Writing for video games was a surprisingly different beast (read: much harder) than the linear story-telling Abby and I were used to when we first started work on Scarlet Hollow. And, like most writing projects, the game saw more than its fair share of rewrites and major edits before we released our first episode.
We thought it’d be fun to share some of the decisions we made (and then unmade) in earlier drafts of our script. So without further ado, and listed in no particular order, here are six of the biggest changes we made that we think helped make our final product better than it would have been.
1. The introduction
Much like the final product, the earliest draft of our game started on a bus en route to your aunt’s funeral, but that’s where most of the similarities end. Instead of having a tense conversation with an oversharing and mildly sociopathic stranger who won’t leave you alone, the bus originally stopped at a gas station, where you had a chance to use the bathroom and talk to an ominous clerk who had a few choice things to say about your destination.
A harbinger, if you will.
Overall, the old version of the opening mechanically accomplished what we wanted to do with the introduction:
- It gave the player an opportunity to learn the basic gameplay loop without worrying about making mistakes with lasting consequences.
- The player got to organically learn some basic information about the setting and their character.
- They could get a bag of boiled peanuts.
Points one and two were particularly important, and were some of the biggest writing challenges we faced early on. Unlike a traditional narrative, we needed to find an immersive way to give our players the knowledge and tools necessary to exist in our world — it’s hard to roleplay if you don’t have context.
While our first draft succeeded at those things on paper, it was also cliche, extremely boring, and lacked any emotional hook. In a word, it felt like a tutorial, which is almost universally a bad thing. It’s a forgivable sin if your game is complicated enough, but then, of course, the start of any second playthrough winds up being a tremendous slog.
And in the case of our game, a visual novel where engaging with the world through dialogue options is the core and only gameplay loop? Unforgivable.
So we went back and mined our life experience for something more interesting and then we remembered this pretty horrific encounter we had at a bus station.
Not only did this rewrite make our introduction more fun to play, but the scene’s emotional extremes gave our players an opportunity to figure out who they were going to be for the rest of their run. Something neat and anecdotal we’ve noticed from watching a bunch of streamers is that many people pick reserved and nervous options in our opening scene, and those playstyles often carry over to the rest of the chapter.
2. Running errands
Because Scarlet Hollow’s concept got its start as a dating sim (more on that later) we originally wanted to give players an opportunity to meet each of the romance options early on in the first episode.
We did this by writing a sidequest before going into the woods with Stella to gather a few key supplies for the trip — supplies that conveniently had the player stop by the home and/or workplace of each major character in the story.
Much like the first draft of our intro, we had a very hard time making this side quest interesting. Nothing exceptionally bad had happened to the player yet so there was really wasn’t much to talk to those other characters about; it wound up just being a series of conversations that boiled down to “Hi, I’m [player_name]. Yep, that’s right my aunt is dead, can I please have some pretzels for our hike?”
After a couple of tries, we decided we were better off cutting that scene — after all, wouldn’t Stella already have everything she needed for her pre-planned hike? Instead of meeting every romance option on day one, we decided we’d be better off dropping hints about those characters and only introducing them to the story when the narrative called for it.
So you only meet Kaneeka after you have something traumatizing and interesting to talk to her about. You’ll meet Oscar in Episode 2 now that there’s a compelling reason to spend time in the library. You’ll meet Reese…. eventually.
3. Adding a dynamic relationship system
One of the things that makes interactive fiction compelling is players’ ability to influence a narrative and directly see the impact of their decisions on the world around them.
As we mentioned in our previous post, we originally started with a +heart -heart relationship system that didn’t leave a lot of room for nuance. This meant that we’d have to use an obscene amount of flags to make our players feel heard.
4. Re-designing Gretchen*
Gretchen was originally the most hideous, wretched old yappy dog Abby could have possibly drawn, to the point that one of our roommates, for whom loving dogs is personality-defining, told us that he wouldn’t hesitate to save Duke in the woods, since it seemed a mercy to spare Gretchen from another moment of existence.
* Granted, some folks LOVE how disgusting Gretchen was in her original design (including Abby), but toning things down and helped players more easily attach themselves to her.
5. Removing “worst possible worldstate” ending for Episode 1
Spoilers for Episode 1 abound below. Careful if you haven’t played yet!
In the third act, Episode 1 splits into three major branches:
- You dive for the flashlight in the woods, saving Duke’s life, but allowing Gretchen to run off.
- You dive for Gretchen, leaving Duke to a grisly fate, but saving Stella’s beloved pooch.
- If you took the Powerful Build trait, you can dive for both Gretchen and the flashlight, and make it out of the woods with everyone alive.
Our first draft had a fourth branch, which we liked to call the “worst possible worldstate.” Originally, players without Powerful Build also had the option to dive for both Gretchen and the flashlight, but if they went for that choice, they would fail horribly — not only would Duke still die, but Gretchen would run off as well. We cut that path for a few reasons:
- It felt like we were lying to our players by presenting that option. One of Scarlet Hollow’s core design philosophies is to always convey clear information to our players to the best of our abilities. We might not tell players beat for beat what will happen when they select an option, but the context around each choice should provide a vague but immediate sense of what will happen and who it’ll effect. Adding an option where you try and fail to do something so important would have undermined our players’ trust in our system.
- Another of our core design philosophies is that there are no strictly “bad” options. A lot of our major decisions are choices between two bad outcomes, but ideally, there’s never a strictly “better” choice. There’s an exception to that with the way traits can provide outs to major decisions, but those are also balanced such that, over the course of the game, no trait is strictly better than any other. A fourth branch where both Gretchen and Duke die would be strictly and obviously worse than other branches, and again, would have served to undermine the trust we’d built with our players.
- People would just reload their save. As an extension to the above point, in an ideal world we never want our players to go back and reload a save after making a decision. They have the option to, of course, and despite our best efforts, some players are going to try and min-max an optimal playthrough. That being said loading an earlier save is immersion breaking, and also destroys the tension we work so hard to create — there are no stakes if you play as though no decisions are permanent. We’ve noticed a lot of players and streamers not even think about saving before major decisions, which we’ve taken as a sign that the game is on the right track.
- A lot of extra effort for very little payoff. Because this option would be strictly worse than others and most players would just reload after picking it, it’s a path that very few folks would wind up playing that would require an exceptional amount of extra writing, coding, and art. None of that nonsense.
Minor but interesting changes
We made some other pretty significant changes over the course of Episode 1’s development, but it’s hard for us to tell if these changes necessarily made the game better. Still fun to share, though!
- Tabitha used to be named Beatrice, until we remember that that was also the name of the grumpy alligator from Night in the Woods. Whoops!
- Ditchlings used to be called Wreckers.
- Avery underwent a major redesign. They were originally much more shy and much less confident, and we were having a hard time doing the character justice. Avery’s in-game iteration is much closer to my personal expression of gender identity, and therefore much easier to write.
That’s all for now! If you like this content and want to help support our two-person indie studio, head on over to our Patreon where you get can early access to new posts and suggest topics for us to cover in future updates.
And if you’d like to see us work on Episode 2 live, we do an art stream every Saturday on our Twitch channel.