Tagged: senp.AI

What Do I Think Of Visual Novels?

I’m a big fan of the Visual Novel. I don’t just mean that I have a fondness for the form borne out of a period of my life where they were a way to both get anime and smut at the same time when those were two things I very much needed in my life to feel connected to the world around me, I’m actually a fan of the structure.

There’s a lot to talk about here so let’s just dive in.

Accessible Moviemaking

I know a lot of people who want to make movies. For some of them, the cinematography of a videogame camera gave them that option, and you saw early demo-editing Quake levels and replays being made to play with those same ideas of echoing subtitled cinema.

I see the Visual Novel as a less kinetic, but more framing-based example of this same basic idea. Good cinematography, an appreciation of good cinematography, makes visual novels a good avenue to construct scenes as if one is thinking in terms of movies, in terms of what keeps people compelled.

So first of all, they’re a way you can do a movie-sized story on a very, very small budget.

The Point-And-Clicker

When it comes to the game elements of the Visual novel, I feel that as a game type, it inherits well the basic structures of the point-and-click adventure, those low-impact, low-action kind of games that wanted to give you the time to quietly and patiently sort your way through problems that are presented to you. Things like Monkey Island and Beneath A Steel Sky, where the point was not some focus on exciting action setpieces but rather a much more slow, wandering kind of puzzle hunting.

Now, there are four basic puzzles in point-and-clickers you can deal with, and there are two that Visual novels can do just fine, and two they do Not So Fine.

Good: Use Key On Door

Something impedes you from a path ahead. For good conveyance, you want that path to be in some way visible to the player. You need an object that you can carry to the impediment, and then that will get it to go away. Ie, it’s a key and door situation. There’s a door, you unlock it, you can continue on your way. This type of design is incredibly common. Sometimes you’re obscuring the keys, sometimes you’re obscuring the doors, sometimes behind the door there’s just another key, sometimes you’re just collecting completely obtuse keys – but that’s the basic thing. Take an object to an impediment.

Bad: Put Apple In Box

So there’s this thing that’s really tricky to do in Twine, and Visual novels as well, where you have containers. Containers are something some games can handle just fine – I’m told Inform can easily handle it when you put an apple in a box, then pick up and move the box around and then set it down wherever you like – the apple will still be in the box. In visual novel coding… this is trickier.

What this tends to mean is that visual novel games often feature a bit less of characters interacting with the world as a place with a lot of material objects. This is also reflected in the genre – note that most games are not about carrying around tools, as much as they are about inner experiences.

Good: The Language Maze

This one’s a little more common for older games, back before we were voice acting everything. A language maze is – very simply – a series of conversation choices where you need to choose a particular sequence in order to find a point in the conversation that an opponent does something that the conversation would not normally do. Some mazes are really simple – you just ask a character a thing, and you’re given an object. Sometimes, you ask a character a thing, and that gives you some knowledge you’re now able to use later. Sometimes you ask a character a thing, and that gives your character some knowledge you’re now able to use later – like teaching your character to pick locks or something.

Bad: Freedom Of Movement

And now here’s where the point-and-clickers of the past are a little different. The typical form of the Visual Novel is a linear flow of time from A to B. You’re very much moving along a line of time, rather than necessarily having the means to travel between locations.  This isn’t to say that’s how things have to be, but it’s not uncommon for people writing visual novels to present them as a single long line of time with you moving along it.

This isn’t to say that visual novels are bad at this – you can definitely set them up to do it. But the default code structure of something like RenPy reflect the genre, where it is very, very easy for the game to just see each play as a series of sequences that check variables, rather than necessarily going to places and letting you move around them more freely.

Scaling Up

Then there’s the things Visual novels can do well that you can lean into and build on in your own projects.


Hey, you know how you have all that writing about your game world, or your characters and you want to give people a place to go look for it and read it if they want to? But if you dumped that in the main space for people to read it’d slow everything down and be super boring?

Well, the Visual Novel is a game form where reading a ton of stuff is a thing. In Hate Plus, there’s a reference codex for any character. You find yourself confused by a name? Click on it and it’ll take you to a place where you can look at that character and look at what they’ve done and what you know about them so far.

The Inner Life

It’s almost a stereotype that visual novels have a first-person narrator, often a narrator who is ‘you.’ Doing things this way gives you an almost unprecedented level of access to a character’s inside thoughts, meaning you can see how they think before they act, how their inner dialogue contradicts their behaviour, their anxiety, their stress.

It can also be a fine opportunity to learn that a character is a total butthead, which is a problem that Roommates has.

Day To Day Life

You know that thing about how going to a place is something that VNs don’t tend to do? Weirdly, they do handle schedules well. Because that’s how you use your time, and you can even trigger or chain events based on what you’re doing with your time, today. These can get super complicated, too!

The Wrapup

I really like Visual Novels. If I was better at designing interfaces or had the knowledge of where to start designing interfaces, I’d probably have made some of them by now (Sorry, senp.AI). They let artists do small numbers of works they like, they allow for clean use of arts and assets, and they don’t require a lot of technical knowledge to get started on. They do need you to be somewhat clued in on structure and planning, which is pretty frustrating stuff if you’re not familiar with it – but you can find your plan in the making, too, and restart.

Look into Visual Novels, they’re a great little genre, and lots of fun to think about making.

senp.AI 0.3

Thanks for your patience, everyone. Last time I talked about how the game was meant to allow you, the player, to select senp.AI’s gender, your gender, and trans/cis/other status for them, and how this changed basically nothing for the narrative. Then my exams and final assessments for University happened for the year, then Nanowrimo, and now here I am going back to this.

The biggest problem I’m having right now with senp.AI is a lack of good documentation for what I’m trying to do. See, the game that inspired me to start on all this was Analogue: A Hate Story (Steam was down when I retrieved that URL – if it’s wrong I’ll look into it later). Analogue is both a wonderful game with an interesting, nonlinear pace and all sorts of other goodies, but it’s also a game with a pretty interface. I’d grown very used to my childhood Visual Novel work being about fairly ugly, serviceable interfaces – and Analogue and Hate Plus have very pretty, elegant interfaces.

Now, Renpy promises it can help me make a visual novel, but it doesn’t promise ways in which I can make that visual novel look like something other than Awfulness. What I’ve been doing lately is working on creating dialogue and character for the game (another problem: It’s going to be a game with Lots of Words in it). The Renpy forums suggest, first things first, working on creating an interface mock-up in a graphics program. Here’s what I have:

senp.ai mockup example
What’s planned is for the first part of the game, the top left window lets you select the thread of commentary you want to read, and the window on the right lets you scroll through it and see what is and isn’t there. That’s the very, very basic ideal at this point. The bottom left ‘wheel’ is how you access different parts of the interface.

At this point, trying to find good documentation on how to create this – rather than just crying and begging Fox, who has some experience with Python, to do it for me – is very hard.

senp.AI 0.2


In typical otome games, the player character is presented with a variety of options in partners and chooses the one they like. Some games try for a more life-simulation approach, where you choose activities and that makes you more appealing to people who have common interests with those activities. There is some degree of control over the character one plays, but there’s also crucially not. These games don’t have room for the truly massive amount of writing necessary to give you an exhaustive list of options for conversation and conversational style. In the end, while some of these stats can work together (imagine an otome game where a higher ‘intelligent’ stat leads to a character that uses more precise vocabulary), there is still a basic ‘core’ of the protaganist, and they have a very basic ‘line’ of action. Most of the time in the games I’ve played, your choices amount to simple binaries – yes/no, do I/don’t I, – or selections from a group – which boy of these five?

I’m okay with a game putting you in the shoes of an existing character. I spoke about Dishonored using this device, and how while you may play Corvo being murderous or Corvo being merciful, Corvo is still a man who solves his problems with violence, one way or another. Such is true in the Otome games, and I feel that that is necessary as a writer. Some ambitious developers invest in the idea of making you, the player, free to do as you like, but typically just provide you with a fairly broad handful of options. I understand there’s a desire for that, and maybe one day games will manage it. For now, though, I feel it best to admit that the player and the game are both bringing things to the table. The character you play in senp.AI has some pre-determined things about them. I need these things to be able to write. The AI will say things, implying that that is how the player is thinking about them – and there will probably be some choices to determine your particular thought patterns.

Where this gets challenging for me is gender.

Recently, in a series of tweets I saw retweeted by a very intelligent someone I know only as mcc, I was caused to think about how videogames that offer character options touch on gender and sexuality. In senp.AI the AI has no base gender. The AI character you play is just an AI. You can call yourself him or her or they or whatever you like, but mostly people don’t refer to you as a person. The game is set in the fantastic space year of 2899, where humanity live in glass towers and use tidal power and solar satellites and have advanced chemistry. What I had planned regarding gender was pretty simple: The player character can eventually choose a gender, and the gender of senpai.

and Senpai’s cis/trans status.

Senpai can be male, female, or asexual. Senpai can be cis or trans. The former choice changes some pronouns. The latter stat changes nothing. There is no reason for that choice to exist (and I’m happy with how the game handles it) because it does not impact senpai’s school life, nobody talks about it or comments on it, and trans status – in this miracle world where transitioning is a matter of a short drug treatment and that’s it – is only noted on medical records for when doctors are checking for specific things.

I like this idea. I am also worried by this idea.

I’m not trans. I don’t know, in my day-to-day life, any people who have informed me about their trans status. As far as I know I’m just a big dumb cis boy in a big dumb cis world. Much like racial issues, where the vast majority of my friendscape is white, I feel hampered in my ability to realistically represent a complex web of interconnected issues simply because they are removed from my experience. I do not want to disrespect the state of trans people with my videogame idea. I honestly want to paint a future where the problems of our modern world’s trans people are non-existent because as a culture, we’ve moved past silly things like caring about the state of a stranger’s junk.

Another person I like a lot but only really know as a handle on a tumblr, Mammon Machine, wrote “Would you rather your story be written by the clueless or erased by them?” and I don’t want to do either with this. I am the clueless and I know it. Becoming an expert in this topic would require a lifetime I don’t have, and I don’t want to treat the trans people I do know as some sort of censorship barrier, justifying my choices to myself.

This whole idea is very hard for me to grapple with. In this setting, being trans is so irrelevant it’s a footnote on your medical history, but it’s still there. Is that a bad thing?

Note that whatever gender combination you choose with senpai, you can romance. Senpai is not gay or straight or bi or pan. Senpai is just senpai, and, if you play your cards right, senpai falls in love with you.

senp.AI 0.1


Let’s talk a little about this project as I work on it. That works for getting me to do things like short stories, and to push myself onwards. I think the genre of game that senp.AI can be considered is a otome game, one where the player assumes the role of a protaganist in a romantic narrative focused on a single, desireable member of the preferred gender. There are lots of games like these; some of my readers may be more familiar with eroge variants, where you usually play a clueless, feckless, ambiguously appealing generic grey lump somehow seducing his way through a cloud of gorgeous women, or choosing a particular lady to really get it on with. While I am all for the knocking of boots and stuff, I’m pretty sure I don’t want to do that. If nothing else, there’s a lot of hard assets to create, and writing erotica has reminded me that there is absolutely nobody as thankless as an audience you actually know.

Rather, senp.AI is meant to be a cutesey love story (no, actually) about playing a (girl||boy||other) at a fancy futuristic high school in Harbour City’s District 13,, who develops a crush on a (girl||boy||neither) in your class. It is your class: You’re the AI designed by the school to keep an eye on the students’ online behaviour, to check for problems and ensure that nobody is showing signs of social problems. You’re the cute AI sidekick of the class’ Guidance Counsellor as they go through the 2898 school year, which will feature studies like 21st Century History, Pre-Cataclysm Literature, and Ethics.

The challenges I’m facing as a writer/creator for this project – ignoring that the engine I’m using is designed for idiots and I’m still struggling with it – come in this general form:

  • Can I write a character that a reader realistically will fall for?
  • Is it wrong to assume a player will meet the game half-way?
  • What kind of homework can I assign the player?
  • How the hell did Christine Love get the interface for Hate Plus to behave?

senp.AI, Motivation, and Action

This video, which I stumbled upon while wasting time on the interetubes was what has finally kicked me into action. I want to be part of the conversation creating videogames. I want to at the very least use the tools that I have bought (or, more realistically, that interested and wonderful friends have bought for me) that entice and excite me. When the time came to write things, to get practice writing things, it was just doing it that made it happen. After some fiddling with different story ideas, I have settled on a project I can – somewhat – work on on my own, using an engine I have, and so, I am going to try and produce it. I don’t have art assets or music assets worth a damn, but we’re going to cross that bridge when we get to it.

The 2013 Short Story project has morphed into something else. It’s definitely a serialised story at this point, and only a small number of its components are legitimate short stories in their own right. They are set in some greater place and have a single over-arcing narrative (with some big flaws I wish weren’t there, but as needs must). I’m going to rename the category to reflect that more accurately and see how it works out for the rest of the year.

This has been my realisation this past few weeks: Do More Stuff. There’s plenty of time in the day to do lots of stuff, and I was just letting days melt away under things that didn’t really matter or help or create. I want to Do More Stuff.