When Stars Collide - Q&A Part 3 (Patreon)

The final part of the Q&A talking about various challenges and takeaways from working on this game...SO FAR.

Challenges and Takeaways

How much harder did you find it to offer 4 body types?

Honestly it depends on what you're referring to.

Writing wise - not that much harder. There are small variations in a few places, depending what sprite body you have, what nouns you're using, etc.

But that's also true for the hairstyle and even whether or not you use the MC's default name. When you have customisation of any kind, these sorts of variations may come up.

Often the variation is just a sentence or two that's different - sometimes just a couple of words. It doesn't come up that frequently either. Maybe 1-2 times per route.

Art wise - I don't know yet. I haven't done the CGs, and that is where the art difficulties will be - particularly with managing the two face shapes and the chest.

The MC is the same size through the neck and shoulders. There are minor differences through the hips and thighs too, but you can't see those most of the time in the game, much less (probably??) the CGs.

It's really the face and chest where I'll have to draw the differences, but honestly I think the hair is going to be more work. But we'll find out.

Decisions, decisions - The biggest issues surrounding the sprite bodies were always around decision-making in terms of how to handle biological morphologies, essentially. There were a lot of options, and managing expectations and scope and what I *want* to do has been tricky.

Will you do it again?
Whether I do all 4 bodies again just kind of depends where things stand after this game is done. A Faerie's Tale had two bodies and yeah, it was a lot easier but it's a difficult comparison since that game also didn't have CGs.

So this is still my first time doing a full game with CGs that have different body shapes.

I think I'll continue to do multiple bodies for sure...

I added it to WSC because, at the time, it felt that it better reflected the player base I had and that hasn't changed. It's just a matter of whether it will be 2 or 4.

And that'll come down to workload regarding the actual CGs - which I haven't really gotten to yet.

What challenges did you find with pacing [during the writing]?

One of the big challenges with writing a game like this - for me - is that it's impossible to grasp the pacing and flow while drafting. Because you're writing so many variations of certain moments or writing multiple conversations that, while different, do cover the same topics...you can feel like a topic is being mentioned too much or not enough or the pacing is too slow or too fast.

If you write six variations of an LI worrying for the MC about something...you might have this idea in your head that people are worrying about the MC an abnormal amount. But it's just because you've written it six times. The player is only going to encounter one of those instance per playthrough, though, so it won't come across the same way to them.

This element of the overlapping routes can really skew your idea of things that are happening fast or slow or a lot or not enough.

So I would add content in, thinking that things were happening too fast and needed to slow down - only to realise later that it wasn't happening too fast and it was my perception that was off. Or I'd take things out, thinking that I was overdoing it and it was too much...again only to later realise that it was my perception that was off again.

The result of writing in this structure - at least for me - is that the image of the game in my head as I write is not at all similar to the experience of playing it.

This is true, to some degree, of a game like Gilded Shadows as well - all games *play* very differently than they read when you have the entire script in front of you.

But it's very different when this issue is being applied to each route individually and as a self-contained unit than when you are applying the issue across all the routes at the same time.

Even if you don't make a game of this type again, is there anything you learned from this "big common route" style game that you will take into future games?

There are a few things.

I learned a lot about using the skip function as an actual game mechanic as opposed to just a thing you can do - a function that exists.

Players use "skip" in all games. In GS, in Changeling, everything. But in the linked route structure like WSC, skip is much more integral to the gameplay, and it is expected it will be used more.

This fact led me to really think about what the experience of the player would be while on a second or third playthrough where they are likely going to be skipping through content a lot and stopping for "unseen" text when it is encountered.

What is that "unseen" text in this scenario - does it justify stopping the player from skipping? The fact that it stops skipping signifies "This is something new!" so when they player reads that content, will they think it was worth it? Those nuggets of new content are what the player is replaying for.

So they have to be worth the process needed to acquire them.

This was a question I asked myself a lot.

I can't always avoiding stopping the player for a really minor change, but thinking about what the experience of consuming original, duplicate, and similar content will be like as it relates to how the player will be trying to skip through content they are already quite familiar with was a bit different than usual.

I had to do this in Gilded Shadows as well but not nearly to the same degree. I learned a lot about how to analyse that part of the player experience.

It's something I'll carry with me to future projects too.

Another thing I learned a lot about for this game was really just...choices. In general.

What they really do in games, how we use them, how players perceive them, how we communicate about them, what people think choices should do vs what they actually do, etc, etc, etc.

That is partially because choices in a game like this are used differently than choices in a game like GS.

I learned a lot about game structure in general and things like feedback loops and the like. You can think you understand a concept quite well but the moment you try something new, you have to ask yourself a ton of questions you have never asked yourself.

And you realise your understanding was quite surface level.

When working on Changeling, I only thought of choices in terms of "mattering." And "mattering" meant changing the plot or the story.

In GS I started to gain a deeper appreciation for choices "mattering" in ways other than changing the story - and gained a more realistic view of what things like "change the story" or "change the plot" even really mean in a visual novel.

But then I get into a game like WSC where all the routes, of necessity, require the plot and story to remain the same because each route follows the same storyline due to the how they overlap and link up. The choices can't change the story that much.

The branching would be monstrous and basically unmanageable.

For instance, let's say that at one point I give the player the option to save the ship using their "space wizard" powers. They can do it themselves or let the crew work out a way to save the ship as a group.

In a GS route, this might lead to two or three possible "world states" for that route. That is fairly easy to manage.

But in Stars, if there two outcomes, that translates to...12 possible variations. Because I have to look at each LI's reaction to MC saving the ship vs MC not doing it themselves. There are six LIs. How do I make all six LIs have a unique and interesting reaction to MC saving the ship? How do I make all six LIs have a unique and interesting reaction to MC not saving the ship?

There just isn't a way to manage a lot of those kinds of story altering choices across six overlapping routes while making them interesting and distinct and maintaining continuity with each other.

So...then what do the choices in this game do? Do they not matter? Are they all "flavour" choices?

They matter.

But they matter in a different way than what I'm used to. This led to a lot of contemplating about all the functions choices perform under the fold (in the code), the outward behaviours of those choices, and how they are used in different ways for different aspects of storytelling.

What becomes important is not making sure the choices "matter" but actually making sure the player's expectations align to what the game is actually going to deliver to them.

What kind of challenges did you discover with this style of game, as opposed to your previous games?

I want to do a full breakdown of this at some point - a sort of post mortem of sorts maybe - but for now I will say this:

In my experience, people perceive this "linked route" style to be the "easier" game structure when compared to a game with self-contained routes like Gilded Shadows. That is because of a lot of reasons, but I think the main thing is that the assumption is...because there's a single plot, it's much simpler to write the story.

And I would argue that is absolutely not the case for a game this size.

My experience is that this game structure is inherently more complex than separate routes.

I know part of the perception comes from the fact that LI scenarios in a game like this are similar to each other by necessity - if all the LIs are following the same main plot beats, you can't have their scenes deviate too much from each other and still fit within that main plot.

And even if you can do that, there are logistical issues that may stem from having characters deviate too much.

If you have Love Interest #1 go on a cute date with the MC in one chapter, but Love Interest #2 stays home with the MC...

Some people are likely to respond well to seeing characters do different things while others may feel like Love Interest #2 missed out on getting a cute date scenario.

So sometimes scenes are similar because you're thinking about things like equity and balance and making sure characters feel like they are being all treated with the same care.

Many games structured like this default into copying/pasting content across routes so there are duplicated "name-swap" scenes.

But I would argue that duplicated content is a symptom of the complexity and difficulty of this structure and not an inherent quality of it.

It is extremely difficult (and tedious) to write six unique interpretations of a single moment in time, where every variation must be :

- Interesting.
- Fun to write.
- In alignment with the specific character's personality.
- Tracking the same major points as the other variations for plot continuity.
- Similar enough that future events will make sense no matter whose version you got.
 - Without having to rely on 100 variables
- Different enough for players to feel it added something unique to that character's route experience.
- But equitable across the routes so it doesn't feel like someone is getting preferential treatment or other characters are being left out.


Just tracking the continuity is a nightmare. By cross-referencing the variations between characters for this purpose, you tend to make them more and more similar because if they are hitting the necessary plot points at the same time, it's so much easier to ensure that each scene is doing and saying everything it needs to do and say.

The problem with that is that all your scenes now align for plot continuity, but...they also all read very similarly.

This gives the impression it was very easy to draft because you were just copying the content, right? But it's actually pretty difficult. And pretty tedious.

One of the struggles I had with this game was how *long* it's taken compared to how long I thought it would take. A lot of that is just because I have to constantly baby-sit continuity and similarity and dozens upon dozens of variables.

At some point, you are exhausted and burnt out and may start copying/pasting because that's how you keep your head above water.

By never referring back to choices or having all the variations be nearly identical, you reduce your workload. And this isn't reducing a small workload to even smaller out of "laziness".

This is reducing a massive workload to "manageable" for your sanity.

Another aspect of this game being intended for episodic release is considering how it's going to be consumed and what that experience is like for players.

Content really has to track much more similarly in length and pacing. If one character gets a 4000 word scene in chapter 3 while everyone else gets a 2000 word scene - that's going to be very noticeable to players. And will potentially make some players feel their favourite isn't being given due attention.

If one character gets a passionate kiss in chapter 5 while another has to wait until chapter 10...again, this is going to make people feel their favourite is being neglected.

Being able to write each route independently in a self-contained space that doesn't affect other routes is such a huge and amazing freedom. It's easier to mentally keep track of things, it allows each romance to flow at its own pace.

In GS, people certainly notice that Yuu kisses Morgan much earlier in his route than Ari but because the routes are experienced independently rather than concurrently, people are able to appreciate the variation.

Where in a game like WSC people often respond better to balance.

For me, the linked route style felt stifling and limited considering how I normally write. The simplicity of separate routes adds so much more flexibility to the storytelling and what I can do with branching and choices.

At some point, though, I really do want to gather all my thoughts about game structure and talk about the different types, the differences and similarities, the strengths and weaknesses, etc, etc.

So yeah. That's all on that for now.

Admit it. You are already working on the next project, aren't you?

...

Maaaaaaybe. 😌