Blizzard Talks Overwatch’s Balance, Progression, What Makes Tracer Rad

We’d be lying to you if we said that we weren’t absolutely stoked for Overwatch, the upcoming hero-shooter from Blizzard. The combination of tight shooting, frenetic team-based combat and some downright awesome design decisions have made Overwatch potentially the most promising multiplayer shooter of the decade thus far. With Early Access to Overwatch’s Open Beta beginning at 4PM PST tonight (the flood gates officially open to the public on May 5), we caught up with Assistant Game Director Aaron Keller to learn the answers to all of our pressing questions.

[Hardcore Gamer] Out of curiosity, how many people are working on Overwatch? I’m often shocked when games developed by larger teams manage to maintain any semblance of personality or flair as a result of the sheer number of voices piping it at once.

[Aaron Keller] I don’t know the exact headcount of the team. It’s under one-hundred people, and it probably started closer to thirty or forty people before we ramped up to where we are now. The cool thing about the team is that everyone on it is such a high performer and so great at what they do. We have a core value at Blizzard called “Every Voice Matters,” and a lot of people on the team are really passionate and outspoken, not just about [Overwatch] itself, but about things they’d like to see in the game. I think we have a really good collaborative environment where people feel like they can throw ideas out there and they’ll be listened to. That’s how we get a lot of personality into the game; our heroes have come from all over the place.

When we started on the project, we actually sent out a team-wide brainstorming email where we just asked people to send hero ideas. We maybe got one-hundred or two-hundred of them from across the team, and some of the heroes in [Overwatch] now are based upon the original ideas that people sent in.

Was Tracer one of those ideas? That person deserves a raise.

Oh! Then that was me! [laughs]

No, Jeff Goodman had a big hand in designing Tracer. He’s our lead hero designer, and she was the first hero that we designed on the project. There’s just something about her abilities that just worked together, and how different she feels when you’re playing as her compared to a lot of other games. When we implemented her, and got her into our first map (which we were developing at the same time), we started getting guarded enthusiasm about the game. With one hero and one map, we were having a ton of fun.


I can’t think of any other instance in a game, outside of maybe Super Time Force, where you have to think about your actions moving forward and consider how they would replay backwards. What went into her design, since you have to think about balancing your signature character’s accessibility with higher-level play?

She’s pretty interesting because her two primary abilities do not work without each other. What kind of makes her Tracer is how the Blink and the Recall work together. We have some heroes where the abilities might compliment each other, but they might not be as tightly tied together like that. From the beginning, she originated as a “whole” hero with those three abilities (the Blink, the Recall and the Pulse Bomb). With other heroes, we’ve gone through iteration where we thought, “Hey, we really like these two abilities, but we’re going to work on this third one and change it all later.”

Bastion is a good example of that. He’s a hero where the original ultimate we had on him, which we actually showed at Overwatch’s first BlizzCon, just wasn’t really working. We actually had an internal joke: what’s the ultimate of the week for Bastion? It felt like we were doing that for half the project.

One of the more unfortunate things about multiplayer games this generation is that a lot of them seem to have their moment in the sun before fizzling out. What do you think it is about Overwatch that will prevent it from being the latest game to fall into this trend? From a design perspective, what is going to make you more like a Team Fortress 2 than an Evolve?

First and foremost, what we focused on the entire project was just making the moment to moment gameplay of Overwatch as exciting, polished and fun as it can possible. What I hope people come away with when they play the game for the first time is the sense of having incredible fun and joy, as well as a sense of wonder for this new universe that we’ve created.

Maybe it’s a little naive to say it, but, for me, I’ve been playing the game now for two years at home, and I come back to it every night because it’s fun. We are trying to make, first and foremost, an incredible fun and exciting game, and on top of that, we have other systems in there. We have a progression system and it works together with a loot box system that we have so that you can unlock different cosmetic items for your character. We also have a competitive mode, which is sort of like a ranked mode. So we definitely have a lot of systems in [Overwatch], but at it’s heart, we just want the game to be incredibly fun.


What are some of the biggest things that you’ve learned from the Closed Beta? Are there any things that made you think, “Wow, we’re in such a vacuum designing this game and I never would have thought that this could happen?” I’d imagine there’d have to be something, since there’d be no point in having a Beta otherwise.

That’s a great question. We’ve definitely changed a lot of things since the Beta; we’ve made a ton of decent tuning changes to the heroes. One of the things that I really wasn’t expecting was the community that formed around the game, and how quickly they were able to get good at the game. We had daily tournaments with extremely good Overwatch players playing the game, and the design team watches every single one of them and we shocked at how good they were. We like to think we’re pretty good at the game, since we’re developing it, but these people are on an entirely different level. It made us realize that the tuning that you need to do for the low level on a hero isn’t always the same as the tuning that you need to do for the high level.

We really had to analyze what made a hero balanced for the entire population rather than just one part of it. When the really good players started getting in, we actually had to start making changes to some of the heroes. We made changes to Zenyatta and Symmetra, which were both spurred on by really high-level gameplay.

One of the things that I’ve heard about Overwatch is that people have gotten really good, really fast. How does the back-end matchmaking take this into account? The last thing you’d want is for someone who waits until launch to jump in and get absolutely throttled as a result of all of the really high-level Beta players giving them no chance to get their feet wet.

While I don’t have all the details of the underpinnings of the matchmaking system, because it’s incredibly complex, we have a really talented engineer on it. He did all the matchmaking systems for all of the other Blizzard games, so he’s really good at doing what he does. However, our values as a team are to avoid that as much as possible. We want the game to be as balanced and as fun for everyone as we can possibly make it.


How does Overwatch strike a balance between being really colorful, fun and joyous while still maintaining an intense competitive nature. In other words, how do you prevent Overwatch from feeling so upbeat and vibrant in its personality that it gets branded as a “kids game?” After all, it is a really intense, competitive first-person shooter at its core.

That’s a really interesting question! Early on in the project, we knew that we wanted [Overwatch] to have a competitive nature to it, but in a sense that two teams would come together and have a really great match against one another. That was the goal on that side, but we are always driving to, first and foremost, make it a great shooter. There was a point in the development where we conducted our own internal tournament with members of the team, and we had a hunch that the game was really fun to watch and become a great competitive game or a great eSports game. That was the first moment where we really felt confident about it, since everyone on the team stayed late every night to watch every tournament match, and they were so fun to watch. Once we got to that point, we started really focusing on the competitive side of the game.

It’s kind of interesting, since the way you phrased the question almost makes it seem like we were driving towards making an eSports-caliber game, and that actually wasn’t the case! We just wanted to make it great, but it’s awesome now that it’s there. The big thing that we need to do is make sure that it’s balanced between all heroes. We need to make sure that all maps are as balanced as possible, and we do a ton of work on that. I think what sets [Overwatch] apart from other games, and where the true competitive magic really happens, is the interaction between all the heroes on the field and the way that you coordinate with your team. It’s a team-based game, so as long as you’re coordinating with your team and you guys are picking the right heroes to not only accomplish your objective, but also to counter the enemy team’s composition, then you’re kind of playing it the right way.

What would you say the biggest challenges were during the course of Overwatch’s development?

We had a couple of them, but overall the project went pretty smoothly. Our team is pretty remarkable and can iterate on ideas really quickly.

I wouldn’t call it a “misstep,” but one of the things on which we iterated more than we normally would was progression. Early on, [Overwatch] had a totally different progression system than it does now, and it would make changes to your character’s abilities, so there was a little bit of power-progression in it. What we found was that one of the great aspects of Overwatch is how easy it is to read what’s going on on the screen and what’s happening in combat. We put a lot of work into what we call “combat clarity,” so once we started messing with people’s abilities, it suddenly became harder to tell what was happening on the battlefield. You’d see the Reaper and wonder, “Is that the Reaper that heals while he’s in Shadow Walk or is it a different Reaper?”

We put a lot of time and thought into that and decided that it just wasn’t true to what we thought the game would be, and that’s how we started on the road to the progression system we have now.


That’s a huge relief to hear that. Obviously with a progression system you’re trying to create a treadmill of sorts, but when so many basic gameplay tenants are locked behind progression, it becomes a treadmill you don’t even want to get on in the first place.

Yeah, we realized that we didn’t want to have power progression in any part of [Overwatch]. We want everyone to be even on the battlefield, and it’s not going to be about how long you’ve played Tracer or how long you’ve played on your account. The whole point is to hop in and find out how good you are.

Here’s a very specific mechanical question coming from someone who hasn’t played every character: do all characters lack the ability to aim down sights? If so, what factors into straying away from including such a genre staple in Overwatch?

Widowmaker, the sniper, does and that’s it. We had a couple of meetings early on that had to do with things like aim down sights and sprint, where we deliberately chose not to put them on all of our heroes. Those ended up becoming abilities that we could give to a hero to make them feel more unique. The other thing is that we wanted our game to feel like it had more of a “rumble” to it when you’re fighting on the battlefield. To us, we’re hearkening back to the Quake and Unreal Tournament days where the movement was really fluid. Something like aim down sights, which is really great in Call of Duty – and I love Call of Duty and still play it – but we strayed away from it intentionally to do our own thing.

We want combat to feel really fluid, and we want there to be a really cool rumble or brawl when you’re fighting other characters. Aiming down sights typically doesn’t give you that.


With any character-based title, there’s always a “meta” that forms, be it from the developers or the players. From a development perspective, how do you work with the meta that is inevitably going to arise in Overwatch?

Well our goal is to have all of the heroes viable, especially in the competitive scene; we want to see each of them played. I think that when this happens, the matches obviously get more dynamic, but you also feel as though there are more available counters for you to pick from at any moment. That’s our value, and we’re working towards that.

We’ve had a few metas develop on the tournament side. One of them was the Dual-Tracer/Dual-Zenyatta meta. We took a look at it, and found that it was a bit too strong, so we made some changes to Zenyatta. Now, it hasn’t entirely disappeared, but it’s no longer the dominant meta for tournament play. If we could, we’d like there to be ten to twenty really strong team compositions out there that you’re kind of picking between based on the players on your team, the other team or the map you’re on.

Obviously there are a million different team compositions and ways to play Overwatch, but what are the core trends you’re seeing with the best players. In other words, if someone is looking to get really good at Overwatch, what do they generally have to do?

The single biggest thing, if you’re new to the game, is to coordinate with their team. One of the things you’ll see with new players is they’ll die, they’ll respawn and then just rush right back into the battle. A lot of the time they’ll be rushing into a one-on-six, and they’ll get annihilated and they’ll do this time after time. The biggest thing is to say with your team, coordinate with them; when you need to do a big push, coordinate that big push (or coordinate that big defense when you need to execute a big defense). For newer players, that right there will, far and away, make them better than most of the teams I’ve played with.


What goes into making maps that feel appropriate for all of these heroes at the same time. In terms of creating maps and creating heroes, what is the chicken and what is the egg? Which comes first, and how do you design Overwatch’s maps to maintain the balance that’s so critical to every moment of its gameplay?

It’s tough to make maps during development. Now it’s easy to make maps, since we have all of our heroes. The first map that we built was Temple of Anubis, and it was built while we were designing our first hero, which was Tracer. We were probably four or five maps into the game by the time we went to our first BlizzCon and had twelve heroes. It can be really tricky designing maps for heroes that are unknown.

Early on in the project, our hero designer Jeff Goodman came to me and we had a couple conversations. During one of them, he said, “Just so you know, there’s going to be a flying hero in the game, so just be prepared that people are going to be able to get anywhere on any of your maps.” We’ve tried to keep that possibility open. Even when we were designing Temple of Anubis we knew that we really wanted there to be strong teamwork. In the first part of that map, there’s a choke that each team needs to get through, and there are these big gates that are there before you get to the first control point. We very intentionally made it so that every hero has to funnel through that area. Certain heroes can bypass it pretty easily; Tracer can squeak through different areas and Winston can jump over the top if he wants to, but the majority of the team has to go through the front of that thing. We really wanted to force a team to work together to basically crack the nut that’s on the other side.


Overwatch is currently set to launch on May 24 on PC, PlayStation 4 and Xbox One. Be sure to stay tuned over the coming weeks for more impressions of the Open Beta, as well as our final review towards the latter part of the month.