This is the first post of a sequence on plotting and interactivity, each taking on a traditional fiction-writing task and then talking about how that task is altered by the presence of player choice. The series is agnostic about whether interaction is through challenging gameplay (solving puzzles, shooting, etc., as in old-style text adventures and modern video games), is expressed through multiple-choice options (as in choose-your-own-adventure books), or is communicated in some other
way.
A set-piece is a big scene the reader can see coming and can look forward to a while, either in fear or in hope, before it’s reached… seeing a scene like that coming, watching it build to crisis, is one of the major ways of creating tension, drama, and suspense in a story. — Ansen Dibell, Elements of Fiction Writing – Plot
A set-piece needs, more than any other scene, to be tightly paced and move forward quickly. It needs to meet player expectations — deliver information we were expecting to find out, bring in conflicts or connections we anticipated — and yet it needs to provide a spin on the narrative that sends us off looking for something new. Those are the story requirements.
Use Your Existing Mechanics
Whatever mechanics you’re using elsewhere in the game, continue to use them during the set-pieces. If you have a game or story that’s mostly about shooting things and then the big emotional set-pieces are about traversing a dialogue tree or choosing which of two buttons to push, they’re going to feel fundamentally out of place. The impact built by any writing work you’ve done up to that point is undercut by the fact that the player feels less of the thrill of agency when he’s separated from the system he’s already learned for controlling the world around him.
That doesn’t mean you can’t use dialogue as a key component of a set-piece, but if you do, the dialogue needs to be important throughout the game, not just at those moments.
Conversely, if you have a game whose mechanics are mostly about shooting and yet your story demands confrontations that are to be resolved with interpretive dance, there is probably something wrong with either your story or your choice of mechanics.
Portal 2 does an excellent job with this rule. There are several important confrontations with the game’s antagonists. Each time, the confrontation is solved using the portal gun and other skills the player has already acquired — and yet the context of the player’s actions, and the reactions of the characters, makes these moments much more narratively compelling than other times when the player uses the portal gun.
If you’re writing a traditional video game, it’s tempting to correlate a set-piece with a level boss: the level design warns you that something big is coming. The tricky thing here is that really hard gameplay — boss gameplay — may disrupt the pace of the experience because the player has to restart the scene over again. That’s fine if the purpose of the set-piece is to show just how hard it is for the protagonist to get through this situation. But if the set-piece is about something else, such as an important choice or a narrative revelation, it may be better to dial back the challenge a bit and leave the player free to concentrate on this other issue.
Introduce No New Mechanics
The set-piece is not the time for the player to learn how to do anything new. It’s the time for him to apply what he’s already learned in conditions of great emotional stress. There are three significant reasons for this.
First, any time the player learns a new mechanic, the game has to explain it to him. That means that time is taken away from the excitement of the set-piece and put into tutorial guidance or at the very least character dialogue that gives the protagonist some clues. In a set-piece, though, we don’t want to be wasting any of the small amount of time available on tutorial experiences. (It is permissible to use an already-taught mechanic in a new way, with bonus points if that results in some exciting twist to the story as well. Both Spider and Web and Portal 2 do this, with terrific effect.)
Second, if the player is still learning the ropes during a set-piece scene, he’s likely to make mistakes and need to redo things. That kind of interaction is fine during less intense moments, but in a set-piece with a lot of narrative significance, it can kill the pacing.
Finally, a set-piece is sometimes a point in the story where we want the protagonist to be making a difficult choice or confronting some aspect of his own character. If the player has an important decision to make (which of my teammates am I willing to sacrifice? Would I die for the sake of my country? etc.), we want him to be fully concentrated on the choice, not on the mechanism by which he’s going to communicate that choice.
Offer Fewer, More Expressive Choices
During a set-piece scene, it’s likely that every character’s attention will be focused on some important crisis situation: a shouting match, a face-off between the sheriff and the gunslinger, a small child in danger, a tense moment between a man and his estranged wife.
This has several important implications. First, this is not the time for exploration-style interaction. If it would normally be possible within the game’s commands for the player to change the subject, talk to someone else, etc., we need to close off those options in some unambiguous way. That might be by working within the story’s fiction, by having the character trapped in a closed room or tied to a chair; or it might rely on an extra-fictional device, such as simply not offering those affordances for the duration of the critical scene. Either way, we keep the player present and focused on whatever is happening in front of him, or the momentum of the scene is completely lost.
There are compensations, though. The set-piece scene is the perfect time to add nuance to the choices the player is allowed to make.
For example, during an exploration scene we might have options like
— Search the bookshelf
— Open the safe
— Head over into the kitchen
These are broad, general options that might lead to movement and discovery. But they’re not subtle and they don’t communicate anything to other characters; they don’t let the player role-play a particular personality; they don’t in themselves advance the plot or change anyone’s attitude towards anyone else.
By contrast, during a set-piece, we might have options like
— Angrily accuse the suspect
— Quietly accuse the suspect
— Lull the suspect into a false sense of security
Even if we’re not using multiple choice affordances, everything that the player does during a set-piece scene is framed by the intense emotional experience of this moment. That means that other characters in the room should be paying attention, if possible interpreting the player’s actions as a form of communication that’s relevant to whatever conflict they’re currently having. Because the feedback from the game is always focused on the crisis-in-progress, the player is also directed back to paying attention to this situation.
The effect can be obvious or subtle. Heavy Rain applies this general principle to a number of its sequences. During quiet, atmospheric scenes, the player is free to move characters around large areas, opening refrigerators, watching television, and otherwise loosely exploring. If he doesn’t like the affordances associated with one area, he can move on to another — drinking orange juice at the fridge, then going to the other room to look for a teddy bear, say.
But at times of tension, the player’s attention is fixed on specific situations. Instead of moving around to find new affordance sets, he’s presented with one consistent — but important — set of affordances: whatever is necessary to advance this particular scene. Listen to the suspect or shoot him? Show a sympathetic posture to the injured woman, or stand aggressively across the room from her?
Making sure that the player’s possible choices are expressive and that other characters pay attention to them during set-pieces makes these scenes stand out as emotionally engaging: they’re times of heightened contact and communication with other characters.
Don’t Waste the Characters’ Breath
The earliest set-pieces will be the hardest, because they’ll still have expositional chores to do… the later ones can be more streamlined and direct. You’ll have your little world set up fairly completely in all its complexities by then; the reader will already know your characters and appreciate what’s at stake… — Ansen Dibell, Elements of Fiction Writing – Plot
Don’t let your characters explain too much during a set-piece scene. They can tell each other things, sure — but only if doing so is actively advancing the crisis. “I had an affair with your sister!” advances the crisis. “Hi, you don’t know me, but I’m a spy for an enemy government…” explains why the crisis is happening, and in this context it’s boring.
This is a more common problem than in static fiction. In interactive stories, just having a character present in the room is often an expensive proposition, because of all the reactions that have to be coded and the assets that have to be created; even in pure-text IF, where there is no voicing, animation, or illustration, there’s still a lot of work inherent in having a character around who might need to react to player behavior.
So it’s especially common in interactive stories for characters to be kept away from the protagonist until the moment of a big crisis scene where they absolutely have to be around. And once they’re finally in the room, there’s the temptation to have them explain a lot, clearing up any mysteries that the player has been wondering about up to this point and establishing what the crisis is about. This can turn potentially intense confrontations into tedious info-dumps. It happens in a host of games, though some of my least favorite examples come from the Myst series (especially with Brad Dourif in Myst III).
I fell into this trap in a big way with City of Secrets. The first 2/3 or so of the game sets up the character of a mysterious rebel leader who may or may not be on the side of right. The player meets her briefly, just once, at the very beginning of the game. He knows that at some point he’s going to have to run into her again, and there are questions about whether, when that happens, they’re going to be enemies or allies. What happens in practice is that they meet again and their set-piece devolves into a long saggy conversation, a scene without narrative drive in which the player has the opportunity to ask every question that pops into his head. It took a lot of work to build, but it impressed no one. I had the idea that the player would want to be rewarded for his hard work getting to that point in the story with the resolution of various mysteries, but it lacks impact.
It would have been much better to cut their interaction short after a few moves and move any further “reward” exposition elsewhere.
Similarly, if the player is scheduled for a late-game set-piece with a character they haven’t met before, make sure that the groundwork for that is laid down in advance in such a way that the player knows who that person is, what their stake is in the situation, and what they hope to achieve in the scene, as soon as that scene begins. The less explanation you need, the faster you can get to the meat of the action.
And if the action is talking, keep it focused. During a set-piece, any conversation available to the player needs to return to the main topic — what’s at stake? who is going to win? — as quickly as possible. That’s easy to control with a menu or conversation tree; it’s a little trickier if there’s a more systematic, freeform dialogue mechanic available. If there are a lot of conversational possibilities and we don’t want to simply make the affordances unavailable, one option is to treat everything the player says or does during these scenes as attempting to advance the protagonist’s main story goals at the moment. If he tries to kiss an enemy during a fight scene, respond to that as an attempt to throw the enemy off, rather than as a genuine attempt to start a seduction instead.
Know Your Ending Conditions
A crisis usually has a limited set of possible outcomes. The sheriff gets shot or he survives. The outlaw escapes with the town gold or he gets taken down. Mechanically, the task is to design a set-piece so that the player has no affordances that will let him end the scene without reaching one of those conclusions. He can’t wander away. He can’t derail the train. He can only go forward.
Conversely, though, thinking about ways that your mechanics threaten the integrity of your scene can be a useful way to brainstorm about twists. So your character has the ability to transport herself instantly to a distant location. Does that mean she can’t be threatened? Or does it mean there’s a way for this escape to go terribly wrong?
Tie Setbacks to Mechanics
A lot of the time in traditional narrative a set-piece before the very end of the story will leave things in a worse state than they were in to start with. Ron’s arm is splinched and the three heroes are on their own alone in the woods. Frodo and the ring have been captured by the enemy. Maybe it’s that the stakes are raised (it’s now possible for things to go even more wrong than we initially anticipated) but maybe it’s that our hopes are lowered (it now seems less likely that we’re going to be able to win).
At the same time, most game structures present the player with an escalating set of abilities (more gadgets for Batman, new portal gun features for Chell, new sneaking skills for the hero of Assassin’s Creed, whatever the player happens to invest skill points in during Fallout). That means that a lot of game narratives tend to focus on raising the stakes, or else on external menaces. The player is moved to a new, more dangerous location. A cutscene or a voice over the loudspeaker lets you know that the villain has something more in store.
Applied judiciously, however, the most powerful of all setbacks is one that changes what the player is actually able to do — one that uses the mechanics and affects the play experience in some way from that moment forward.
This is risky design business because players enjoy becoming more and more powerful during the course of a game, and mechanics that seem to take away earned rewards or make challenges less fun to play may trigger people to quit in frustration.
Echo Bazaar handles this inventively: in addition to positive stats, the player can gain menace points — Nightmares, Wounds — when things go wrong. Accumulate too many of these, and you go into a section of the story that’s specifically about how you deal with the damage you’ve taken. You can play through that and come back; it’s not a death condition. But rising menace points mean you want to play a little less riskily if you want to avoid the detour.
But it’s also possible simply to take away some advantages, as long as the gameplay afterwards remains fun and fluid, and the player feels more like he’s being challenged than like he’s being punished.
—
Other related posts: denouement, part of a series by “blogs of the round table” about denouement in games; Plot, Scene by Scene, which categorizes scenes by the kind of interaction the player is involved in during each; and the complete set of posts in the plot and narrative structure category.
Emily Short's Interactive Storytelling 
awesome! thank you!
This was a very helpful post. Although I was not really familiar with the term “set piece” once you gave some examples it was very illuminating. I understand much more clearly how such a high-stakes / limited options scene could be used to drive drama and player engagement in IF. Thanks very much.
For those unfamiliar with the term, here’s a link to a textbook definition:
http://www.ldoceonline.com/dictionary/set-piece
–Zack
I’d like to point out that balancing the lack of choices is extremely important. You indeed need to have a set of actions that’d feel balanced rather than just limited. Because games do what you describe a lot – during important scenes you are cut off from your usual abilities starting with as simple ones as the ability to freely travel around. And often, very often, instead of making it more dramatic and focused it makes everything extreemely cheesy and cheap. “You can’t leave the room right now, because OMG something so important is going on! [OK]” It doesn’t make it more believeable, it makes it much less believable. It makes the whole scene artificial. If I can’t screw up a scene in a very silly way, say, just start walking away like a complete moron and get shot in my back, because of artificial “you just can’t do it, so there” reasons, why would I find it believable when it’s possible for me to screw up things in a clever story-friendly way, like getting shot by the sheriff during a fair duel?
So I’m not saying your suggestion here is completely wrong, you’re making good points, but it’s just so many games did what you’re describing so terribly wrong that it turned into a its own painful cliche. Why would you mention Heavy Rain as some sort of a positive example is beyond me. Its habit of trapping you in each protagonist’s stupidity made me wonder why I’m so devoid of rational, obvious or just appealing choices and forced to assist the character in his/her weird approaches to fix the situation rather than made me feel being a participant in any sort of drama. Sure it can be taken as characterisation, but we’re talking about interactive experiences here where we can actually take some part in an immersive way, aren’t we? And I’m actually so afraid of that kind of stuff at this point that I’d rather play a game that allows you to run in a lot of inconsistencies than a game that cheaply limits my actions to a small set just to keep it all pretty. In fact you yourself address the issue by saying “keep the mechanics”. If the mechanics are limited without a good reason it’s a change and a damn bad one. And there’s nothing wrong with limiting your choices with your character and what your character considers important and appropriate, it’s perfectly fine and well. But it must happen exactly that way, not backwards – when your character suddenly becomes defined by some unexplained artificial lack of choice that came out of nowhere.
So as a conclusion I’d say that you not only need to prevent the player from bumping into “distracting” elements but you also need to prevent the player from even discovering that you’re preventing him/her from anything. Whether you do it mechanically or by providing story/interaction so captivating that the player wouldn’t possibly care about anything other than what’s happening in your set-piece is up to you.
Anyway, thank you for an interesting article, it was a pleasant, thought-provoking read.
Why would you mention Heavy Rain as some sort of a positive example is beyond me. Its habit of trapping you in each protagonist’s stupidity made me wonder why I’m so devoid of rational, obvious or just appealing choices and forced to assist the character in his/her weird approaches to fix the situation rather than made me feel being a participant in any sort of drama.
I think Heavy Rain is often a mess, but with some specific moments worth studying. So I take your point. But it seems to me that what you’re describing here is not about the narrowing or broadening of choices, but the fact that you hated the specific choices you were offered. These characters failed to convince you. I agree, as it happens — but I don’t think that it was the wrong idea to restrict the player’s options a bit. It would not be an improvement if, during the middle of Jayden’s intense showdown with some criminal, you could make him wander off and have a look through the guy’s refrigerator instead of dealing with the crisis at hand. It would just be dramatically slack as well as implausibly characterized.
Anyway, I think we essentially agree here. I’m not crazy about having some of the mechanics shut off without explanation either; it’s best if there’s a valid in-story reason for it. (This is also one of many reasons why writing and design aren’t separable fields, because what you can reasonably do with the pacing as a writer is tightly interlocked with the affordances opened or closed by mechanical choices.)
Tangent: there’s a lot to be said against IF’s opaque affordance handling, but one thing I’ll say for it is that it gives the author a chance to explain to the player why some particular action isn’t available right now, including in-character reasons such as the protagonist being afraid of the dark, too polite to walk out during a conversation with a superior officer, etc.
Oh, one other point. Designing a scenario with few choices doesn’t always have to mean that the player sees a screen full of greyed-out options or that mysteriously the A button no longer makes him jump. A lot of the time environmental design — controlling which nouns there are available in the space — is enough to narrow things down.
Portal 2 does this to a massive degree, since your powerful world-changing device is only powerful and world-changing on the surfaces that they chose to include in a given area, and that gives them tremendous control over where you can go when.
I was facing a situation like this just yesterday! In my Indigo speed-comp entry there is a scene, which could arguably be defined as a set-piece (though I didn’t know the term until I read this post a while ago), in which the player has a vision where he witness (or foresees, that’s deliberately unclear) a close dear friend dying in a traumatic way. For the post-comp release I have currently in the works I thought It could be a nice detail if the player character could just ignore any typed command for a brief period of time (2 or 3 turns would be nice) due to beeing in a shock state. Trouble was I’m to newbie into Inform 7 to have a clear picture of how preference order actually works for rulebooks, so I couldn’t avoid the dramatic moment beeing spoiled by simple “examine” commands which would tone-down the whole thing with answers like “Oh, yeah, the table was like this or like that…”. To my luck a workaround was easily found (and now I know a bit more about I7 rules!). I just found fun reading a post about how to deal with set-pieces just the day after I was struggling against them without even knowing they are called that way! :-)
Re setbacks: interestingly, there are quite a few games that put a significant setback, often in the form of loss of equipment, near the beginning of the story rather than near the end. Doing this provides a framework for escalating abilities, as you gradually recover the powers you had in the prologue over the course of the game. Castlevania: Symphony of the Night is a prominent example.
Very nice and thought-provoking read. Bbbbut I have a nitpick and a related wonder.
Nitpick: In the first sentence you seem to equate interactivity and player choice. But are they the same? I suspect in Portal 2, and I’m pretty sure in Spider and Web, that the relevant choices are “Do something that will let you progress in the game, or don’t.” (Spider and Web has choices later, but I think not at the point you’re discussing?) And I’m not sure that that is what we should mean by player choice.
Related wonder: Could you introduce a new mechanism in the set piece, if the player isn’t required to do anything particularly complicated with it? This won’t work if there’s a choice involved, because the player can’t effectively make a choice before she’s mastered the mechanic involved (I remember you complained about Heavy Rain interpreting your failure of reflexes in a QTE as a decision to abandon a maze), but I can imagine getting the gun/jetpack/other new capability being a big dramatic set-piece that introduces a new mechanic. I imagine the mentor figure saying “Now fire away!” as a big X flashes on the screen and you laughingly destroy the enemies you’ve been running from, or even something less lowest common denominator. I don’t know if any game has done that successfully, though maybe something like that happens at the climax of The Ebb and Flow of the Tide or in a related moment in Photopia (Ebb and Flow isn’t very tutorially about it, though, Photopia maybe leads you through it a little more). Maybe in Metal Gear Solid (which I haven’t played), where one of the apparently most memorable scenes features a villain telling you in so many words to hammer the action button to survive his torture, and then another NPC tells you to put the controller on the part of your arm that hurts? But that game clearly kicks down the fourth wall and jumps up and down on it.
Really what I want to say is, I want to play the interpretive dance game. Maybe A Single Word in Her Beautiful Calligraphy is the next best thing.
[This is basically something I already tried to post, with some links stripped in an attempt to get out of the spam queue.]
Very nice and thought-provoking read. Bbbbut I have a nitpick and a related wonder.
Nitpick: In the first sentence you seem to equate interactivity and player choice. But are they the same? I suspect in Portal 2, and I’m pretty sure in Spider and Web, that the relevant choices are “Do something that will let you progress in the game, or don’t.” (Spider and Web has choices later, but I think not at the point you’re discussing?) And I’m not sure that that is what we should mean by player choice.
Related wonder: Could you introduce a new mechanism in the set piece, if the player isn’t required to do anything particularly complicated with it? This won’t work if there’s a choice involved, because the player can’t effectively make a choice before she’s mastered the mechanic involved (I remember you complained about Heavy Rain interpreting your failure of reflexes in a QTE as a decision to abandon a maze), but I can imagine getting the gun/jetpack/other new capability being a big dramatic set-piece that introduces a new mechanic. I imagine the mentor figure saying “Now fire away!” as a big X flashes on the screen and you laughingly destroy the enemies you’ve been running from, or even something less lowest common denominator. I don’t know if any game has done that successfully, though maybe something like that happens at the climax of The Ebb and Flow of the Tide or in a related moment in Photopia (Ebb and Flow isn’t very tutorially about it, though, Photopia maybe leads you through it a little more). Maybe in Metal Gear Solid (which I haven’t played), where one of the apparently most memorable scenes features a villain telling you in so many words to hammer the action button to survive his torture, and then another NPC tells you to put the controller on the part of your arm that hurts? But that game clearly kicks down the fourth wall and jumps up and down on it.
Really what I want to say is, I want to play the interpretive dance game. Maybe A Single Word in Her Beautiful Calligraphy is the next best thing.
In the first sentence you seem to equate interactivity and player choice. But are they the same? I suspect in Portal 2, and I’m pretty sure in Spider and Web, that the relevant choices are “Do something that will let you progress in the game, or don’t.” (Spider and Web has choices later, but I think not at the point you’re discussing?)
Yeah, I take your point. I went back and forth about what terminology to use here. One might say “affordances” instead, because I am largely talking about the range of things that the player is able to do at any given moment, but
— I thought that might be off-putting to someone coming at this from a writerly angle, or indeed to many people in the IF community, since the term isn’t used there as much as it is elsewhere;
— I wanted to make it clear that what I was talking about applied equally to a multiple-choice game or story produced by ChoiceScript, Undum, et al., where anything the player can do is called a choice;
— there are a lot of games with options that fall into a sort of grey zone, where they’re expressive (either of two actions will move the game forward and will not branch the plot or leave any significant state change behind, so the options are there purely to help you define the character to yourself in the moment) or reflective (in the Echo Bazaar sense of asking the player what he thinks about something without actually recording or using that information; see also Blue Lacuna).
And I’m not sure that that is what we should mean by player choice.
If you mean that it would be useful to distinguish decisions that significantly alter the story content subsequently presented to the player, then I agree, but I’m not sure “choice” is specific enough. And expressive and reflective choices are important to storytelling as well, so even the phrase “choices that affect the story” doesn’t really seem specific enough to mean what I think you mean.
Could you introduce a new mechanism in the set piece, if the player isn’t required to do anything particularly complicated with it?
All rules have exceptions; I take a bit of a hard line in this article. But I’m not crazy about introducing a big flashing X during a set-piece for aesthetic reasons (aesthetic reasons on several levels, even). At a surface level, the tutorial elements are distracting from what’s going on and they take up valuable dialogue time when that’s likely to be at a premium. (If you have something aggressively breaking the fourth wall, then the instructions about what button to use can legitimately be part of the story, but that’s a fairly uncommon case.)
At a deeper level, set-pieces tend to be narratively about paying off a lot of set-up: you’ve been primed about a conflict that’s coming up, you’ve learned what the stakes are, and you have to figure out how to cope with it when it arrives. It’s great if there’s a twist in the middle that lets you use knowledge or skills you already have to resolve the problem in an unexpected way (Portal 2 does this beautifully). It’s somewhat less satisfying as story if the protagonist turns out to have lots of previously unrevealed options — “and Mary Sue was of course secretly a black belt in karate” is a surprising and somewhat annoying way to resolve the long-dreaded arrival of the serial killer on Mary Sue’s front porch.
Obviously it doesn’t have to go as extreme or stupid as that, but I almost always find it tighter and more satisfying overall if the big crisis scenes resolve using only pieces I’ve been shown in advance by the authors and designers.
(Sorry the spam queue ate your original message.)
The spam queue thing is just that I got greedy with the links (not being logged in may not have helped) — anyone who wants to find them should search for “Press the Action Button, Snake” and “A Single Word in Her Beautiful Calligraphy.” (Which is a Christine Love production.)
On the terminology question, I definitely understand why you didn’t use “affordances.” I guess I’d just use “interactivity” throughout. The point about expressive choices is well taken, but I think you can have interactivity even in the absence of that sort of choice. Thinking of the hoary old “Playing photopia is just like reading a short story with command prompts” canard, which is demonstrably untrue to people’s experience with it — you interact with it, even if you don’t have much choice about how to interact with it at key points. Still the interactivity is important, because it turns you from a reader to someone who’s assuming a role in the story. Judith might be a better example here (though you do have the expressive choice of how much you want to dither around in order to delay the next plot trigger).
Good point about the narrative payoff; I was thinking of a case where you know you’re going to get the power eventually, and the set-piece involves your getting it and using it. Can you name names as to who messes this up by trying to combine setpiece and tutorial? (The IF examples I mentioned are both of the “Mary Sue was a black belt” form, but as I said they aren’t really tutorials — you’re entering commands as you’ve always done, it’s just that you’re using a different command. Which you never use again, since entering the command ends the scene or the game.)
…IF might be able to pull this off more smoothly than other media — if Mentor says “Lean left to turn left” as you finally mount the hoverboard, that’s diegetic and also tells you to type LEAN LEFT.
Probably the most memorable (video game) set piece from my childhood involved a sudden, unexpected power upgrade — the end of Super Metroid. But then, that one that built upon the most fundamental verb of the game, so it was pretty clear what you had to do!
Yeah, that sounds more like a new application of an existing mechanic — which is entirely fair game.
Thinking of the video games I have played: In many of them, the set piece coincides with the cut scenes. The story proceeds as an alternation between intense interaction (with a predictable set of actions, used repeatedly, but with little development in the storyline) and moving the story on (where all interaction is taken away from the player, but many things happen in the story). Come to think about it, this is almost universally true. Lara Croft has her interesting dialogs and gets her new assignments in cut scenes. The various Sierra games (King’s Quest, etc.) work that way. Command & Conquer develops its story line between the “levels”. And so on.
Most Interactive Fiction does not escape this pattern. Telltale signs are the longer paragraphs of text, sometimes concealed in a sequence of ‘wait’ commands or in a forced dialog with very few options.
Story lines in games, with their tension, relaxation, twists and turns, often develop as a justification for the more or less strange mechanisms and puzzles in the game. This ranges from lame but efficient excuses (you are in an abandoned castle, the door is magically locked, so now run through mazes, kill monsters, and find the magic key); to exciting but shallow (you happen to be an international agent and there is always a crook who runs off to a foreign country with temples and mazes and all kinds of guns); to elaborate but far-fetched (somehow two brothers have the power to write about fantasy worlds and bring them to life, so now go and visit and fix them).
My main interest in IF is actually related to this: Interactive Fiction has the ambition of story over mechanism, of interactivity where it matters most. It turns upside-down the historical mindset of game makers and players. In spite of all the ambition and attempts, I feel that there is much, much to do. My assessment: It takes a very skilled and inspired author to write IF in which the story is told and moved forward using the standard mechanisms rather than textual cut scenes or concealed versions of them. Is it possible to help, to teach, or even to “force” the average author to reach that level of sophistication? Concretely, can engines authoring tools be designed that not only correctly imitate the relevant mechanics of the physical world and dialog, but also naturally implement the relevant mechanics of narration?
Most Interactive Fiction does not escape this pattern. Telltale signs are the longer paragraphs of text, sometimes concealed in a sequence of ‘wait’ commands or in a forced dialog with very few options.
Mmmm. Well, certainly some IF has cut-scenes, in the sense that there can be long passages where the player loses control at the granularity he’s had it previously. I disagree that passages with waiting or narrow dialogue choices are always divergences from the core mechanic, though — sometimes dialogue choice is a core game mechanic throughout.
Interactive Fiction has the ambition of story over mechanism, of interactivity where it matters most.
I’d say the tricky part is often about creating mechanisms that are capable of sufficient expression to be the centerpiece of the story scenes. (Floatpoint is probably the game of mine that does this most explicitly, in that the whole premise of the game was about giving the player the ability to communicate via objects, and then responding to a wide range of possible messages.)
That can sometimes be easier in IF than in other media because
— there are fewer formal expectations about exactly what kind of interaction will be found in a given game (contrast a first person shooter or driving sandbox game);
— in text, the special effects budget is always the same, which means that it’s possible to make mechanics that would be very expensive to produce in other media;
— because it’s possible to express many types of verb, it’s possible to include aspects of the human experience that would be difficult to express graphically at all, even with a very large effects budget — actions to do with memories, thoughts, or emotions as well as interactions with a physical environment.
Is it possible to help, to teach, or even to “force” the average author to reach that level of sophistication?
Force? No. Nor would I want to.
As far as helping goes, that’s the reason I write up things like this in the first place. (Whether it’s actually helpful or not, I don’t know.)
Concretely, can engines authoring tools be designed that not only correctly imitate the relevant mechanics of the physical world and dialog, but also naturally implement the relevant mechanics of narration?
The only way I can imagine something like this functioning at all, even in the sketchiest way, is by restricting the author tremendously. For instance, by allowing only a small number of actions in any one game and requiring that the author always work with those and disallowing more than a couple of paragraphs of output text at a time (so it’s just not possible to parse any wait or multiple choice input) or by forcing all text to be procedurally generated in a way that made it very hard to hand-craft unique text that didn’t reflect something actually happening within the world model. And none of that describes a system I would want to work with myself.
Ah, the fear of restrictions! I would imagine that you, experienced author, dislikes them as much as I do as a computer programmer. The problem is, of course, that the less restrictive authoring tools are the most difficult to master for novice authors.
By “force” I do not mean to make it impossible to do certain things; but rather, to make it more natural to write a game a certain way. (Just as a tool like Inform “forces” you to unlock a door before you can open it, but if you want you can overwrite it.) I am mindful of the complaint that tools often appeal to the mindset of the programmer, not the author.
A mechanism not explicitly found in Inform or TADS, but a nice idea in ADRIFT, is the use of Tasks. I like the concept because it “forces” the author to think more about goals and sub-goals, and less about maps and things. (The ADRIFT “Task” is not sophisticated enough, and often functions more as a generic way to give non-standard responses to commands, but the idea of “task”-oriented game design seems nice.)
By “force” I do not mean to make it impossible to do certain things; but rather, to make it more natural to write a game a certain way.
The scenes mechanism in Inform is meant to help the author to think in terms of multiple-turn narrative chunks with discrete ending conditions, unique-to-that-scene behavior, and the possibility of branching. It’s loosely based on code I wrote for some of my plot heavy earlier games — particularly City of Secrets — where I needed a consistent way to control things like dialogue choices and player movement based on plot structure rather than objects and rooms on the map. (I also like scenes for certain types of puzzle, where there are multiple ways of satisfying an ending condition and it doesn’t matter how the player goes about it.)
Thinking in terms of a sequence of player tasks is also possible, and Aaron Reed’s Intelligent Hinting extension does some interesting stuff with this, building knowledge of puzzles into the game.
Neither of those things really guarantees a harmony between mechanics and story advancement, however. I think it would be impossible to build a tool that did, because creating that harmony requires strong design sense and aesthetic judgment. What we’re talking about here is not some sort of minor craft issue that could trivially be corrected; it’s not comparable to having a world model that auto-opens doors for you. It’s the core of the art itself.
That said, I think I’m not too distressed by that conclusion. I’m in favor of making things easier for novices where that’s reasonably possible, but I also think that writing a great interactive story is so hard that no novice is ever going to do it — or to put it another way, by the time he’s done, he’ll have to have put in enough work and thought and sweat that he won’t be a novice any more.