Jeff On Games

This is in response to Darius’s post on Agile development. I post it on my blog because, well, Darius only allows people with Blogger accounts to post comments to his blog and I refuse to get a Blogger account. I don’t know why, I just do.

Anyway, Darius talks about how the Agile process used in the book Getting Real doesn’t work for game development. Here’s why he says Getting Real works for web apps:

Getting Real works for webapps, because it's about delivering a limited feature set, launching early, fixing bugs as you go, and adding or subtracting features as users demand them. Which is kind of useless if a lot of your players have already finished your game by the time the first major update comes around.

My response would be that “users” aren’t always end customers, though in the case of the book I’m sure that’s true. The system of early “launch,” fixing bugs as you go, and subtracting features is really just a more formalized version of rapid prototyping in my mind, something I know Darius likes and I’m firmly behind as well. Instead of thinking of deployment as deployment to the outside world, you can think of deployment as a locked build to your testers. Programmers develop requested features quickly and get it to the testers, who are then able to actually work with said features and decide if those features work in the grand concept of the game and request that they be changed or removed as necessary.

That’s just the way I see it though.

I will say that I think he's right about how games are different from web apps and I'd like to add a little thought of my own to it. Although it’s not the central point of his post (in my mind, I could be wrong) I have a feeling most of the responses on his blog are going to focus on it. Anyway, Darius says: "I play a game. I beat it. I'm done, and I'll probably never play that game ever again, unless it's my favorite game in the universe, or it has infinite replayability."

I personally want to emphasize that I think this is how we should think about games. I’ve heard way too many people (in academic papers and books about interactive story and narrative) assert that truly interactive works must be played through multiple times to be appreciated, and I think that’s ludicrous. As I stated in my thesis, an interactive work should be able to be appreciated the first time, as if it is the only time it will be played. The fun should come from (either implied or actual) agency in the work, not from the exploration of multiple branches during multiple times through.

I think that the pure enjoyment of agency as a form of fun on it’s own is something that most interactive story does terribly (and I haven’t seen that many people working on it that actually recognize this as the actual problem), and I think it is something that more “classic” games (meaning non-story based games) do very well.

I need to say more on this, though in a sense I already have. What I really need to do is pull out the most important parts of my thesis and write a much short document for people to read. A statement of purpose if you will. 5 pages tops, as opposed to the 140 that is my thesis. What can I say, I’m just long winded.

By the way, I’ve been extremely busy over the past few weeks preparing my room for a new tenant. As such, updating anything (let alone my blog) has been an uphill battle. Once things settle down, I will get around to updating.

So, I have a nice 6 hour flight on my way to GDC (to my layover anyway), 2 hours of battery life in my laptop and lots of thoughts running through my head concerning story in games. These thoughts have been mulling around in my head since Craig made his original post way back in the day. As I mentioned at the time, I had an introduction worked up for a formal response, but in the weeks following I decided I didn’t agree with myself (this happens when I think too much). Finally, (however many weeks later) I think I actually have a real post on the subject in my mind. Note that this pos will be in four parts, published over the next few weeks just to give it better flow and to give you a break between long rambling thoughts

Right now (after finishing the first post), this is how I see it fleshing out. This first post will concern whether or not story exists in games, and how to tell if there actually is a story to a game. My second post will be concerned with how story and discourse (aka narrative) work together with game states to make a concrete “story game” whole. The third post will probably be on redefining interactive story and narrative, since I am convinced this line of thought will probably prove the definitions in my thesis incorrect, and my last post on the subject will probably concern patterns and the differences between playing a game with a story, versus playing one without. So, without further ado, here’s the first post.

When I originally started my response to Craig, I was going to post about how all events that occur in a game constitute a story. Feeling rather confident, I opened up my thesis to find passages to quote to prove my points, looking in particular for quotes from Narratology and Story and Discourse. When I finally found the definition of narrative from Narratology, I was surprised that I hadn’t remembered it exactly right (and in fact and probably not even thought about it correctly in my thesis). Here is Bal’s definition of narrative:

A story is a fabula that is presented in a certain manner. A fabula is a series of logically and chronologically related events that are caused or experienced by actors. An event is the transition from one state to another state. Actors are agents that perform actions… To Act is defined here as to cause or to experience an event.(p5)

The thing I’d forgotten about is the way Bal defines fabula (and therefore story) as involving “chronologically related events that are caused or experienced by actors.” Now, I remember reading in Chris Crawford on Interactive Story that stories are about people, and, at the time, I didn’t think he was right since, in Poetics, Aristotle (and anyone that based their own narrative theories off of Poetics) states that he can imagine a story without characters but not one without events. This drove me to focus my discussion of story in my thesis on events, not on character (though I did admit character is a fundamental principle of story). In the years since I wrote my thesis, I’ve insisted that actors are not a central part of a story, and that a greater focus should be placed on what happens rather than who did it. In retrospect, I’ve been very, very wrong. However, I will say that, in my defense, what happens to characters and actors is just as important as their existence.

So why did this make me rethink the alleged dichotomy of story in games? Well, I realized that although I wanted to make the point that all games have stories because they are simply a series of events, under the definition I’d just re-discovered, events must either be performed by actors or be experienced by actors; actors being defined as agents that can cause action, and action being defined as causing or experiencing events (kind of circular I know). So, I thought about Tetris as an example. Events are occurring in Tetris, but because they are not being caused or experienced by anyone but the player, and (as I’ve stated before) because the player is not actually an actor in the game, there is no intrinsic story.

So, of course, this got me thinking about the following question: “When is story intrinsic to a game?” This generated the following answer: “Well, whenever there is a agent intrinsic to the game that is causing or experiencing action.” This, however, immediately spurred on other questions: “When is a representation in a game an actor? I don’t think of a block in Tetris as an actor, and I think of Sly cooper as one, but what about my dog in Monopoly? My king in Chess? Kerrigan in Starcraft? Professor Plum in Clue? My Sims? Dr. Mario? Even after I figure out which of these ‘avatars’ can be considered an ‘actor’ in the traditional sense of a story, how do I figure out which elements of it’s behavior can be considered story? When events occur because of my input, are they being ‘caused or experienced by the representation, therefore making it an actor? Crap, this is going to be harder than I thought!” Okay, so that last part isn’t a question.

Thankfully, Bal supplies the answer right there in his (semi-circular) definition of an actor. An actor is one who acts, and to act is to cause or experience and event. Distinctly lacking in this definition is the implied fact that the actor must be a force internal to the text or experience. As a player, you are a force external to the “text.” Avatars, on the other hand, are internal to the text, and are causing and experience the events you order. So in the case of Tetris, the fact that the game state is really only manipulated by you and time (and no, I don’t consider time an actor) is fairly obvious, and therefore no story exists. In the case of Sly, it is fairly obvious that Sly as your avatar is “taking orders” but is an internal force to the text causing and experience events.

Some of the other border cases are harder to say, and can sometimes depend on your focus. Here are my opinions for each of the previously stated questions of actors. The dog in Monopoly is definitely a borderline case, and depends on whether you believe your representation is actually buying real-estate and charging others to stay there. Professor Plum offers a similar problem, and it depends on whether you think of yourself making the accusations or the Professor. In both cases, I tend to believe that the “avatars” are merely placeholders, and that since the player is the one changing the game state, no story is actually intrinsic to those games.

The king in Chess and Kerrigan in Starcraft appear similar: both are “essential pieces” that you move around the “game board.” Do they cause or experience action? Most definitely, so I would have to say that both Chess and Starcraft have stories. The problem is that Chess has mostly lost its story in favor of looking at it as a state machine, making it easy to overlook the fact that the pieces are indeed representational, that taking pieces is really a small battle, and that each match is its own miniature war story. However, I will admit that looking at Chess from the perspective of its story makes little to no sense.

The last two, Dr. Mario and Sims, are interesting cases. The former I will talk about now, the latter will be its own post, because I think the whole game can be used to really nail down how stories actually work in games, and how I’ve come to believe that the concept of the “player’s story” is intrinsically flawed. Dr. Mario is interesting because he is definitely an anthropomorphic “actor” in the world. He changes the game state by throwing pills into the world, but the player is external to this, simply changing a game state that is offered up by his apparent obsession with throwing multicolored pills onto a game board. However, because of the definition I have all worked up, I would say that the game Dr. Mario definitely has story. However, like Chess, studying it from this perspective makes little to no sense.

So there is my opinion on the existence of story in games. Story only exists when there is an internal actor that is able to change the state of the game in some way, whether they are directly under the control of the player or not. That said, there are many games that have story that do not benefit from the story of study in games and merely exist to help players find a way to attach to the game. There are only two problems that I can see with this. The first depends on whether you can consider time an actor. However, since in Narratology time is considered completely separately from actors, I will not consider it an actor in games either. The second is rules of the game that affect the state of the game. The reason for this is because all actors intrinsic to a game can be considered part of the mechanics. Unfortunately, I’ve yet to figure out exactly how that works out.

In my next post (which will come out at some point), I’ll talk about game states, back story, and the Sims.

This post comes after seeing the Gamasutra post of the first chapter of Game Design Patterns, a subsequent conversation with Darius, and after reading two posts from Craig on the subject. I still can’t say that I have a firm grasp on anything about the book, but I am now willing to offer my opinions on the subject a bit more freely.

As I said to Darius, I’m not a fan of the “patterns” as I assume they are expressed in this new book (I have not read it as I said to Darius). I will fully back up and elaborate on something I said to Darius though:

I think theorists jumped into patterns before really thinking about what they want to solve. I think that patterns should be able to show problems in the mechanics that will negatively affect the dynamics and aesthetics… In a game, we don’t know the equivalent of [the base concepts we find in computer science]. Maybe “frustration,” but frustration usually comes from feeling lost or cheated; and those things stem from other more meaningful problems which we have yet to fully understand. So how can we move to making patterns?

In computer science, a design pattern is a “general repeatable solution to a commonly-occurring problem in software design. … Object-Oriented design patterns typically show relationships and interactions between classes or objects without specifying the final application classes or objects that are involved. ” [Wikipedia] I think this is the solution we should be going after, and I’m not sure that this is what is being done in current solutions. As I stated to Darius, I just can’t see the mechanic / dynamic problem that a boss monster “pattern” really solves, save those that are more story driven than game driven.

Furthermore, computer science design pattern grew out of a need to solve conceptual problems. Problems like (I’m sorry to use it so much but it’s currently near and dear to my heart) coupling. Students of CS know (or should know) that certain patterns will reduce the reliance of some objects on other objects, thus allowing whole sections of code to be pulled out and replaced without affecting other systems. What problems do we aim to solve with the design patterns offered in the book? What fundamental concept of game design do we seek to improve? Since we don’t know these concepts, it might be very hard to identify them.

I’d also like to jump on the large notation bandwagon for a minute here and offer up the fact that my thesis made a short mention of the topic near the end of it’s life which I still stand behind. When I was first working on my thesis, an interactivity / interactive narrative notation was what I wanted to be the end result. This wasn’t going to happen for an undergrad, but I still wanted to make mention of it, as well as several concepts that I felt were important to creating a usable game design notation. So you don’t have to read the whole thesis, here are the relevant conclusions.

A good game design notation should:

• Avoid using the branch or choice as its key interactive element. Instead, it should be able to notate fuzzy concepts cleanly and efficiently.
• The notation should be easy to read so that a simple glance will suffice to be able to interpret all but the most complex interactive systems.
• It should work on systems of dependency and universality instead of branching.

I would also include that because most interactive systems can (and usually are) made of much smaller interactive systems, the notation should have a way of defining and reusing smaller interactive loops in a variety of places. A simple example of this would be most game buttons that have rollovers. A hover over the button brings up a highlighted graphic or animation, which is one interactive loop. Pressing the button also causes a simple action, another interactive loop. That button may set into motion a series of steps that will end in an affect on the game world, which may in turn be part of set of steps required to do something like build a stronger tank. More loops. I’m not suggesting that you would notate the rollover or button press itself, but you would certainly notate the last two steps as separate loops, one dependent on the other.

The problem I see with most notations is that they first assume the linear interaction and causality. They end up notating things that would be better left up to narrative theory and notation, instead of actual game theory. In my opinion, the flow of levels and tension in a game has more to do with stories than mechanics, semiology and interactivity. I don’t think we need to understand this any better. The problems we have are understanding why certain interactive loops work great, and why others completely end up sucking, and way certain combinations of interactive loops work, but others end up frustrating the user.

A notation that is able to show loops easily would have immediate value. Patterns would emerge from good and bad notations, and we could start understanding just what loops work, and what loops don’t, then actually decide why and create more solutions from those. Armed with such a tool “designers [would] be able to start developing patterns for interactive systems, common ways of structuring interactive works that increase agency, reduce frustration, and enhance the willing suspension of freedom” [Thesis].

I have to agree with Raph's comment that eventually there will probably be levels of notation, and people will use whatever levels they need and are most useful. I think working at too high a level, especially with notations that inherently portray linear and causal systems, will prevent you from seeing the full potential of interactive systems.

Alright I think I’m done for the night…

I’ve been re-reading a book I read for my thesis, If On a Winter’s Night a Traveler by Italo Calvino, and I’ve found some of the passages that never made it into the thesis, but heavily influenced my current assertion that when playing a game, you are not playing yourself, but a character based in some way on a mixture of your expectations and what is presented by the designers. First let me quote*, then let me explain.

I am not at all the sort of person who attracts attention. I am an anonymous presence against an even more anonymous background. If you, reader, couldn’t help picking me out amongst the people getting off the train and continued following me in my to-and-fro-ing between bar and telephone, this is simply because I am called “I” and this is this only thing you know about me, but this alone is reason enough for you to invest a part of yourself in the stranger “I.” Just as the author, since he has no intention of telling about himself, decided to call the character “I” as if to conceal him, not having to name him or describe him, because any other name or attribute would define him more than this stark pronoun; still by the very fact of writing “I” the author feels driven to put into this “I” a bit of himself, of what he feels or imagines he feels.(p14-15)

Calvino writes the actual main plot of If on a Winter’s Night almost entirely in the second person, referring to things you are doing, including telling you that you are now reading the novel If on a Winter’s Night a Traveler, and while the assertions Calvino makes about you may or may not be true, they are easily accepted as fact, because the portrayal of “you” as the reader character is very easy to get lost in.

The actual novel beginnings are combination straight novel writing and the “experimental writing” seen above: breaking the barriers between the various levels of reader and author to try to expose just *how* novels are actually read. Another example from another novel beginning later in the book describes the what you are actually reading, and yet when engrossed in the book, you can actually believe you are actually reading a book rife with description, even though Calvino himself offers very little description of the scene.

An odor of frying wafts at the opening of the page, of onion in fact, onion being fried, a bit scorched because in the onion there are veins that turn violet and then brown and especially the edge, the margin, of each little liver of onion becomes black before golden, it is the juice of the onion that is carbonized, passing through a series of olfactory and chromatic nuances, all enveloped in the smell of simmering oil. Rape oil, the text specifies; everything here is very precise, things with their nomenclature and the sensations that things transmit, all the victuals on the fire at the same time on the kitchen stove… Here everything is very concrete, substantial, depicted with sure expertise; or at least the impression given to you, Reader, is one of expertise... (p32)

Much of the book is like this, and parts that aren’t are still referring to you, as reader, making assumptions about your life and telling you just exactly what’s happening to you. This is ludicrous obviously; you’re not in a book store returning a book by Italo Calvino called “If on a Winter’s Night a Traveler” because it’s a shoddy printing, you’re sitting in bed, reading the book itself. Yet, as reader, I had no problems accepting that this was exactly what was happening. I was not really the “you” of the book; the “you” was a character.

Basically, Calvino was taking advantage of, as well as pointing out, is the structure of communication as it flows from author to reader. Here’s a diagram from my thesis (stolen from Narratology I believe):

The “you” is really the Narratee, being addressed directly by the narrator. The book also interestingly blends the narrator who speculates about the implied author, the person or caricature who wrote the narrator as “I” from our first quote; this author may have imprinted a piece of himself on that narrator known as “I”, but the “I” (and the reader for that matter) can only speculate. We form our own opinions about the implied author, and they may match up with the real author.

Or they may not.

So what does this have to do with games? And why did I say in my last post that game designers should probably read this stuff? Because think about the implications this same model has on even the most simplistic first person shooter. As a player, I am not Master Chief. When I tell my story about a game session, sure I may refer to what happened by saying “I did this” or “I did that,” but I didn’t really. What I did was manipulate a controller, and a character on screen responded. But, couldn’t that character refuse? And would there be anything wrong with that? I say no, but why I say no is a much longer blog post than I’m prepared to write, so instead let’s look at it another way; let’s look at it from the perspective of aesthetic believability.

Let’s say I’m playing Sly Cooper. Sly is a nimble thief, I am not so nimble, even with my thumbs. If I almost cause Sly to run off a cliff, I’m not going to be totally miffed if Sly, as a nimble thief, is able to catch himself, and in reality I may be miffed if he doesn’t. Even though I technically gave the direction to run off that cliff, Sly “refused,” because, really, running off a cliff is just not something Sly would do. Go the other way. Sly is a nimble thief, and everything about his art style says he is a nimble thief. Does the animation style and characterization of Sly make me play the game a different way? Is the animation a type of self fulfilling prophesy (communications theory / psychology coming in)? What about in Ico? Do I (as player) play Ico differently because of the perceived affection between Yorda and Ico? I know I did, and it made an immense emotional impact.

So when I play a narrative game, am I me? No, I’m not. Just like Reader from If on a Winter’s Night is not me, Ico is not me, Sly is not me, and Master Chief is not me. Even in a role playing game where you are encouraged to “form your own character,” do you really play a complex social being such as yourself? Would you even want to? Or would you rather play a character? The mage. The thief. The warrior. In Morrowind, which aside from leveling was really classless, did you play as you? Or as your class?

Granted, the concept as a whole isn’t that simple, especially because this post is really centered on the linear model presented above, but I’m working on an interactive version of the model and the complete “player as character” concept. It’ll just take a while.

*note: quotes are from the version linked earlier on Amazon, and page numbers reference pages in said version.

This post is almost a direct response to Andrew’s post on GTxA, in response to Janet’s response on Andrew’s panel at GDC. As Andrew point’s out, the responses from the panelists are available.

Where I disagree with Andrew is in his comments on interface in the first point. I’m not convinced that natural language is the best idea as an interface for game based interactive stories. In IF or purely interactive drama, I might agree, but there are many problems with such an interface. Interestingly, GTxAuto posted parts of a response that I agree with a few days later. What’s interesting is that Andrew doesn’t quote what I feel is the most important part of the original post, where the real problem with natural language systems is revealed.

Frustration happens when interactions fail. Either the participant didn’t know enough to form expectations as to a result (illiteracy), or the interface led the participant to form bad expectations (inconsistency). Either way, the participant guessed, and guessed wrong. Guessing leads to failure, and failure leads to frustration.

For those familiar with my work, I refer to this as the “Willing Suspension of Freedom” (a term I believe I made up since I can find no mention of it anywhere ot the web), something I believe is central to all interactive works and applications, not just games. Both that and the willing suspension of disbelief need to work together in order for an interactive story to work as a whole. In my opinion, natural language is always going to cause involuntary breaks in the willing suspension of freedom, no matter how good the parser is.

What it comes down to is this: using natural language processing makes people believe that anything typed should get a proper response, and although this can be true at least 50% of the time, from a game perspective this is not good enough, especially when it’s understood that games are based on rules and regulations. In the case of most mass market games, adding "invisible walls" isn't necessarily a problem (despite rants to the contrary) so long as it's done well and thoughtfully as part of the design, instead of completely arbitrarily or because of technical limitations.. The loss of complete control is okay, so long as it’s consistent with my understanding of how the game should work.

Instead of adding more verbs as both Andrew and Chris argue for, I am now firmly of the opinion that our verbs simply need to be more meaningful. For example, walk and look are very simple verbs that are in almost every modern computer game, but where you walk and where you look are not usually considered meaningful. Why not? In real life, if I walk up to a bar and look directly at the bartender, eventually, he’s going to come down and ask me if I want a drink. If I walk up to a bar and look directly at a woman sitting there, she’ll probably look back and (depending on her character) start a conversation (“May I help you?”). Here, we’ve added no verbs, just made verbs already freely available more meaningful. From a game perspective this doesn’t complicate the issue for the main stream audience. I'm still all for adding verbs, so long as their addition doesn't overwhelm the interactive story and the game you're trying to make.

This probably came out all wrong, but hey, at least it’s a response of some sort. Also note that I’m pretty sure I can’t go a single blog post without linking to my thesis.