Game Concept
The concept for Sometimes came from me wanting to further explore and process my personal experiences with mental health with a particular focus on depression and anxiety. I wanted to use it mostly as a creative outlet for how I had been feeling, to better understand myself, and as a way of continuing to develop my skills. Starting off I had not intended to release this game publicly as it was about a very personal subject and even from conception made me feel slightly vulnerable when discussing it with other folks.
I started off by brainstorming some ideas for the kind of game I thought I might be able to create to convey and explore the ideas I wanted to. Early on my focus was definitely on the way in which I internally talked to and about myself. I was incredibly critical of minor slip ups and could send myself spiralling into self destructive habits. Some of my initial ideas included a game where enemies crawl out of mirrors and other reflective surfaces and attack the player, an exploration of a town with a twisted community centred around a lake, and exploring a neglected and broken house.
None of these concepts felt right to me. I wanted to explore more than just the self critical talk aspect that I had experienced. There were other feelings associated with my experiences that I felt were important and missing from these concepts, and I wanted to more directly address the themes I was aiming for. So I went back to brainstorming trying to find a form that the game could take that I felt would allow me to reasonably explore the ideas I wanted to. I eventually landed on the idea of smaller and very focussed gameplay vignettes after replaying The Beginner’s Guide as well as some smaller local games like Spiders.
This vignette approach instantly clicked for me and I quickly began designing the vignettes. The approach I took was to focus each vignette on a specific emotion. The emotions and ideas I initially came up with were:
Nowhere - I wanted to make the player feel like they can never move anywhere. To feel like no matter what they always end up back where they began. The mechanic for this feeling is having the player be able to move around but is constantly dragged back towards the center of the screen.
Alone - I wanted to make the player feel like that whenever they get close to another person that they run away or withdraw from the player. The mechanic for this feeling is having the player surrounded by others but as they get close the others fade away.
Talk - I wanted to make the player feel anxious and a little paranoid, like everyone near them was talking about them behind their back. The mechanic for this feeling is having the player surrounded by others who are talking, but stop as the player gets close.
Trapped - I wanted to make the player feel claustrophobic and locked in. The mechanic for this feeling is having the player surrounded by walls that slowly close in upon them
Dragged Down - I wanted the player to feel weighed down, heavy, and forcibly moved towards a place they do not want to go to. The mechanic for this feeling is having the player constantly pulled down towards a void.
Left Behind - I wanted the player to feel left out and abandoned. The mechanic for this feeling is having the player move along a group of others who move faster than the player, eventually leaving them alone.
Leaving - I wanted to create a vignette based around suicidal thoughts. The idea for this is having the player in a hallway that they can leave through a door, leaving them surrounded by complete darkness.
After I had come up with the concepts for each vignette I jumped into Unity and prototyped the primary mechanic for each vignette. Below are some gifs from this initial prototyping phase.

An early prototype of the mechanic of Alone where NPCs would avoid the player as they approached

An early concept of the Leaving vignette where the player would walk down a hallway towards a door that opened to a void

In addition to the mechanics of each vignette I came up with the idea of having an intro to each vignette that explicitly states the feeling presented in the vignette. I also elected to use language similar to that present on DAS (Depression Anxiety and Stress) or E10 Forms, several of which I had filled out in the course of my own mental health journey. The specific words that I knew I wanted to use were “Sometimes”, “Often”, and “Always” in order to convey the intensity of the vignette as the player proceeds through the game. Each vignette would be preceded with something like “Sometimes I feel <feeling>”.
This also kicked off thinking about the structure of the game as a whole. I wanted to convey how these feelings can intensify and combine as one’s mental health deteriorates. I did this by grouping up a number of feelings and their mechanics. I had 7 individual vignettes for the first act of the game, Sometimes. For the following acts I would combine the mechanics of multiple vignettes into a single one. The pairs for the second act, Often, were Alone + Talk,  Nowhere + Left Behind, Dragged Down + Trapped, and Leaving + Talk.
And again for the final act, Always, grouping 3 vignettes which were Alone + Nowhere + Talk, Dragged Down + Left Behind + Talk, and Trapped + Leaving + Nowhere
Connecting the sections and contextualizing the vignettes was another area where I needed to spend some time. My first thought was to have a written discussion between the player and a therapist but I felt that this would overstep my knowledge and experience, as well as being problematic as I am absolutely not a therapist. My other idea was to have this as a conversation between the player and just another person, potentially a friend. I wasn’t sold on this idea entirely so I shelved this decision for a while as I continued to work on creating the systems necessary for me to make the game.
After creating each of the vignettes (both isolated and combined ones) and testing them I found that two of my vignettes were conveying a similar feeling. These vignettes were Nowhere and Dragged Down. Cutting one would give me a few benefits: fewer vignettes to work on and polish, and at the same time reducing the total vignettes as I could pair up feelings without doubling now and group 2 lots of 3 for the final section of the game. I decided to cut Dragged Down as I felt the more effective section for conveying the ideas was the one where the player has the most agency over their ability to move.
I soon realised that Sometimes wasn’t balanced. Not in the sense of tweakable values and such but in the emotional and thematic sense. I was so focussed on presenting the negative emotions I had felt in my own journey that I neglected the positive emotions. These are incredibly important in this story as they help contrast with the negative, empowering both to better convey their ideas, and at the same time I could more properly convey my own mental health journey by showing how these feelings are affected at the same time. In order to address this I started designing some vignettes based around more positive emotions. The vignettes I brainstormed were:
Wanted/Included - Player is able to walk up to others who then walk along with them
Successful - Player collects some pickups
Supported/Loved - Player is walking near others and when they get close they get a little speed boost
Helpful - Player presses buttons in order to help people
Strong - Player can walk over ghosts to make them disappear
Safe - Player moves around a peaceful and serene area

The layout I used to organise each vignette using Milanote

This allowed me to move the vignettes around to find what felt nice to me in terms of ordering them. I quickly used this template for the positive emotions and moved them among my existing vignettes to see how I could fit them all together. I was considering taking the same approach with these vignettes as I had the negative ones by combining them as the game progressed. This felt wrong to me as what I wanted to achieve by including the positive feelings was to show how they become dulled while dealing with depression and anxiety. This led me to the next idea of twisting the mechanics of the vignettes as opposed to combining them. The twists would become more intense and the positive vignettes would become much shorter as the game progressed through Sometimes, Often, and finally Always. The twists I designed were:
Often: maze between player and others
Always: wall between player and group
Often: player can collect most of the collectibles
Always: any collectibles collected are reclaimed after a few seconds
Often: player receives smaller boost
Always: player scares others when they approach
Often: player is able to help most of the others
Always: player actively harms others
Often: player is knocked back by some of the ghosts
Always: player is unable to do anything to the ghosts and is constantly hurt or knocked back
Often: threats appear near edges of screen
Always: threats break through to where the player is
After I had done the initial designs for the twists I began to move the positive vignettes about and placing them around the negative vignettes. I was thinking about each section of the game and how I wanted to bookend that collection of vignettes. I knew that I wanted to end each one with the vignette that included Leaving, but due to my goals for the game it felt wrong to start each section with a positive vignette. Particularly for the Sometimes section where I would alternate between positive and negative vignettes it prompted me to cut one of the positive vignettes. I decided to cut Wanted/Included as it felt the weakest to me and the mechanics didn’t convey the feeling well enough for me to be happy with it. I then decided on the following order for the vignettes in the Sometimes section:
Left Behind
This felt like the right balance to me as the positive vignettes somewhat mirrored one of the negative vignettes around them which created a stronger contrast. I also did the same for the Often and Always sections of the game.

Determining the grouping and ordering of the different game sections

I decided to add an extra section to the end of the Always section to show the kind of negative self thoughts spinning around in the head when dealing with depression and anxiety. I felt this was vital to the game as it reflected my own experiences much better.
The addition of that section prompted me to go back and look at how I would connect and contextualise each section. I ended up liking the idea of a conversation however I decided it would be an internal conversation that the character is having with themself as recognising and going through the mental health journey for me was a very self reflective and introspective process. The issue that faced me at this point was figuring out how to end the game as I didn’t want to end it on the down note that was “everything sucks” at the end of the Always section, but  then cutting to “I got better” felt dishonest about the amount of effort it took to work through all those feelings and regain emotional balance.
Another important point was that I didn’t go through it alone and I wanted to convey my behaviours of avoiding difficult conversations, and eventually reaching out to other people for help. The result of this was adding an Other that the player was conversing with at the end of the game. This way the player has someone to reach out to for help and it made an ending that more closely aligned with my goals. Below was my initial script for each of the conversations that open and close each section of the game, as well as the newly added Epilogue.

Initial script for the intro and outro conversations that bookended each section of the game

While working on the vignettes and testing my approaches I found that another way I could reinforce how depression and anxiety affect even positive feelings was by distorting the sounds in those vignettes and dulling the colours and effects at the same time. This resulted in vignettes that reflected the ideas that I wanted to present and helped me reach my goals for the game.
It was at this point where I had made each of the vignettes in their entirety and started to piece them together to make the game whole. For the epilogue I quickly cut between a couple of incredibly short versions of the vignettes (maximum 5 seconds) in between the lines spoken by the Other and the Player. After implementing the introductions and conclusions to each end I sent the game out to some testers and got some incredibly useful feedback. The primary bit of feedback I received was that it was unclear that the conversations were happening with another person once the Other started having lines.
My approach to fix this issue was to utilise the existing assets to create some quick new scenes at the end of each section where the player is forced to walk a few steps in a specific direction. For the Sometimes and Often sections I had the player walking away from the Other and in Always I have the player walking towards them. As the sections progressed I also slowed the player down to show mechanically how difficult it got to keep moving forward.
I also shifted the way I approached the Epilogue at the same time. I decided to cut in between the lines with the micro-vignettes as a sort of punctuation and break in between the dialogue. I wrote a few extra lines to fill out the required amounts that helped further my point about moving forward and pushing through to find how to heal, how others help in this process, and how despite going through that those feelings are still felt to some degree.
Visual Design
From the beginning of the project I wanted to keep the visuals simple. I chose this direction as it kept the visuals achievable for my skill level, and the simple forms allowed players to quickly discern the goals and structure of the vignettes.

One of the first concepts I made when going through the prototyping phase

After rigging and creating a simple running animation for this character I felt that it was missing something. It wasn’t quite meeting the goals I had set out for the visual design and I was concerned that these were too abstract or weird for players to relate to. My next iteration of the characters were simple and relatively featureless mannequins.

The mannequins I created for the early stages of production

Mannequins animated and implemented in the Alone vignette

But this came with another problem. The mannequin is representative of a singular body type and once again may exclude some players from relating to it. On top of this I was also changing perspectives between some of the vignettes. In particular Nowhere was displayed in first person, with most of the other vignettes being presented from a top down perspective. At this point I was finding my 3D skills somewhat lacking and the list of assets I needed to make quite large, not to mention having to consider proper lighting and shaders for the models. I felt that I could produce the game much quicker with 2D assets and at the same time I could bring the perspectives entirely in line with each other and keep it consistent through the entire game. I felt this persistent perspective was important as it would allow me to retain the same control scheme throughout the game which would make it easier for players to get into a new vignette.
I still wanted to keep the assets very minimal and clean. I sketched up some ideas for the characters and quickly landed on a design that I liked that also allowed me to use expressions easily. Below is the final design for the characters in the game.

Final character design showing multiple expressions

Final character design walking around

Final character designs showing the Talk vignette mechanic

Final character designs showing the Helpful vignette mechanic

Back to Top