Hack-Man: An interview with Watch Dogs' creative director

Watch Dogs is poised to be a success on the next-generation consoles, but it's clear that Jonathan Morin, its creative director, sees it just as a step toward true "next-generation" game design, not its full flower. 

As he closes a five-year chapter of his life, from prototype to completed game, Morin was reflective about not just what he made, but where games are going as a medium. In this interview, conducted at Ubisoft's San Francisco offices during a recent preview event for the hotly anticipated title, Morin speaks candidly about how he sees games and players. 

He's a big believer in emergent gameplay and in a shift towards inquisitiveness on the part of players. He sees a future in which players pull what they need from games, rather than having it pushed upon them, and in which developers accept the fact that huge swathes of the player base will never see everything a game has to offer. 

Players should not be guided: "It's almost like you're giving a music instrument to the player, and he's playing the music to you, and you should shut up and listen to the music," says Morin. 

You told me that hacking the traffic lights was your first successful prototype, and that hacking is Watch Dogs' core mechanic. Before that, what assumptions did you start with when making the game? 

Jonathan Morin: The team wanted to do an open city game. The team wanted to explore an open, urban environment -- that is what the team was all about. I'm a big systemic kind of designer. Emergent gameplay is just a style of design -- either you like it or you don't. It's hard for someone who loves it profoundly to do something else. So as a team, we were also very into that. 

So we were searching, and at Ubisoft... one of the first things you need to deliver is a differentiator. You need to define something that is not just a gimmick, but can deliver new gameplay in the entire game... The core challenge is always, "Let's find something that will change the game." So hacking became that. The very first slide of Watch Dogs that I would qualify as the real pitch of Watch Dogs, was a hand with the red button, and underneath it was written something like, "Control an entire city at the press of a button." That was the idea. 

The design idea was "hack everything around you -- but it has to be one button." It cannot become super complicated, because we knew already we would have to apply that to driving, to shooting, to mechanics. 

Later on, we did explore, voluntarily, tossing aside shooting, which helped us discover other kinds of mechanics, knowing that in a city like that eventually we would have to have it back. And that was the first inkling of it -- that's why we had hacking quite early, in the end. 

You alluded to reducing things to one button. Can you talk about the process of creating a complicated system that allows for emergent gameplay, but at the same time reducing the complexities of interaction? 

JM: It's a nightmare to do. [laughs] The one thing we had to do, we started by making a worst-case scenario. That's what you want to do. AI is once of the worst things to develop, in the sense that it takes time to have something decent. So what we did is, we did a multiplayer [version] very quickly, where four players could play against each other. We were focusing on fast driving, and then it was all about, "Can I, at 200 miles per hour, against other talented players, use hacking against them?" 

Back then there were traffic lights, bridges, and bollards that were already implemented. Those were the three, and we started this way. We fine-tuned a lot there. Back then, we had two buttons -- one for certain kids of categories, and then another. The goal was always, "How can we bring it to one?" 

So we ended up processing. The hardest part is control inputs. You want to have one button, so that people don't have to swallow "how" on top of "when" and "why" to use those things. The hardest part is there are other mechanics -- there's other stuff that people expect. 

We look at games like GTA and Saints Row -- all of those games. Because we know that they have complex controls, controls that can change conditionally based on where you are. We wanted to fix that. That's why it's the same when you sprint as when you accelerate. We wanted to bring those things so you can sprint and shoot. We had the aim to shoot afterwards, so we changed the purpose of the button a bit. 

You [the player] never change the control scheme. You never have this screen where you go from, "Okay, this is avatar... This is avatar with a gun." Way too much stuff to remember.  That's kind of how you start doing these things. 

How we ended up with one button... It was really fast. We ended up making a statement, where we use the A button for hacking, which is the most important one on the [Xbox] controller. A lot of people were like, "Woah." We put that statement in, saying, "It needs to become relevant. People need to not question the fact that it's there on the controller, which means that it has to become useful in every single situation."

It's one thing at the beginning of the game, when you're playing through the baseball stadium escape. It's clearly a path that you've crafted. 

JM: Yeah. For sure. 

Later in the game, it's going to be free-roaming. You don't know what angle the player is going to come at it. 

JM: Very fast. Yeah, very fast it becomes open. But you need to have that beginning to make sure people realize certain things: How to hack things, the camera, for example. Those are simple systems. 

We went through a big deal in defining what we should teach and not teach. That's a big question for me. When you end up with so many mechanics, there's a big choice you can make. You can choose to go the teacher way, and you end up playing a game, and it works -- quite a few games do that. For example, I would put GTA in that category, where they would really spend a lot of time in the course of the game teaching you stuff. 

Personally, I like, and I miss, in certain games this idea that players become curious. You want to ignite their sense of wonder and try to discover possibilities. That's why the single-button hacking ended up becoming also the icon of the button on screen. It's almost like, "Oh. What does that one do?" 

So the only thing you have to teach is: When there is an X [on PlayStation] you hold it, and it hacks the thing. The rest is about experimentation. And also about the skill tree -- which is, you unlock it, but when you unlock it, you know a bit what you're choosing to unlock, which gives you, already, an insight on how it's going to work. 

So it creates a natural flow where you search for your own ways of expressing yourself, instead of having yet another mission where -- pause! -- "This is the new thing, learn about that," and then it starts to convey to players that this mission is about using this thing, and then this thing. 

So we wanted earlier on to make sure players would arrive in this environment where they hack a camera -- which they learned -- and now they're in a situation where, "I'm in a camera. I'm safe."  And they start to see a new X -- sometimes on enemies, sometimes on stuff. That's why there's a word next to it saying "explode," "distract." So that way players can slowly but surely pay attention to those things and start exploring them, trying them. 

In the baseball stadium part, it's essentially the tutorial, so you can control the way the player encounters things. But in the real game, they could be approaching the game from any angle. How do you make sure that works? 

JM: Well, I think the first thing that you build is the second one. [laughs] 

Right, sure. 

JM: The tutorial comes at the end. You build the complex, 360-degree environment. Especially for level design and also AI. And then afterwards you make sure that people find all of their stuff -- like you make sure that if people try something, and it didn't work the way you want, you can come back and make sure that the systems connect properly. 

That's, in fact, one of the main reasons why, of the delay. We had very deep connection of systems that some players found, and the outcome wasn't as we planned, so we said, "Fuck it, we need to push. We need to take the time to fix those." That all came based on testing those kinds of maps. 

Beginning of maps, tutorial maps... Let's say "the Uncharted way of designing those kinds of things." I think most people know, but players may not realize how much money that costs. Every time you make one of those things you need to script it entirely, you need to animate it, and those kinds of stuff. So we didn't want to go that route, on top of having all of those systems.

So you make your systems and then you define, "Okay, we feel the player really needs to master this thing early, or else they're never going to access this part of the game." And then the stadium became that bit. And frankly, we had iterations where we were teaching too much stuff. It felt overwhelming. So you always come back and say, "Do they really need to know that? In fact, they don't." So that you can just give them the bare minimum.

Developers want players too often -- especially in games like Watch Dogs -- and I fall in that trap. Everybody fell in that trap. You watch somebody play, like today, for example. You guys are playing the game. We're behind, we watch the screen. But we always have the snapshot of us, who've played for five years the game, and we wish everybody understood every single layer of the game, every fact. It hurts. "No, he doesn't know that thing!"

And it's completely the wrong way to look at this kind of game. You want to just let them play. You want to make them feel like what they try works naturally. If it means shooting, and hacking less, those kinds of things, you should be more interested in bringing situations where hacking feels more useful, instead of saying, "Oh, no let's remove that gun," or "let's make the gun [appear] later." Try as much as possible to not force a certain kind of way to play the game.

Instead, it's almost like you're giving a music instrument to the player, and he's playing the music to you, and you should shut up and listen to the music. It's harder to do emotionally, because some people work really hard on tuning those systems in a certain way, but it's the next level in looking at those guys playing and then saying, "Okay, you wanted to go there? I ended up tuning in a way so that this becomes completely frustrating. I need to back off and find a way to address that as well."

You talked a little bit, also, about not overwhelming people with information. I can see why that would be a challenge in this game. 

JM: One of the things that was important was to give to the player a way to toggle it on and off. That's why the profiler is there. A big chunk of the information, especially after the first or second mission, is when you press the profiler. If you tap X, on the Xbox, that's going to pop up the profiler. And that pretty much opens the analysis of information of people. If you don't want that, you put it off. You can use it in cameras -- it's automatically on, in cameras. So that way you remove a lot of the problem.

We also went for stuff like, in the map... In the end, and I think it's something I'll push differently forward in the future, on whatever I do next. One of the solutions I've found is the idea that the player can pull information out instead of you pushing to them.

I think games of tomorrow have a lot of systems that are going to happen, and I think people need to be more at peace with that. I think it's okay to make a game that has a lot more depth than what some players might discover. You need to be at peace with that. And I think in the end a lot of the information turns out to be for the more curious ones.

The one thing that happens sometimes is that people end up forgetting about the profiler, they don't close it, that kind of stuff. So you need to make tweaks in the sound, and stuff like that, to try and bring them to close it if they don't need it, to not have anything that distracts them on screen -- without closing it for them, or they'll just forget about it.

You talked about how, in the future, players are going to be pulling more info out. Can you talk a little bit more about your expectations? 

JM: I think so. I think it's kind of a full cycle. When we were kids and playing games, we were doing only pull -- too much! We were suffering to pass a game, most of the time. In the era of the arcade, and even the PC, it was about trying to beat the game designer. It was really hardcore.

But then we went on this almost educational type of spectrum, where Zelda was once upon a time a thing without fire and a torch, and you could go, like, "Wait a minute -- that works like this!" This same game today has a bird showing up and telling you everything.

So I think we went in a full circle, and now I feel like sometimes the games are dumbed down, because design-wise, it's a lot easier to do. And there's a lot of money involved in the execution these days.

I think in the future games like Minecraft, for example, are examples of how to pull out information. I can see my kid playing Minecraft for hours, and understand deeply the connection of all things. But the way they discovered it is, "I'm breaking a cube." Eventually, "Oh, it's dark. I don't see shit. Problem. I'm going to try and find a way to do that." And they go and navigate the menu themselves.

So I think if you want to make something deeply immersive, yet more open to expression, I think there's no other way. You can find new tutorial elements. You can create new helper systems that are dynamic. But I think that they all have to be driven by the player.

And I think the game should, I don't know, have something like 10 percent of the sophistication is necessary to finish it, and 90 percent of it is there for feeding curiosity and encouraging expression of players so they can find their own way of playing it.

And that's hard to do. But I think that's definitely at least me, design-wise, that's kinda where I want to go. Other people will push for more rollercoaster type of experiences, those kinds of things. I fear that that has almost an end in terms of the cost you put into making a map for that, versus the benefit the player has as a return.

So I guess in the future, as people are getting more and more exposed to games that let them express themselves, like Minecraft, you end up turning players into people who are more inclined to feel that kind of urge than the rollercoaster one. I might be wrong. That's pretty much where I would like to go.