Note: This blog post is a WIP! And technically not really a post-mortem...
I wanted to write about my experience developing a game with a small team for GMTK Game Jam 2026, similar to how I did in 2024. If you didn’t know, the GMTK Game Jam is a huge jam run by a popular game-analysis YouTuber, Mark / “GameMaker’s Toolkit”. This year was the biggest GMTK jam yet, with over 10k entries made over four days.
To avoid spoilers and provide context, you may want to play my team's game before reading this!
I always do art and animation for all of my games, and in some cases (especially in narrative jams), I do everything else as well. In my ideal situation, however, I'm working with at least two others: a programmer and audio wizard. This year, I worked with the same great team I did in 2024 — Patrick / "Patbug," who handled programming, and Tyler / "pleadthefilfth," who handled composing and SFX!
We hadn't completed a game jam together since GMTK 2024, so we did a preliminary game jam, the 2026 One Button Jam, in early July. While it was smaller and less competitive than GMTK, we did well! (Although I'm still mad that our ratings got nerfed by landing under the median number of ratings... in a jam where ratings was completely randomized! In my head, I'm pretending that didn't happen 😂 We would have gotten first place across most categories, including art and overall, otherwise.)
A rookie mistake I made before GTMK 2026 started was making a mood board before the theme dropped. I really wanted to work on something more abstract and magical-feeling:
.png)
But the best-laid plans of mice and men often go awry, and we ended up going in an entirely different direction.
Immediately after the theme ("Countdown") dropped at 1 p.m., my team struggled to come up with an idea that felt feasible with the time we had, while also being something we'd all be excited to work on, as well as connected to the theme. We had a truly insane number of ideas (over 50) on our brainstorming document...
... but none of them felt good to us. What was crazy about this theme is that there were a lot of immediate, obvious interpretations, but it felt incredibly hard to come up with something that felt more unique. Even interpretations I thought were super clever, like "Count Down Bad" (a Count Dracula dating sim LOL) seemed like they had a shocking number of people doing something similar.
We ended up with three ideas we explored more seriously:
We ended up going with that last option, as it felt most exciting to Patrick, who would have to make most design choices! This was intially how he pitched it:
The only problem? I was just coming out of having a carpenter ant infestation in my apartment (which I thankfully rent, not own), as the old wood in our gazillion-year-old walls were starting to rot. Bad timing to have to draw a ton of ants 😭
Patrick and I discussed a few other thematic options, including ducklings following a momma duck (Make Way for Ducklings, anyone?), but after seeing how small the player would be on screen, it was very clear to me that it would have to be ants. The day drew to a close, I went to dinner with a friend, and then started feeling out the art and narrative direction of this game when I got home at around 11 p.m.
For some reason, this jam felt like it had a lot of friction, and honestly, with a less patient and talented programmer, I think our experience might have fallen apart. Friction was present early on as we went back and forth for an hour, just trying to decide just some basic aspects of the game, such as the size of the ant and its grid:
The clock hit midnight, and I went to bed.
I woke up early on Thursday, full of adrenaline and determination, and went into Boston's game dev co-working space (shoutout Critical Mass!) to lock in. I chatted some with Patrick about the assets we'd need, and since I started to realize that the background would need to be less busy than prior games given that all the terrain was playable area, I asked if I could add a vegetation frame to the gameplay screen:
Unfortunately, I had a really hard time figuring out what kind of style I wanted to go in for this jam. I truly feel like the worst part of a game jam for me is in the first 20%-40%, when I'm experiencing a sense of almost existential dread as I iterate on ideas without direction and am so tunnel visioned in that it's hard for me to tell whether my art is good or bad. Since I had taken Thursday off work, I was expecting to be able to lock into a clear direction earlier, but I started sweating by the time lunch rolled around. I told the Critical Mass folks to grab lunch without me. (I ate at my desk!)
Thankfully, by mid-afternoon, I started to feel out the edges of the direction I wanted to go in, even if I was feeling very shakey about it:
Later that afternoon, Harold was reborn from the ashes:
Making seamless assets has to be one of my least favorite things to do in art, just because of how repetitive it is, but after discussing with Patrick, I agreed that we needed a seamless background so the camera could scroll across it some in more complex levels. I did this by bringing the original background into a new Photoshop document, copying it 3x, and displacing each of the four background copies to the four quadrants of the artboard, like so:
I then work to paint over the clear lines where the asset copies meet to make it as seamless as possible, and export the result. Since I have two.jpg textures overlaid on top of all my game art, I then had to repeat this process for both textures to make the final export truly seamless. As I said, it's a pain!
It wasn't until that night, after dinner with another friend, that I finally felt more confidence in the art I was making. Exhibit A:
Exhibit B:
I was finally happy the art direction after making this main menu — it felt beautiful, dynamic, and unique. I exported the day's assets, caught up on text messages, and then went to sleep!
I woke up early on Friday, and immediately got to work. I tend to like making itch.io assets about halfway through jams, after getting all core assets done, so I started on the cover art GIF. It was a fairly smooth process, thankfully! On Thursday, Patrick had suggested some game titles, and we ended up going with "Antempo" (Ant + Tempo!).
I then pivoted to the level select page:
Next, I drew and exported a bunch of obstacles and miscellaneous UI elements. It was at this point that I started to slow down — I was starting to feel some soft burnout after working on this project for 8+ hours each day. I pushed through, but it felt painful. One detail I was really fond of was deciding to make a wilted leaf variation on the results screen for failed levels:
I then went to dinner with my grandparents. When I got back, I drew Harold's bully:
Night crept up on me before I was expecting it, and I went to sleep.
Bonus content:
Saturday was another early day. I realized we had wanted to add some narrative, but we hadn't written anything yet. Tyler had written a lot of the text for our GMTK 2024 entry, but he still had a lot of composing and SFX work to do, so I decided to get some draft narrative written:
Once Patrick had put up a new build with a tutorial helping mechanic in the beginning, I sent it to my friend Jeremy for playtesting. This is when everything started to go awry. Jeremy commented that our first level was wayyyy too easy and needed more interesting mechanics, which he was definitely right on.
So, Patrick and I began a day-long scramble to figure out solutions. I had previously been feeling pretty confident about the direction the art was going in, but I tunnel vision into game jams and have such a hard time judging it my art objectively, so I tend to rely heavily on the feedback of others. As we were uncertain about the direction of the gameplay, I was feeling torn about the direction of the art. Thankfully, some friends (shoutout Mike and Ananya!) gave some encouragement, and Isaac was incredibly helpful with this feedback:
I ended up adding a hand-drawn animation effect to the border, which was a simple move that I think helped a lot!
As we were waiting for a new build with updated mechanics from Patrick, Tyler and I started working on the opening narrative video, which worked to introduce Harold and why he was trying out for marching band:
At around 7 p.m. that evening, Patrick uploaded a build with the new mechanic that increased the difficulty of levels. And thank God — it made the game juuuuuust hard enough that the gameplay felt satisfying!
.png)
I continued to work on some miscellaneous UI elements and gameplay obstacles, Tyler continued to work on music and SFX, and Patrick continued to iron out the mechanics. Later that night, Patrick mentioned that he was going to have to pull an all-nighter to bring everything together in the end. We definitely were cutting it closer to the deadline than we had in the past 😩 I felt badly, but knew that there was little left for me to do, so I went to bed around midnight.
I woke up with adrenaline at around 5 a.m. on Sunday. I pinged Patrick, who sent me the latest build with levels 1-3, which were fantastic! I drew a few more house assets to diversify our decor:
And also used gradients to recolor the seamless background for alternative level appearances:
Since there was nothing left for me to do but I still felt like I had a lot of adrenaline, I got started on writing this post-mortem a bit early! Patrick continued to work on ironing out the levels and adding some wishlist items, like an absolute trooper. He then uploaded the final build, and we submitted!
The ratings period has just begun, and it's hard for me to know how our little ant game will be received. To be honest, I feel like this was a much more friction-filled experience than most prior game jams; it was quite hard to iron out artistic and gameplay direction. I also don't think Antempo has the je ne sais quoi that our GMTK 2024 entry had, which is just difficult. It's hard to feel like any project is going to perform worse than a prior one.
A few other notes:
That's it, for now, I guess? If you've made it this far, thanks so much for reading! Shoutout to my amazing team, especially Patrick for being such a trooper, and special thanks to Mike, Jeremy, Ananya, Nila, and Isaac for playtesting and encouragement.
Signing off with love,
Julia
Copyright © 2023-2026 Roman Hyacinths | Privacy Policy
Website built from scratch with ♥ and Webflow.
.svg){kind=logo}