Sojourn Past

Project Breakdown

Brief

Sojourn Past is a pixel-art action-adventure filled with fast-paced bullet hell combat. Players explore enchantingly beautiful hand-drawn landscapes themed after the five stages of grief, fighting gauntlets of monsters and fearsome bosses while piecing together their abandoned world's tragic history.

The game draws combat and world design inspiration from titles like Hyper Light Drifter, Hades, and Enter the Gungeon.

Team

ReDimension Games is a game studio based in Nagaland, India. They have won international acclaim for their work, including winning the India GDC Upcoming Game of the Year award in 2023, winning a booth showcase at GDC through the Winzo Bharat Tech Triumph program, and being selected for seed investment by KRAFTON through their India Game Incubator program. Sojourn Past is their debut title.

Project Goals

I was brought onto the team at ReDimension Games to address design issues that they had identified when playtesting early demos of Sojourn Past. I also assisted the team in building a public-facing demo for Steam Next Fest with a length of 60-90 minutes, over a period of 3 months.

Core Responsibilities:

Rebalancing player and enemy abilities and movement to make combat more engaging
Redesigning all existing levels in the demo, and building new areas from scratch
Improving studio processes in the design pipeline as well as documentation

Players trying out Sojourn Past at GDC, 2024

Design Challenges

Combat Feel

A lot of the work done on combat feel was the result of many incremental changes - I analysed frame-by-frame footage of games like Hades, Furi, and Hyper Light Drifter, and used the data to find more satisfying values for movement speed, cooldowns, and combos.

Small changes can have big impacts - for example:

Slightly increasing the character's movement speed allowed players to rely less on the dash ability for mobility, solving a lot of complaints related to the dash cooldown.
Making the radius of the melee attack's hitscan slightly bigger than the visible attack animation solved the issue where players felt their attacks weren't connecting with enemies.
Adding a tiny inertia/momentum slide to the player when starting or stopping movement did a lot to reduce player complaints of "stiff" gameplay.

These are a selection of design challenges I worked to solve during my time on the project.

Dagger Teleport

The dagger teleport is a unique feature of the game’s combat system - but playtesting showed that it has a high ‘attention cost’ to use. Players need to first scan the target location for threats, then throw the dagger, then trigger the teleport - all of which takes precious time and focus in the heat of combat.

To motivate players to use the dagger in combat, I needed to make them feel smart and competent for using it skilfully. One of our approaches was to make the enemy behaviour AI lose track of the player for a brief duration after each teleport, allowing the player to circle back and strike them from behind 'like a ninja' if they were fast enough.

We also made it easier for players to set up the dagger in advance, and then use it later to escape from danger - and even save themselves from a fatal fall at the last moment.

Environmental Engagement

In feedback, some players expressed that combat felt repetitive and boring. While our first impulse was to add more enemy variety, scope limitations pushed us to consider other avenues instead.

As I analysed feedback, it became evident that the problem was less with combat itself, as with the spacing and frequency of combat encounters. They were placed one after the other in sequence, creating player fatigue.

The approach I chose was to treat combat as just another type of environmental puzzle.

Each combat area should have unique challenges - if not enemy variety, then using environmental hazards or arena geometry.
Incorporating more interactable environmental pieces helped make the world feel more tactile, both inside and outside combat
Combat sections were interspersed with free exploration and simple puzzles, which greatly reduced fatigue

Level Design Pipeline

Since Sojourn Past's levels are all uniquely hand-drawn, we had to come up with less expensive ways to prototype and iterate on our levels.

The pipeline we chose at the studio involved creating 'greyboxed' levels using tile-based art in LDTK, which was then exported into Unity.

These prototype levels could be used for iterative playtesting both inside and outside the team. Only once we felt absolutely confident about the shape of the level, would the LDTK file be split into sections for the art team to turn into hand-drawn, optimized production assets.

Player Paths

An issue that previous versions of Sojourn Past's demo faced was that players refused to engage with the level design more than necessary. The main path to reach the demo's boss was obvious and unobstructed. Most players had no context to realize that there were secret areas and rewards they were 'supposed' to unlock.

For the Next Fest demo, I opted to make the level pathing less linear and clear. Multiple paths might converge to a single destination, or sometimes players would be expected to explore down a small nook to find a switch to unlock an obstacle.

These seemingly superficial changes altered player behaviour significantly. Because the game challenged their sense of navigation early on, they were much more likely to stumble across secret rewards early in the game, and were then motivated henceforth to actively search for secrets in subsequent levels.