Designing Scarlet Hollow’s trait-based roleplaying system

Players get to pick two traits from a selection of seven.

Hey folks, and welcome to the first in a series of blog posts we’ll be writing about the development and release process of Scarlet Hollow! Each of these posts will be available a week early to $5 supporters on our Patreon, and $10 supporters will get to suggest and vote on the topics of future posts (along with some other goodies!)

With my expert shilling out of way, onto the topic of this post, Scarlet Hollow’s trait system (there are graphs!!)

For those of you who stumbled upon this post without already knowing about our game, Scarlet Hollow is a choice-driven indie horror game set in a dying coal town in the mountains of North Carolina. It’s being released episodically, and you can play the first episode for free on Steam and itch.

When we released Episode 1, we included six different traits from which the player could choose two at character creation:

• Powerful Build
• Mystical
• Street Smart
• Keen Eye
• Book Smart
• Hot

These traits give you extra dialogue options, can give extra information that isn’t otherwise available, and might be able to help you out of a difficult decision at the climax of an episode…

We added traits to give an extra layer of immersion and to help each run feel unique (which, of course, makes the game more replayable!)

When writing traits, we have a couple ground rules:

1. Traits need to be balanced. We don’t want anyone to feel like any one trait is the “correct” choice to get “the True/Good Ending” or anything like that (all our endings are Valid and no trait is going to get you a clearly better ending than any other.)
   We had a similar approach to traits as we did to our relationship system — the relationship system is complex and its flags are hidden from players to keep them from asking “what decision is best”, and instead getting them to ask “what would I say or do here?”
   So each trait will have an equal number of major decisions it can save you from (at this point, probably one per trait,) and we try to give every trait its own set of secrets to unlock, so no one winds up thinking they need to pick any one to get the best outcome. A himbo should have just as interesting/satisfying a run as a bookworm, and vice versa.
2. Using a trait has to feel good. We initially toyed with the idea that trait dialogue options could sometimes have terrible consequences, but decided against it after some early play-testing. A player’s trait selection defines their character, so when a trait comes up as a dialogue option, they should always be tempted to take it. So when a Powerful Build character tries to intimidate someone, it works. When Street Smart tries to con their way out of a situation or open a door, it works.

Even if that door just opens to reveal a brick wall…

There were a few ways that we failed on this second point in the initial release of Episode 1 — there’s a segment where you explore the house where Street Smart couldn’t open a door to the forbidden wings (they were instead given extra information about the lock), and Book Smart often made players feel like a jerk and know-it-all.

We patched both of these traits in December, so Street smart can now open the door (which leads to a brick wall…) and Book Smart was rewritten entirely to be less Reddit Moderator and more Reads A Lot and Knows Things.

Initial Trait Distribution Across Players

A few days after the release of Episode 1, we released a survey on the Black Tabby Games twitter asking players what traits they chose on their first run (sample size was 116).

We’d noticed from our earliest playtesting that Keen Eye was almost universally chosen, and this carried over into the release, with over 60% of our players selecting Keen Eye. (Less dominant than it was in playtesting but still extremely dominant.)

We’ve had some theories about that: maybe the trait most directly matched to solving mysteries in a mystery game seems like the “right” choice, and players felt like they’d be missing out on critical details by not taking it. Or it’s just the one that feels most “neutral”, so it seems like a natural second choice after picking something more character-defining like “Book Smart” or “Mystical” — an “unflavored” trait that wouldn’t muddle the character they wanted to play.

Other key takeaways from this first survey:

• Book Smart, despite being the trait we struggled most with writing, was a lot more popular than anticipated.
• People really weren’t drawn to Powerful Build. This was something that came up in playtesting (maybe one or two of our 30-some testers took it), and probably meant a lot of folks didn’t figure out there was a third option at the Big Climax… a special surprise for them to learn as more episodes are released ;D

We think in part because of the trait system, folks wound up playing through episode one a lot. Average playtime for a first run of Episode 1 was around ~45 minutes (though some were much longer.) Average time played has inched up to 67 minutes, meaning the average player has completed slightly under 1.5 runs. Thanks for playin’ so much!

Talk to Animals

As most of you probably know, one of the stretch goals of our Kickstarter was to add a seventh trait to Scarlet Hollow — Talk to Animals. While we didn’t hit that goal by the end of our campaign, between pre-orders, add-ons, and winning the two of the prizes from the AT&T Unlocked Games competition, we managed to raise enough money by the end of 2020 to justify adding it, and we released the Talk To Animals patch on Christmas Eve!

Naturally, we wondered how Talk To Animals would impact the popularity of other traits, especially Keen Eye. Would Keen Eye and Talk to Animals just be the most popular pair of traits? Would Talk to Animals replace Keen Eye as the go-to trait?

We ran a second poll (76 respondents) after players had a chance to replay Episode 1 with Talk to Animals asking what trait combo they intended to choose for their canon run. We also asked them who they were planning to romance, so we were able to do some fun crosstabs too.

One of our immediate takeaways was that, holy crap, aside from Talk to Animals and Keen Eye pulling away from the pack, the final distribution of this survey was shockingly balanced.

Looking at individual pairs of responses, we have quite a few takeaways:

• Every single pair of traits was represented in our poll, even if three of them only had one response apiece. (Mystical and Street Smart; Powerful Build and Book Smart; Talk to Animals and Book Smart)
• Talk to Animals and Keen Eye was by far the most popular pairing of traits — over three and a half times more popular than mean popularity.
• Every single trait pairing that was more popular than the mean had either Keen Eye or Talk to Animals.

Romance Options

Pictured from left to right: Avery, Stella, Oscar, Kaneeka, Reese

Before we go into crosstabs, here’s the results for preferred love interest.

RIP Oscar and Avery — both less popular than “Nobody.” We expect these results to change quite a bit the more people play the game and get a chance to actually interact with non-Stella characters. It’s absolutely wild than 20.3% of our respondents went for Reese, though, given that nobody’s even met him yet. I guess a lot of our players have a type, and that type is Adam Driver in PJs.

Before we get into our tables here, it’s worth noting that looking at crosstabs was tricky with our sample size, especially for Avery and Oscar. So is the data in this section remotely scientific? No. But is it fun to look at? Extremely. Because of this small sample size, I didn’t bother to spend much time with how trait combos relate to romance choice, and instead just looked at what individual traits were selected.

Avery

• Avery was the least popular romance option on this survey. (We’ll make our respondents eat those words (votes?) in Episode 2, if all goes well.)
• Every single person who wants to romance Avery picked Talk to Animals.
• Not a whole lot to say outside of that, given that there were only 4 Avery-philes.

Kaneeka

• Almost half the people who want to date Kaneeka picked Hot as one of their canonical traits. 🔥
• We’ll see more of this later on as we look at other characters, but Kaneeka’s respondents definitely fall into an “opposites attract” bucket. She’s an extremely rational/scientific person, but only one of her 12 respondents picked Book Smart as a canonical trait, and despite how hard she pushes back on the idea of otherworldly beings, we saw a slightly above average # of mystical responses.

Oscar

• We haven’t met Oscar the librarian yet, but the players who picked him are into being smart one way or another. And they do not like talking to animals.

Reese

• Reese daters fell fairly close to the average, but it’s probably usually worth a note when romance option received above average Talk to Animals votes, given how popular that trait is.

Stella

• Just like we saw with Kaneeka, Stellamancers were consistently in an opposites-attract bucket. They’re extremely below average for Mystical and significantly above average for Book Smart, and slightly below average for Street Smart.
• We also saw (slightly) above average responses for Keen Eye. Maybe folks dating Stella think being observant will help with her YouTube channel.
• Stella also saw below average Talk to Animals representation. Did Gretchen make some players feel like a 3rd wheel?
• Her slightly above average Powerful Build votes are also probably worth noting. Players that wanted to save her from extra trauma in the woods? It’s probably worth exploring whether players pick a romance option and then pick canon traits to fit that option, if it’s the other way around, or if the two aren’t consciously related.

Single for Life, Baby

• Not a whole lot of note here — above average Book Smart and Mystical representation, and Hot is unsurprisingly underrepresented. Overall, these responses conform pretty tightly to the average.

We’ll be doing this poll again at the end of each episode to see if the winds blow a different way once folks have spent more time with our romance options and find themselves in even more peril than Episode 1.

In the meantime, if there are other polls you’d like to see, or if there are other subjects you’d like to see us cover in our blog, be they about game development, marketing, or anything Scarlet Hollow related, please be sure to leave a comment and let us know!