Bonus Content - The Personality System
Added 2024-01-03 02:00:02 +0000 UTCHey again, everyone!
This month’s editing-oriented bonus content is going to talk a little bit about the system I’m using to track player choices with respect to romances. It’s something of a combination personality/romance system, because while it certainly has a lot to do with how the player chooses to behave… well, I’ll just let the below explanation do its job, shall I? Read on if you’ve an interest in trying to game-design a romance without “romance points.”
----------------------------------------
Rationale
As with the skill system, my affection system started in the classic manner of IF everywhere: you’d play through the game, indicate romantic interest in a character by flirting and doing things they probably found generally agreeable, and romance would blossom.
The thing is, this fell apart pretty quickly. First of all, I didn’t really end up tracking whether the NPCs agreed with the PC all that much or not, because the kinds of disagreements these characters might have just didn’t feel like the kinds of things that would prohibit a romance, and a general pattern of agreeing or disagreeing didn’t seem to carry much narrative weight. Also, a lot of the disagreements are matters of small differences in nuance, and it’s generally assumed that the player character is a goodhearted person who isn’t unreasonable, and that means that, well, frankly all the characters in the Underworld would just be chill with it even if it happened.
In some games, agreement or disagreement on the issues that come up can be pretty make or break, like in my other game, Diaspora. But it wasn’t the right mechanic for this game, which is overall a cozy slice of life about people getting to know each other and working together to solve problems.
Nor, really, did romance end up being something connected to whether the player character flirted (overtly or more subtly) with any of the characters. In fact, I began to realize overt flirting actually made some of them uncomfortable, and this presented me with quite the dilemma. Generally speaking in games like this, if a response has a heart next to it, it should result in the romance moving forward, at least a little. But here I was, thinking there were instances when it should have the opposite effect. I could just remove those opportunities to flirt, of course, and that might turn out to be part of the solution. But really, what I was grappling with what just one particular instance of a much larger problem in game design: how do you design a romance that feels real? One that feels like the NPCs have agency, and that they connect with the PC on a genuine level?
So I decided that any semblance of a single ‘relationship bar’ needed to go. It wasn’t working, and the two most common mechanics by which such a bar would increase (flirting and agreeing), just weren’t cutting it. Besides, there’s something that just feels… kind of shallow about this sort of system. As Michelle Clough points out in her book Passion and Play, which is all about designing romance in games, at a certain point, systems like this become reductive, turning NPCs into vending machines, where the right kind of token (certain actions, or something very generic, like “kindness”) are fed into the machine enough times and a romance pops out.
This has its advantages, certainly; it’s easy to implement, can be done either with small subplots or in games that are all about romance, like dating sims, and it’s always relatively clear to the player what will get the result they want.
But there are also problems with it. My main three issues with systems like this are 1) It makes the NPCs feel less like real people, 2) It reinforces a very uncomfortable trope where enough “niceness” or flirting or agreeing with a companion will get you a relationship, and 3) It limits players, by encouraging them to pick the options that seem like they will initiate the romance, even if that’s not what their character would do. In more extreme cases, people might “game the system;” this happens mostly in RPGs with approval systems, where you can selectively leave companions out of the action for a bit if you’re going to do things they disagree with, and so on. It all makes the process much less organic, and in this case, for this game especially, that was bad.
I ended up (functionally) just letting everyone choose their route at a certain point and then proceeding forward from there, no flirting necessary, but this feels strange in a different way, in that wildly-different PCs will be treated exactly the same by the NPCs. It’s less stress on the player to choose “right” answers, because there aren’t really wrong ones for the romance (except living on Olympus most of the time, which I think is fairly obvious). But it also just… makes the choices not matter in a different way. Instead of forcing people to choose kind-of cringey flirt options to advance a romance, I wound up with a game where, beyond immediate flavor text, it didn’t really matter what the player chose at all. Things weren’t remembered, there were no differences in anything based on the PC’s personality, and I think this made things feel pretty lifeless and lacking in chemistry.
So. I need a different way to do things, and I think I’ve come up with one.
A Responsive Romance System
My solution to this problem is to make the romances responsive to, but not dependent on, the player’s choices. This sounds extremely obvious, I know. Let me explain.
Generally speaking, in an IF game, what distinguishes one run, and therefore one PC, from another is twofold: first, the choices they make, and secondly, their personality. Sometimes skills or the equivalent also count, but in my particular case, traits are not just skills, but also parts of a PC’s personality. So in order to make one run feel different from another, you can either give the different choices different results, or you can use personality to flavor things. I think the best IFs in Asphodel’s genre do both.
Usually, though, differences in what choices do are more significant. They’re where you want to change whole scenes, and so the number of times you can offer choices that really do this is limited. Otherwise you’re writing a thousand branches and you never finish. Differences in flavor text can come from choices as well, of course, but a lot of the time those really come down to personality, even if it’s not tracked. You get different flavor text if you said that thing nicely, said it cruelly, or didn’t say it at all, but this doesn’t fundamentally affect the flow of the game.
And that’s fine! Flavor text is great. So great in fact that I would like there to be more of it in my game. But I would also like it to do work beyond just one choice’s difference. This is usually where personality systems come in. Unsurprisingly, these are bars that track how much of X quality (kind, stoic, flirty, or whatever) the player character is. And then the other characters react to this, or it’s used to write PC dialogue not directly chosen by the player, and so on.
My goal is to stretch this typical model just a little further. Here, I’m taking great inspiration from not only Michelle Clough’s ideas about “chemistry,” but also from the design of the game Scarlet Hollow, a horror visual novel currently in early access on Steam. Scarlet Hollow does this really interesting thing where your relationships with the other characters are dependent not one some objective meter for how kind you are, but on a series of vectors describing how each individual character perceives the PC. Are they clever or kind of dull? Reliable or flaky? And so on. One of the devs has written a really great article on this, which I’ll cite at the end of this; it’s really worth a read, and the game worth playing if you like what you read about (Scarlet Hollowalso uses a trait system, as it turns out, though unlike mine it lacks drawbacks).
Anyway, the combined idea here is that one way to make character interaction feel more organic and truly different on different runs is to tie it not only to the PC’s personality, but to really think about how the PC is perceived by the other characters. For example, if the PC is generally known to crack a joke about everything and is seldom serious, then a serious remark may be interpreted on that run as a joke, whereas on another run it would be taken seriously from the beginning. Likewise, the ROs will react differently to different combinations of personality traits; they may use gentler approaches with a PC perceived as lacking in confidence, for example, and double-check more often that things are okay, or worry that the PC wouldn’t say so even if they weren’t. This could, at least in theory, make the tone of certain scenes completely different, rather than essentially the same on all playthroughs (as for Asphodel they are now).
Scarlet Hollow has its own set of personality traits that most impact the story. Mine are different, and admittedly I’m still on the fence about a few of them, but here’s the basic list:
Warm --- Cold
Confident --- Unsure
Forthright --- Reserved
Humorous --- Serious
Selfless --- Self-Caring
Each of these, I think, has places in the story where the traits can be expressed, and each would significantly affect the way the other characters approach interactions with the PC. None of them is inherently bad or good, but they would shape interactions in ways that matter. I think its most likely that I just use one set of variables, rather than a separate one for each character, because the Underworld crew talk to each other a lot and are likely to have roughly the same information about the PC’s personality. The thing is, they’ll react differently, and that’s the part I’m excited about.
Of course, one way to do this is to have these function as true “opposed pairs.” In CS games, this is often done. Gaining points in “warm” effectively subtracts from your percentage of “cold,” which is not a truly independent variable, just a label for the part of the “warm” bar that expresses its negation.
But I think there’s some benefit to tracking all of these independently. Some people, for example, have a ‘hot and cold’ demeanor, and that’s very different from a ‘lukewarm’ one. In an opposed pair situation, both of those would be expressed by a “warm” stat at around 50%. There’s no way to tell the difference. But if you track “warm” and “cold” separately, a hot and cold demeanor is when both of the values are relatively high, and a lukewarm one when both are relatively low. There’s more room for nuance.
Of course, there are more personality traits in the world than just these. To a certain extent, you just have to narrow the focus in any given IF to the ones that seem most relevant to the story, but there are also other chances in the narrative to express different things about the character in a more one-off way that might come back. For example, in the present game, there are a few opportunities to express that the PC is afraid of heights. If you take one of the early ones, the later one automatically reflects this. I plan to include more of these, as they’re relevant, and there will be a separate, small-but-significant chunk of miscellaneous, but story-significant variables, like one for the PC’s general view of Olympus, another for how they feel about Demeter, and so on. These should allow me to write the relevant passages a little less neutrally, which was always a challenge I faced.
All in all, I’m looking forward to trying out this system. I think it will be considerable work to implement, but it’s actually helped along by the fact that I planned to do some work on the romance arcs anyway. Changing a bunch of things at once should allow me to really get stuck into this, because I’m expecting some fairly heavy rewrites anyway. Which brings me to the question of implementation.
Implementation
As with the trait system, the implementation of the personality system is largely going to be about going through what exists and asking myself questions about it that should help me figure out where the changes ought to go. It’s also going to need to be implemented not only where there are currently increases for personality values, but in the back half of the game. (I knew about halfway through that I didn’t like the personality system I had in the first draft, so I stopped including increases for it.) Outside of choice blocks, personality needs to flavor the text a lot more than it currently does, for that more immersive experience.
So here’s my list of questions:
1. If this is a choice block, can it be adjusted to include possible increases to a personality trait?
2. If it’s not a choice block, would it make sense to change the “default” text if the PC had a strong or weak score in any particular personality trait?
3. Is this a place where a certain combination of personality traits might be relevant?
4. Particularly in scenes whose focus is character interaction, such as romance-path scenes: does it make sense that the NPC(s) involved would say something different/act differently/the scene would vary based on the PC’s personality?
5. For other chosen attitudes, such as opinion of Olympus or Demeter, is that relevant here? How should I change the text accordingly?
Despite being questions, they’re not terribly unlike the list of principles for the trait system. In fact there’s going to be a lot of similarity in implementing them, which is one reason I might actually end up doing both at the same time.
Conclusion
So, this is the big idea! Of course, implementation here is going to be a bit trickier than with the Trait System. That mostly checks straightforwardly for whether you have a given trait (or given combination) or not. Personality is by nature quite a bit mushier. I am hoping that, rather than being a huge pain, the mushiness and large number of possible combinations will allow for some interesting nuance, and make every playthrough feel really unique. (At least, if you don’t use exactly the same PC every time!)
By the way, here’s that article from the Scarlet Hollow dev that got me started thinking about this type of system specifically: https://blacktabbygames.medium.com/creating-a-dynamic-relationship-system-in-scarlet-hollow-eb175aa899a8