Games as Interactive Storytelling

MAN: We're going to start immediately because there's not a full 15 minute break here. We are on a tight schedule. I don't like to be fascistic about it but I want to make sure everybody has a chance to be heard. This next panel is Games as Interactive Story Telling. I think we've seen the issue of story telling runs through any discussion of games. But I think we now have a chance to really focus on that dimension of it closely. I was just asked to make an announcement that for people who would like to follow up on the conference in a post mortem there's going to be a get together for game developers in the Boston area Tuesday night at 7:00 pm at Sam Adams Brew House in the Lenox Hotel off the Boylston and Prudential T stops. Along Boylston just past Prudential Center. Feels like the beginning of one of those early tech space games. [LAUGHTER] For more information kent@cognitoy.com for information. I now want to introduce our first speaker. For those of you who are following along the questions they were given are now at the bottom of page ten, the top of page 11. Our first speaker is Hal Barwood. Barwood is a project leader at LucasArts where he was designer, writer and team leader on Indiana Jones and the Fate of Atlantis, Big Sky Trooper, and Indy's Desktop Adventures and Yoda Stories. In addition he directed the live action video Rebel Assault II, a Star Wars combat game. His most recent project is Indiana Jones and the Infernal Machine, a real time 3D action adventure. Before Barwood began building games he spent 20 years in Hollywood as a writer for Sugarland Express, writer-producer for Dragonslayer and writer-director for Warning Sign.

BARWOOD: I'm going to give you my take on the here and now as opposed to the future. I'll tear through this. I'm Hal Barwood. I'm a game designer. I write for the games I design. I'm a team leader that builds these things at the LucasArts entertainment company. So the question is, are games a story telling medium? Not necessarily, of course, because sports games, combat games, vehicle sims, contests, muds [SP?], pure puzzlers like Shanghai and Tetrus, transplants like backgammon, cards and chess, have no need of stories. They're just wonderful games. People play them, people like go on playing them. But games and stories, I believe, do go well together. Like books go with stories or television goes with stories. And I believe there's a narrative tendency that we see in a lot of games that aren't narrative that suggests that there's an affinity which is rather deep. Let's look at the oldest known game that was actually designed. Backgammon which has the feature of doubling off the men at the end of the game. This has the effect of bunching up the losing team with the winning team and the result is that it feels kind of like a story being told. It's a deliberate part of the design. Hearts is like this too. Hearts, you try to keep your score low, don't take any of those little red cards and stay away from the queen. But every now and then you reverse your strategy, you try to grab everything. When that succeeds, or more hilariously, when it nearly succeeds [LAUGHTER] you have a wonderful moment when it feels like a story. And the same thing is in chess. Chess is a contest, a battle of wits. But every now and then a lonely pawn makes his way up the ranks, is promoted to queen. And when that happens you feel like a narrative has unfolded. You feel something like the story is starting to happen. So I believe that games and stories are natural partners. Stories inform game structure and game play. Story elements even started to appear in games like flight sims where you have missions and real time strategy games where you usually have an overlying scenario and in martial arts punchouts where you always have elaborate biographies for the characters. People like stories. I believe that stories are deeply embedded in the human psyche. For example, split brain research shows that animals, which don't confabulate, predict random light patterns better than humans, who do. What's the meaning of this odd experiment? I think it's a psychic illusion. I think that evolution has found it utilitarian for the human to go ahead and try to make up a story about the data he sees in the world. And this is so valuable that it's maintained by evolutionary pressure even when it can be tricked, as it is in this case. So what is a story anyway? I like to think of it as the entertaining recapitulation of unusual human events. It's more than simple narrative, I believe. It's not journalism, it's not history. And there's a metaphysical aspect, I believe, that creeps in. If narrative conveys us into the woods the most interesting thing we will find there is not the trees. It's not the undergrowth or the animals. The most interesting thing will be a path we find there that's laid down by somebody else. There are a lot of unsolved problems in computer games. The state of game AI is abysmal. I think of it as artificial stupidity. [LAUGHTER] Algorithmic plot and character hardly exists. Natural language production is theoretical at best. And yet a lot of us in this room are in the business of making a living at this. And we must ship finished goods. So the tyranny of commerce demands that we muddle through to ship titles using whatever techniques actually work. Judged by the revenue in the billions these days we're already a major entertainment medium. So how do story games operate? Do we use narrative collaboration as one of our major techniques where somehow the designer of the software figures out some cool things so that the user can then use these features of the program to create their own story? Do we use branching plots where user choices push the plot in one direction or another? Do we use believable AI agents as the non playing characters? And by and large the answer is at present none of the above. Interactive story telling is understood by scholars and academia and theoreticians. From a practical point of view the form doesn't exist. People want others to entertain them. People don't want to be trapped inside their own imaginations and the limitations thereof. So stories, commercially, are generally preordained and authorship lives on. How do we do what we do, whatever it is? I believe we use principally two techniques. First one is plot advancement. Players must act to thrust the stories that we tell forward. Without player action nothing happens. Scenes that in a movie you would observe between characters now become puzzles. As you take the place of one of those characters and have to actually solve the problems you saw so adroitly solved by a character in a film. So that the psychology and the purpose have changed. And the player, although not the author, is the star. The other thing we do is character evolution. In movies character is a set of unsatisfied inner psychological states that drive the choices of the drama. In games I believe character is capability. [UNINTELLIGIBLE] and revolver determine the scope of action of any game that he will appear in certainly. And what you can do is more important than who you are. Players must work to enlarge their capabilities in order to succeed at the game. So go ahead and collect those [UNINTELLIGIBLE]. What kind of video games result from story telling? We have one set of games. They're all closely allied but the adventure games are stories that are usually very strong stories with lots of complicated plot. Because we're an extended medium and not a brief medium like movies. And the story unfolds through the unlocking of puzzles. Role playing games usually as a player assumes the persona of somebody and then works for long hours to enlarge the capabilities of that character in alliance with other non playing characters and usually fighting a lot of enemies along the way. All of those activities of which enlarge your own character. And then finally there are the kinds of games you see represented in the little backgrounds that I'm doing here. Action adventure where some of the elements of both of those other things are going on. Usually there are puzzles. There usually is some character evolution. But the whole thing is a much tighter enterprise. So what's left to be interactive. I believe players manipulate fixed story elements all right. But they're usually large chunks instead of the atoms. We don't have it in our ability yet to be able to have stories just constructed on the fly. But we have elements of stories, chunks, scenes and locations which can be put together in various ways depending on how the player plays. And when that happens the player strives to reach his goals in a satisfying way. Our critics think that we're engaged on a trivial task in the absence of evidence that we can stir the emotions. And I think this is a fair worry. Because stories really are about humanity and hopes and fears of the players are what's in the balance. Can a story game evoke laughter? Yes, I think so. I've certainly played many games where I laughed outright. Theoretically a deeper question, can a story game evoke tears? I've never encountered such a thing but you know what? Who cares? There's a sporting element in games. Games must challenge. INteractivity minus goals equals boredom. And games evoke emotions of struggle, of competition. The kinds of things you feel aren't often given common names in our usual everyday parlance but they are important emotions that we feel and go through and enjoy and find in some mysterious way enlarge our spirit, I believe. How about the anxiety that you feel as you approach a game and don't know whether you have what it takes to get through it? How about the vexation you feel when you can't find the path even though you know you want one? How about the expectation, the hope that you feel when your chest suddenly swells as you realize you are going to be a master. How about the sense of self that develops as you concentrate all your being and the various parts of your being upon the task of overcoming the obstacles. How about the dejection you feel, the despair when you fail utterly? And how about the exultation and the sense of triumph and the joy you feel when you actually succeed? And sometimes a little bit of awe as you maybe find that path out there. There's another name for these emotions and game developers call them fun. But the story isn't everything. Game design goes way beyond that. And my job at LucasArts is certainly more of a game designer than a story teller. You have to define the modes of play, you have to create a workable subworld. We're not gods, you just try to imitate the big guy. You have to attach various kinds of handles. Some prefer this as the interface. But I think of it as something more complicated idea of how does the player get a grip on this world and how does the world influence the player? Is it going to be by conversations? Is it going to be by shooting? Are you going to lay bombs? Are we going to work combination locks? All of these things attach the player to the world in a way that engages him. And then you have to orchestrate the whole thing. Are you going to create a game which is based on a world where the player has freedom to run all over the place and you gate it through some subtle methods? A good example of this would be the action adventure game Zelda where Myamoto always creates what seems like the whole world but if you really investigate you find out after awhile you can only go certain places until you do certain things. Or are you going to make the whole thing chapter based and just be overt about it? However, story, I believe, does aid design. The premise is what attracts players in the first place. The plot focuses your design efforts. And the story arc lays out the short and long term goals that you need in order to keep a player involved in the face of a challenge. This is something that normal story telling doesn't have to get into. So that for example, exposition is more important than it would be in a film. Film exposition, you want to do as little as possible. In a game you might have to do quite a lot of it. Because if you don't the player won't know the information he must later act on. If you watch a movie and you don't quite get it, it won't matter. The actors know and they're going to go through it anyway. But that's how you maintain player involvement. Finally, I think that story forms over time learn to explore possibilities that are inherent in them. So that for example in the warm confines of a single room in the presence of actors physically there and the creepy stage machinery, it turns out that that's such a nice environment and it's so attractive from a human point of view that mere verbal behavior is all you need and you have drama. In movies the actors are two dimensional. They're on celluloid or acetate or hard drives. And as a result there's a cooler emotional temperature in a film but yet movies don't have to stay in that room. They can go outside and they can point the camera just as easily at a speeding car going off a cliff as they can at actors. So because of the slight flatter emotional state of a movie and the greater possibilities for moving images movies have settled, I think, upon melodrama as the natural form of expression. There are other kinds, of course, of all these things. And I think in games, where now we can do all of the above but in addition to that we give the player control, you take the controls and you want to do stuff. And that doesn't mean you necessarily want to twiddle a fingernail clipper. You might want to go out and conquer the world. So I think that games over time have moved further and further toward their natural form, which is action. That's it for me.

MAN: We'll take audience questions in just a moment. Let me remind you that we're not allowed to have people stand in the back or sit on the floor. Do we have somebody eager to ask a question of our panelist?

MAN: Hal, it seems to me that when people talk about any limits on the emotional qualities in games it's just because they haven't played enough games. It feels like, I can even think of a game, although I can't name it right now, that I did, I'm not sure tears rolled down my face but that I was very sad at. I think that's happened more than once. The leaving scene at the early part of Zelda I thought was very complex emotionally for me. When you first leave your girlfriend. I think that I had all sorts of complicated emotions in Fantasmagoria or Gabriel Knight II. I mean it feels to me like the emotions are there. I don't know why this keeps coming up as if they're not.

BARWOOD: I think the emotions are there. I just think that emotions you tend to experience are a different set and not as familiar in common vocabulary as the ones that people think we ought to be dealing with. But I think those are interesting emotions. And I think that part of the attraction of games is to put those emotions into operation. I think that somebody said, in a question used the term rehearsal for life. And I think some of that is what's going on. When you want to experience emotions in an indirect way that's when you turn to fiction and literature and drama. And I think that games provide a set of emotions which we all know about and are important to us. Especially it's hard to remember back but when you're growing up and you're testing yourself against your peers these emotions are actually in the forefront of your being, I think, for a period of years. And I think they stay with us over our lives. And I think it's important to deal with them.

MAN: I'm going to remind people to identify themselves. And I could ask the last questioner just to give us your name for the record.

CLANTON: I'm Chuck Clanton, designer at Bullfrog.

PEISER: Hi, I'm Patricia Peiser. And similarly, I have heard an awful lot of talk at different conferences about a lack of emotion in story telling and a lack of emotion in game and that we need to bring this into games and we've finally got the graphics thing and now we need to introduce this element. And the part that always rather bothers me about this is that one of the very first computer games that I ever really loved was Planetfall. And I can confess, I cried when Floyd died. [LAUGHTER] And the six guys who were in the room with me did too.

BARWOOD: Oh, god, I hate these confessionals. [LAUGHTER]

PEISER: But in that medium there was the ability to get you involved with those characters, get you that deeply emotive and the immense sense of relief that you had when he came back was fabulous. And that to me meant a fabulous gaming experience that I'll never forget.

BARWOOD: Hear, hear. [LAUGHTER]

MAN: Any of the other panelists want to respond to this issue about emotion in games?

MAN: Yeah, I want to take a piece of this because a lot of these saws that we're peddling here in addition to that games don't provide the kind of emotional releases that we're talking about is also, Hal talked about. I mean he mentioned several times that game characters are necessarily more cardboard than film characters are. I think one of the things that Hal brought up is that character in games, that character is capability. Character is action. Character is what you do. That exact same argument is made in terms of Hollywood film making. If you're a screen writer, you're told to define your character by what he or she does by action. And so we've got that same capability. I remember different things operating in the art cinema or perhaps some certain independent cinema things. But Hollywood is defined around action. And so the structural differences between the Hollywood film and the game is that Hollywood defines character motivation oftentimes by conversation. If you have a movie, say Sleepless in Seattle, where the goals are Tom Hanks is trying to fall in love with Meg Ryan and vice versa, those goals are very clear and very much defined by action. However, how we find out about motivation is that there's always a side character who's there for Tom Hanks to talk to, or a side character there for Meg Ryan to talk to about those sorts of things. If you do that in a game then gamers go oh, talk, talk, talk, let's get out of here. And one of the structural differences here is that it's not whether or not they're more complex or whatever. It's just that with the action understanding of the game it leaves out some of the flexibility that film does. Just basic structural stuff.

MAN: Hal, you've written film. Would you reply to that?

BARWOOD: Let me give you an example of what I think works in a film and what I meant by the distinction I made. Let's take the film Robocop. I think it's a successful drama. Wonderful main character. The character, I believe, can be boiled down to about three characteristics, one of which is his background, he was a cop. Second is his credentials. HE's a metal bodied guy who's unbelievably strong and powerful. And the third thing is his dramatic trait. He only has one. In a novel or a play you might have like half a dozen. But he's got just one. What is his one dramatic trait? By that I mean something in himself which is unsatisfied and for its completion needs the assistance of something outside himself. That's what's going to drive the drama through a story. And that characteristic is he wants to know his name which he cannot remember. That's all it is. It's very simple. In a game that by itself really won't work. Because there you ar running this guy and you know your name and you know his name and that kind of stuff doesn't really work. What really works much better in the game version of this is the fact that he's got that metal body. That's really cool. [LAUGHTER] The fact that he can run around and romp and stomp. That's what I was talking about.

ICE: My name is Matthew Ice. I'm from University of Wisconsin Milwaukee. I was wondering how familiar you are with Japanese games and just your general knowledge.

BARWOOD: Is this a test? [LAUGHTER]

ICE: No. I guess my point is that when you said have games ever involved tears and you said who cares, what I was thinking is that no, I've never gotten that from an American game. But I can give you, I can go on the Internet, I can give you hundreds of e-mails, letters, things people have written where people cried at certain scenes in Final Fantasy 7. And this is unanimous.

MAN: I cried in Final Fantasy 7 too. Tears of boredom. [LAUGHTER] [APPLAUSE]

BARWOOD: I was being a bit of a wise guy when I said who cares. Of course people care. But what I'm trying to do is direct attention away from what I believe is a very conventional view of emotional life in a human being towards something else, which I think is a richer, broader experience. And games tackle an end of the emotional spectrum which is not commonly dealt with in normal forms of drama. But I think it's important. I think it's interesting. I think it's what we all go through when we play games. We ought to recognize this and realize that just because we can't make people cry doesn't mean we don't have an emotional affect. Or for that matter that we don't, in some mysterious way, enlarge the spirit of those who play our games.

ICE: You said we can make people cry? Is that what you said?

BARWOOD: Is there an echo?

ICE: All due respect but maybe you just can't make people cry.

BARWOOD: I don't even try.

ICE: Why not?

BARWOOD: Because I think that I kind of explained it before. So I'll let other people here take over and go through it.

MOORE: Quick question, or comment rather. One thing I've noticed lacking in this discussion, by the way, I'm Eddie Moore from Irrational Games, for the record. One thing I've noticed that is lacking in this discussion is the emotion of fear. I happen to think that video games excel at creating the emotion of fear. Why do you think that video games are better at doing that than say creating sadness or evoking tears or such?

BARWOOD: Because there are just a lot of easier ways to create that in the gamer. It's a lot easier to startle somebody than it is to make them feel noble. It's a lot easier to arouse somebody than it is to make them sort of contemplative. I mean it's just one of the easy things that you can do. So it's the lazy way of getting people excited when they're playing your game.

MAN: As somebody who's done quite a lot of work on emotion from the academic side of view, I can tell you that an enormous preponderance of academic discussions of what emotions are deals around the notion of fear. Because fear fits that central paradigm of what an emotion is. There's an object out there that you're responding to. And you're responding in very precise ways, either fight or flight. And that is so clear and so visible on the outside that that makes it easy for those people who talk about emotions on the academic level to talk about as opposed to fuzzier ones like depression which are harder to talk about. Same thing operates, I think, for games. Because it's so clear, because it's so legible visibly, for the same reasons.

MAN: And when people talk about emotions and games they're talking about a bunch of different things going on, some of them at the perceptual and some of them at the level of cognition. So it tends to be a somewhat confused conversation. It's easy to startle people by moving something quickly in their field of view. Are they really scared? That's kind of debatable. Because when people talk about scared they talk about something that, I think Doom was one of the scariest games because what it did to you occurred over many hours that usually ended about 4:00 in the morning. [LAUGHTER] And I've been really curious to understand how they were able to do that. You have these two things which sound the same, I'm being scared, but one of which is very easy to create and the other which appears to involve a process over several hours and a lot of cognition and claustrophobia and a bunch of other cues. So I completely understand why you're saying that it's easier to talk about some things simply because they're more approachable states in the player than some of the more interesting ones.

MAN: Do you think as graphics hardware advances, such as with Playstation two and G Force and pushing more polys [SP?] and being able to render more realistic believable human characters do you think it will be easier to do complex emotions such as depression or sadness as time goes on in video games?

MAN: No. [UNINTELLIGIBLE] And the answer is I think that we're leading ourselves down the wrong path by thinking that being able to render faces more accurately is the key to engaging emotion. I don't think that's true. I think there's a certain extent that when early film began to engage those emotions that part of that came from the ability to use close-ups and get closer into that face. I do think that's true. But I think that most of what that capacity to engage emotions came from was it came from narrative. Came from being able to import melodrama, as Hal mentioned, as the dominant way of doing that. By importing that set of narrative conventions [UNINTELLIGIBLE] grew as a medium that's able to feel. I think it has less to do with whether or not you can see the faces, though that's obviously of some import.

MAN: I would disagree but in a pretty specific way. There are a couple of things you do when, we spend a lot of time looking at higher and higher quality representations of humans in games. It does in fact have significance in terms of the emotional reactions that people have to them. And then there are like very specific things. Llike do they blink or not? If you have your characters in the game start blinking all of a sudden it changes people's perceptions of what those people are. And then if you go to the next step where people's heads turn you get different responses from people, even though you're not really changing anything other than these really specific things. And then the latest thing we're working on right now is making sure that characters in the game focus on you. If you have characters in the game who don't focus on you, in other words the orbs of their eyes don't rotate, people tend to characterize those as being more aloof or disinterested or mechanical than they do. But I'm talking about a pretty technical aspect, not your fundamental point about [OVERLAPPING VOICES].

MAN: It's not necessarily that technical. For example, if you go to Disneyland and you see Fantasmic across the rivers of America there's Mickey Mouse in his little glittery suit and he has to act for an audience which is like 50 yards away. So he acts like this. That's acting at that distance. In a theater you have a slightly subtler version. But people do a lot of stage blocking and walking around because that's what acting is theatrically. And gesture of the body is what's important. When you get to movies you have film acting. And film acting is like this and the eye blink is a significant part of what you're doing, how you're conveying emotion. So I agree with Gabe.

MAN: I would nuance this. I think that my answer about narrative being the key is that when we get to that landmark ability to then move emotions deeply I think the answer is going to be narrative. But I would certainly not be somebody who has done facial research would say that faces has nothing to do with that.

MAN: Once again I'm going to point something that I'll reiterate in my presentation. A lot of problems come from what are we exactly talking about. It's like narration is something that occurs at a cognitive level. Most of the effects I'm talking about are not cognitive at all. People can't even tell you why they're more attracted to this character, even if you show it to them back to back. They won't pick up on it. So that's not a cognitive process. So that's why I'm concerned a lot of times in discussions about games and entertainment that you be relatively careful that you're not talking about things at different logic levels. And that's a point [OVERLAPPING VOICES].

MAN: I'll give you a little anecdote. I mean actors in movies, of course, are discussed among film makers. And it's not necessarily what you think. For example, Karen Black, an actress of 20 years ago who was well known certainly, was wall eyed. And this is very difficult when you're shooting close-ups because it produces the effect that Gabe was talking about.

WOMAN: She was what?

MAN: Wall eyed. [OVERLAPPING VOICES] As they would say in Britain, she had a squint.

MAN: It's the opposite of cross eyed.

MAN: Her eyes are going like this a little bit. So when she's looking past the camera you're not quite sure where she's looking. And so that aloofness that Gabe described does happen. And there was a worry for people who would work with her. You have to do special things to make sure that you aren't going to produce that effect.

MAN: I have to be hard nosed and move to the second of our speakers on this panel. I apologize again to those waiting for questions. Hopefully we'll get you in on the next round. Gabe Newell is the founder and managing director of Valve, an entertainment software company based in Kirkland, Washington. Valve's first game, Half-Life, has won more than 50 game of the year honors and was called best game of all time by PC Gamer in November 1999. Before starting Valve Newell held a number of positions in the systems application and advanced technology division in Microsoft where he worked for 13 years. His responsibilities included running program management for the first two releases of Windows, starting the company's multimedia division and most recently leading the company's efforts on the information highway PC.

NEWELL: I'm going to go back to something I touched on a little bit earlier. The way I'm going to talk today is I'm going to talk pragmatically, as if I'm talking to another game designer. Or if I were talking internally within our company. I thought of a couple of different ways to present here and I thought that this would be sort of interesting. So I'm really going to talk to you as if you're in the same business I am and we're just talking about how to build these things better. One of the first comments that frames a lot of what I'll talk about is that when we talk about games and game design and the processes of building these things inappropriate cognitive tools get applied to design aids. Inappropriate design aids are getting applied all over the place. If you talked about the story of Tetrus you could actually be pretty funny. The story of Tetrus is you die and then you die and then you die. [LAUGHTER] If you talked about the character you could go [SOUND]. And so on. And you might actually sound like you're making sense. But the problem is you aren't really helping yourself make any interesting decisions. It's like talking about the pacing and arc of film emulsion. You're not really helping yourself at all. So that's one of the things. Trying to talk about things when you're actually talking about things at a different level and inappropriately applying design tools from another field. I absolutely believe that story telling is an incredibly important part of game design. I think that game designers have to be really careful not to view their goal as telling a story. Because pragmatically what ends up happening is you end up getting people who build a really bad movie or build a really bad story. You don't get people who build good games. Is that necessarily the case? No. I actually believe that some of the most compelling stories are going to be communicated in games in the next 20 years. But just once again, the people who set out to tell a story and view that as the goal of their design almost always go astray. So let's talk about the things which do help. What are the tools? Right now my company is faced with the daunting task of not having a horrible sophomore slump. PC Gamer said that we were the best PC game of all time. And somehow in that context we're supposed to figure out how not to disappoint all of our customers around the world with our second product. So we're really somewhat frantically trying to figure out how we can do this better. Some of the things that are really helpful in thinking about stuff are reinforcement schedules. It's pretty startling in the game space, or in the game creation space, the number of people who have never heard of reinforcement schedules. Who don't understand it as a tool for analyzing or designing any of the experiences that they create. Another tool that actually was probably the single most useful tool we had in designing Half-Life was this notion of narcissistic gratification. By that what I mean is, I'll come back to the heuristic that we derived from that. But this notion that the world is reacting to me was an incredibly powerful one. That if I shot at the wall, in games before Half-Life there were no bullet holes that appeared on the wall. The wall was ignoring me. And every time that you made the world react to player behavior or extend player behavior we basically used that as our operation definition of game playing. Another example, that combines a couple of things, was that you'd throw grenades at people in the game. And in a lot of games the incoming threat is just ignored. So we did all this stuff where if you're shooting at people they'd run away. We actually did presentations where we said we've done this amazing thing that nobody has ever done before. You shoot at other creatures or whatever in the game and they're going to try to get away from you. And that was pretty cool because they were doing something other than just coming into the hails of lead that your character was throwing at them. And we had this problem that some of the marines could get into a state where a grenade was at their feet but they couldn't find a path, because of the way we're doing path finding in the world, that would get them away from the grenade fast enough. So what they'd do is they'd just stand there. That was really bad. They went from being really clever to being really stupid all at once. And we struggled, we'll work on our path finding algorithm. The reality of it was that we had a perfectly reasonable thing that we could do at that point. Or what the marine was doing was perfectly reasonable but what was annoying about it was he had stopped reacting to me. And what he had him do is say we can't find a path so he'll just duck and cover and basically go fire on the wall. At least continue to provide the narcissistic gratification even if he couldn't do something that was successful. This leads to the next point. Which is, we had a bunch of local heuristics that we used. They weren't grand theories. They weren't unifying concepts. They were just simple rules that we applied. We actually went into it thinking that we would have some grand theories. And we ended up finding that we ended up getting lost in our own conceptual models for the process that we were doing. So one of the things that we discovered was expose game state to the player always. In the case of the marine it was like we kicked ourselves in the head because we had this heuristic lying around that we knew was really useful and all we had to do was make sure that his state of being trapped because of the path finding algorithm was exposed to the player. And the player would suddenly go from thinking that your game was really stupid to thinking that your game was really smart. Another heuristic. The more ways the game reacts to the player the better. And the more ways you can basically quantitize up and categorize player behavior and have things in the world respond to it the better. We just had a whole bunch of heuristics. No weapon in the game should basically be just a more powerful version of another weapon. Each weapon had to have a unique behavior. Otherwise there was no need to put it in the game. Another thing that's really useful. User interface design papers. And I have yet to, once again this is an area where game designers seem to be relatively myopic. There's a tremendous amount of literature. Game design is the process of building a user interface into a data set. The user interface isn't the menus and stuff like that. It's like you've got this representation of a world and what are the interesting ways of exploring it? Is a 3D presentation of that world data useful? Is a 2D presentation of the world? What are the filters that you apply to your data sets that allow the player to do it? So there's a tremendous literature out there that's very useful. I almost always came away from reading one of those things with a concrete feature we would change in our game. Simulation is actually useful. The procedural generation of behaviors in the world for specific things. Procedurally generated combat AI. We can do that now. Procedurally generated physics. We can do that now. I'll come back to that later. Negative feedback. Making sure that, Tetrus does a great job of getting you to the point where you're being challenged. So making sure that in the feedback of the design that you're rewarding the better player with more challenge as opposed to the better player just getting ahead of your game and just getting so much of an advantage that it's no longer challenging. I'm going to speed up a little bit here. Explicit economies. In other words, knowing in your own mind how resources trade off. And listening to psychologists and behavioral and cognitive scientists are very useful things to do. Things which don't help. Listening to most game designers actually doesn't help you. [LAUGHTER] A couple of reasons. They aren't software designers. Coming from my background I was really surprised by the lack of professional training. Game design isn't a mature field at all. They weren't software designers and they weren't graphic designers. They weren't people who had a design background. They weren't people who, a lot of game designers, certainly in the PC space, are recapitulating other people's designs with minor twists. They have no idea about the intentionality of design or any of the things that, a lot of times we found that better game design work comes out of really talented software programmers or graphic designers than out of people who are calling themselves game designers in the game space. Genre distinctions are totally useless. You learn nothing by saying oh, we're a strategy action game. That never helps you make a single decision. It's what other people are going to call you after the fact. But you can't learn anything by staring into those terms. Realism is the bane of game, we banned the word from game design discussions at Valve. [APPLAUSE] It isn't helpful. It just doesn't get you anywhere. And I could spend two hours talking about that.

MAN: You've got about two minutes.

NEWELL: OK, I'll speed up even more. Things which don't help. Trying to simulate plots. Right now, can't do it. Things which don't help right now. Heuristics help. An elaborate theory of game playing, we weren't able to do it. We would have never shipped. Things which sound more helpful than they actually are. I was in love with this concept for a while of we're co-authoring the experience with the gamer. It sounded really cool. I could have done lots of slides about it. But fundamentally it never helped us make a really significant decision when we were designing the game. A similar concept is the player is the main character of the story. I used to say this with tremendous urgency in design meetings. And then realized I couldn't think of a single decision we had made where that was useful. Once again, I'm talking about the pragmatics of building these things. Immersiveness. Immersiveness is a concept, Half-Life has been lauded for its immersiveness. But we never sat down and said let's put some more immersiveness into the game. [LAUGHTER] In terms of where I'm personally as a designer, the things which I'm trying to understand better right now, is the role of expectation on reinforcements. I think that that's something I just need to understand better. It's like the anticipatory impact on rewards and punishments seems to be something that we really need to understand better and is a stepping stone towards integrating some level of behavioral understanding of a game with narrative. Maybe I'm wrong and the gulf is way too wide. But I think that the role of expectations is incredibly important and I need to know it better. I really wish that we had a formalized statement of what we already know. I mean I'd love to sit down and, is Warren here?

MAN: Yeah.

NEWELL: There are a bunch of us who have been saying we've got to write this down as a book, just hit pause for a while and just talk about what we already know. The other thing that we're deeply embedded right now is we created a set of heuristics for Half-Life and right now we're generating a set of heuristics for social play design. And they're really interesting. It's the thing that I'm working on right now. In Team Fortress Two we have very realistic looking people, you can talk to other people on the Internet. But it's really interesting how much variability in outcome there is based on the context into which you put people. And the self explanatory nature of certain contexts has a huge impact on how quickly people organize themselves an dhow well they organize themselves and how well they self report afterwards their enjoyment. So let's say we're making progress. What are the tests, I would say to our own teams and to other teams, about whether or not we're making progress in the space? Can we stimulate a broader range of emotional reactions? If we can I think we're moving forward and becoming better designers. Which emotions you pick is a separate issue. But if we could broaden the range that we can consistently arouse in players I think we'd be doing better. I think we're doing better if we start seeing spontaneous playing goal setting that is driven by some overarching plot and characters. In other words, if you're creating an environment where people are saying I'm going to go do this thing because I really hate this guy in the game. And it's growing out of, in other words are we at least connecting character and plot to something other than this linear pathway through a world and instead starting to cause the player to do other things? Spontaneously deciding to go do something because of what's happening in the overarching plot, I think, is a step towards reaching the point where we can procedurally generate plot. The other thing is being able to articulate a bad, broken, grand unified theory of games I think would be an interesting experiment at this time. And if we are able to come forward with that and say how plot fits in and get enough of a cognitive scientist's or a behavioral scientist's input so we're at least up to the 1970s in terms of the thinking in the gaming industry about how those things relate to what we're doing. So that for me is where my head is at right now and what I think the interesting things are for us moving forward.

MAN: Thank you. We still have people piling up in the aisles. So if you could try to move down a little and let people in. Then we'll be ready for questions. It looks like there's a question over here. Give your name first.

HALL: Hi, Gabe. My name is Justin Hall. I'm from gamers.com. It's a really nice site. Gabe, I just wanted to say that I may not have cried but I was very sad when you didn't define reinforcement schedule. [LAUGHTER] Could you share with us what you mean by that?

NEWELL: Is there anybody in the audience who wants to do that? Go ahead in the back there.

MAN: [UNINTELLIGIBLE] wants to hear it.

NEWELL: No, who wants to define it. There's nobody from behavioral science here?

MAN: No, I think you're going to have to do it, Gabe.

NEWELL: Reinforcement schedules, there are a bunch of reinforcement schedules. You've got a pigeon, you've got some pellets. And the question is, how do you maximize a behavior given a set of rewards? I mean the classic example was done just down the street, or I guess a lot of the original work was. It's this notion of looking at behavior, looking at rewards and punishments for that behavior, and then having a clearer expectation of future behavior. And there are different kinds of reinforcement schedules. Llike every time you do a behavior the punishment occurs. The more interesting reinforcement schedules are related to frequency. Varying the frequency of the reward or varying the amount of the reward in a non linear tends to cause a much higher rate of behavior expression.

MAN: How do you usually [UNINTELLIGIBLE] company this?

NEWELL: We don't manage behaviorally. They'd get mad at me and throw shit at me. [LAUGHTER]

MAN: I think he's asking more about the games you design. [LAUGHTER] Where are we as gamers given pellets by your system?

NEWELL: Remember, it's a way of thinking about your game that helps you make better games. It's not a way of thinking about people. So I don't want anybody to go away thinking that we think of gamers as pigeons. But you basically say how often do you punish the player, how much do you punish them for an action? For example, in Half-Life if you shoot a civilian there's a cost associated with that. And we basically don't want you to do that. So when you ask a behavioral scientist how do you extinguish a behavior quickly, there's a reinforcement schedule that you use. And if there's something we really want you to do what's the reinforcement schedule to use? Llike the rewards in Half-Life are getting to see new monsters, the plot is moving forward, getting to have a new fun weapon, getting to see something really cool. Those are the reinforcement schedules. And you want to make sure that throughout the course of the game that they're getting rewards. You want to make sure that your pacing of those rewards is consistent with a reinforcement. You want to look at it from the point of view of a reinforcement schedule and say OK, that makes sense. I mean there were points in the game before we shipped where there were long lulls where basically all you were doing was doing stuff that you'd already done before which in our view didn't represent a reward. So we said we've got to put more fun stuff in here or eliminate that section from the game. Does that help clarify?

HALL: It's great.

MAN: I would add just one other, or perhaps two other, observations to that. Behaviorism tells you that positive reinforcement is just as strong as negative reinforcement for changing behavior. So reward and punishment are equivalent. And it also says not just variable rate but variable interval variable rate is the best reinforcement schedule. What that means is not knowing whether or not you're going to get a reward from an action is powerful. As well as not knowing how much reward you're going to get. If both of those vary in ways that you cannot predict your behavior will increase more than if you can predict either of them.

MAN: [UNINTELLIGIBLE]

NEWELL: That is a property of games.

MAN: You want to say that into the microphone just for the record?

NEWELL: Who was that, by the way?

CLANTON: Chuck Clanton, Bullfrog. [OVERLAPPING VOICES] learning company. Just from my undergraduate psychology, if you have inconsistent reinforcement schemes the behavior that you're trying to stimulate, an inconsistent positive reinforcement scheme, variable interval, variable rate, the behavior that you're trying to stimulate accelerates very quickly. But also if you observe the subject you get very high senses of stress, helplessness, loss of sense of control. Your mice go into a frenzy smacking the lever trying to make something happen and freak out. One of the things I want to say in terms of reinforcement schedule is, having worked in the kids game space for a number of years, this is something that we've really grown to pay meticulous attention to. Particularly if you're doing educational stuff where the activities themselves may not be quite as cool, as engaging as we'd love them to be. Sometimes we can nail it, sometimes we can't within production constraints. Managing those reinforcements has become a critical design consideration. So I feel that's a place where that segment of the industry has started to take some leadership.

NEWELL: The thing that I love is that the language of this is so much more precise. I mean you were sort of alluding to there's frustration that gets created with inconsistent reinforcement schedules. That is such an incredibly precise way of talking about something that most game designers have no idea of how to talk about. That's why it's a really useful thing for designers. It gives you a framework for much more efficiently categorizing and communicating a bunch of your design decisions. And the most frustrating things are the random punishments of game players by bugs. [LAUGHTER] That, I think, is the source of the greatest frustration right now.

MAN: Let me introduce our third speaker on the panel. Greg Smith. A former system analyst for IBM and other firms. Smith is now an assistant professor in the communication department at my alma mater, Georgia State University, where he teaches media studies and history. He earned his doctorate from the University of Wisconsin Madison. Smith is the editor of On a Silver Platter: CD-ROMs and the Promises of a New Technology from New York University 1999. And co-editor of Passionate Views: Film, Cognition and Emotion, Johns Hopkins University Press, 1999. Smith is interested in how new media try to incorporate existing content from other media and in new ways these adaptations can often expand the aesthetic capabilities of new media. I'll turn it over to Greg.

SMITH: Thanks. When we academics here at this conference, what can we potentially bring to this conversation that's going on here? I think one of the things that we can bring is a different perspective. Unlike game designers we're not necessarily caught up in that meeting the deadline pressure and meeting the pressure of working on a particular framework. And that gives us, I think, a different latitude. A little more latitude to think about some of the possibilities and to open up the possibilities of this medium. I think it at least provides us with the opportunity to think about the unused potential of the gaming medium. I think it provides us a vantage point that we can think about different kinds of stories that can be told by this medium. So I want to talk about one particular kind of story just as a suggestion along the ways. But I also want to talk about not just different kinds of stories but perhaps some of the assumptions that are relied on when producing a game, assumptions that are invisible at times. But if we can bring them out into visible light then we can perhaps look at them and tinker with them a bit. My perspective, from where I'm coming from, is I'm coming from this as a student of media history. As a student of the way that new technologies have been taken up into society. If there is a lesson to be gained by media history that lesson is that there is no such thing as a new medium that's truly new. New media always borrow from other preexisting media. It's an important dynamic that goes on with any new media. Though we often ballyhoo them as having amazing revolutionary potential what you actually put on that new medium is sometimes recycled content from other media. But I don't want to say that they just borrow content in new media. Because when we adapt things from old media it sets up the possibility of transforming the very form of new media in a way that's very, very interesting. Henry yesterday spoke about one of the quintessential things that we consider cinematic. He said the chase scene is one of the essences of what we consider cinematic. And that's very true. It's part of the definition that we think of as the movies today. But the chase scene came to us as an adaptation of 19th century theatrical practices. There were chases and they imported those into movies in their early days. When they imported them they changed and they transformed and they became part of the definition of movies. Same with any medium that you pick. Television begins and when they try to get content they pull it from radio. And the kinds of things that we now think of as quintessentially television, things like game shows, things like soap operas, have been transformed from old media. And so I think it is this process of borrowing and transforming which really does define a new medium. So if we're looking for ways to enrich this medium in which we're working and talking about, if we're looking for ways to enrich games, I think the place to look is to look in existing media. Places to borrow. Knowing that when we borrow those things it will enrich and possibly make this medium grow and help to define what is new media at this point. So I'll give you two of these ideas just as an example of how you might go about doing this. So you may like these ideas, you may not like these ideas but hey, the price is right. They're cheap. So if you want to run with them, fine. We'll see where they go. One place to look clearly is television. And clearly there's an intuitive fit with a lot of television. With things like game shows and Who Wants to Be a Millionaire and You Don't Know Jack. Clearly there's a kind of intuitive fit there with that. But I don't want to talk about the ones that have quite that intuitive fit. I want to talk about the potential of adapting a different form of television, one that I think is now at the heart of prime time viewing. And that's the serial narrative. The serial narrative is when you don't get the entire story within one episode but it's continued across episodes. When I look at television and at the really good television out there, pretty much all the good television that I see on prime time follows serial narrative form. Just to make a quick list, X Files, Practice, Ally McBeal, West Wing, Felicity, ER, The Sopranos, Oz, all of these television forms are serial narratives. And of course, there are other experiments like Steven King's Green Mile, for instance, in literature that's doing the same kind of stuff. Why? What are the networks onto with this? I think what they're onto with the notion of serial narrative is it's a great way to get brand loyalty. And certainly brand loyalty is one of the big things that's being gone after in this point. If you can keep people coming back to The Sopranos, if you can keep people coming back to ER, that's a very powerful thing. And clearly the networks have gotten onto something there with that. So I'm suggesting that one of the things that we might consider as a design possibility is considering designing a game as a set of separate serial chapters or a set of episodes delivered one at a time. I can sense that any designer in here wincing at this notion because god knows it's scary enough to produce one game and get all the constraints dealt with. What's the notion of producing multiple ones one right after another? It's scary, I know. However, I think we should also recognize that in some sense games already do this, or at least many games do this in a way. There's a way that different levels of a game operate as kind of different episodes in a certain sense. I mean going from level one to level two to level three. These are different chapters if you want to consider it in that way. So in considering that you noodle around with this notion of serial narrative I want to suggest that maybe it's not as radical as one might think. It's perhaps just a notion of different packaging, a different delivery arrangement. Instead of delivering a game with 40 levels as one package to consider what might be different if we consider those as ten packages, or 40 packages or however you want to break that up. So at one point I want to say that wouldn't be that different in terms of design capacities. On the other side, I would want to say that that might potentially quite very much change the form of games as we now experience them. It would change the form because at the end of every level that exists now, every level, every chapter, every episode, you provide the gamer with a payoff. There's a clear payoff at the end of every level that you go to. And if we went to a more serial form for certain kinds of games you'd need to provide a payoff and a cliffhanger both to hook them into going to the next chapter as well. It would change the form. It would also change certain of the pleasures of games. And I think that one of the major pleasures of the serial narrative as we experience it now is the pleasure, oddly enough, of waiting. Of anticipating what's going to happen next. If you're in a game where you go from level one to level two as soon as you go to level you immediately hop onto level two. There's no anticipation about what's going to happen next. If you're denied the access to that next chapter, to that next level, then that sets up a very different dynamic which is very, very powerful, as has been demonstrated in the history of serial narrative. I recognize that this runs counter to the notions of interactivity as we talk about them. That interactivity, as we talk about it, gives you the freedom to go wherever you want to go. And that is very powerful. But I also want to suggest there are other pleasures. There are pleasures of waiting. There are pleasures of not being able to go. There are pleasures of anticipating where you want to go with that. As every game designer knows, that restraint and constraint is as important a portion of the equation as anything with that. And so when you are waiting for the next chapter it opens up a space in which players could potentially chat online about what happens next. It will create a very, very different dynamic. Nobody says I wonder what's going to happen in level two because you can immediately get to level two immediately after level one. And so I think the notion of trying on the rethink of this as serial delivery might open up a different way in which we can define gaming communities from that. Ways that could be quite powerful and could change the way that we think of some of the kinds of narratives, some of the kinds of stories that get told in games. So that's one chunk. I think that by looking at other forms we can find new kinds of stories to tell in ways that are powerful to adapt. I also think that if we look at some of the assumptions that are made going into the game, that can also be very freeing and very interesting. And so I want to look at, for example, one of the assumptions that games carry over from film, and we heard Gabe mention this as well, the notion that you have a first person identification with the hero. That seems to be a part and parcel assumption of the way that many games are played. In fact, it's institutionally built in to many games. You're given a certain character's point of view in many games with that. And of course, there are certain games where you can be different characters. You can choose your character, you can design your character with certain qualities, certain costumes, those sorts of things. But those are experienced one at a time. You can walk that character through the space. And then you can become another character and walk that character through the space. What we're learning, and I think Chris alluded to this in the previous panel, we are learning that what people do with these things that are built is very different from what we envisioned that they'd do with them. One of the things that we've learned in film studies that people do is that people do not ride piggyback on one character through the whole film and identify with them solely through the film. Even though that seems to be the way the narrative is set up that's not really what happens in terms of the way audiences deal with that. What happens is that audiences tend to engage in multiple identifications. They tend to identify with the hero and then switch off and identify with the enemy for a while and back and forth and constantly switch with this. It's much, much, much more complicated in the way the identification happens in film than the way we conceptualize it. And at times when I'm gaming I can feel the limitations of that. It feels, at times games can feel more restricted to me than film because I feel locked into that one hero that I am taking one at a time through there. And so as a suggestion one of the things you might think about doing is the notion of allowing yourself to trade perspective midstream. I know that you can take a character and run that person through the person. But you kind of reboot the game and run that person through. What about the possibility of changing character, changing bodies, midstream? It would give a very, very interesting dynamic if we take that assumption out that you are taking one person through the space throughout the game. It opens up very, very different kinds of pleasures. If I can be the person who hits and I can also then suddenly be the person who is hit sets up a very, very, very complicated emotional dynamic with this. We're so tied in the notion of games that the pleasure of gaming is the pleasure of doing. And it certainly is. But I think that also ignores the possibility, which other medium do activate, that there's pleasures of doing but also pleasures of being done to as well. And if we can open up both of those then I think we've opened up other emotional registers. Over and over we've talked about can games make you cry, can we explore these emotions that are broader? I think one of the limiting factors is that we are so tied to this notion that the pleasure of game is the pleasure of doing. And so if we switch it to the pleasure of doing and the pleasure of being done to we can, I think, potentially get us closer to the notion of games that make you cry.

MAN: Why don't we see about, if you go to the mike, Christopher and I'd like to ask the panelists if they have some responses while Christopher is walking to the mike.

SMITH: I just want to reiterate that while I've put forward a couple of suggestions here the main point I do want to emphasize is that it's enriching when you borrow. It's enriching when you go outside to other media and pick those up. That's the main point I want to come back to.

MAN: Either of you want to --

MAN: I'm one of the people who have done something like this. I mean in Fate of Atlantis you get to play the leading lady as well as [UNINTELLIGIBLE] from time to time. There are branching paths and so on. But ultimately I think that the reaction was not even a blip on the radar. So from a practical point of view of just a working stiff game designer I'm not very attracted to these ideas. Because I just find that they immensely increase the amount of labor that everybody on the team will have to go through. And I don't see this immensely increases the pleasure of the game. So it's a tough equation for me, I'll say.

SMITH: One of my fandoms is that I operate inside online wrestling fandom stuff. And one of the things that I know that is spoken of frequently in wrestling games is that the dynamic of wrestling fandom is not just the dynamic of being on one character but is the dynamic of being both and going back and forth with that. And so one of the criticisms that happens with wrestling games in particular is this notion that you are tied to a single person. Clearly you can walk different wrestlers through the ring but you are tied one at a time. And I can say that that's one of the significant criticisms that I hear in online wrestling chat about the notion.

MAN: One thing is that multiplayer games have a very different range of emotions that people experience. And part of that is a function of being done to as opposed to doing. And I do think it's a really interesting problem. I think that you can exploit that by creating interesting shared play spaces. It would be an interesting problem to try to solve. But it might be easier to try to do it in a multiplayer space where you're going through a shared experience than to try to do it without those other people helping you.

MAN: It also works well when you've got a very limited world, which is like a wrestling mat, and a great variety of characters. It's somewhat simpler to just do the mechanics, to have the experience you want. So obviously I mean that's a major feature of all the fighting games too for that matter. Everybody wants to be Chung Lee in addition to the other guy.

MAN: Christopher.

WEAVER: I'll address the first point that you made, Greg, because Hal and Gabe, I think, have successfully touched on enough of the points I was going to bring up from the second. You started off by making an interesting point of the usefulness of mixing academics with people who are actually doing. Because it basically brings out something which is hopefully hybrid, and I agree. The issue of creating cliffhangers, a la radio programs of old, has in fact been tried. The problem has absolutely nothing to do with the idea because the idea is an excellent one. The problem is that while the distribution mechanism is limited to the current distribution mechanism it is an uneconomically viable decision. I won't bore you with the exact reasons why but I will tell you straightforwardly that it will not happen while the mechanism is as it is. When you have broad band distribution and other things that allow you to take advantage of what was previously freedom of airwaves, such as radio, then you have a totally different type of paradigm and you can start thinking that way. But as long as --

SMITH: And I agree that the notion of the ability to download these chunks will make the whole notion a much more viable delivery system with that. Yes, I agree. I also note that somebody mentioned yesterday that perhaps the way that these innovations come about is not from below from some young turk building a system but from somebody with a name making this innovation. I think here that the Steven King Green Mile example is a good one. Given the name Steven King he can go to a publisher and say look, why not do this as a serial novel and published Green Mile. I don't know if they went to him.

MAN: Most of the games buyers [UNINTELLIGIBLE] Tom Clancy. [UNINTELLIGIBLE]

SMITH: Right. But there are certainly names that could be involved within the genre that can work for this. So maybe it becomes the [UNINTELLIGIBLE] or whomever, whatever name you want to work with that may be the way that this stuff gets done. Because that person would be the elephant on the block in the way that Steven King has become an elephant on the block in traditional publishing.

MAN: This is Dave Wole [SP?] from Mattel Interactive. Chris said about 90% of what I was going to say. And about 900% more elegantly so I won't recap. But two quick seeds for thought. One in terms of looking at the beginnings of the future of episodic games, I think, if you spend a little time looking at the webosotic [SP?] adventures that are being delivered, the chaptered stories accompanied with illustrations or flash, I think that's the very beginning of serial gaming although it is a quasi linear narrative form. I've been creatively exploring some ideas of how we can add more user involvement. Naturally amazon.com has done some interesting exploration where they had users submit continuation for Updike's beginning. Also just wanted to commend you, if you haven't had a chance yet to spend a little time with Day of the Tentacle in terms of multiple perspectives. It doesn't fully get into the joy of being the doer and the being done to but I thought was a breath of fresh air along the design channel that you suggested.

SMITH: And one of the things I didn't bring were some of the attempts at web serials that have happened. Because there are have been some rather spectacular failures along those ways. Llike early example like the Spot, for instance, being an example here. I don't think that necessarily torches the value of the idea because I think the Spot didn't provide something that was interactive enough or that didn't provide something that was distinct enough from the pleasures of the traditional soap opera. So I'm aware of some of the web based serial stuff. That kind of stuff is still so bound in the restrictions of porting it across lines.

MAN: One of the things, and it's an ill formed idea in my head, but right now Half-Life is generated in about 220 million player minutes a month online. Our online community is quite a bit larger than a lot of existing traditional media in terms of its reach and its frequency. We've been learning a lot about how to, first of all serialized content helps a huge amount in that space. If you keep releasing new things for people to do. They're happy that it's in reruns because they keep doing that. But if you starve them for new content they eventually go away and go do something else. And one of the things that was really interesting that we learned was that you can use narrative hooks. I'll give an example. There are these two forts and you run back and forth and grab and a flag and you take it back to your team. If you have a bunch of experienced players who are in the same room with each other, talking to each other, it's amazing, you can watch it and it looks like a war movie. You go out on the Internet, it's total chaos, it's just a bunch of serial killers running around. [LAUGHTER] But then we did this thing which we called the Hunted. And basically there was one person, this fat guy in a business suit and an umbrella. And at the other end there was a white Bronco. And then there were you, depending on who you were. You basically were the secret service or whatever, the protection people. And then other people could be the sociopaths, the snipers who are trying to kill this person. And the amazing thing was that even with that really limited presentation to people we saw much higher levels of spontaneous story telling than was going on. You didn't need to put people in the same room, they just got it. Ah, I get it, I'm, and what you saw was a lot more local behavior and a lot more communication between people. By giving them enough cues and creating a scoring, or a set of game rules that really made it easier for people to say this is my role in this story. And you could pick who you wanted to be. But that was a much more successful, and by and large a much more enjoyable, experience when you're bringing a bunch of anonymous people together who don't have physical co-location and the use of voice. So that's something that we look at and it teaches us, it taught us a lot about the way to design these collaborative social experiences, sort of a twisted form narration. Narratives emerge out of this whether you want it to or not. And you could use it or ignore it. And using it seems to give you a lot more popularity online.

MAN: My name is Philip Tyne [SP?] and I'm an undergraduate in Comparative Media Studies. And I'm sure Gabe and people who play Quake and people who play [UNINTELLIGIBLE] know about the culture fans creating new levels. Which is a phenomenon that exists very similar to your episodic thing. Because there are some people who just create new ones every month or something like that. And it seems like it already exists, this episodic system. Maybe not with a narrative tension. But it already exists. And since game designers, they're already creating the engine, you must find some way to make money off it instead of having the fans get all the glory. So I mean it's true that current technology and maybe time pressures may make it difficult to implement this but I think that idea of like the episodic thing should not be discounted just because it's very difficult to implement right now. We have to keep going back to all these concepts, ideas that may have seemed a bit too far out some time ago now may seem perfectly easy to carry out. And may provide us also revenue.

MAN: Right. And my hope is that admitting that there are difficulties with the delivery system but that we shouldn't give up that notion because I think it does keep with that. And I probably want to say that I would argue for serial narrative something stronger than what we have, what exists in the ability to tweak Quake, for instance, by adding different kinds of characters and changing the world. I would probably want to add something more to that than the current capabilities. I would want something stronger than that in what I would consider to be truly serial narrative.

MAN: There were a set of movies created by a clan called the Rangers. And it was just sort of the adventurous, I mean they had enough hooks into the Quake engine that they were able to, you weren't doing anything, you were just watching these guys. And they were bad guns and some people changed sides and betrayed the other guys. Where's the Half-Life T-shirt guy? Where is he? Through a whole bunch of episodes and was enormously popular. And it wasn't its lack of popularity that caused it to go away. It was just that these people had to go back to college and finish their degrees and stuff like that. But it was interesting that one group of people decided to use the tools that were there and create this very, very popular noninteractive serial narrative using these tools.

MAN: My name is Justin Coby. I'm from Cobain [SP?] entertainment software and we're actually doing episodic content with our game, Ashram's Call, which is an online game. Part of our basic plan is that every month the plot advances for the world. Two months ago the entire island went under a sudden season, it started snowing everywhere. And the players actually ended up killing the creature which had been sucking the heat of the island. And so as a result this month things are thawing out, suddenly there are these shadow fragments popping up everywhere. We've got a couple of years worth of story arc that we've planned out. These sort of things are doing it, are happening. But it's not in your traditional retail model that we're on an online game.

MAN: So what's not working? What are you learning that's frustrating?

COBY: Ongoing content creation is a pain in the butt. If you're writing a game and you ship it you can apply a couple of patches. But online gaming is an addictive thing. That players always want more. It would be really cool if we had horses. Or what if we could have our own castle? Or I want to have siege weaponry. Or new clothing or new weapons or new spells. That it's always about more content. So you have to be able to deliver it so as not too much so as to completely overwhelm the player. But at the same time give them something so that they feel as though they're continually getting something. There's a reason to come back to the world and to continue.

MAN: You aren't seeing any problems where the distance of the player from the narrative events are so great that they lose significance? It's like oh, the guys at Ashram's Call, at the company, are making it snow now. As opposed to significance seems to be an important thing that grows out of narration. I'm really curious. I think it's a great experiment and we've got a bunch of Ashram Call's players at Valve. I really wish you were up here and were about to give the next presentation so I could hear how this was going to go.

MAN: We can talk about that later. I don't want to bore the rest of you. But the other thing about your point about doing things and being done to, I was curious if you played the Sims at all. Because I think that that is such a radically, our office is addicted to it. It's bad. Just for those of you who haven't played it, basically you get to create your own soap opera. You control some people and you can buy them furniture and you get them jobs and you have them pick up the mail and you have them take out the trash and you have them have parties. And you really feel as though you're interacting with these people. They become real people. And it's a very powerful experience. And it's not 3D accelerated. It doesn't have volumetric fog. [LAUGHTER] A couple of presentations ago somebody said is it difficult to get a game released if you don't have the latest and greatest graphics art work? Absolutely. No question. You may have the best story idea of all time but if your technology doesn't look flashy it's never going to get on the store shelves. I was curious your thoughts on this.

MAN: I think that getting to delivery systems that that's what I I'm talking about. Is only available perhaps in online delivery systems. Or at least it's more available in online delivery systems. That's the first part of that. The second part is, I guess, perhaps the way that I look at the Sims is in terms of my former interactions with all the Sim programs and all the Sim cities. Is that I seem to still inherit that kind of I'm in god's perspective notion which I still feel underwrites all the Sim city sorts of ideas. And so I still feel that notion that I am not allied in a way that you get. I still am back from that. That just may be the way that I read that.

MAN: I think I'm going to take a lesson from Greg Smith's book. We've just had a happy resolution. I'm now going to give you a cliffhanger. Come back in 15 minutes and find out the future of games.