The Making of Trollmother

Trying to explain my process for making Trollmother. Spoilers, if that's a concern.

I knew I wanted it to feel Scandinavian, and I knew I wanted something primarily menu-oriented, but for a long time I wasn't sure how to approach the competition theme.

The earliest version of the game was just the main menu, adapted from a troll-themed Figma phone menu mockup I'd done for Halloween the previous year. At that point it didn't even have “mother” in the title.

The music was another early addition. I'm not a musician, but because I'm not I figured I'd get it sorted early, so I wasn't bogged down when I had some momentum and a deadline approaching. I just kept putting bits of old folk songs into the GB Studio music maker, looking for something that sounded right to me.

Had a breakthrough with “Trollens brudmarsch” by Pelle Schenell (“bridal march of the trolls”, bridal marches notably a separate phenomenon from wedding marches), whose music should be well within the public domain. The march formed the basis of most of the soundtrack – edited and elaborated upon, but the central hook intact.

Early sketches already establish Näcken and the Skogsrå/Huldra as major non-troll characters, both being iconic creatures of folklore in Scandinavia. 
   It was only at this point that the idea of a troll raising a human child started crystalizing, inspired by old stories of changelings and of children being sent off by their parents to die (like Hansel and Gretel). I figured it would be an interesting angle on “You are the Monster” to have the “monster” be a nurturing presence.

I considered exploration elements, wandering around the world with the child to interact with NPCs and unlock new activities. The idea was abandoned early – there was just no way I'd have time for it and I'd risk making the overall design feel unfocused.

Then came stats, activities, and UI-design.

The opposing stat pairs were inspired by the Choice of Games style of interactive fiction. Figured it would make the stats more compact (for the sake of the Game Boy), and introduce an interesting strategic element to stat building (since you can never max out more than half your stats).

Sidenote: It's funny how Ida shows up in early notes as the “boring human”. I think she turned out genuinely likeable. 

I decided early on to try to fit all the information you needed for the planning stage in a single screen. Year, season, stats, available activities, locked activities, and chosen activities.
   Halfway through development I still hadn't conclusively decided on the 5^th and 6^th activity types, and I felt like I needed 6 of them for game balance. Initially you just picked an activity in the menu and was sent there immediately. I found it more satisfying to plan out three in advance, especially with how the month-slot icons for different activities flow into each other.

As a narrative game, it was always going to be text heavy. I also felt that character building games like this require multiple end states to feel entirely satisfying, so I started at the end, planning out potential endings before I'd written anything else.

I spent about half the development time just writing and still had to cut a lot of planned content. The witch, for example, never appears in endings in the Jam release, despite a lot of endings still having unused placeholder scenes for her. The Lord of the Forest ending is also, unfortunately, entirely absent from the Jam release.

I drew most of the portraits by hand before translating them to pixel art. I find that my pixel art feels less derivative when I do. And yes, the princess was originally planned to be a sheep. The main reason that this changed, honestly, was that I found the portrait size-difference with the frog to be funnier.

The activity backdrops went through a number of revisions and are heavily inspired by the work of illustrator John Bauer. Above all, I tried to emulate his approach to illustration, the distinct flatness and composition balance that he favours, which give his illustrations an almost sacral quality.

This kept everything neat and easy to navigate, made adjustments quicker, and likely prevented a lot of problems later in development.