top of page
Search
Writer's pictureQuain Holtey

Creature Comforts - A Postmortem

The open dryer in Creature Comforts after shrinking a folding chair and spitting it back out.

Overview

Creature Comforts was a game jam made in 96 hours for the Game Maker's Took Kit (GMTK) 2024 Game Jam. The theme of the jam was "Built-to-scale." So, in Creature Comforts, you play as a mouse with a desolate den who must carry human-sized furniture to the dryer to shrink it down to mouse size so you can take it back home and build up your den with the creature comforts humans enjoy.


In a spark of inspiration (or insanity, take your pick), I decided to do a game jam with some friends the weekend of my birthday. 27 never looked so good, I'll tell you what. I reached out to several friends, some of whom said yes! Coincidentally, there was a lot of overlap between the people who worked on the last game jam I was a part of (being Juniper), and this one, with several newcomers I was happy to meet!


I think because I started getting everyone together, and because I was one of the contributors with the most availability during the jam, I took on a sort of project manager role, helping to manage the discord server, creative conversations, schedule and conduct meetings, make the development schedule, set up and help with version control, and kind of manage our Trello board. Though, it's important to note there were several contributors with production experience who helped with all of that, too.


I also did a fair amount of developing. Diving into blueprints, especially front end and in-game UI, as well as helping with interactable item functionality, game managers, scoring, and narrative system delivery.


What went right

Waiting for the dryer to finish shrinking an object in Creature Comforts

Compared to my last game jam, this experience was silky smooth. We set aside time on the first day to meet in person after ideating on the theme all day asynchronously to decide on which idea we would pursue, and what our scope could look like.


Somehow, and this is wild to me even now, we over scoped despite being very conservative in our approach. You might think something like this should be put in the "What went wrong" category, but I disagree for one specific reason: we prepared for it. The original concept was to be an entire house, giving the player multiple days/tries to gather as many items as they could while they had less and less time, and more and more obstacles. We were immediately worried about this. Despite its small scale, we knew we only had 4 days (technically 3.5 after the initial meeting), rather than a full week like we did with Juniper.


So, we narrowed. Let's start with a studio apartment, or better yet, just a basement! Let's focus on getting just the 1 day figured out and refined, then we can focus on expanding to 3 days with a persistent level. Let's focus on the item interactivity, shrinking, and scoring before we worry about creating and placing obstacles. Etc, etc, etc.


There's always talk about focusing on the "Minimum Viable Product," but I've always found that term to be vague and unhelpful in development. Everyone always has different ideas on what's considered "minimum" in an experience. Instead, for this jam, we narrowed our focus on the "Core Gameplay Loop." What was necessary to make the game functional from beginning to end? Grabbing items, shrinking them, bringing them back to the den. A time limit, a score screen, and a fail-state. That's it. As long as we had all of that, we would have a game, and one we could expand on if time allowed.


Time, shockingly, did not allow.


We also scheduled. With Juniper, I remember being very frantic as we were getting things in incredibly late, like art assets and gameplay items, and the result was a lack of polish on a lot of things that felt good in concept, but that felt bad in gameplay (the parallax background, or the inconsistent jump plants for instance). We also struggled, HARD, with getting a build of Juniper submitted on time because we waited to build the game until the last day.


So, starting out the jam, I made a schedule. One very reminiscent of what I learned at Epic and their development cycles. I introduced deadlines every day, including goals for feature complete, hard lock, pencils down, daily and nightly builds, and playtests. I'll talk about some of these that we didn't hit in the "What Went Wrong" section, but merely having these deadlines, and knowing that regular builds were going to be made and distributed, resulted in an entire team aligned with when they needed to get things into the game. Which leads me to the final point I'll make in this section.


We communicated. Overcommunicated. Loudly communicated. We told each other every time we were checking out a level or a blueprint or an asset, we managed the few merge conflicts quickly, we shared designs early and often, and we let each other know when we were blocked and why. Were we perfect about our communication? No. Not remotely. But there are times when I look back on projects like Alien Boogaloo, where miscommunication and lack of communication led to days of work getting overwritten again and again, and I think about how amazing the team on Creature Comforts was by comparison.


Additionally, I made myself available as much as possible. It led to LONG days, and a fair amount of interrupted work, but it meant that I could answer questions, resolve conflicts, and align team members on goals quickly and regularly, which become crucial on the last day of the jam, where everything finally started coming together.


But, like I said, not everything went smoothly.


What went wrong

Picture of the den in Creature Comforts after returning several shrunken items, such as a couch, table, lamp, and television.

This will be, I think, the easiest section to write here. There are some things that went wrong, but for the most part, I feel like there wasn't much that got in the way.


First and foremost; we missed deadlines. One in particular: feature complete. We wanted to have the core game loop finished by the end of day Sunday (day 3 of the jam) and we missed it. The game was not playable, as several core systems were 90% there, but not functional. Art was still being implemented into the level, the dryer/shrink functionality wasn't working, and there was no den to bring items back to even if we could shrink them. We were working HARD, and most things were almost there, but by the end of the day, the build looked no different than it had the previous day.


This was a morale hit, for sure. We had planned to spend all day Monday polishing and implementing some of the stretch asks, like multiple days, a Roomba enemy, and an expanded level. Instead, Monday was spent finalizing core gameplay features, and any polish and bug fixes came in later in the evening, and for some, early morning.


We didn't paper design. Sometimes it feels like there isn't time to paper design, or that functional code is greater than written ideas/documentation. As a writer, I think about this a lot. Outlines are only useful as long as you stick to them, but a rough/first draft is where the "magic" happens. That's when you have a story!


In game design, it can be tempting to jump from ideation to implementation, and in a game jam the timeline feels like that temptation is a requirement. But just like in writing, if you don't review and fix mistakes early on, you'll be spending more time fixing them later. We didn't paper design, or at least we didn't set aside time early on to review each other's paper designs. The result? We missed feature complete. The den, dryer, and interactables all interacted with the player controller in unique ways, and EVERYBODY wanted to make changes to the player controller blueprint, resulting in conflicts and weird hand offs so someone could do one thing real fast.


We encountered this issue everywhere. Game menus, gameplay levels, game manager, first person controller, just to name a few. It was rough, and entirely avoidable if we had gotten together on what we were thinking in terms of design, especially on core elements like interactions and items, and aligned on how we were going to do things earlier.


Another shortcoming was task work. Several times contributors asked me what they could work on next while they were blocked or finished with their core tasks, and I didn't really have an answer. I could make suggestions, but scoping out exactly what would be considered needs, must haves, nice to haves, and want to haves so that we could work down the list could have helped us get more features out with the game and added to the core loop.


And finally, something I think went wrong was an understanding of contributor availability. This is more of a personal failure as a project manager than anything. Several of us were able to spend all day, every day of the jam on the game. Others could only contribute a few hours each day. Others could spend all day on the weekend, but come Monday they had, you know, work.


I kick myself for this because I knew this would be the case, but I didn't think it would be a problem. My mindset going into the jam was to scope out the work that needed doing, and to let everyone choose which work they would contribute to with the understanding of their own availability. Basically a "scope yourself" mindset.


While I don't think that mindset was a complete failure, it led to unrealistic expectations throughout the project. In hindsight, I wish I had reached out to the team members individually to ask about their expected time commitment to the project so I could better help assign work, align expectations, and get what needed getting done finished earlier, if possible.


What I'll change in the future

Creature Comforts Credits list: Lauren Adzima, Stephen Forsyth, Rebecca Hertsel, Natasha Hollander-Ho, Quain Holtey, Okra, Olga Collazo Perez, Clifford Rich, Val Rubio, Sameer, and Kate Zgaga
  1. Setting scope expectations - working with the team to figure out where exactly our talents lie, what/how we would like to contribute to the game, and how long we can actually spend in development. Doing so will help discussions from brainstorming and ideation to task assignment to deliverables. Establish these expectations EARLY and be conservative in the estimates. Encourage team members to round down, not be optimistic, to avoid over scoping and end-of-jam panic.

  2. Paper design & review - Have regular meetings during the jam focused on establishing core designs and functionality. Take notes and post for those who couldn't attend the meetings. So many late jam issues could have been solved if we had been aligned on how things like "picking up an object" were going to work beyond just concept.

  3. Narrow down - Focus on the core gameplay loop. What minimally is required for the game to play from start to finish. The rest of the game can be filled out around that core loop, but without that core loop, the rest of the game has nothing to rest on. Be merciless: what exactly is needed, not what would be nice to have.

  4. Task out - task out core features to those that care able and available to work on them. Prioritize them first, but have ancillary tasks and features planned out for the contributors for when they are blocked/finished with their core tasks.


How we did


Overall, there were 7,640 games submitted to the jam, and we ranked 1,819th overall, putting us in the top 23% of games in the jam. Pretty good, but not great. Compared Juniper, Creature Comforts was a better development experience by miles (though admittedly more stressful under the tighter deadline), and I think the final product is ten times the game Juniper was.


We were rated on enjoyment, creativity, and style, all of which were rated around 3.3/5. I think if we had a tighter understanding of our core systems, we could have polished them up a lot more, as well as added more features to add to the enjoyment of the game. Higher creativity could have come from an expanded gameplay loop. Our original idea was for the mouse to hide/break line of sight from patrolling humans, with each shrinking object taking away hiding spots for the mouse, and I think that would have been a more creative concept that wouldn't have needed a timer; but scoping out that part, since we didn't have character models, animations, or someone versed in AI controllers, was the right call in my opinion. And style...I'll be honest, I'm not sure. More sound effects, animations, and generally more "juice" would have made a more stylistic experience I think, but I don't think we could have been much more stylistic without more art and sound design elements, which we didn't have time for.


Overall, top 23% of over 7,600 games is pretty good. Definitely room for improvement, though.


And speaking of improvement, we'll be polishing up Creature Comforts for a steam release in the coming weeks. Adding in our initial wants and fixing what's existing, so watch out for that!


4 views0 comments

Recent Posts

See All

Comentários


Post: Blog2 Post
bottom of page