Heavy Dreams: Pushing Interactive Narrative

Quantic Dream made a major splash with its 2005 game Fahrenheit (also known as Indigo Prophecy.) It certainly had its flaws, but its interactive drama was not quite like anything being done at the time, particularly on the major consoles. Quantic Dream's upcoming game, Heavy Rain, eschews the supernatural themes that marred Fahrenheit's realism and explores the incipient dramatic/interactive form further.

Recently, Gamasutra had a chance to speak with Guillaume de Fondaumiere, co-CEO of Quantic Dream and executive producer of Heavy Rain about the game. What is it? What is the intent behind it? And how do the design decisions support this intent? These are important questions for a company that sees itself trying something new, and de Fondaumiere answers them frankly.

Heavy Rain has been much feted by the press; it was the subject of a dramatic announcement by Sony and remains a flagship exclusive title for the PlayStation 3. "If you want to drive cars and run and shoot people, then you shouldn't buy Heavy Rain; this is not a game for you," de Fondaumiere says.

On a platform where the majority of commercial successes revolve around that activity, can Quantic Dream find commercial and artistic success?

Do you officially call the game an "interactive movie"? I thought that I had seen that.

Guillaume de Fondaumiere: Actually, it's always difficult for us to brand what we're doing and to explain what Heavy Rain is about. I guess that the best short-form definition we found was "interactive movie", which is a double-edged sword quite simply because it reminds people of those games at the early '90s when you were basically in a movie and you had the choice between going to the left or opening the door to the right -- which of course Heavy Rain is absolutely not.

But to a certain degree, it is an interactive experience in which your actions have consequences on the story, and, because it's so cinematic, I guess calling it an "interactive movie" is probably the proper definition.

Do you feel that games approaching reality is a good thing?

GdF: We didn't set out to develop specifically realistic games. However, we thought that, with the story that we had -- it was really grounded into reality -- we wanted to create something that would not necessarily mimic reality but look relatively realistic. It's also because we wanted to create an experience where players would really be immersed in the environment, and I guess it's easier to immerse them in an environment that they understand and that they know.

This is why we went this route. But that shouldn't necessarily be the case... with the new technologies that we have today, with the capacity to create realistic characters, I guess you are going to see more and more games that look pretty much like real life more and more so. But I hope that developers are not only going to develop highly realistic games; our studio in particular -- we can do different types of games and venture into different graphical [territory].

The immersiveness of realistic graphics is something that David [Cage] and I have argued about before because the more realistic the graphics become, the more the human mind will judge the results against their perception of reality.

High-end graphics make you scrutinize the entire world much more; since it looks like reality, you check it against your own reality, so that it puts the script and the production design and the scenario under a microscope. Do you know what I mean?

GdF: Um, yeah, maybe. I think what was very important for us was to create the means to be realistic. We set a very high bar in terms of graphics, and I think we set a very high bar in terms of scenario. David and his team have been working for many months on writing a story that would be at the level of some Hollywood movies. I think, in my personal opinion at least, that we're doing a pretty good job on this.

What kind of measures do you take to overcome that challenge? Have you had a lot of, for instance, playtesting with different demographics to see how they react and how they feel -- like, is this realistic, or is this plausible for me?

GdF: First of all, that's work that we've done on the script. We've had scriptwriters who helped us. David wrote the entire story and the entire script, but we had a couple of scriptwriters who were really doctoring the script, so that was very interesting.

Now, of course, we're using also playtest to see what people's reaction is, and we're perfecting this game for many months now; and it's going to be perfecting up until we release. All elements are scrutinized both from a graphical standpoint and from a story -- a dialogue standpoint. So it's a lot of work.

Attention to detail is one of the essential elements in this game. Of course, because of the technology, because of the graphics -- and this is why I totally understand what you mean -- we need to have a very high standard in all compartments of the production. That's a real challenge, but this is what we wanted to do.

Have there been any times where you were playtesting something and the players just totally didn't get what you were trying to do, and you had to change something?

GdF: Thankfully not, but there's been a lot of little adjustments that have been done, and this is really, again, what we're saying about detail. I don't think that what's been written at the beginning was totally wrong; there's really small things that we added, small things that perfect and make the whole experience believable, because this is really what we want.

What can you do to keep -- or maybe you can't -- people from kind of ruining the mood by screwing around? 

GdF: There isn't, to be very honest, there isn't. I think at the beginning of the game there is a certain kind of a pact between the game and the player. We're setting a context; we're setting a story; we're setting characters. We're giving you the choice of doing a number of things.

The sets are very interactive; you can talk to people... I think we're trying to make sure that, whatever the context of what the player has to do, we're giving an awful lot of choice to players so they can really go their own way. But of course, if someone wants to just ruin the experience -- but who would want to do that?

In Fahrenheit, when it said that I shouldn't combine alcohol and painkillers, well, it was the first thing I did, and I died immediately...

GdF: But this is giving you a possibility. We didn't even have to warn you about that -- well, in this particular case we do -- but sometimes we don't even warn you about that. It's just a choice. You can have this possibility, but of course it's like in real life. If you do something that you shouldn't do, there is a consequence.

I think that this is a central theme of Heavy Rain; you have to bear the consequences of your actions. I think that what's really interesting and pretty much unique about Heavy Rain is that your choices have consequences and you can see what happens when you make a choice.

When I talked to David and he did a postmortem for Game Developer, he was disappointed with how the combat worked out. Have you come up with some solutions for this game?

GdF: Totally. On Fahrenheit, we had a system that was more a success or failure system, to start with, and we really wanted to get rid of this. In Heavy Rain, the combat sequences are -- or I would more say the action sequences, because sometimes it's combat; sometimes it's driving; sometimes it's dancing. So we have a system that enables us to offer all kinds of actions. It's really not a success or failure mechanism, so you don't have to try and try and try again.

Simply depending on your actions, something different is going to unfold; something different is going to happen. So we came up with a system that I think is perfectly, from a graphical standpoint, totally immersed. We don't have a big Simon Says thing like we had in Fahrenheit in the middle of the screen; no, it's totally integrated within the scene.

Also, the moves that sometimes you have to perform, because we're not only using the four buttons, we're also making use of the analog sticks, so you really have to unfold the animation. You really have to do the movement that the character has to do on screen. That all works far better now, and we're very happy about that.

How many conditions or results can you have for actions like that? Say, if you're dancing, I can imagine that it would be most likely the results of your partner being disappointed or neutral or happy, or something like that. How much gray area can you have between these?

GdF: I think that's a good example. It's a little bit like in real life, and this is really what we wanted. If you're dancing with someone -- let's say, imagine there is a scene in the game where you have to dance with someone. If you do the correct moves, you're going to dance nicely, and you're going to engage in a more interesting dance.

Stepping on the foot of the other person, he's going to start looking awkward, and at one point he may tell you, "You know what? I'm not so much into dancing -- let's do something else." And this is exactly what would happen in-game.

So again it's not the failure-success mechanism; we're trying to think, "Okay, if I would be in this situation, what would happen?" and we're trying then to create all the animations. I couldn't give you a straight answer how many different possibilities there are in the action sequences; there are so many, in the thousands!

Let's say per sequence.

GdF: There's no rule. There is no rule. Take, for instance, the combat in the mud between Mad Jack and Norman Jayden. You have multiple possibilities in this particular scene to die, but you have a number of possibilities to overthrow him; to continue the fight; to succeed over him.

We try to be as realistic as possible. If you're getting five, six punches that look deadly, you must die. This is it; we always ask ourselves, what's realistic? How much leeway do we have? And the same applies actually to dialogue. It's the same basic mechanism: giving choice and seeing what consequences you have.

Something that I've been really interested in recently is conditions of failure that don't feel negative -- like "failure" that yields an interesting result, so you don't feel like so much that you've failed, but just that you found something different.

GdF: In Heavy Rain, there are two things. First of all, you are in a story, and so your journey is prompted by your actions. There's no point system, so you cannot reach a certain status. To a certain degree, you just evolve in the journey, and depending on your actions you are going to experience a story that is going to be different.

So we have absolutely no mechanism to gauge the successfulness of a story. What's very important to us is, whatever the outcome, it must be interesting. I'm always using the example of the fact that in Heavy Rain, you have no game over. You have four playable characters, and you can lose the characters; now, if you lose all four characters, you're still going to have a game over.

It's the end of the story. There is a very, very subtle but important difference between a game over, something that implies that you have to do it again to succeed, and the end of a story. All story threads give you a satisfying ending -- satisfying in the sense that it can be sad; it can be happy; you may know or not a number of answers to the questions that you have while playing the game.

But it's a fitting ending to the story. You have a satisfying experience. If you lose all four characters, it is going to be the end of the story; it's a sad ending, but we hope it's going to be a satisfying experience because it ends not just on an end screen asking "why?" No, you perfectly understand why it's so, and that's very important to us.

What are your own personal goals for this project?

GdF: I think that the most important goal for us is to show that video games can be more than just shooting, driving -- that games can be meaningful. We really think games can be more than that; can be a true form of cultural expression, like movies or books. So I think we, as a developer, feel that what we did is successful if people, even just slightly, change their mind about video games and think, "Yeah. Actually, a game can be as meaningful as a movie."

I don't have my own personal opinion on this, but I've heard some people say that it's so much like an interactive movie -- why not just make a movie?

GdF: Those are two radically different things. In a movie, you are passively looking at the story. You see a perspective to certain events; you see one single possibility. What we're doing, we're really giving you a choice. We're setting a strong context; we're setting a number of possibilities. We're presenting to you different kinds of characters, but what really happens depends on your choices. If Heavy Rain was a movie, it would be ten, twenty, thirty movies, and that's what's interesting.

It seems like in this scenario there are a number of things that could potentially kick you out of the narrative. For example, HUD elements must be carefully crafted, or a saving system might be quite difficult to implement. How have you addressed these kinds of things?

GdF: The game is saving all the time, so your progress is stored. We hope that players are not going to play like this: We hope that they are going to bear the consequences of their actions and play at least the first time through the story; however, if at any point you want to go back and you want to see what would have happened if this and this, you have the possibility to do so. But yeah, the system works, and it's smoothly implemented, so I have no worries about this.

And what about things like the stick images that come up on the screen and stuff? Obviously, you have to show players how to do it, but do you find that it is at all jarring for people, or do they start to forget that the icons are there?

GdF: What you have to bear in mind is that we've shown, so far, four scenes. Two of them are relatively in the beginning of the game; two others are pretty much in the last third of the game. In Heavy Rain, when you start the game, the first scene is also a tutorial. It's not only a tutorial; you're really in the story. The story starts, but we are making a tutorial so people slowly get to understand how the control works.

We've seen that in user tests, once you've gone through this first scene, you totally forget about the interface. We need to show something on screen because otherwise people don't know what to do, and we're trying to integrate it as best as possible in the environment -- in the 3D -- so it has a minimal impact on your immersion.

But really, that was our experience on the play tests; people really forget about it. At one point, you're just doing the moves and interacting with the characters without even looking. You know at one point how it works, and then it's totally non-intrusive.

In a recent design column in Game Developer, the author spoke about managing player expectations. If the world looks real, a player will see something and feel like they should be able to interact with it. How far can you really go?

To re-use the example, if there's a slide in a playground, and the player thinks, "Oh, well it's a slide; I should be able to slide down it." Perhaps it doesn't make sense for the character to do it, but the player may feel like, in this realistic world, they should be able to do that sort of thing. How can you deal with that?

GdF: The important thing for us is context. The second thing is, whatever in the environment within the context should be interactive, then it's interactive. People were quite surprised, I think, to see how interactive Heavy Rain is and the fact that you can really interact with most of the objects that are there.

I mean, if there's a television you can switch it on and look at it. A radio, same thing. Objects, books, newspapers, doors -- if there are doors, you can open them, you can venture into different environments.

Even if it is not driving the story as such, but of course we believe it is extremely important for people to feel immersed in the environment. However, players shouldn't expect to go against the context.

If your character is on a playground and there is a slide, but it would make absolutely no sense for the player now to slide down the slide, then it is not interactive because it simply makes no sense. That's the limit that we have, but we try to minimize those moments where you eventually could think of interacting with something... most of the time, you can.

In Fallout 3 -- I don't know if you've played it -- you come out of this Vault in the beginning of the game, and you're searching for your dad. But if you want -- and this is what most people do -- you can spend hours and hours and hours playing all these sidequests, just helping somebody research a book or going and defeating 12 monsters of a certain type or something like that.

If you care so much about finding your dad, you're probably not going to do these things. It's good the options are there, but, ultimately, if someone were really analyzing it, it could undermine the main narrative. Balancing player expectation and their desire for freedom with the desire to push the narrative proper forward seems like a very tough thing to do, so I guess we'll see.

GdF: It is a tough thing to do, but again, it's always the same thing; it depends on what the expectation is on the game and what we as designers define is the experience that we are offering. If you want to drive cars and run and shoot people, then you shouldn't buy Heavy Rain; this is not a game for you. If you want to be the hero of a grand thriller, if you're looking for something that is meaningful, then you're up for the Heavy Rain experience.

What kind of camera are you using? It's not clear to me based on what I've seen.

GdF: It's cinematic cameras, because at any point in the game you always have two cameras tracking you. We didn't want to have a camera in the back or first-person; we wanted to give a very cinematic feel to the game, and so we're using this system that works pretty nicely, actually. It looks far better, if you ask me.

In addition to the artistic implications, it also seems that, in terms of the hardware, if you can control the viewpoint then you can also increase the number of polygons and things running on the SPUs on the screen at that time. Is that something that you are employing?

GdF: Absolutely not, actually. The whole game is being rendered in real time all the time, and we actually have a camera team that is working in post-production to set all the different cameras -- but everything's running. So we're not using the camera direction to help us render more polygons. No.

How large is the camera team?

GdF: We have four people working on the cameras, which is quite a big team.

It is!

GdF: Every moment in the game, someone thought of the two camera angles. I think to a certain degree it's giving perspective; it characterizes the playable characters even more.

Do these people come from the game industry, or did you also consult film?

GdF: No, most of them come from the movie industry. They have movie experience; this is really what you want.