Increasing Agency and Decreasing Linearity in Choice Narratives

Yes: this is exactly why I migrated to choice-narratives. I know the appeal of parser is the player can do anything but sometimes I don’t want the player to be able to do everything, I want them to engage with my story within the parameters I’ve set. And that’s not a bad thing. If you’re frustrated that it feels like you want to write regular linear stories that’s fine. In choice, you get to control every aspect of what can be done in your world. This requires re-thinking what a “puzzle” is or eschewing puzzles or roadblocks altogether.

And choice narrative are exceedingly easy to beta test - you can get one or two readers and they don’t need to try breaking windows or making sure the player can’t flush the plot-critical antique pocket watch down the toilet because you didn’t think of that.

I will disagree with “The whole “freedom to explore” (like traditional parser games) is out the window though.” As one authors choice narrative the game has exactly the amount of exploration the author wants to build into it. I started in parser and love the non-linearity, so I want to bring that to choice and try to take every opportunity to make sure the player never just has one choice they can make. Even if it’s just several different ways to role-play how to say “No, Bob, I don’t care about your lost wallet.” that is still interaction. And doesn’t need to drastically alter the story to make the player feel involved. Even if behind the scenes I’m leading the player around and their sense of agency is simulated, I never want them to feel like they’re just clicking for the next page.

But I heartily agree with “creativity is born from working within (or overcoming) limitations.”

There are dynamic fiction games that offer no significant choices but still let the reader interact. While this extreme low-agency format isn’t my favorite, you can write something like Xalavier Nelson’s Screw You Bear Dad which relies on low agency primarily for its incredible comedic timing and pacing. Or Michael Lutz’s *My Father’s Long, Long Legs" where the low agency serves to build dread about the Monster at the End of the Book.

Choice Narratives offer more author control over pacing and focus. Is the wallpaper important? If not it’s just text. If so, I can make a link so someone can examine it, pull it, or lick it. You get the opportunity to write completely different kinds of interactive narrative that isn’t just about finding a key to get through a door. But you can do that if you want!

This exactly. The joy of choice narratives is the reader doesn’t get to go off on a tangent you haven’t planned for them. You don’t ever have to worry about telling parts of the story you don’t want to because your player would rather look under all the rugs then talk to the main character who’s tapping their foot waiting on the stairs to direct them to the room of Enchanted Things to Do.

11 Likes

And choice narrative are exceedingly easy to beta test - you can get one or two readers and they don’t need to try breaking windows or making sure the player can’t flush the plot-critical antique pocket watch down the toilet because you didn’t think of that.

Even better, in many cases you can automate it with a player who just clicks choices at random (or semi-random), and then run thousands of playthroughs to check for things breaking or getting stuck. (I’ve never written something complicated enough to warrant this, but I like that it’s a possibility).

1 Like

In my experience, I don’t even need the illusion of choice, even meaningless choices add to the immersion: E.g. “You eat ice with your friend. What flavor do you choose?”. I immediately know that this choice probably won’t affect anything, but getting to make that choice still matters to me. What’s worse (and where many non-IF games also get on my nerves) is when there is a choice that feels like it matters, but when I replay the game I find out that it actually doesn’t.

Exactly, choice games can still have a world model, it’s just not built in. In theory you can convert any parser game into a choice game by just presenting a loooong list of all possible actions (though that would probably destroy the puzzles).

Choice of Games has 2 tools to automate this IIRC. One examines all possible branches and sees if the end can possibly be reached, but it makes some simplifications and is inaccurate because of that. The other tool is what you suggested: It uses random choices to get through the story, and does that many, many times to find dead ends and such.

If you can work in the rigid confines of ChoiceScript, it’s a good system with a big community around it.

2 Likes

This is perfectly fine. Not every parser game needs to allow every single action the player might try. I personally tend to bounce off of games like that because it can feel overwhelming. When every action is on the table, it’s harder to intuit how the game wants me to act. A well-placed rejection message can help the player stay focused on what’s important, and it can help characterize the PC or add flavor to the world. My parser game did a lot of this, and everyone who commented on it seemed to enjoy it.

6 Likes

We’re definitely getting into a tangent here, but I tried that for Miss Gosling’s Last Case (in the current IFComp), and it takes some rethinking to figure out how to make puzzles that aren’t obvious from the choice of verb but imo it’s definitely doable. Or look at limited parser games like last year’s The Vambrace of Destiny, where there are exactly 26 possible actions corresponding to the 26 letters of the alphabet.

4 Likes

I speak in generalities to cut to the chase. I was speaking from Lance’s perspective. I don’t believe he should build a choice-based world model (and complicate things) so I was basically saying, “get those thoughts out of your head” and embrace the idea of stereotypical choice IF. Don’t be a trailblazer at this point. Perhaps that sentiment didn’t come through.

I tend to get a lot of pushback when I generalize things. It’s not my intention to paint things into a corner. Talking be hard for me. :wink:

3 Likes

There are examples of this. Lynnea Glasser produced Coloratura in both parser (I7) and Twine (I believe Harlowe) versions. And I re-created the setting and baking mini game from parser Baker of Shireton in choice-narrative Cursèd Pickle of Shireton.

And implementation isn’t nearly so difficult. There are ways around just making “a long list of choices” - since you don’t need to include choices that do nothing those can be left out.

I found the best solution is to create your world model as a series of hubs that you can drill into at different levels and return to the original location. The thinking to break is that every choice has to move the story forward. If you consider you have a passage representing a room, the text is the normal description. Anything in the text that you want interactable you make a link. Clicking it brings up more text that is similar to an EXAMINE function. This may just be flavor text “The wallpaper is flocked but otherwise unremarkable.” or if there are new commands you can include choices on this screen that weren’t obvious before. You can think of it as being focused in on this “object” with all its interactions and provide a “Step back…” or “Return to the living room” link that sends you right back to the original page with the room description.

Most people think that there’s no way to do puzzles in choice because “You can only list links and the solution must be among them” but it just takes some thought to figure out how to use variable/qualities and plant things to discover that are several links deep. If you click the wallpaper link in the description it may give you an option to “Lick the wallpaper” and doing so sets a variable - maybe you’ve just ingested psychedelic wallpaper ink and when you return to the room description as usual everything is different and now there are elves romping around you couldn’t see before and can now interact with.

The other habit to break from parser is that you need to actually mimic every single element the parser world model. If a player needs to open a chest to get a key to unlock a door, the first impulse is “I need to make a chest, and make variables for whether it’s open or closed and flip the choice correspondingly, and since it’s a container the player might want to put other things in the chest so I need to create an array of inventory items that might be in there and create an inventory system so the player can put any item in the game in there…” - no you don’t! You streamline the action you want them to do. There’s a chest. Click to open “You found a key!” (set a variable indicating the player has a key). The whole reason for the chest is to give the key so it can happen automatically - or if you really want you can give the player a choice whether to take the key or not, but why wouldn’t they? so you can streamline. There’s no plot reason to interact with the chest anymore so it can now just revert to scenery and no longer have an open link. Or you can even just remove it from the description since its story function is fulfilled. The next time the player encounters the correct door, they’ll see a new choice “Unlock the door with that key you found in the chest.” That’s why choice narratives are focused and streamlined - the interaction only needs to be as complicated as you make it.

5 Likes

Something that would come quite close to the parser way of puzzles is to have the puzzle links hidden in descriptions: Set the color and hover cursor to normal values with CSS. The player then has to find the word to click with clues, like having to find the right command. Or click on every word, which would just be typing every command with everything.

I don’t know if this is done anywhere in practice, but that also seems like a good solution to me.

Concealed links is a strategy, but in practice I think that would be a frustrating endeavor - basically having to “pixel hunt” through text to find anything that is a link but doesn’t look like one. I’ve considered prototyping a choice game where every word in the text is a link so there is some consideration involved. Maybe not an entire game, but that might be a good occasional situation where if you were trying to find clues in a room description, or figure out how to take apart a complicated watch that is described in great detail; every unimportant word gives you a “nothing here of interest” type message so the player needs to pinpoint what they’re focusing on, replicating close search and scrutiny.

One thing I have done…I’ve made a choice link that is mostly normal, but inside is one word that is a different link if the player happens to focus on that word.

6 Likes

I think a lot of puzzle types (those that involve setting up a specific set of conditions for something to happen, interacting with machines etc, dialogue puzzles, mostly exploratory object-finding) work totally fine in choices. On the other hand, “uncommon use of object/verb”-style puzzles are harder to fit into the choice paradigm.

In the point & click game that I made in Ink, I tried to have a “use item” option that appeared contextually only on things that could actually meaningfully have an item used on them, while still leaving the choice of item up to the player. But I’m not sure whether it added much in the end.

4 Likes

One way to do this is to basically assume player knowledge and awareness is the prerequisites to solve a puzzle. If the solution to get a key out of the drain is to bait a fishing pole with a chewed piece of gum, you can infer a lot about this by the player’s actions instead of making them literally type in the command.

You have your passage about the drain, mention something shiny down there. They examine and see a key. No idea how to do it. When the player finds the fishing pole, they’re aware of its existence. When they find a piece of gum and chew it, they are advised that the flavor quickly fades but it might be a good idea to hold onto the chewed gum instead of leaving it for the butler to find. After they’re aware of both these inventory items and investigate the drain again, they get a new link “Put the gum on the end of the fishing pole hook and fish out the key.”

This actually eliminates a lot of “guess the verb” frustration.

3 Likes

The problem with that is that it’s hard to prevent the player from blundering into the correct solution without figuring it out themself, especially if they’ve reached the “I’m stuck so I’ll just try everything again in case something has changed” point. In my point & click game, for example, I had a puzzle that was supposed to work similar to this (dialogue options appearing once the player theoretically has the requisite knowledge to say the appropriate thing), but when watching playthroughs I saw a few people who didn’t put 2 and 2 together themselves but still stumbled into the solution once it appeared as an option.

I get that impulse, but I highly encourage a perspective change of being on the player’s side. Having them blunder through a puzzle solution by accident is a better situation than them knowing exactly what they need to do but not understanding how to do it.

So much opens up once the author gets out of the mindset that their purpose is to prevent the player from advancing in the game. In fact, I routinely have made puzzles look difficult but helped and allowed the player to solve them quickly. The player gets a warm fuzzy believing they’re a genius or just lucky and have somehow “outsmarted” the game while I’m behind them, “Sure you did sport, great job! Gosh, however did you knock down that devious series of logic leaps I set right in front of you with blinking directional arrows? :wink:

10 Likes

Right – I think the key thing is to make sure the player doesn’t feel like the best/easiest way to make progress is just to lawnmower through the list of possible actions via trial and error.

(I will just link to my article on this stuff again, since it’s hopefully relevant)

7 Likes

I don’t really understand this “lawnmower” issue for choices;

Given actions A, B or C, it would be likely that after choosing A, then B and C might no longer be available and replaced by new choices. In some cases they might still be there, but with compound consequences. So it’s quite important to choose the right one or at least get the order correct.

  1. kick beggar
  2. donate coin
  3. walk on by

After (3) i could maybe come back and do (1) or (2). (1) followed by (2) will be different than (2) followed by (1). And maybe (1) hides (2) and vice versa.

The only time I’ve found that you can feel confident that A, B and C can be chosen in any order without affecting the outcome is during conversational dialogue. In this case, you’re usually asking a character some questions and can ask these pretty much in any order. I don’t see anything wrong with that, because the player is actually expected to do this, and to choose the order they think is most interesting first. Often there are some “padding” questions to add irrelevant background.

Edit: it’s like claiming for a parser game; Hey you can just lawnmower the map by going N, E, S, W, everywhere until you get the answer.

1 Like

Given actions A, B or C, it would be likely that after choosing A, then B and C might no longer be available and replaced by new choices.

But in this case we’re specifically talking about choice-based games that otherwise mimic a parser-style game (with persistent rooms and a world model and puzzles), not the kind where choices go by once and disappear. More akin to a point & click graphical adventure, which does sometimes devolve into “try everything until something works” when you’re stuck.

EDIT: at least, that’s what Hanon and I were talking about.

3 Likes

Yeah, exactly - I enjoy these kinds of games a lot but without forethought they absolutely can devolve into USE BANANA WITH PIRATE. USE BANANA WITH LOCK. USE BANANA WITH FIREPLACE. The article I linked above goes into a lot more depth on this if you’re interested!

7 Likes

This happens to me all the time in parsers. I fundamentally don’t understand why the lawnmower strategy keeps coming up in these threads.

Are y’all really saying that you stare at the screen until you think of the correct answer and type it in? Because I am typing in all kinds of stuff and creating a massive pile of failed actions until I finally land the command the author had intended.

12 Likes

I think preventing this kind of play is a challenge, but it is just the challenge of making a good game. It’s hard to give the right amount of information. It’s hard to have challenge without frustration. I don’t know how to prevent lawnmowing at the mechanical/or system level, because it is very rarely a mechanical or system problem.

I don’t think of it as a strategy but rather a resort. What was missing that made it necessary? Sure, sometimes things just don’t click with people, but hopefully design can minimize rote experiences.

(and like Hanon says, player advocacy is a good design POV.)

9 Likes

Dang I should have waited until you replied before I posted because you hit the nail on the head. Neither parsers nor choice games are safe from the lawnmower. Once the lawnmower starts being used, it’s indicative of a game design problem, and not something the player wants to do in just choice games, specifically, somehow.

6 Likes