The Designer's Notebook: Three Problems for Interactive Storytellers

Last month's column on adventure games brought such a strong response, I thought I'd discuss an important related issue while I still have everyone's attention.

Interactive storytelling has been a subject of hot debate since computer games were first created. Many of the early game developers were programmers with no experience at writing fiction, so there was a real shortage of talent at creating things like character and pacing and plot. Since then professional writers have entered the industry, and the quality of our storytelling has improved somewhat.

Despite that, however, there's still a larger philosophical question looming over the subject: "What does it mean to say that a story is interactive?" It's a question that remains unanswered. You could argue that no answer is needed - adventure games tell stories, and they are interactive; therefore they constitute interactive storytelling, and no further discussion is required. The problem is that most adventure games tell rather poor stories. We've never yet seen an adventure game that was the caliber of works by Dickens or de Maupassant.

I believe that interactive storytelling suffers from three very serious problems, and they're clearly visible in adventure games today.

The Problem of Amnesia

This is the simplest and most obvious of the problems. In a normal, non-interactive story, the characters belong in the world of which they're a part. They understand that world. They know what's in all the drawers in their apartment and what's in all the shops in their town. When they first get up in the morning, they don't start their day by opening up every single closet to see what's in it, nor do they pick up every object they see and stick it in their pockets in case it might come in handy later.

But that's not true in adventure games, is it? When you play an adventure game, you have no idea what is going on. You have amnesia. Even if start the game in your own home, you have to explore it. You don't know what's going to happen to you, so for safety's sake, you pick up everything you see, and you end up carrying around a collection of objects that make you look like a demented bag lady. (Consider the original Adventure: a lamp, a birdcage, a wooden rod, an axe, some gold coins, a bottle of oil...)

A few games have actually been written to incorporate this problem into the plot. There was a game simply called Amnesia, published by Electronic Arts; and there was a game based on Roger Zelazny's series of fantasy novels, The Chronicles of Amber, which started with a character who had amnesia. But let's face it, this isn't a major genre of literature. There are very few novels about amnesiacs. In most stories, the characters just charge ahead and have their adventures, and it's up to the author to make sure they're carrying whatever they need to survive them (if they're going to survive them).

There are three types of stories in which the characters start empty-handed and ignorant, and have to figure things out on their own. One is the rookie-in-a-new-situation story - the new recruit who's just joined his ship in the Navy, or the gunslinger who's just been made sheriff of the western town. In these cases it makes sense that the protagonist has to do a lot of exploring before he can accomplish anything. The other two are mysteries and heroic quests - both situations that involve a lot of talking to strangers and examining unfamiliar objects.

It makes sense, then, that most adventure games are, in fact, mysteries, heroic quests, or new-kid-in-town scenarios. There's nothing particularly wrong with that, but it does mean that the genre is limited by the amnesia problem. We may be able to create interactive stories, but we can't create any kind of story we want.

The Problem of Internal Consistency

When we judge a work of fiction, we judge it on a number of things: are the descriptions clear? Is the dialog believable? Does the writing flow smoothly? And so on. But we also make a more fundamental sort of judgment as well. If you walk out of a movie, having seen it, or if you put down a book, having read it, and you say to yourself, "I don't think he would have done that" or "I don't think she would have reacted to that situation in that way," then we say that the story has a flaw. There's something wrong with it; it doesn't make sense. Any story must be true to its own inner laws. It has to be coherent. At any point in the story, the circumstances at that point have got to be consistent with everything that went beforehand.

Mysteries are an interesting example of this, because in a mystery, you have a lot of different possible explanations for the crime, and right up until the detective gets everybody in the room at the end and reveals which is the correct one, each explanation has got to seem plausible. But the rules of the genre require that only one of them may actually work; the rest must be logically impossible, and furthermore the author must have shown all the clues to the reader. It's a very difficult task to create four or five apparently consistent possible explanations, and introduce them to the reader in such a way that the clues are all there, but the reader is still surprised to learn which is really the correct one.

This requirement for internal consistency isn't a matter of pure logic, of course. I don't mean to suggest that at every point in a story the circumstances should be rigidly derivable, like a mathematical proof, from what came before. But if you look back at a story, it should be consistent. Stories shouldn't be predictable, but they should make sense in a satisfying manner.

So what does all this have to do with interactivity? The answer is, nothing. Interactivity is about freedom. Interactivity is about giving your player things to do and letting your player do them. The whole point of interactive media is letting the player do something on her own. What that means is that a lot of times your player is going to jump off the rails and go do completely weird, unanticipated stuff. That doesn't work very well in stories.

Consider Superman. Superman is a character who is congenitally incapable of ignoring a baby who's crying in a burning building. He never says, "You know, I'm gonna let somebody else deal with this for once." But what if our player is being Superman in a computer game? Here's the burning building. Do he run in and save the baby? Well, he has to if he's Superman, and if he doesn't do it, then he has violated Superman's basic nature. There's this conflict that arises between the player's desire to do as he chooses, and your desire to impose a plot and characterization on him. It's a tough one. How can you be sure that the player is going to do something that is coherent, that goes along with your story?

The Problem of Narrative Flow

As we all learned in junior high school English class, every story is supposed to have an introduction, rising action, a climax, falling action, and a conclusion. It's the business of the story's author to structure it in such a way that it builds to a dramatic climax - an action, confrontation, or other event which resolves the story's inner tension. One of the problems an author faces is making sure that all the characters involved are ready - psychologically and physically ready - for the dramatic climax to take place. If he doesn't, then we read the story and say, "Wait a minute - where'd that knife come from?" or "How did he know the villain would be hiding in the hall closet?"

With ordinary fiction, this is a challenge, but at least you as the author are fully in charge. The characters have to go where you tell them, to know what you want them to know, because they're all part of your picture. You set up the pieces, interlock them like parts of a jigsaw, and when the puzzle is complete the picture is formed; the dramatic climax takes place.

You can't do this in interactive stories. There's one character who's outside your control as an author, and that's the player. The player is doing whatever he wants, and taking as long or as little time about it as he likes. How do you make sure that when the dramatic climax takes place in your interactive story, your player is there and ready for it? This is the Problem of Narrative Flow.

There are three traditional solutions to this problem in adventure games. One very simple one is to limit the interactivity. You either cut down the interactivity so that the player can't get away from the plot, or you give them a lot of interactivity but you make it all meaningless - the interactivity doesn't really affect anything.

I don't think this one is an acceptable option. Reduced to the minimal case, the game turns into "Hit ENTER to see next screen." Limiting interactivity is not what we're supposed to be about here. A few games have actually done this, but they were universally acknowledged to be bad games - certainly not the ideal example of interactive storytelling.

The second traditional solution is that you say, "Too bad. If the player's not ready for the dramatic climax, that's tough." In this case, you can create a world that's alive, that goes on around the player, regardless of what he's doing. This makes for some really interesting adventure games. Night falls, and people come out of their shops and go home, and the muggers come out, and so on. It's interesting to watch things take place around you in one of these kinds of games. The difficulty with them is that you tend to lose the game a lot. You end up having to start over all the time, because you weren't ready for the dramatic climax when it occurred. But that's no way to present a work of fiction! Nobody reads a book by reading page one; then starting over and reading page one and page two; then starting over again and reading page one, page two, and page three, and so on. It would drive you crazy.

There is of course a workaround to that problem, and it's called "save game." But saving the game utterly destroys my suspension of disbelief. If I'm fighting off the evil trees in the enchanted forest with my magic sword, I don't want to stop every five minutes and have a little interaction with my hard disk drive. Saving the game makes it unnecessary to restart over and over, but at the expense of taking me out of the world I'm trying to belong to. I don't think that's a satisfactory answer either.

The third traditional solution to the Problem of Narrative Flow is the classic adventure game solution, and that is to make the plot advance along with the player's advances. This absolutely guarantees that the player will have everything he needs when he gets to the dramatic climax. If he needs the magic sword, then he'll have the magic sword, and if he doesn't have the magic sword, there's no way he can get to the dramatic climax; the plot simply doesn't go anywhere. It's easy. You just link up the player's actions to the advancement of the plot.

The difficulty with this solution is that it's mechanistic. It turns the game into a series of puzzles to be solved, and once you've played two or three of these games, you can really see it. If nothing seems to be happening, you must be doing something wrong. When you do something right, then interesting things happen. The flow is jerky, stop-start. You as the player can do what you like, but you don't have the sense of being carried along by the story; in fact it's quite clear that you're not in the story, the story is an external mechanical object that only progresses when you do the right things. It's rather like trying to operate a VCR with unlabeled buttons.

Conclusion

You might think at this point that I'm going to offer some solutions to these problems. But I don't have any solutions, and I'm not certain that there are any solutions. I won't go so far as to say that interactivity and storytelling are mutually exclusive, but I do believe that they exist in an inverse relationship to one another. The more you have of one, the less you're going to have of the other.

In its richest form, storytelling - narrative - means the reader's surrender to the author. The author takes the reader by the hand and leads him into the world of her imagination. The reader still has a role to play, but it's a fairly passive role: to pay attention, to understand, perhaps to think… but not to act. A good story hangs together the way a good jigsaw puzzle hangs together when you pick it up, every piece locked tightly in place next to its neighbor. But it ill tolerates any fiddling. Remove a few pieces, and it's likely to fall apart.

Interactivity is not like this. Interactivity is about freedom, power, self-expression. It's about entering a world and changing that world by your presence. In most games the world is static and dead until the player arrives; the player is the only thing that makes it move. Interactivity is almost the opposite of narrative; narrative flows under the direction of the author, while interactivity depends on the player for motive power.

This doesn't mean that I'm backing down from my call for the game industry to create more adventure games - far from it. But I recognize that adventure games, at least at present, tell only a limited kind of story: the mystery or quest. We can't yet make an adventure game about a troubled family or a young man's slow descent into madness. Adventure games have to sacrifice some of the best things about stories for the sake of interactivity.

I think adventure games should be just that: games about adventures. They should give the player a sense of achievement and accomplishment. They're about doing, making a difference. This does not mean that they have to be shooters or twitch games, only that the player and her actions are the most important things in the game. In computer gaming, you subordinate the player to the plot at your peril.

It's not our job to tell stories. It's our job to build worlds in which players can live a story of their own creation.