For what it’s worth, I don’t really agree with the criticism that the story said nothing. Things happened, our vampires went from place to place and did things, I was able to join a faction (or not) and annoy some people (or not) and some other things. I do agree that the choices often feeling mostly cosmetic, and the undefined lore terms, and, yes, the lingering “AI-ness” of the text, all can conspire to make the story feel more hollow that it otherwise might have, and I would say there’s plenty of room for improvement. But at least for me personally I still got the sense there was something there, just in need of refinement.
1 Like
I struggled with the fact that I didn’t know to what degree you or the AI had influence on the story before critiquing. But yes, while the presentation is quite strong (and that is no small feat), the story is weak. Relying on visceral, animalistic descriptions of fleeting thoughts does not a story make.
I get the feeling that you want to build a world to live in, more than tell a story within it. Prove me wrong or prove me right, but pick a direction and commit to it.
If the world is more important, then you need to expose more mechanics and attributes so the player can create their own story and fill in the blanks. The mechanics need to be satisfying and apparent. However, if you want to tell a story, then you have to have something to say to make the story satisfying. I caution trying to be great at both, but at least focus on one over the other initially.
David Sirlin’s three Writing Well articles [ 1 ] [ 2 ] [ 3 ← particularly this one ] were a huge inspiration to me because I’m not a creative writer. I don’t really find pleasure in writing for writing’s sake, but I absolutely see the value in the clear communication and expression of ideas. And I think writing a story is largely about having something to say with a delivery that touches upon our emotions and intellectual thoughts.
Rhetorically, I ask, what are you trying to say with your game? That answer should be the driving force behind your story and efforts.
As a side note (or dagger in the back?): I think AI is holding your story back, influencing it in a negative way even, and you will be fighting it constantly. It will not elevate your story. Only you can do that.
@sparkletwist Read the third article I linked to. This is what I mean when I refer to having something to say.
3 Likes
I do agree with this.
To expand on this a bit: have you seen games like Fallen London or cyberpunkdreams? (The technical term that we tend to use is “quality-based narratives” or “QBNs”, but the more general name for the technology behind them is “storylets”.)
The general idea is that the game is made up of a bunch of little vignettes (“storylets”) that are presented at random, usually contain one or two choices each, and those choices affect stats which control which vignettes are available next.
Is that the sort of game you’re aiming for here? I get that sense from the demo, but I could be entirely wrong on that.
3 Likes
Interesting read! I can definitely relate to being the “STEM kid who got an A in English class but also kind of hated it,” and I had my own “Professor Cooney” of course. (And in English class, naturally!)
Irrelevant digression
She got extremely sick in the middle of the year and we had a substitute teacher for like 2 weeks straight. We all loved that sub and pretty much wished she could stay our permanent teacher. I would like to think that woke up the regular teacher to how disliked she was, but not all that much actually changed, so probably not. I don’t have all the details but I heard later on that the illness was some chronic thing that came back and got much worse and she had to quit teaching… probably a worse outcome than even she deserved, although at least nobody had to suffer her particular brand of pedagogical malpractice any more.
Anyway, I like where you and @Draconis are going with talking about mechanics that build a more emergent story. An interactive work doesn’t necessary need to “say something” in a literary sense, because experiencing it for what it is the whole idea. I’ve played a lot of (tabletop) RPGs and so I like these kinds of organic if not quite as neat and clean stories a lot. (Of course, in that case, clearer mechanics and more meaningful consequences would help a lot, but I have already been over that so I will not belabor the point)
2 Likes
@sparkletwist Yes! Absolutely.
Edit: I mean I agree with you… to a degree. It doesn’t have to speak deeply about the human condition, like classically defined literature does. Absolutely.
Having something to say is akin to picking a strong essay theme and supporting it with every detail and nuance written. So when a passage of text feels like filler, it doesn’t really say anything important to the game.
So in a story sense, exploring the struggle of understanding your vampiric affliction could be chapter one, and now you have a more focused direction with authoring the story. The focus can shift as the story progresses, but you have to pick a something and actually explore it. The game is a story.
However, if the crux of the game is about managing your affliction through stats, then expose that aspect and make it obvious as to how your choices are affected by those stats. Then the story takes a passenger seat and you are seeing how your hand-crafted character navigates the world and what options open up to them. The story is a game.
I’m not sure where Sangwright fits between those two ends of the spectrum… and that’s something that needs to be rectified with the game.
@robserpa I have to apologize. I was overly harsh and that was uncalled for. I didn’t need to trash your story. I think the reason for my ire was that when someone uses AI to generate their story, I feel like they are taking a shortcut and that perceived lack of effort can turn from “this is great and has potential” to “they are actively trying to waste my time”… and I can get aggressive. I guess this is where giving the benefit of the doubt comes from. Sorry about being an ass. I still don’t know if I’m being critical of you or an AI algorithm, but for what its worth, I apologize.
3 Likes
No apology needed, but I sincerely appreciate it, and understand where you were coming from. For what it’s worth, the frustration you described (“are they taking a shortcut”) is the right instinct for this community to have, and part of the reason I joined
The Sirlin articles were a good read, so thank you for those! I fell down a bunch of rabbit holes as a result; lessons learned incoming in another post.
You asked where Sangwright sits on the spectrum between “the game is a story” and “the story is a game”; the honest answer is story first. The mechanics are there to make you feel something specific, like the Waning ticking up; it’s meant to be the slow dread of realizing you’re comfortable with things that should bother you (that loss of morality, your former self). The stats serve the narrative, not the other way around.
I hear the critique that the current version doesn’t commit to that clearly enough. If it’s story first, the story needs to actually say something, not just in the premise, but in the individual passages. Call it new writer’s burden 
On the AI question: I won’t pretend I’ve figured that out. The pipeline has moved toward hand-written prose, and I’ve built a detection process to catch the structural tells. Whether that’s enough remains to be seen, partly from conversations like this one, and I take it seriously.
2 Likes
I know Fallen London by reputation only, but haven’t played it (my husband suggested I should). cyberpunkdreams I’ll look into.
The honest answer on the QBN question is: The current arc is linear-branching, so your choices flavor the prose but the backbone story is the same per path. That was a validation decision, not a design conviction. I needed to test whether the writing and the nightly format worked before investing in a more open structure. Marketing and finding audience to get that validation has been tough.
Storylets are interesting to me because they’d address the cosmetic-choices problem that @sparkletwist mentioned. If the world is assembled from vignettes gated by stats, the player’s choices genuinely shape what they encounter, rather than just how the narrator describes it. But I haven’t committed to that direction and I’d rather be honest about that than pretend I have a design doc for something I’m still thinking through.
If you have thoughts on where the QBN approach works well and where it falls apart, please share.
The RPG comparison is actually really helpful. My husband loves his RPGs, and I’ve bounded ideas off him. When he’s a GM, not just a player, sometimes he’ll set up the situation, and then let the story happen from the characters’ choices, not from a script. That’s closer to what I want Sangwright to feel like than a traditional branching narrative where the author controls every beat.
The tension right now is exactly what you and @HAL9000 are circling: if the experience is supposed to be emergent, the mechanics and info presented need to be clear enough that the player can make real decisions. And if it’s supposed to be authored, then the writing needs to earn the player’s eyes. I think it’s both, but the current version doesn’t commit hard enough to either side to make the blend work. It’s useful to hear when stated so plainly.
I’d love to hear your thoughts on that.
Regarding storylets, I’ll just say my favourite boardgame is The Captain Is Dead. Each turn requires a scenario card to be drawn from a deck that brings about detrimental things to deal with on your spaceship. The cards are like rudimentary storylets in this case and many cooperative boardgames employ this mechanic. Anyway, this might help you envision storylets in a simpler light, but I’m also interested in what others have to say about them with IF specifically.
What I do know is that whatever Emily Short has to say about IF, carries weight.
→ https://emshort.blog/2019/11/29/storylets-you-want-them/
1 Like
Out of curiosity, what’s the motivation for u
Out of curiosity, what’s your motivation for generating AI prose instead of writing it yourself? Is this about speed/convenience, or do you feel your human writing skills don’t measure up?
This community as a whole has a strong bias against AI writing (as do I), but if you’re determined to use it then it’s good to know what specifically you’re trying to accomplish with it in order to give appropriate feedback.
4 Likes
A combination of the two, weighing more towards my lack of skill, desire not to “fail” (or fail faster), a dash of immediate gratification, and excitement to make my vision a reality. I feel like AI has enabled me to bring my ideas to life that I’m genuinely excited about.
2 Likes
…a dash of immediate gratification…
…an em dash of immediate gratification…
There. Fixed it for ya.
3 Likes
Some food for thought: we’ve had a couple people come through here insecure with their own writing and using AI to compensate. Every single time I’ve found their own writing is better or at least more interesting than what they’ve generated with AI.
My opinion is that chasing down AI tells is a Sisyphean task – OpenAI and Anthropic and all the rest are certainly doing the same thing, so odds are the tells will have changed by this time next year. But what AI companies haven’t been able to do yet is make an AI that generates writing better than “passable”, frankly. It makes sense once you think about it? LLMs are generating a statistically probable response to any given prompt, and better-than-average writing is by definition improbable. There’s a theoretical future where the identifiable verbal quirks have been ironed out of LLMs but they are still incapable of producing great writing.
Insecurity is a tough nut to crack, and frankly has been one of my personal biggest challenges in game development. But one of the ways I’ve dealt with it is by putting my work out there, getting criticism, and using that feedback to improve! Writing well is a skill you can learn like any other if you practice at it. Using an AI to generate text won’t give you this practice and it has a much lower skill ceiling than you do. I have a STEM background and was completely new to creative writing when I got into IF in 2020, and while I had a lot of help (my wife/coauthor is a very good writer and also an editor) I can look back at my earlier games and see how much better I am now than I was then.
That’s my two cents. I hope it’s helpful!
7 Likes
ZING! Thank you for the laugh! Well earned!
2 Likes
Thank you for sharing your thoughts and your own personal experience.
@HAL9000 was gracious to share David Sirlin’s Writing Well articles (I’m still working on my write-up of lessons learned. Needless words will DEFINTELY be my challenge. I’ve been lovingly accused of being like Billy from Family Circus: I eventually get to my destination, but I might wander, pet a dog, play on the swings…); might I ask if there were articles or techniques that helped you improve your writing?
2 Likes
I can’t point to any articles but I can say what worked best for me was:
-
Write a lot, and try to finish what you write before you start editing. “Finish” is relative – it can be a paragraph, a scene, or a larger section of your game, but the important thing is that you want to get a chunk of progress done before you start rereading what you wrote.
-
Give yourself space between writing and editing. If I try to edit what I’ve written immediately I will cringe myself to death and delete everything, but giving it a day or two allows me to look at it with a fairer eye. You can also get someone you trust to read it over and give you honest feedback. It sounds like your husband might be a good person for this? (My wife was/is this person for me and she’s not afraid to tell me when I’m doing something that isn’t working.) If not him, then a friend who reads a lot will also be able to tell you if something’s not coming off like you thought.
-
Read widely, in the genre you’re working in and just in general. Make note of what books/games/etc have writing you enjoy. Start thinking about why that might be and what they’re doing. You can do some research into the craft of writing here if you’d like, but that’s more effective if you can then look at your favorite books and see how the craft was applied.
-
Be silly about it sometimes. Have a goofy side project. Write short stories full of in-jokes for your friends. But most importantly, give yourself a space to experiment and practice away from the Big Project. You’re going to be harder on yourself trying to write something that you’ve been thinking about for ages! You’re much less likely to overthink and get in your own way on a smaller and lower stakes project and that makes it much easier to finish a chunk of writing in the first place (see point 1).
There’s a lot of fantastic writers on this forum so if anyone else has advice feel free to chime in!
5 Likes
- Down-draft: just get it down. Make it shitty. Don’t stop and think. Don’t edit. Just write, mistakes and all.
- Up-draft: fix it up and make it coherent.
- Dental draft: go over ever tooth and make it all shine.
3 Likes