The hero Baptiste is getting a ton of love in Overwatch right now: Blizzard Entertainment has given the support character a short storyand a special skin for the Baptiste Reunion Challenge, which ends Monday.
In honor of Overwatch's 30th hero, I caught up with principal game designer Geoff Goodman and lead writer Michael Chu to talk about exactly how these characters come to life.
Heather Newman: How do you go about designing a new character?
Geoff Goodman: It's actually kind of funny with that, because there's not actually a super-set way we generally start with a hero. Lately, most of the post-release heroes have been very gameplay driven. We're looking for either some sort of mechanical answer to something in the metagame or something like that.
But a lot of times in the past, especially way back, some of the original heroes were actually drawn first as cool concepts, and then we were sorta like, "Ah, we should get that guy." Winston was just kind of a concept. It was out there. Definitely wasn't a gameplay-designed-first, for him. You know, it came out backwards.
Newman: Can you give me an example of a character that was designed gameplay first?
Goodman: Brigitte came from a place where we were hoping to get a tanky kind of healer in the game. It worked out pretty well for Brigette, because we already had that character in mind, maybe she'd make in the game somewhere, she's already in the story. So we were like, "She would be really cool as a hero," [and] we're talking about this paladin-type hero. So it worked out really well.
Newman: What about Baptiste in particular?
Goodman: For Baptiste, it was more gameplay mechanics we were playing with. He started from a high-concept of, what if we get a healer who had longer cool-downs and deployed some gadgets that he could throw down to kill people or do different effects? [He] had an internal name, and he's still actually referred to internally, as "Gadget."
Newman: Do you usually name characters during the development process?
Goodman: A lot of the stuff like names comes a little bit late. We're already pretty well in development, so we have to name the files and everything something. So we usually just come up with some weird tag. But he's still known as Gadget healer. So that's where that came from.
Newman: So how did he develop from that concept?
Goodman: We started thinking about what would be really cool devices he could throw out to kill people, or feel like he's got this bag of tricks. Originally, we didn't know what kind of character he's going to be, so he didn't have that kind of solider edge he does now, like a mercenary.
So his primary weapon was mostly just as a generic healer. I don't remember if he had secondary fire at that point. Might have just been the grenade healing. We thought that'd be kind of a fun way to heal.
Newman: So what happened when the storyline started coming in?
Goodman: Once he started developing as a character with a more hardened, more military background, it opened this opportunity. It was like, "You know, we don't actually have sort of a hitscan [high precision, twitch-shooting] rifle in the game. It would be kind of fun, really fit with that character really well. Let's just try that and make the healing grenade a secondary effect on the weapon."
Newman: How important is the story as you're working on character design?
Goodman: A lot of times that will happen with a character, where it's like, I'm coming at it from pure, raw gameplay. Nothing from the art end or anything.
But as we start to develop the character on the art side, the story side, we try to integrate a lot of those parts as well, like, "Oh, yeah, this makes more sense now. It makes more sense now if we added some effects from this new background." So it's kind of organic in that way, and I think it helps to have that kind of flexibility.
Newman: Are there things, both with Baptiste and some of the earlier characters, that you can walk through as examples of, "This particular ability we introduced as a way to counter this other ability or this other play that was becoming a little too dominant in the meta?"
Goodman: Sure. I think the most notable one is Brigette. At that point the meta was heavily the dive meta, with all those super fast heroes jumping around, high damage, Tracer, Genji, kind of characters that are really hard to counter.
For a while, we tried a decent amount of changes to help curb that a little bit within the game that's already live, tweaking Tracer down, things like that. But it wasn't enough. We'd get to a place where we're like, "Well, I don't know if we want to just completely halve Tracer's damage or something like that, to just try to like reduce the impact of this meta."
Newman: So what do you do when that happens?
Goodman: That's when you start getting to the point of, maybe we just need to introduce better answers and put them in the players' hands. So if the enemy is running a particular composition you'd be like, "Well, at least I have this fallback hero I can use, a mechanic I can use to help counter that."
In Brigette's case, having the ability to stun enemies really does slow Tracer and Genji and hopefully allow your team to capitalize on that stun. Also, her ability to quickly heal somebody in a burst of healing and actually even overfill their health and get armor, makes it really hard for those fast heroes to burst down a target. So she was definitely built with the metagame in mind, to help give players an answer.
Newman: What was Baptiste countering?
Goodman: For Baptiste, it was less of that. The immortality shield is really strong against some prominent [ultimate abilities], such as D.Va's; it helps with Junkrat ults as well. His doesn't have as popular of D.Va's ults, but certainly is very strong at getting a lot of play. So it does help in those kind of situations.
But he was mostly built for allowing more options to our players. I mentioned the hitscan rifles, really new to support [characters]. So if you're a hitscan player, and you don't really like to play support, or maybe you just want to stick to your guns that you like. And maybe you like this guy now, because he has this cool gun and you really like that gameplay style.
So there's a little more in that direction rather than a specific answer.
Newman: How do you set Baptiste and other characters apart visually from others that might share a fairly similar silhouette?
Goodman: It's actually been something we've been pretty careful of. It comes up as ideas, suggestions around the office of like, "Let's do a really small tank. It's something we don't really have."
We try to make sure that at a glance, hopefully, you get an idea of what the hero can do and how tough they are and things like that. So all of our tanks are pretty big. That also helps in the raw gameplay sense, because you can actually step in front of [other characters] and stop bullets and stuff like that. So because you're a first-person shooter, you get that benefit as well.
Newman: What's more important -- the function of the silhouette or the uniqueness of the character?
Goodman: Honestly, the bigger deal is just the read. You want to be able to understand, again hopefully at a quick glance, because the game moves pretty quick, exactly what's in front of you and what they're going to do.
Specifically, when we break down even further than that, having a character like Baptiste, he's not super huge, not super small. He's roughly average height. We have to be careful that he doesn't look too much like Soldier 76 at a glance, or Reaper.
Newman: So how did you set him apart?
Goodman: The artists spend a long time on color. Like Baptiste typically has these orange and teal colors, and it applies to all of his visual effects. When we see an effect that fits the colors, it helps you understand that it came from him. Stuff like that.
We spend even more time on a lot of the game play mechanics to make sure that they're readable on the field. That becomes really difficult, because we're always trying to balance you understanding what you're seeing, it's very clear the radius of something, or if you're inside an area or outside an area. We're trying to balance all that with the raw kind of screen clutter, making sure that game itself, there are not too many effects going on all the time and blinding you.
Newman: How difficult is that balance, now that you have 30 characters?
Goodman: Strangely, it takes quite a long time to nail that down. A lot of people have different opinions.
We'll go through a play test, and it's like, "Oh man, I was playing with Baptiste, and you know, Winston's dome was there, and then Symmetra ulted, and I just couldn't see." And it's like, "Aw man, okay," so maybe we pare stuff back. And then we get the other feedback, "I keep throwing down the immortality field, I didn't even know where it was. I didn't know where I had to step in. I didn't know that I was in it."
So there are a lot of visual effects, plus sound effects are a big deal too, trying to balance that, how much weight of information is on the sound or the effects of something. And that typically can take quite a long time, it takes us basically until the end.
Newman: What tweaks did you make to Baptiste in particular?
Michael Chu: I remember when we were designing him, there was a lot of iteration on the exact pose. Soldier 76, Ana, Ashe, all the gun-carrying heroes, there's a different attitude, we make sure those are different. And then I remember we also tweaked his color scheme a little bit, which is how we ended up with his current one, to make sure that he really popped from the rest of the lineup.
Goodman: Also even something as simple as the rough outline of his weapon. Originally, it was more rifle-like, almost a little more of a Soldier 76. But we really wanted to push away from that. I mean, one thing was we wanted to make sure it was different than Soldier so it was very clear.
But also we wanted to emphasize the grenade launcher aspect. He is primarily a support and a healer. Even though the grenades are his secondary fire, it just felt more natural as a control scheme for the game. We wanted those to be more prominently featured as a character-defining ability, so we kind of pushed the gun as having a bigger drum for the ammunition, so people want to pick grenade launcher than actual rifle.
Newman: You're designing character No. 31 now, versus character 12. Do these considerations become more difficult?
Goodman: Yep. That permeates kind of every part of the design of the character at this point. From the raw game design stance it's like, "Yeah, we want to make sure that the characters feel like they're doing different things, we want to feel like they're different variations of each other."
But also, all the visuals, make sure they're all very clear, you know the difference between each person's abilities, who they belong to. Originally with the game it was a little bit easier, because we didn't have that many heroes. Whenever you saw a certain color effect or something, you could pretty much associate it with a certain hero.
But we have so many heroes now that there are only so many colors. It's definitely a really big challenge and something we're always fighting with.
Chu: In the audio aspect too, making sure the character sounds different enough, so when you hear them, you're like, "Ah, I got it." Right? There's a lot of parts that go into that. Vocal texture, vocal tone, manner of speech. But we're definitely reaching that point where, well, we've been in it for a while. Making sure they're different is really important.
Newman: Is there a hero that was especially difficult from that perspective, that you spent a huge amount of time iterating on?
Goodman: I think each hero kind of has its own big challenges in that regard. Genji is an example. We knew who we wanted Genji to be right away. It was just cyber ninja guy, cool sword, dashing around, you make the elevator pitch to somebody, everyone's on board, you get it.
It's harder to do a hero like Baptiste or Orisa, or something that's a little more high-concept. There's not a lot of direct reference you can associate it with. So you sort of have to, you know, lean on the visual a little more on those.
Newman: But did Genji's development have any challenges?
Goodman: Genji is an interesting example, because while he was pretty easy on the art side -- he had this sort of green color that we weren't using anywhere, besides maybe Lucio -- so we had the color scheme thing, which was nice.
But from a game design standpoint, it was super rocky. We tried to make melee work for a long time as his only way of attack, so he had to dart out all the time. And that just really didn't work. We thought it was going to work, we'd already made Reinhardt at that point. It seemed like we'd already got melee covered, this should be easy. It just was not easy.
Theoretically can do a lot more damage in the tank, so he's supposed to be able to kill people easier. It was just, in a [first-person shooter], using a melee weapon like that is really challenging.
Newman: How did you figure out the problem?
Goodman: We wrestled for a long time -- he even had spells and backstab, all these different kind of moves. We ended up having this epiphany, when we're just like, is there any way we can really make the sword feel really cool, to make it do a lot more damage, and just make him limit it?
So that's when we ended up making his ultimate, and getting him shurikens. He wasn't originally designed to use shurikens actually. We had to redesign his arms to be able to hold these shurikens and deploy them. From a game design perspective, he was really, really difficult.
Newman: So we've talked about heroes that were driven by gameplay. What characters were driven by the story?
Chu: I think Ana. We were like, "Hey we want to do a sniper, skill support character." And we're like, "Oh, we have this character named Ana. She'd be great." We'd always thought that Ana could never make it into the game, because we already had one pretty sniper who was locked in.
Goodman: The prototype was very different. Imagine a guy with a bunch of bottles and he drinks a potion and does things and then injects himself. But once we began, it was like, "Oh, we should do Ana." It definitely changed the whole direction.
Chu: I feel like once we're sort of in the final direction, everything comes together very quickly. Brigitte is a character who existed in the story for a long time, we knew we wanted to do more with her, and then we got this idea: "Hey, what if we did like a Paladin archetype?" Immediately, we're off.
I feel like Baptiste had a little bit of that. We've made a lot of heroes. And so when we're looking at making a new support hero, it's how will this hero be different? One of the things we really latched on to was, could we come up with a personality profile that was different than the other heroes? I think we have a lot of altruistic heroes. I know Moira is obviously not an altruistic hero, but she's not a soldier type. And Ana, maybe one day was, but now doesn't feel that way.
I think we were really interested in trying to find a new archetype for a healer. That's how we ended up with Baptiste. He is more lethal, he's willing to go out and kill people, he doesn't have as happy a background. Well, I don't know if any Overwatch characters have happy backgrounds. [He laughs.] But that's how we settled on his personality and who he was.
Goodman: Soldier 76 might actually be the most direct from the story. Soldier wasn't at all looking at the play concept. It was just literally like a character, and we're like, "Eh, just make him like a rifle run guy," more of like being in Call of Duty. And we started experimenting with that and it came together.
Newman: How much of a character's story comes from their abilities, and how much from what you need in Overwatch lore?
Goodman: We really like to look at the gameplay mechanics as a bit of an outline for how the character should be. I think some of our most successful characters have a good relationship between their personalities, the way they see the world, how they behave, and how their gameplay is. Reinhardt is a guy who protects people, and he's a little brash. So he's got his charge and goes in there--
Goodman: Exactly. I think Baptiste is one of these where we really want that sort of synchronicity. Because I think it helps make the heroes more interesting, engaging. I think then there's a lot more looking at the universe, defining a place for him, figuring out how he connects to different story lines. And then there's always the simplification task for the character.
When we were in development, we talked about having Baptiste be in a new mercenary group, but ultimately, we felt like it was just like too much. He needed to be a little closer to the core of the story. So we altered his backstory so he's a member of Talon. I think it makes for a stronger character.
Newman: Are there other examples of heroes whose backstories changed?
Goodman: Pharah used to have a really crazy, much more in-depth story. Her suit had artificial intelligence in it. She was actually -- we didn't do this until Wrecking Ball -- but Pharah was going to be the character that had two voices. She was going to talk to her suit.
She had this whole story about figuring about the mystery of what was going on, and she was like adventuring around the world trying to find the truth. Ultimately, we made that a lot more simple.
It happens with every hero. Every hero starts much more complicated than where we end up. I think it's actually to the benefit of the heroes, because one of the nice things about Overwatch and its characters is that we can always add depth to them later. At the outset, it's helpful for these characters to be understandable and simple and graspable.
Newman: How do you decide which characters to take a deeper dive into?
Goodman: There are some characters that we delve into because we feel like developing their story will help to push forward and elaborate on the overall story of Overwatch. There are characters who get development because they are adjacent or involved in events that we're going into for other reason.
A lot of characters are simple, but you have questions about them. That's very much by design. A lot of the time what we want to do is explore those characters and add that layer of depth onto them. That was very much the idea behind the Baptiste story. There are some characters who are often asked for, but we haven't gotten around to. And sometimes there's just a story that we feel is awesome and we are passionate about telling, so we do that.
Newman: What hero backstories do players ask for?
Goodman: I think it would be really interesting to get into, and I think it's often requested, to get into Mercy a little bit more. I think Zenyatta is commonly asked for. I think he is a really interesting character that really fits into the world in a way that could show a lot about the world and robots and Omnics and how they're treated. So those are two that really spring to mind. I wish we had time to do all of them. I think all these characters are itching to have their stories told.
Newman: You have 30 heroes now. How many is too many?
Goodman: That's a real concern. There are a ton of perspectives on that exact point. On one hand you have some people comparing us to more MOBA [massive online battle arena] games like Heroes of the Storm, and they typically have really huge rosters, upwards of 100-plus. Hopefully we won't have that many heroes. Or you compare us to more of a fighting game style, more of a Street Fighter. We're a lot closer to that I think. They have pretty big rosters too.
But the big concern is, we want to make sure that we're not creating just color variations of mechanics and characters we already have. We want to make sure that every hero stands alone both in gameplay and as a character.
From a pure gameplay standpoint, it's a little bit inevitable that we'll start reusing some mechanics. I remember when Orisa came out, a lot of people were like, "Ah, you're just doing Zarya ult again, as a normal ability to pull people in." I think once people play they realize it's pretty different. There's a little bit of overlap in the mechanics there, so we have to be careful that we don't push that too far and characters overlap too much.
Newman: What lessons have you learned in the process of developing these 30 heroes?
Goodman: We played the game for a while before it came out, obviously, play testing. We thought we had a pretty good idea about how the game was played, and what our own metagame was. Then of course once it comes out and the audience at-large starts playing it, we found out we were pretty wrong.
I think that was probably inevitable, and probably most game developers have a similar story about whatever game they're working on.
Newman: How did your view change?
Goodman; We used to have roles of defense, offense, support, tank. That showed how we thought the game might be played. Since we have a lot of attack and defense objects, we probably would say, "The defensive heroes on the defense, and the offense heroes on offense," and since we weren't restricting, you might mix and match a little bit. But we thought the good defense was like a Bastion and Torbjorn there. It doesn't really work out that way, it turns out. Those heroes we ended up reworking.
We learn over time, lessons about how the game is actually played, and how players want to play the game. So we make adjustments to old heroes to try to keep up with how the game is going, rather than try to direct it: "No, no, no, we want it to work this way. Go this way." We're going with the flow.
Newman: Does the similarity of characters used on offense and defense, and between maps, in competitive play disturb you?
Goodman: We would love to see more diversity. I will say, so far in this Overwatch season there's been a decent amount of what they call "ghost meta," which is three tank and three support. But even beyond that, we've seen Torbjorn picked and Symmetra picked. Granted, they're not picked very often, and they're often trick plays, but we've actually seen quite a bit of diversity this season. But from a balancing standpoint, it's definitely always the goal. We have discovered over time, it's less about the game modes that cause people to play different compositions of heroes, it's more about very specifically the level design and the engagement distances and the flank route and stuff like that.
Newman: Can you give some examples of maps that change comps?
Goodman: Areas like King's Row are very tight. You're able to play Reinhardt very effectively. Even in the previous seasons when Reinhardt wasn't played very often, he was consistently played on King's Row, because the choke points are very tight and you can guard your team very effectively. With wider points like Numbani, it's a little harder to play him. You have lot more high ground to attend to, so you kind of want to play Winston, D.Va in those cases.
In some of our latest maps we've done a lot better with developing different areas of the map for different heroes. Hollywood, granted not the newest map anymore, but Hollywood was an example where we were trying to take those lessons and apply them. The middle point of Hollywood through the streets is very different than the first points.
Newman: Are there other maps or heroes you're keeping an eye on?
Goodman: We're still learning a lot about how players want to play. One thing we talk about now -- Junkertown is probably maybe one of the only places where Bastion is allowed to play. And he gets a lot of play in Junkertown. Probably 90 percent of his playtime is on that map. We're trying to figure out why that happened, and also on Route 66. There's a lot of similarities between those two maps: both payloads, they both have a lot of the same sight lines. We have a lot of stats coming in, and hopefully the more we figure out from that, the more diversity we'll see.