it would be ideal if those weaknesses were more easily predictable. Fighting a dragon? Yeah, SA isn’t going to work while she’s flying. The dragon moves behind a bench and your apparitions can’t path? That part just sucks and we want to fix it. We did fix (we think) the problem during encounter transitions where a boss becomes immune just for a moment and the apparitions fail to path and just give up. We’re fine with them being less effective against, say, flying targets, so long as Shadow has some other advantages. Shadow still does really well in multi-DoT situations as just one example.
Why was haste reducing Mind Blast's cooldown removed in the Mists beta? As it stands, other classes have resources that gain bonus regen from secondary stats, but Shadow Priests (via Shadow Orbs) do not.
Haste is still Shadow’s best stat I believe, and having Mind Blast benefit just put haste that much further ahead of crit and mastery. Priests who want variation in Mind Blast usage do have the option to take Divine Insight.
Can you recall if there been any surprising rotations or ability uses for Shaman, or has everything worked out roughly the way you expected?
Elemental and Enhancement have extremely different rotation design styles. Elemental has a relatively straight-forward rotation and it’s easier for us to predict how it will work out. About the only unexpected thing that has gone on there recently has been occasional Chain Lightning usage in single-target situations, which is unintended, and not something we’re super excited to see, but also not causing any real problems at the moment. Enhancement, on the other hand, has a rotation built more around priorities than using X then Y then Z, with a lot of clashing cooldowns, random procs, and so forth. Players haven’t done anything truly unexpected, but because the rotation isn’t very rigid, our expectations aren’t either. If someone comes up and says “Hey, I found you can do more DPS if you prioritize LB at 3+ MW stacks over UE, instead of at 4+ MW stacks,” we’ll say “OK. Cool,” not “Unacceptable! Time to nerf!”
Will we see changes to Frost Shock to make PvE kiting easier? Thinking back to Lady Vashj in Serpentshrine Cavern and kiting striders in phase 2.
Usually the question here is whether shocks can have unlinked cooldowns, and the answer to that is probably no. We like that there is some decision making about when to use each shock, rather than just using the DPS ones on cooldown while still having utility always available. If you’re just saying that kiting striders was cool, we agree, and maybe there’s room for more of that, but there are also a lot of specs now with CC abilities.
What was the original intention behind Thunderstorm? Is it intended to be used more for AoE, knock-backs or mana regen?
There was a time where we were trying to make all DPS casters care about mana (not just healers) and so we were handing out Evocation-like tools to get mana back. The original idea for Thunderstorm was that you’d weave it into your rotation when mana started becoming an issue. We’ve since abandoned that idea and made mana not really much of a consideration for Elemental. That leaves its biggest current benefit as its knockback. At times, its damage has been relatively strong, and we don’t want its damage to be terrible, but we also realize that it’s not particularly fun if it feels like you HAVE to run into melee range as an Elemental Shaman, in order to usefully AoE. We’re currently erring on the side of it not being an AoE DPS gain, but it might be too far in that direction.
How well has the Unleashed Lightning glyph worked? If you had the time would you change how Elemental works while moving so you can replace this glyph?
Speaking frankly, we regret it. This is one of those changes that opened up the slippery slope for damage while moving. If we had not made the change to Unleashed Lightning, we probably never would have invented Kiljaden’s Cunning or Glyph of Penance or all the other “cast while moving” mechanics. Prior to Unleashed Lightning, there were a few cast-on-the-move effects, but they were limited (short duration on Spiritwalker’s Grace) or really crappy damage (Scorch). Removing all of those would be really hard to do at this stage—by which I mean it would cause a lot of legit negative feedback—so we’re unlikely to make changes here anytime soon. But if we could go back in time and make different decisions, we would. Elemental doesn’t have to be great at casting while moving, so long as they do have moments where they shine.
With the anticipated decoupling of PoH from groups, are you taking a look at Resto's spatially constrained AoE healing (via HR and CH)?
With regard to the spatial constraints on Resto specifically, we think it promotes some intelligent gameplay since you need to take the actual game world and the contours of the battlefield into consideration. You might pre-drop Healing Rain at a spot where few people are currently standing, but where you know the whole raid is about to converge in a couple of seconds. You avoid using Chain Heal on the Hunter that’s standing 40 yards off to one side. By contrast, PoH having a party requirement is a purely logistical construct that has no real connection to the actual events that are unfolding during the fight.
At the view from ten-thousand feet, the right way to design healers for WoW in 2013 is for every healer to have a variety of AoE heals and very few single-target heals. You just don’t need to cast single-target heals very often, while AoE (even in 5-player dungeons, even in PvP in some cases) is more common. (Now, in some ways, a Druid casting Rejuv on a lot of different targets is a form of AoE healing–it doesn’t strictly have to be a PoH clone.) However, if we decide in some future expansion to remove some of the single-target heals and add a couple more AoE heals, then it seems fair for Resto to have some kind of efficient, low-throughput, long-range, no-cooldown AoE heal.
Any consideration to granting Resto Shaman a tank cooldown (since they are currently the only healing class w/out one)?
This is the kind of thing we try really hard not to do. When every healer has a similar single-target cooldown and a similar AoE cooldown, that just leads to all healers feeling really similar. We agree there are design tradeoffs with healers feeling unique and possessing strengths and weaknesses.
Why was the Telluric Currents active-regen model for Resto Shaman put aside in Mists?
It was never intended to be a mana-restoration mechanic. It was intended as a way for Resto to feel a little more like a hybrid class (a Resto SHAMAN) and less like an independent class (a RESTO Shaman). Shaman cast Lightning Bolts, so here was a way for Restoration to do so without having to worry about sacrificing precious mana that might be needed later for heals. As sometimes happened, we missed the mark, but it was kind of a fun thing for savvy Shaman players to do to try to sneak in LBs as a mana-regen mechanic. As we got to the later Cataclysm tiers, it became less of a fun thing to game and more of a mandatory mana-return mechanic, more like Mana Tide Totem. With the re-emphasis on Spirit-based mana regen in Mists, we decided to return Telluric Currents back to its original goal of “free DPS.”
That brings us to the end of this Q&A. I'd like to thank both Ghostcrawler & Zarhym for this opportunity to ask a few questions.