Temporal Thief: GMTK 2025 Postmortem

It is that time of the year again, when the mass of aspiring indie developers on Itch.io band together to accidentally DDoS the site by uploading their jam projects.

The 2025 GMTK game jam was even bigger than the last one, and it seems that it’s breaking records again, as Itch just couldn’t keep up with the bandwidth once more. I, of course, also took part and I made Temporal Thief – a puzzle game, inspired by The Sting! (2001, neo Software/JoWood) and Stolen in 60 Seconds (2007, HeroCraft).

So, let’s dive right into dissecting it, shall we?

What Went Right?

First and foremost, I managed to do a lot more polishing than on previous jams. This was mostly because I had the chance to work uninterrupted through the whole four days of the jam. The theme was announced at 8:00 PM local time, and while I had a house party at the time, I was free to start working and keep my head down until the end of the jam if need be.

Also, while some of the mechanics didn’t make the final cut (there was supposed to be an undo feature, for example), the ones that did were quite fleshed out and bug-free for a jam project.

What really helped the game stand out from my previous project was lighting. I’ve always undersetimated the power of lighting in 2D, but no more. After seeing what I could achieve just by lowering the global light and adding some focus lights to the props and characters, I will never make an unlit 2D game ever again! I am exaggerating, of course, but the use of lights really gave it the feel that you are sneaking though dark corridors and not just a generic pixel-art environment.

Speaking of art, I made all of it myself and, honestly, I am quite proud of it, considering the limitations of time and talent.

What Went Wrong

Honestly, not a lot. There are only a handful of things that could have gone better.

The bugs from the undo feature are actually due to my biggest back-end error. The architecture of turn-based time travel must be done right, or else you have a ton of edge cases that you need to address. Unfortunately, I got it wrong. So, if future development occurs for this project, a lot of the logic will be rewritten from the ground up.

The most obvious problem is the lack of an undo feature. While this was part of the game for the majority of the development cycle, it was cut in the closing stages. It was just a quality-of-life feature and it introduced a ton of bugs, which I had no time to address.

Late core mechanic changes also threw a spanner in the works. At first, player characters spawned at time portals and had to return to them by the end of the level. However, that severely limited the puzzle design options, and I realised that too late, because I left level design for last. Once I did, though, I had to spend two or three hours uncoupling the spawning from the exit logic and making sure players can spawn anywhere. This limited level design time significantly, as it was all done in the last 24 hours of the jam.

Speaking of level design, I underestimated the difficulty of designing puzzle levels. I shouldn’t have, but I did. And for a puzzle game, that is a make-or-break moment. I believe I could have done a better job with the puzzle design if I had left myself more time for it.

Lastly, objectives were supposed to be a part of the game at the start, and when I cut the scope down, I should have removed the money objective. It just never occurred to me. I could have, and should have, left just a single “collect all gems” or something. One of the twelve levels even has an optional objective, because it was the first level ever made, and it was used for testing all the mechanics.

Summary

The results of the jam are not out yet, but I wanted to write this postmortem, unbiased by my performance after the ratings.

I believe I did a great job overall. There are areas where I can improve for future jams, but I strongly believe that if I had more time, I would have avoided the biggest pitfalls. Regardless of the results, I am happy with my performance. I am planning to reuse a lot of it and begin working on a spiritual successor of sorts to Stolen in 60 Seconds.