Information Flow and Gradual Characterization

Ran across an interesting post from Krystian Majewski on design problems with multiple choice dialogue, which, among other things, draws on some play-testing reports on Emerald City Confidential:

[Playfirst] did a user test of the indie point-and-click adventure Emerald City Confidential and they described how casual gamers reacted when they first encountered a multiple choice dialogue. You might think that point-and-click adventures are a good match for “casual players”. Well, when faced with their first multiple choice dialogue, most players simply froze in panic. They assumed that one of the answer as “correct” while others would lead to failure. From the kind if information they received, they couldn’t really anticipate what would happen. Even worse, after they decided, they didn’t receive a clear feedback on what effect their choice had. They were used to the transparent feedback schemes of most casual games and weren’t able to cope with the uncertainty.

I found this really interesting, because I would have assumed, in general, that a simple multiple-choice presentation would be more accessible to casual players than some other mode of interaction.

Majewski goes on to argue in favor of consistent verbs representing standard strategies usable at every dialogue point. At any given time (for instance) the player might have the option to respond intellectually, sensually, or in a religious way — a strategy that reminded me a little of the dialogue system in Forever Always.

I’m not sure how much application this has from an IF perspective. Certainly the core IF audience tends to have different expectations about how dialogue will work, so is probably not completely frozen by a menu. Moreover, it’s possible to write menu dialogue in such a way that it gives more of a hint about the likely effects; the writing and cluing of the player determines, to a large degree, how much agency it feels like one has. At the same time, I generally agree that menu dialogue feels more distancing (to me) than parser-based dialogue, even if the latter is heavily clued with hints about what the player can say.

One of Majewski’s commenters adds:

Real dialogue flow should be such that no specific choice matters too much, but the sum of the choices does. But as you’ve mentioned this is just a nightmare from a design perspective.

I’m not sure it has to be so bad a nightmare as all that, and it’s been done. To pick the obvious example, in Blue Lacuna… (very mild spoilers, and discussion of a WIP of my own, follow the cut)

…my relationship with Progue was clearly reacting to a wide range of different inputs including both dialogue and action. (The relationship with Rume seems a little more cut-and-dried, but even there there are a number of events that allow the player to express some kind of reaction.)

What’s more, many of the choices I made were things that at least seemed (from the context of the story) as though they might well be significant, such as treating Progue well or badly, taking something that belonged to him, or breaking into his private space. Finally, the game at several points articulates the causality after the fact, by having Progue mention, “Oh, I realized this about you because you did X”, and so on.

So the interaction density here is higher, because the NPC pays attention to both action and dialogue; the dialogue itself is not exactly menu dialogue (though because one is usually picking keywords from a list of available options, it bears some resemblance to menu dialogues); and the sense of agency is heightened, if after the fact, by the way the story picks up and comments on past actions.

I’ve been thinking about these design issues a lot, because one of my WIPs (not Alabaster, but using the same conversation engine) is also structured around the idea of gradual rather than abrupt articulation of character relationships.

Many of the scenes are meant to put a particular aspect of character relationships in tension. For instance, in one early scene, a sister is irritated by her brother because he doesn’t show enough respect for her privacy and independence. She can choose to put up with this or challenge him on it, but that choice is expressed over the whole course of the scene rather than in a single A-or-B option. It may be articulated in actions as well as words: she could choose to verbally challenge him, but she could also choose to hide or show a private possession. Or she could choose neither to challenge nor confide in him, but instead express resentment by being rude to him without saying why. By the end of the scene, the relationship between the characters has been refined in a way that will affect later play.

This is an experiment and I have no idea how well it will prove to work. It’s definitely an approach that comes from thinking about the design as narrative + interaction, rather than gameplay as such: instead of considering winning or losing, or challenges to give to the player, I thought about

  • what are the major conflicts between the characters?
  • how might these be expressed (and resolved) over the course of the plot?
  • in each particular scene within the plot, which conflict is most important? (sometimes more than one, but some care is required not to make this too confusing)
    • how do I indicate which interpersonal issue dominates during a given scene? (mostly a writing challenge)
    • how do I signal to the player what options exist for expressing a reaction to this conflict? (mostly a design challenge, though the scene also has to be written in such a way as to create the cues)

Usually there are no right and wrong answers about how to resolve conflicts, and becoming closer to one character often means distancing another. Moreover, there’s no way to lose the game. Even so, there are occasions when one course of action might seem wiser than another, so I also think about

    • how do I square the player’s urge to “win” with the protagonist’s urge to make emotional decisions? Say the protagonist is tempted to be rude or teasing to another character even though that might have negative repercussions. If the interaction is poorly written, the player will experience no temptation to follow the protagonist’s irrational impulses. So part of the challenge, through writing and interaction design, is to create motivation to act in character and explore the feelings of the protagonist instead of focusing on gaming the system.

I suspect, incidentally, that the casual playtesters of Emerald City Confidential would hate this, because though there is feedback from characters about how they feel about your actions, there is no message that says “YES YOU JUST CHOSE RIGHT!!”.

So another critical point will be to package this in such a way that it’s clear it’s a story, there’s not a single right answer, and no choice you can make will prevent the completion of the game.

17 thoughts on “Information Flow and Gradual Characterization

  1. Ironically, one of the only ways of enforcing narrativist behaviours over gamist ones may be by implementing gamist penalties for deviations from potentially dramatic actions.

  2. Casual players have a lot of difficulty dealing with uncertainty. They are used to games where every action gets immediate positive or negative feedback, and where the game has almost no hidden state: every parameter is displayed somewhere on the screen, either as an aspect of the world, or on the head-up display.

    So if you want to make a game which has hidden state or actions without immediate feedback, and if you want to reach casual players, then you have to train them to accept uncertainty.

    Or accept that games with dramatic narrative choices are not very casual.

    • Or accept that games with dramatic narrative choices are not very casual.

      Sure. I’m not especially devoted to the idea that heavily narrative IF has to be suitable for casual players; for one thing, my current WIP has a fairly complex plot, and it isn’t intended to be played in 15-minute bursts over a period of weeks. So trying to fit one genre into the procrustean bed of another isn’t likely to help.

      One thing that bugged me about the beginning of Emerald City Confidential was the hand-holding tutorial mode that tells you exactly what to do at every stage. This is clearly an import from the casual game world, where players of Cupcake Mania Rush 2 expect to be walked through using the cupcake stamper, the froster, and the delivery chute (or whatever). But it felt all wrong for the point-and-click genre, because it gave the (false) impression that what followed was going to be primarily about predictable process.

      All the same, I was intrigued because this challenged my assumptions about what novice players would find simple or accessible. (Maybe the deal is that these aren’t novice players; they’re expert players in a different field, and actually novice players would have a different reaction still.)

      With what I’m working on now, I’d be hard-put to say what genre it belongs to: it’s not a text adventure, but it’s also not exactly a conversation game like Galatea, Urban Conflict, Snowblind Aces, etc., because it’s not one-room only, and instead of a big sprawl of possible interaction, there are much shorter and more directed scenes. Maybe it comes closer to a visual novel, only without the visual — but there’s a lot more information flow, in the sense that Majewski has in mind, than there is in an average Ren’Py production, and I think the experience feels very different from that, too.

  3. She can choose to put up with this or challenge him on it, but that choice is expressed over the whole course of the scene rather than in a single A-or-B option. It may be articulated in actions as well as words…

    I think that’s an excellent way to go about it.

    I’ll repeat myself from raif a bit:

    On the problem with point-and-click decisions:

    When faced with a choice, a player ought to have some idea of the stakes, of the options, and of the risks. Basically, the player usually ought to have some idea what kind of move he or she is making and what rules govern the outcome.

    Very often with point-and-click conversations, I’m not given enough information to make an informed choice: I say something conspicuously different than what I intended; I commit to a course of action I didn’t want to; and so on.

    Usually, people nose around and test the waters before making choices. Before agreements there are unspoken agreements; before unspoken agreements there is rapport. We get a sense of the situation. This nosing around is a crucial part of the decision making process: without it, people have trouble choosing.

    Lawnmowering is driven by a kind of buyer’s remorse. If you know that by making a certain decision you’re (for example) going to betray or not betray an NPC, and you have some vague notion of the consequences, then after the decision you’ll want to see what happens: you’re committed to the decision, you’re done with it.

    But if you don’t have enough information to build anticipation in your mind, you’re wishy-washy. No sooner do you choose than you want to re-choose: you’re still in nosing-around mode.

    Similarly, choice-points ought not be points. They ought to be networks of moves, each move clearly working toward the chosen effect. The gauntlet-node structure, where we play through a series of mini-games broken up by choice points, is avoidable: give each mini-game more than one solution and make each solution a plot choice.

    That structure can be used in conversation just as well as organizing plot contingencies, if we think about conversations as social “mini-games.”

    Plus, being able to move around inside the exploded view of a decision node is just more interesting than push-button contigencies.

    Conrad.

    ps – Looking forward to your WIP, Emily. Do you have a title for it yet?

    C.

  4. I’m unconvinced that multiple choice dialogue continues to be jarring for the player beyond the first instance in a game, unless it is rare in the game. If it comprises a significant part of the gameplay, anecdotal evidence on my end shows that it is accepted as simply another game input after the initial shock, particularly if it has proper feedback.

    So I think this might be an issue of proper feedback. Conrad’s comments higher up sum this up pretty well. Most works that do multiple choice well seem to understand that there is a “nosing around” stage, or they wait to introduce the important dialogue until the player understands what the stakes of their comments are.

    As an aside note in regards to IF, this same idea (that multiple choice causes the player to view stakes when there may be none) makes it harder for me to work with ask/tell when I *want* dialogue to have some kind of stake to it, since it is a system that inherently lends itself to nosing around with few consequences. If it has been done with consequences, I haven’t seen it done well yet (or seen it be any less jarring when it happens).

    The linked article does provide an interesting hypothesis regarding the interaction to feedback ratio. Seems like dialogue choices should be kept short while still conveying what they will communicate to the other party. This is probably less true in a text-based medium where someone is already reading everything, but as far as games go “not text-walling the player” is probably an admirable design goal.

    • As an aside note in regards to IF, this same idea (that multiple choice causes the player to view stakes when there may be none) makes it harder for me to work with ask/tell when I *want* dialogue to have some kind of stake to it, since it is a system that inherently lends itself to nosing around with few consequences. If it has been done with consequences, I haven’t seen it done well yet (or seen it be any less jarring when it happens).

      Certainly this is something people have complained about with Galatea: they didn’t realize that asking about X or Y was going to have a significant effect, so they were surprised and taken aback when it did.

      For most IF purposes, I’m more or less persuaded around to hinted ASK-TELL (which, like a menu, gives the player more of a clue about what specifically his conversation options are, while, unlike a menu, leaving a certain amount of navigational freedom).

      How this might play with a novice audience, though, I don’t know.

  5. Thanks for linking my article! What a coincidence and an honor: I really enjoyed Galatea back when I was researching IF for my thesis project.

    Much of what I wrote was specific for point & click multiple choice. Dialogue in IF is a different beast. Even though the choices may seem similarly limited, they are PERCIEVED by the player as wide open. In IF players have a poor overview over their options so they will typically overestimate their options. You can type everything you want after all. So at least it appears like more information is exchanged between player and game.

    On the other hand, I believe especially because of that, IF may be even more troublesome for the casual audience. Not only is it difficult to anticipate and understand consequences – your options are uncertain as well.

    But I agree with chris: everything has a lot to do with feedback. I wonder: is there a IF that uses a feature similar to command line completion?

    • I wonder: is there a IF that uses a feature similar to command line completion?

      Most ADRIFT games can do it, if you’re using the “ADRIFT runner” interpreter on Windows.

  6. For the game versus … “story” tension, one could just say upfront in the introduction “This is not a game, but rather a interactive story; there are no pre-determined right or wrong choices, you can not win or lose.” Set expectations right up front.

    However, this means you may not want to confuse the player by going back and forth between win/lose game and open exploration.

    On the other hand, as noted by Horatio above, there is also the approach of combining game with story by making the “right” choices correspond with dramatically appropriate or “in-game-fiction logical” choices.

    Games should seem “fair.” In particular, they should provide information about what the right choice is before making the choice, and, as noted here, feedback about whether one is winning or losing. A non-game doesn’t necessarily have to be fair. It might need other virtues, like “natural” or “illustrative of the human condition.”

    • Games should seem “fair.” In particular, they should provide information about what the right choice is before making the choice, and, as noted here, feedback about whether one is winning or losing. A non-game doesn’t necessarily have to be fair.

      I suppose not, though I think in fact there’s a common principle that applies to both games and non-game interactive narratives: we want to give the player/reader agency, and agency requires having some idea in advance of what an action will accomplish, and some feedback on what it actually did accomplish. The game vs. non-game distinction might then be about whether there are clear goals, or whether the value attached to a given outcome is up to the player to determine.

  7. To bring the question of fairness back around to the topic of gradual (player character) characterization…

    The problem as I understand it is that you want the player to be able to shape the character of the PC, and yet in order to make that meaningful, character decisions ought to translate into final outcomes. And you don’t want all outcomes to be equally good (or bad), since this in effect removes the player’s influence over the outcome.

    And yet, making certain character decisions lead to bad outcomes in effect punishes the player for their characterization preferences. And it can be preachy.

    I think that’s a solveable problem.

    The thing that in literary analysis we call “character” we can instead see as the preference for a certain kind of strategy.

    In Emily’s example of the sister-brother conflict:

    She could choose to verbally challenge him, but she could also choose to hide or show a private possession. Or she could choose neither to challenge nor confide in him, but instead express resentment by being rude to him without saying why. By the end of the scene, the relationship between the characters has been refined in a way that will affect later play.

    We can call these four options assertive, revealing, concealing, and passive-aggressive. Each demonstrates a different strategy preference in interacting with the brother. Let’s say the revealing demonstrates trust and concealing demonstrates suspicion.

    Now you know how passive-aggressive, how trusting, how suspicious, and how assertive the gamer is playing the sister character. And you can further refine that picture in subsequent scenes.

    Now, the game does not have to decide which of these are good strategies, leading to good outcomes, and which are bad strategies leading to bad outcomes. Rather, there can be another dimension of player agency.

    People do not pick a social strategy and stick with it. Rather, they vary their strategies with the situation, *and* they prefer to enter into situations where their favorite strategies work well.

    So, besides the dimension of determining the sister’s character, the player also must make choices to enter into contestation which are inside or outside of her chosen best strategy, and also has the opportunity to shape situations into something more amenable or less to that strategy.

    Having a happy or sad ending, then, would be a matter of the player making compatible selections on these three dimensions.

    Just a thought. There are probably other ways to do it too.

    Conrad.

  8. Pingback: Types of Action and Types of Agency « Emily Short’s Interactive Fiction

  9. The only casual gamer I know is my 6 year old daughter. She likes to play flash games about cooking or dressing. They are all the same thing, but since characters and graphical styles change, she perceives them as pretty much the same game.

    Funny thing is, all casual gamers are 6 year olds when they are playing. Expectations are pretty low and it’s all just a bit of silly fun, not something they’re willing to commit their intellects into far much.

    It’s silly to try to sell text-based entertainment to a casual public.

  10. Pingback: The Plush Apocalypse » Blog Archive » You got some narrative in my system. Hold on, I’ll get a paper towel.

  11. CYOA’s (and all games) need to fundamentally understand what choices they’re offering players, and realize when they’re offering players “fake” choices, such as “Choose between identical doors, A and B”.

    “Good design of choices” is like the “No passive voice” rule for English writing.

    More thoughts:

    http://www.mxac.com.au/drt/Choices3.htm

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s