In the Company of Heroes: An Interview with ‘Overwatch’ Designer Jeff Kaplan


This story is over 5 years old.


In the Company of Heroes: An Interview with ‘Overwatch’ Designer Jeff Kaplan

We discuss new characters, the importance of the meta, and the social responsibilities of game developers in the 21st century.

It's been almost 12 months since Blizzard's class-based first-person shooter Overwatch burst onto the scene, its cast of colorful characters possessing both uncommonly inventive weaponry and an array of catchy quips. Since its launch in May 2016, the game has grown in popularity to today reach some 25 million registered players, and its heroes have become one of the industry's most prolific meme machines. Overwatch isn't simply played a lot—it's become a landmark of online gaming culture.


A single figure has guided Overwatch fans through the ups and downs of each new hero's release, and the buffs and nerfs that keep the game as balanced as it can be: lead designer Jeff Kaplan. Through a series of online developer updates, he has remained in constant communication with players, and become a celebrity of sorts in an industry which generally eschews such recognition.

To some, Jeff is Blizzard Entertainment. I caught up with him in London for a conversation about Overwatch as it stands, and where it could be striding next. Note that this interview was conducted prior to the leaking of the new Overwatch: Insurrection event.

Orisa. All screenshots courtesy of Blizzard Entertainment.

How do you think your latest hero, Orisa, who you introduced in March, is faring? She's suffered criticism for not being "tanky" enough.

Orisa was our best hero release to date: In regard to the announcement, the engagement players had with the announcement, and the timing of the announcement, I think it went really well. But I think it's too early to tell balance-wise with her. We're in what I call the "Ana state" (Ana being the game's first post-launch hero, released in July 2016), where the community is claiming that she's underpowered and I'm not sure that she is. If you remember it took two or three months and multiple buffs to get Ana to really be recognised to the point where we were buffing her and didn't really think she needed to be buffed.


With Orisa, what I think is going on is that she's very new and there are a lot of non-tank players using her. Some of these non-tank players have a perception of tanks that is not true, i.e. that you can survive anything. Take any tank—if you play Reinhardt, if you play Winston, if you play D.Va—and if you jump into a one versus six situation, then you're just going to die. There's no way to survive that, it doesn't matter what tank you are.

I'm watching a lot of the Orisa players be a little bit overly aggressive. I think it needs to settle down a little bit and for us to see people really master her and understand that just because we label her a tank, that doesn't mean you can just sit and take Soldier: 76's gun to the face.

It's also the case that our community is heavily influenced by the pro-scene and not all tournaments have had Orisa available yet. What will be very interesting in particular is when the Korean players start playing with Orisa. They're very creative and tend to make everybody work in very unique ways. I really want to see the pro-scene and how they influence the rest of the community in terms of how should Orisa be played. Her kit and her mechanics seem really strong. If anything, she seems like a character that will just get a few numbers tweaks here and there that will put her in the right spot.

You've added another tank into an already tank-heavy meta. Why does Sombra still not feel relevant?


So we buffed Sombra, especially the hacking in order to help with the tank situation. If Sombra gets played, especially if you're playing one of these tanks, then it's devastating. Not being able to put up your shield is pretty much the worst thing ever. Being D.Va and not being able to fly away is really devastating.

We also were tweaking Bastion to be a counter toward tanks. We think Symmetra with her right click, her ability to get through the barrier, was really helpful. She's not a tank-buster necessarily, but she can help a little bit.

We're seeing a lot more prevalence of Pharah these days. She's really coming into the meta, and Pharah is a wrecking ball to people like Reinhardt. People like Orisa. She gets behind them and even if the Rein is blocking then he's exposing himself to everyone on the field. We also made some changes to Junkrat to help with the situation as well. We've always felt that Junkrat is a very strong anti-barrier character. He wasn't getting as much play as we liked, so we made a few tweaks to him, like not taking his own grenade damage.

"I love being held to high expectations. And part of the thrill of what we do is trying to meet them." — Jeff Kaplan

Like the rest of Blizzard, your team is held to stratospheric standards by its community. Everything you release inevitably gets torn to pieces in some quarters. How do you cope?

I love being held to high expectations. And part of the thrill of what we do is trying to meet them. I know we won't always meet them, but I think that the type of developer that's attracted to be at a place like Blizzard wants that challenge and wants to try to make the best thing possible. We think it's totally fair that people criticize us—it just spurs us to do a better job next time.


All of the nitpicking comes from a place of passion. People who really don't care about Overwatch, or don't play Overwatch, don't talk about it; they're not part of the community at all. It's not like there's this guy who loves some other game and hates Overwatch so he spends all his time on the Overwatch subreddit talking smack about us. There's just not enough free time in the world, so we know that when people are aggressive or critical toward us, even if they don't know it, even if they're not communicating in a positive way, I look at them as a really dedicated, passionate fan. They might think that they're hostile and over it or whatever, but I'm like, I think you're really into what we're doing.

Any creative person knows that you learn more from that critical feedback than you do simply from praise. It's nice to hear praise and what you're doing right, but you learn way more from people being critical, and it gives you not only a thick skin but a great perspective on your work and how to improve upon it. There will never be a point where Overwatch is perfect and we can all go home and pat ourselves on the back—it's always a work in progress. The only way we're going to keep pushing ourselves is if we hear that criticism.


Turning to where the game's set: To me, utopias feel few and far between in games now, because they just aren't all that interesting. Why is the world of Overwatch still so compelling?


Well, we don't think of it as a utopia. We think of it as a bright, hopeful future that's larger than life. The way we always describe it is the tourist's view before the tourist has actually travelled to the location: what we imagine things to be. The one I always reference is Hollywood. Actual real-world Hollywood is not the most exciting place to go, and it doesn't look beautiful and art-deco like Hollywood did back in the glory days. For us it's more about somebody's fantasy of what a place might before they've been there.

Also there is conflict [in Overwatch] and that's what differentiates it from a utopia. If you look at [the map] King's Row, there's a lot of strife going on between humans and omnics and we're hinting at it. You have to have some conflict in a game world for the game world to be interesting. It also justifies that phrase we use within the game, which is "it's a future worth fighting for", so it's not a perfect future, but I understand why I want to live in this future and why we need to make this it work. If you take a post-apocalyptic world and just bomb everything out, beyond just fighting for survival I don't know what else there is. It's just a grim reality.

Do you think there's a need for more darkness in the game, going forward?

I think it's always interesting to have a contrast. We don't want a bunch of lawful good paladins only. It's hard for a lawful good paladin to be lawful and good if we don't have a chaotic evil to contrast against them.


One thing we see that the fans want more information about is [the in-game terrorist group] Talon. They want to know more about this and they want to know more about the Omnic Crisis. These are things that we're going to focus on more and bring to the forefront, but I think there's also some interesting moral ambiguity in some of the Overwatch heroes. Obviously Reyes (Reaper) is the most dramatic example of that: he's gone from one side all the way to the other in a very dark way. But there are some other characters like Genji, like McCree, who have wavered. You can't just say McCree is a just a good guy, as he's very questionable. He basically hung out with Overwatch, because he was given a choice: Join Overwatch or rot in jail, so his motivations weren't particularly altruistic in the first place. I think think this is what makes the game world interesting to players and lets them engage a little bit more.

"Overwatch at its core is about inclusivity; it's about this vast diversity in humanity. Once you have broad reach like that, I think it's very important to think about the people that you represent." — Jeff Kaplan

You introduced Tracer's girlfriend through the online comics, in December. What social responsibility do you think games have, now that they exist in a more culturally prominent space?

It's a very interesting and tricky topic to talk about. If it's going to be a great intellectual property, you have to stay true to the creative integrity of what you're doing. So, I feel like you should never do anything in a pandering way. Overwatch at its core is about this inclusivity; it's about this vast diversity in humanity. And that's not even just in reference to gender or sexuality or race, or anything like that—it's also in terms of gameplay.


With that said, I think about the place that games have now taken in popular culture and how influential we are. It used to be that games were looked at as a toy. Mom and dad would buy a kid a NES and say, "Go play." Our generation has grown up with games being more than just a valid form of entertainment—it's a dominant form of entertainment. A lot of people of our time would rather engage with video games than movies, books, or music.

Once you have broad reach like that, I think it's very important to think about the people that you represent. With Tracer it was interesting because we knew that she had a girlfriend and we received a lot of pressure from our fans. They knew that we had characters that represented the LGBT community, and they said, "Just tell us who it is." But we thought that felt really unnatural—we don't tell you who Torbjörn's dating until it's relevant to the story.

Until Tracer, the only character we had ever revealed a significant other for was Widowmaker, because it was core to her origins. So we were really conflicted—we felt like we didn't want to pander, and we wanted to be true to our characters and only reveal more about them when it's a natural part of the game story's evolution.

Related, on Waypoint: The Sad Truth of the 'Overwatch' Plateau 

At the same time, now that we've revealed that Symmetra is autistic, you should see the amount of mail that we received from either autistic people or parents of autistic children, and how important it is to them. And we sort of realised the same thing about the LGBT community, that there is an importance of knowing, of validating. Overwatch is a game with 25 million players, and Tracer is on the front cover. There is a statement in saying, "Y'know that awesome superhero that's on the cover of the box that we all wanna be when we grow up? Guess what. She happens to have a girlfriend and that's just normal and let's just get over it and move on." The amount of positive feedback we've had on that has been sort of amazing.


I don't think games should feel that they are forced to do anything, but I also think that with a game that represents as much diversity as Overwatch does, it's extremely important to the people that you're speaking for that you do these things in tasteful way that is related to the game. It shouldn't just be, "Oh, by the way, did you know that Reaper is a vegan?" Just so that all of our vegan players out there have something that they can relate to.


How does it feel making a game like this, about internationalism, about tolerance, in a time that, so the news informs us, is increasingly about protectionism and intolerance?

We didn't set out to make a political game, but in a weird way, I don't think that we've ever felt more proud about the stance the game has naturally taken. It's so easy to be scared and skeptical these days with everything that's going on in real-world situations and real-world politics. The magic of Overwatch is looking at all the different people in the world, and instead of looking at them through that scared, skeptical lens, going, "Well what's awesome about this person?" It's easy to turn on that skeptical lens and focus on difference negatively, but there's so much that's cool about each and every one of us.

It's weird that we're more accepting of place than we are of people. Foreign locations are very exotic to us, and we want to take that same sort of easy passion that is evoked when people think of a foreign place and apply that to a person. We've got our lady from France, our gentleman from Sweden, or our lady from Korea. What is their world situation, and why should I love her and want to be her? Or why should I love him and want to be him?


It's unfortunate that in a game about diversity the meta-game then makes certain heroes unviable. Will there ever be a solution for that?

Two things. One: the diversity of the characters and the player fantasy actually helps the meta in a weird way. If every one of our characters was a generic soldier person—picture the past 10 years of shooter covers—then I don't think you'd have Torbjörn and Widowmaker and D.Va mains. I think you'd have whatever the flavour of the week is, the "AR-15 main" or something. But weirdly we have people who just love that girl from Korea, and love her voice lines, and love her skins and love her fan art. They're going to be her whether she's meta viable or not.

Obviously at higher levels of play, people take the meta more seriously and aren't just making a purely emotional choice. But I feel like that emotional choice has lead to more diversity than we would have in the meta to begin with. The meta then becomes more of an issue of game balance—if we see something too dominant, if we see something in the hero picks. Most recently we're making changes to Lúcio, those are on the PTR (public test region) right now. Because Lúcio, especially in competitive play, is almost regarded as a must-pick hero for various reasons.

"I would say that we're at about a 70% of where we want Play of the Game to be. It catches a lot of cool stuff, but it's nowhere near as awesome as I think it will be." — Jeff Kaplan

You can have a huge degree of control over heroes, but players often prove trickier to handle. How do you go about keeping them in line?

A lot is in the core design of the game already. We don't have a system in place that is like a karma system where you're voting people up or voting people down. An idea like that is something we're interested in, but it's difficult to make one meaning, so we look to the core game design of systems, like the scoreboard. We took a lot of heat from the hardcore shooter community early on for that. They wanted to be shown everybody's kills, everybody's deaths. We call it "the spreadsheet" internally—people just wanted the spreadsheet. We feel like that decision directly reduced toxicity in the game, but ironically that wasn't our prime motivation behind doing the scoreboard that way. Our prime motivation was that in Overwatch with the hero diversity and the role diversity there's no way to measure everybody's success on the spreadsheet, and have it make any sense. If you're Mercy, for example, you're going to have zero kills.

The end-of-round cards are something I think a lot of people also sometimes question: "What are the cards? They don't really do anything." The cards are what we call a diffusion moment, where it diffuses the intensity of the match. You're commenting both on your own team and other team members, and we're just trying to make people more aware. For instance, if a team gives their Genji a hard time during a match they may then realize on the cards that he contributed more than anybody, even though he was hardly seen. There are also cases I have seen at the end of rounds where the cards come up and the whole enemy team will compliment the other team's Mercy, and she'll get the epic card. That's what I call those diffusion moments, where you're just trying to cut the edge off.

Play of the Game is the same way. We've had a lot of players say that they wish they could skip it, or that only their team should get it, or that only the winners should get it, or that every player should have their own, or that it's just really stupid. Sometimes it will pick a dumb moment that was just Orisa wandering around the map, and we're totally okay with it because what players don't realize is that there's a social moment that's happening there. Let's say that Play of the Game is just broken and shows Orisa just wandering around in a circle: the 12 people commenting "Blizzard's dumb, they should fix this" are actually having this social moment, whether they realize it or not.

But really, are you happy with Play of the Game at the moment? It often feels underwhelming.

There are a lot of changes we would like to make to Play of the Game. We'll continue to iterate on the algorithm to make it smarter and pick better things. We also have these ideas right now, like Sharpshooter and Life Saver, which are actually pretty cool: Sharpshooter, for example, demonstrates multiple long-range shots. The problem is that it's they're not shown very cinematically right now, so we would like to make some changes so that the camera tracks the action better. Same goes for Life Saver—a lot of the time you don't even realize what's happening, but somebody is doing something that saved somebody else's life. Without some changes to the cameras and the cinematography of the moment, those aren't great right now.

I would say that we're at about a 70% of where we want Play of the Game to be. It catches a lot of cool stuff, but it's nowhere near as awesome as I think it will be some day.

Follow Justin on Twitter.