Heart Machine
patreon
A Heart to Heart with the Possessor(s) Level Design Crew
🕑 Added 2025-04-16 20:27:36 +0000 UTC
Can you share a little about Possessor(s) overall level design principles?
“One of our main level design goals in Possessor(s) is balancing excellent room flow with grounded, naturalistic 3D spaces.
Making memorable, interesting, large-scale 3D spaces was almost second nature for us by the time Solar Ash shipped. When we started out working on Possessor(s), we continued our momentum from Solar Ash and built large 3D spaces where the “game plane” (the playable area of the 3D world where the player character is) was just a slice through memorable, expansive 3D spaces in the style of a Solar Ash level. We avoided having too much of the world aligned to the game plane, favoring spaces that were rotated at odd angles so that the player wouldn’t be able to see the grid/”gaminess”. This resulted in a very unique, naturalistic feel to our levels. However, it also resulted in poor gameplay readability. It was hard to tell how far platforms were from each other, and some objects that were close to the game plane but didn’t intersect would appear to be meaningful platforms. Players couldn’t tell where the limits of the playable space were, so the game wasn’t comfortable to navigate.
After a lot (A LOT) of iteration, we landed on a balance between the grounded, naturalistic spaces we were building early in pre-production and the excellent gameplay clarity that’s characteristic of the side-scroller genre.” - June Patterson, (former) Game Co-Director
What about your own core philosophies around level design?
“The approach really depends on the game. I’ve worked on a Metroidvania like this before, but the philosophy for that was very different. Here, we’re telling a story about bad relationships—our goal is for the narrative to help characters grow and understand healthy relationships better.
Looking at the world itself and how it ties into this theme is crucial. We’re creating a playground where the spaces, the characters, and their connections all feed into the levels. Lore tidbits, cutscenes, and environmental storytelling help reinforce that. A big focus for me has been weaving platforming and combat encounters seamlessly into this experience.” - Mars Ashton, Senior Designer
Talk about your level design approach from the start and how it led to where we are now with Possessor(s).
“Our approach to level design for Possessor(s) wasn’t necessarily what you’d call typical. We did what we called ‘Designer Deco’—no paper prototyping, just bringing things straight into the engine, testing over and over, and then layering in concept work, art, and audio afterward.
Instead of starting with rigid tile-based layouts, we jumped in and got super loosey-goosey with it. Early on, we were just building these believable spaces in a world that’s been ripped apart. A lot of trial and error—Jordan, June and I would build levels, play through them, and figure out what worked and what challenges came up.” - Mars
“Since Possessor(s) is the combination of 2D and 3D elements, it took a while to figure out that balance.
In traditional 3D level design, you’d block things out with featureless cubes—just the basic gameplay geometry, making sure the player can walk from A to B, without worrying about aesthetics. Initially, we tried something similar here, blocking out levels with 2D tiles. But because of the fixed camera, the 3D elements had a huge impact on how the player perceives the space, even if they’re just background elements.
For example, you might see a 3D building in the background long before you ever see the door you can actually approach and interact with. That’s a big deal. If you build everything in just a greybox, players get confused, lost, and disjointed. But just adding some set decking—small environmental details—dramatically improves navigation and spatial awareness.
That approach really opened up the range of things we were able to build and get away with.” - Jordan Fanaris, Level Designer
Room Size matters
“There’s a lot of variation in room dimensions game-to-game. We tried a wide spectrum of room sizes before landing on what we now consider to be our standard.
First, we had no rooms. Every zone was a single contiguous space. This made navigation very difficult. Clear room boundaries and transitions between rooms are essential to players understanding the space and being able to confidently explore. Room transitions specifically allow the player to empty their mind of what they were focused on in the previous room and start taking in new information fresh. We struggled with giant continuous spaces because it felt like everything was relevant all the time — the player was constantly accruing new seemingly relevant information without ever having a mental break, and eventually, players gave up on making intentional navigation decisions.
Next, we dramatically reduced our room size. We targeted rooms closer in scale to the smaller range of rooms in Metroid Dread. This made our spaces feel bite-sized and manageable, but too much so. It made the gameplay flow too rapid and we lost a lot of the intrigue of larger spaces.
We slowly built back up to a room size closer to Prince of Persia: The Lost Crown. These larger rooms allow us to have big memorable spaces like we had early in the process, but with the benefit of periodic fade transitions to allow players clear bookends to chunks of gameplay.” - June
“From there, we did a rough environment art pass while Sarah Morris (Senior Concept Artist) worked on concept art under Danny Moll (Art Director), managing themes and atmosphere, alongside post-processing and VFX. Wolfgang Traenkle (Senior Environment Artist) and Matt Graczyk (Senior Environment Artist) then took stabs at key spaces—checkpoints, boss rooms, the starting area—setting the standard for what Jordan and I built into.
It was a lot of different bubbles floating around, occasionally bumping into each other, but we landed in a very good spot.” - Mars
We went from 2D to 3D for Solar Ash, and then Hyper Light Breaker. What’s it like to go back now?
“2D-level design is inherently easier because there are fewer factors to worry about, but at the same time, it’s harder to make interesting. You can make a room that you walk through from A to B much more easily in 2D than in 3D—there are just so many fewer components. In that respect, it’s easier. But since you also have less to work with, it’s harder to come up with things that are new and different and interesting.
It was interesting to pull 3D tricks and apply them back into 2D. In 3D, it’s easy to add a windy, spiral path that turns a straight ramp into a twisting, meandering journey—where you look back and realize there’s been this whole experience. In 2D, it’s easy to lean into zig-zagging back and forth, but that can quickly disorient players when they can’t form perspective.
I’d never worked on a 2D game before—never made a platformer in Game Maker or anything like that. Fortunately, Solar Ash is a platformer, so a lot of things were directly transferable. And you can make a lot more gameplay way faster. The maintenance cost, or cost per square foot of building 2D levels, is much lower than in 3D.” - Jordan
I was tapped to move over after doing QA on Solar Ash near the end. I handled Breaker’s initial QA pipeline setup, and then from the get-go, Alx said, ‘I want a smaller team to work on this project.’ So that transition happened pretty naturally.
I was already used to working in 2D, and since this is 2.5D, it was a pretty easy shift. We’re still able to think about the way things exist around the player. We’ve also experimented a lot with the camera—how we want to manipulate what the player sees and how that impacts gameplay.
Before any of my work with Heart Machine, I was working in Flash all the time—on tons of 2D side-scrollers—so in a way, this has been a return home for me. - Mars
Some of the team’s biggest game design influences and references
Collective considerations while playing these games from the team’s perspective were:
How the different levels in each game are interconnected and their shape (if a room is a box, what is its capacity? How much can a character move around?
Framing, composition, and camera positioning
What various things are done with traversal to allow the player to have bigger, more involved 3D spaces, while also having fast-paced action gameplay?
Dark Souls, Hollow Knight, Prince of Persia: The Lost Crown, Metroid Dread, Legend of Tian Ding, Hyper Light Drifter, Blasphemous, and Inside were all games we looked at asking ourselves those questions.
Further analysis by June on Flow and Grounding
Flow
Allowing players to continue their momentum along a cardinal direction is essential to clear navigation in a side-scroller. Players struggle to navigate levels that constantly have them changing direction. Good Metroidvania design typically promotes instinctual navigation by allowing players to choose between continuing their current momentum or branching to build momentum in a new direction. Generally speaking, room dimensions line up with the intended navigation experience of the room (e.g. a room whose main direction of motion is left to right would be wider than it is tall).
“Maze” rooms with lots of winding paths that don’t have a clear cardinal direction are interesting punctuation in the level flow, but shouldn’t be used too often because they jumble the player’s internal compass
Room type analysis of Hollow Knight:
Unlike in a linear game with one-off levels, too many true dead ends can be very frustrating in a Metroidvania. Inter-connectivity is essential to good, non-linear level flow. Good Metroidvanias are a net of valuable places — the player is constantly planning the route to their next location, and figuring out efficient ways to move through an increasingly familiar map. Very few rooms are “one-offs” that the player only ever needs to visit once.
Grounding
Similar to HLD, we want our levels to feel realistic and grounded given the rules of the game’s world. Obviously, we don’t want to fully constrain ourselves to the real world, but consistency is very important. Many Metroidvanias sacrifice this sense of realism in favor of interesting gameplay challenges, but we want a solid balance between good gameplay and a believable world!
An example of this is stairs: we tried many different approaches to staircases before landing on a consistent set of rules for stairs that we apply across the whole game.
And finally, how do you want people to experience this game? What do you hope they will notice?
“I would hope that players notice we’re able to get paths and connections into things that are very complicated in a very large world. Our levels are really big, but we’ve done a really good job of building those paths in a way that isn’t as complicated or convoluted as it actually is.
We had a playtest where one of the testers said, ‘I keep wandering around thinking I’m lost, but then I realize I’m where I needed to go!’ That’s exactly what we’re going for. If you can go on a full adventure and feel lost but not be lost, that’s something a lot of people mess up because they listen to what the player is saying instead of observing what they’re doing.
BEING lost is dead ends, backtracking, walking in circles. FEELING lost is the player saying, ‘I’m not sure where I am or where I’m going,’ but they’re still making progress and getting to exactly where we want them to go. - Jordan
Here’s a walkthrough of the player’s experience of new areas:
Expeditions: find the bounds of the playable space. Where are the a) gates and b) suspicious locations? Gates and suspicious locations should be easy to find. The player should “stumble across” these as they run around the level
Investigations: look for secrets (keys, abilities) in the suspicious locations found in step 1. Secrets, as the name implies, should not be easy to “stumble across”. They should require active attention from the player to find
Solutions: use keys and abilities acquired through investigation to pass gates identified in step 1
The “investigations” step should have multiple essential secrets to discover. Each new area should have multiple threads to pull on to progress the game (optional side paths don’t count, I mean multiple ways to progress the “core path”). - June
I want people to notice that there’s intent in every space. There’s a crate on an enemy—I’m hoping players will see that and have a reaction, whether it’s ‘That was cool’ or ‘Oh, who would do that? Horrible designers!’ Any kind of instinctive response.
The ‘aha!’ moment isn’t always about a big discovery. It’s that feeling of ‘I see what they’re doing here’ or ‘I see what they want us to do.’ That’s what I hope players notice. - Mars
Comments
Rixen
Super neat to see the thought process behind everything!
