Augmentation Evoker Interview with Graham Berger

You’re a member of the combat design team, which I understand is the driving force behind how Evoker specs are playing, right?

Graham: Yes! The combat team takes care of all of our classes, anything directly combat related, so we’re always working with the Encounter team, Quest Design, Content Design, etc. They’re building the monsters, we’re building the tools to help fight them. So yeah, we oversee all the classes, PvP balance, all of that.

Good to know! We have a bunch of questions for you today surrounding mainly Augmentation Evoker, but also the combat log integrations. Let’s jump right into it with how did the original idea for Augmentation Evoker come about? What was the thing about it that most excited the team?

Graham: When we were working on Evoker before launch, one of the big things for the Dracthyr is that they can use all five schools of draconic magic. With Devastation, we went in the red/blue direction, and Preservation green/bronze, so we still had black magic left. We gave them a couple of core spells in the class kit because we wanted them represented. But we didn't give them the "this is the main thing" that black magic does. Deep Breath is super iconic. Putting up dragon scales feels very black.

But the idea of exploring pure black dragon magic was really appealing to us because we've seen a lot of corrupted black dragons and shadowflame magic. But what did Neltharion do before he felt a corruption? And wanting to explore that more. And we also knew that bronze magic had so much more to give, like time magic. Getting to play around with that is a lot of fun for us, a lot of fun to give to the players.

So we're thinking what can these two do together? That led us sort of in the direction of Augmentation. That's something that the team has been interested in exploring for a while, this idea of a support playstyle. It's an audience that isn't super well served by any of our existing specializations and has a really unique playstyle.

So everything's really coming together for this. We have themes we know we can build on for this type of playstyle, we've been wanting to do it, and we want to keep telling the story of the Dracthyr through Dragonflight, in patches. This is going to be a weird thing to add to the game, so adding it kind of midseason has some advantages and is lower risk than "show up at launch and see what happens". That gave us time to answer a lot of question we knew we had to answer about how do you introduce this to the game? How do you balance this in all the different kind of group sizes we have in the game? How does it work solo when your main thing is buffing other players? So yeah, a lot of things came together to kind of make it happen.

How did the idea change over time? You've mentioned previously that the approach was to buff a single player first and we've moved on from that, as we see. Was there something else that was fundamentally different compared to now and what drove the change?

Graham: I think that's the biggest change that the original designs were. We played around a lot with sort of a Holy paladin, Beacon of Light, type of mechanic where it put a mark on one person, half-hour buff, or whatever, and your things would automatically target them with all your different buffing effects. That's been a goal from the outset, to make Augmentation very approachable. You can show up and focus on enemies, and not clicking your allies, and then transition to a more group-oriented playstyle. If you're buffing a lot of people, you're doing much more maintenance buffs rather than spiky burst effects, that's the big one I think.

Obviously, we tried a lot of different spells. A lot of stuff ended up on the cutting room floor, all kinds of haste buffs, all kinds of wacky stuff overriding other people's spells with new ones. You can imagine with 38 other specs in the game, that became kind of a monumental task.

The kind of single target to group-oriented was the biggest change. We did play around a bit with having it be more ally focused. Augmentation, your buffs sort of pick targets. You can choose them if you want to, but they'll work on their own, and you're focused on enemies.

Another early version had you more looking at your raid frames and playing more like a healer. Let me buff this person and that person. And again, that worked, but it wasn't as approachable and it still had the kind of burden of knowledge issue. How do you know who to buff at what time? How do we give you the information of, e.g. the Mage is using Combustion or the Rogue is using Adrenaline Rush, and then having to switch back to hit enemies, because since we didn't have you buffing 100% of the time meant switching back and forth between targeting, and that was a lot.

This iteration process may actually be reflected in the spec IDs that the game is using today. I've always wondered about this. Usually, spec IDs are just in ascending order, but then there's Preservation at 1468, and then a sudden gap to Augmentation at 1473. Is this just a technical limitation?

Graham: No, that was way before we launched building out. We thought: "Okay, let's just add the records. We know we want to do three. Oh, we actually aren't doing the third just yet." It was very early. It was like: "We're making a class, cool, plug in a couple of specs.". Then it went: "No, we're only doing two, and the third one is going to be later." Okay, pull it out and then re-add it later. The way the database works, the ID changes.

What challenges did designing Augmentation have that other specs typically didn't? Did you have to design any new tech in the background to enable it, or did you have to think outside the box on how to normally approach a spec?

Graham: Yeah, we've done some new things with Augmentation that required support from our engineering, UI, and art teams. The ones I think players will notice first is the floating combat text when you buff other players. When you Ebon Might, you see plus 800 strength over the warriors, when you use Prescience, you see 3 % crit.

On the design side, when we're prototyping things, the editor for World of Warcraft is incredibly powerful. You can do all kinds of stuff. So you can fake a lot of things and cobble things together, but they aren't polished and crisp and nice. We had it working where those bits of text were popping up, but they were very slow. The text would rise slowly and fade slowly, and it wasn't very crisp looking, it wasn't high resolution. We were playing with that internally and people are like: "The buffing is cool, but it feels weird, something is off about it".

Having it run through the actual floating combat text system to have the numbers show up quickly, fade quickly, be that crisp resolution, be able to have them pop as crits if we wanted to, all that, had to be built out. The system is existing, but it only shows numbers or predetermined strings of characters. So making it show whatever the designer wants actually is some additional work. But that was huge for Augmentation because you aren't seeing as many damage numbers, so how do you know how well you're doing? Let's give you the feedback we give to other classes.

Certainly, the Ebon Might bar on the portrait frame showing the timer for the aura. We have other resource bars on classes, but I think Brewmaster Stagger is the only one that's a unique resource that isn't tied to a power you spend. So getting that hooked up to deliver the information players need upfront and immediately without having to install addons necessary to play Augmentation effectively.

Beyond that, the automatic targeting for things like Prescience and Ebon Might. We had some of the tools to do that, but again on the design side... Built it in a very non-performant way initially, and then said: "Engineering, please help I know I'm doing a bad thing, can you make it work well?" and of course, they came through.

Certainly for Evoker, building a power spell is like wanting to push World of Warcraft, try new things, create a new experience for players because we love how many different ways there are to play the game. Augmentation is just another step there. I'm interested to see what from Augmentation maybe ends up on other classes in the future. Is the automatic targeting something that players really latch on to and want more of for other spells in the game, that sort of thing.

When the first hints of Augmentation surfaced, how did the team react to early community speculation and excitement?

Graham: It's always a mixture, right? On one hand, seeing the community be really excited about it is awesome. But it's also, as a designer, you want to show them the version you believe in. We don't really want to put it in front of players until we know we're excited about it too. And if it gets shown too early, we hope they like it, but we know there's a bunch of stuff we want to fix.

By the time we got to the first PTR launch, and having Augmentation playable with the full talent tree, it wasn't done and ready to be shipped, but we are feeling more confident by that point. It was like "Hey, you can see our vision now, tell us what you think of it" vs if it's a leak, or gets revealed too early, it's like "Okay, you're seeing this slice of our vision, but not the full picture." That can lead to more questions we either don't have answers to or not in the direction we're thinking. Which can spark some interesting discussion, but makes it challenging to talk about, I think.

You've previously mentioned adding additional support-oriented specs might depend on community perception. But if you were to do so, would you change existing specs or add entirely new ones? What about sprinkling support spells into existing classes? You just mentioned the auto-targeting, maybe making it into other things, but that's more like a technical thing. What about additional auras, for example?

Graham: Yeah. So I think in terms of adding additional support specializations to the game, or support playstyles, because it is still a damage dealer role, community perception and reaction to Augmentation is going to really inform that. But also, Augmentation as we said earlier, was this confluence of a bunch of things coming together. I think we would need that as well. It's like adding more specs to existing classes. Does that fit with the story we're trying to tell? Does that match with where we're going in future expansions and how Azaroth is evolving? So a pretty high bar in general.

But if we were to, I imagine we'd be more likely to add new specializations rather than change existing ones. If you're playing enhancement shaman today, which in the past has had some of that support playstyle, you play it today because you like lighting your weapons on the fire and doing a ton of damage, and changing that out from under players is a pretty big ask, I think. So unlikely to do that, I think we'd be more likely to add new ones.

In terms of support spells on existing specs, not a hard yes or no. The nice thing about Augmentation and the support playstyle, as a player you're opting into that. You're saying "I want to give up my direct throughput to make others stronger. I'm okay with my damage meter in-game showing a smaller number because my other allies, their bar is going to be bigger. That's cool to me."

Whereas putting too much of that on an existing spec, like Chaos Brand or Mystic Touch, that's a perk. That is something that it's not taking away from how your class is tuned. It's just adding to the group. That's fine. But once it starts being something like instead of you doing 100% of expected damage, you're going to do 90, and making someone else do 10. Unless you're opting into that, that's not what you signed up for when you pick that spec. So I think that's the magic of having the dedicated playstyle to it, you can opt into it versus feeling like you are sacrificing some of your own performance, even though you don't want to.

Noticeably, there is a lack of haste buffs in terms of support spells in the Augmentation toolkit. Historically, external haste buffs are problematic as it's been basically impossible to actually measure the contribution. You just mentioned there were some in earlier iterations, was maybe part of the reason this changed?

Graham: Couple of reasons! That is one of them. Here's the thing. Haste is really fun. When Bloodlust or Heroism goes off, when you get Power Infusion, it is one of the stat changes that as a player you can feel really strongly. Compared to critical strike, if you get 10% critical strike, I don't know that's something you can necessarily feel. If you get 100, you can. If everything starts to crit, you notice that. But haste, you hit the 10, 15% mark, you start noticeably feeling the change. So of course, that's something we wanted to try. You're a buff class. It was a goal that other people noticed you buffing them. Let's do haste.

The issues we ran into, one is the damage attribution portion. Haste is an extremely complex stat. Versatility, % damage done, the math there is relatively straightforward. Haste, what do you do with that? Incredibly challenging. I'm sure that's something you guys have run into.

That was a piece of it. There are also already a lot of haste buffs in the game. The more of them there are and the more they stack up, from a feeling sense, can devalue each one. Numerically, they're still valuable. It also feels samey. It's like, Yeah, okay, you have Bloodlust, you have Power Infusion. A lot of classes have personal haste buff because, again, they feel really good. So adding another one on top actually doesn't feel super special. Let's explore and see what can we bring to the table that actually does feel more unique and feel more different. Those are big factors as well.

Does Augmentation, as a result of that, introduce any new balancing difficulties? Were there any challenges in ensuring it was balanced across all content?

Graham: Oh, absolutely! That was one of the questions that before we even got too deep into the spec, let's make sure we can answer this. What is our strategy for balancing and augmentation in every piece of content in the game? Because you can play solo in the outdoor world. You can do PvP in groups of two and three and five and ten. You can do Dungeons. You can do raids from 10 to 30. How do we make this work?

We talked about a lot of options, but the tools we end up using are the target capping for their effect. Ebon Might hits four players, Prescience hits one. You can keep it rolling on two. Even the major cooldown, Breath of Eons, really only works on four players, five including yourself. We paper-designed a lot of abilities that were more giving the whole raid some damage amp, Heroism 2, but that's a balancing problem. Do we do a second type of Exhaustion debuff? Or do we just expect you to bring as many Augmentation Evokers you can fit in the raid? None of those are super satisfying.

So target capping for one, in a dungeon you have one. Great, you cover everybody. In a raid, there's a soft cap, depending on how many other DPS you have, but you can have multiple and that's okay. The second factor, is in a raid versus a dungeon in particular, but also in arena, in a raid you're buffing other damage dealers pretty much guaranteed. Ebon Might and Prescience prioritize DPS. In a dungeon, you're buffing two DPS, a tank, and a healer. So that's not going to bring the same value in terms of granting primary stats. Yes, granting the healer primary stats helps hit survivability checks, but you are a damage dealer. Your role is to bring damage. If your team doesn't need healing to beat the mythic keystone you're doing, that Ebon Might is going to waste.

That's why they have a passive in the spec tab called Close as Clutch Mates that increases the effectiveness of Ebon Might and Breath of Eons outside of raid content. So when you're in a dungeon, only buffing the two dps and the tank, and the healer, we can make sure you are still providing as much damage to the group as another DPS would. That number right now is 40%, but very easy to tweak up and down once we get more data once Augmentation goes live. Those are the two main factors, the target capping and the Close as Clutch Mates.

We also have the PvP mode systems that have existed for a while. In arena, Augmentation actually deals a fair bit more damage than they do in the outdoor world. Their buff service is slightly less effective relatively. You still want to cast them, maintain them and they're good for setting upburst windows. But PvP is such a different beast than PvE, and wanting to balance them a little differently makes sense.

Between all those tools, I feel like we've got the knobs to make sure that in that type of content, they're not doing enough or doing too much. Let's turn this one and bring them up or down.

Is there concern about an increasing performance gap favoring well-coordinated guilds due to how Augmentation scales with the skill level of the surrounding players? E.g. Ebon Might prefers targets with Prescience, so you can target players with big burst windows and PI on the start of the pull. I can see this swing clearly in favor of better-coordinated groups, will this have an impact on encounter tuning?

Graham: Potentially. I think as we see Augmentation play out, and this is the advantage of releasing it midseason, Race to World First is done, etc. A lot of guilds are on farm at this point. That gives us the opportunity to see how Augmentation plays out once it's in the wider community. PTR testing is awesome and we love it, we really appreciate folks who participate in that. It's great for finding bugs.

In terms of balancing data, the sample set is just so limited compared to when it's in the full game. I imagine we're going to have to do some post-launch tuning, and that's okay. That's expected. But by the time we get to 10.2 in the new raid, we'll be able to have Augmentation really dialed in.

In terms of the skill gap, there's kind of two sides to that. On the high end, yeah, I imagine players are going to be able to really min-max Augmentation. Honestly, I'm excited to see what they do with it, and we have some thoughts of what we might need to tweak or change or bring up or down if that gets too out of hand. But that is sort of the promise of the spec as well. You're buffing others, your knowledge of what they're doing and their knowledge of what you're giving them should have an impact on how well you and they perform. I think that's cool. Just what is the delta?

On the opposite end of it, we've done a lot to make Augmentation more approachable. To raise the skill floor so that focusing on the uptime on Ebon Might, it's a pretty steady state buff. Your uptime is 60-80% ish. Their burst effects are primarily Breath of the Eons, and yes, Prescience, but that is 3% critical strike, not 30%. So making sure as long as you are maintaining your buff, your effectiveness throughout the fight is going to be at a pretty high floor. I think it'll help players on the lower end, yes, your friends around you need to perform well a well. On average for your group, as long as you're doing your job and they're doing theirs, you'll be equivalent to them. And that's really the goal is like that you are competitive amongst your peers.

One last question about the design approach for Augmentation, was there any concern about secondary stats? Leech and Versatility are quite different when a large portion of your actual damage is technically done by other players.

Graham: Yes, it was a known concern going in, and I think it is going to play out a little bit differently than a traditional DPS might. But it ends up being a math problem. If you're giving primary stats to other players, any primary you give them is scaling off their secondary stats. And so if it also scales fully off of yours, well now it's double dipping and suddenly a low-geared Augmentation, if I make numbers up,is worth 80% of a DPS, and a high-geared Augmentation Evoker is worth 120% of another DPS. That's not great, with the better geared you are, the more Augmentation Evokers you want in your group. So trying to keep that gap as close as possible was an important goal to us for the health of the game at large.

Ebon Might scaling off primary means you're still going to want to get better gear and improve your item level because that is such a big portion of what you contribute. But yeah, having the secondary stats not be as strongly impactful is something we're aware of and going to keep an eye on, but a known cost of building out a spec like this.

Moving on to the Combat Log. Ion mentioned things like Chaos Brand, Power Infusion, etc are unlikely to be attributed via support events as specs are tuned independently of them. Are there plans to add support events to less impactful spells? Healing received modifiers, or damage reduction from external or, say, the legendary buff, for example?

Graham: It's going to be on a case-by-case basis. Our focus right now really is on Augmentation. Giving players the tools primarily to evaluate their own performance and ask "Am I contributing enough damage? Am I buffing people at the right time?". That sort kind of thing. We want to make sure players have the ability to analyze that if they want to, using things like Warcraft Logs or WoWAnalyzer. That is super important.

We are focusing our efforts there, and there is still more work to do. We have a couple of abilities that I think we're both aware isn't attributed yet, it's really complicated. We have some plans on how to work through that, and that's going to be an ongoing process.

Past that, it's going to be looking at what is most impactful for us, and for the community, to have attributed and deciding where to spend the time. So things like healing, I think could be a good win. Especially because Augmentation has some healing received effects they can apply. So in the interest of that spec going that direction. Damage mitigation is definitely interesting as well.

The Legendary is a hot topic of debate, I understand that. I think philosophically for us, that is more in the it's a perk category. Because the buff you're giving isn't coming out of the budget of the weapon, it's additional on top of the primary and secondary stats you get, it's more of a raid resource. So attributing that back to the Evoker doesn't really align with our approach to Augmentation. So I think we're unlikely to make that have the same attribution model at this time.

In the past, there have been some lag issues with encounters that fire a lot of sync events, for example Sylvanas pre-fix. Is there any concern that the additional support events fired from Augmentation might impact this?

Graham: Interesting. Not that I have been in discussions, but I will mention to our engineering team and get their take on it. I'm sorry, I don't have an answer for you.

How linked was the design of the legendary to Augmentation? If the spec hadn't been planned already, would this legendary look different, or would not exist at all, possibly?

Graham: We were in constant communication with the team building the legendary, and our rewards team does some awesome work. They knew they wanted to do an Evoker legendary because it's Dragonflight. Evoker is the hero class of the moment. Let's tell their story and give them something cool to go after. It is really driven from continuing to tell their story. We saw in that cinematic when the gauntlet shatters off Neltharion's hand. Okay, but where is it? Let's do more with it. That was really the spark of the idea.

We knew Augmentation was coming in, so talking with them. The themes do overlap a fair bit. The gauntlet was about dominating and controlling, but about affecting lots of people. And the way the Dracthyr reimagining it is similar to what Augmentation does. I think it's cool that they align in that way. We did discuss though, some of the early paper designs for the legendary were Augmentation abilities. Don't do that one, let's brainstorm some other ideas, sorry! So definitely some parallel design. But I think what we eventually came up with was really cool.