My Experience Taking Part of the GMTK Game Jam 2025!

Route ‘n’ Run! by Nathaniel Jones

Create AI truck loops and avoid getting squashed!

itch.io

The GMTK Game Jam was going on between a beach and snow holiday. There was not much work going on between the two events, So I signed up! Here are some reflections on the entire experience as well as some tips I learned along the way!

The Beginning

I slept horribly. The theme was appearing on YouTube in 6 hours, I could barely bring myself to sleep as my mind was buzzing with a cocktail of worry and stress.

This would be the first Game Jam I would participate in. What if I embarrassed myself? What if I was not good enough? What if Mark Brown himself knocked on my door and told me it truly was a miserable showing?

You're not defusing a bomb! Stop and take a breath! Chronic stress may be on the rise throughout the world but this should not be one of the contributors.

How do I know this? Because after my horrible night do you know what I was greeted with? A timer for another 24 hours, I did not adjust for the timezone difference.

It may be Thursday July 31st for me, However America where the competition was hosted was enjoying their Wednesday July 31st. I was 24 hours early!

I couldn't help but laugh, get back to my own work and relax. Which defused the stress. The next night I would continue to sleep like a baby.

The Theme

I stared at the screen still trying to stifle a yawn as I read along, Loop... How we translated the theme was up to us, however the well was already a tad poisoned for me.

For context, I very recently completed one of the best examples of a game with the theme of "Loop" in existence, The Outer Wilds. In response, I was very conscious to ignore time-loop based mechanics as they were still living rent free in my head.

So I sat and pondered about the theme for an hour, no lists just almost meditating on the problem. It was rather early in the morning I did not have many friends to pitch ideas to. I was on my own.

A good idea would come here and there but the trick was to match it with something that could be completed by a Solo Developer in the small provided time-frame.

At the end of my pondering I had settled on an idea. I liked the idea of a game of recording actions and looping them back. Then dealing with the consequences of those actions.

Development Begins

So I settled on the idea of being a logistics manager and coordinating AI pre-recorded trucks around a warehouse factory area. I wasn't confident in my ability to live record actions so I came to the concession of turning the truck moves into a sort of turn-based puzzle.

Your employer would not want to give you overtime, because of this the pre-recorded trucks would playback in real-time as you're trying to run out of the factory causing you to get run over and splattered by your own trucks. The idea was set and the game was in motion.

I began work on creating the pre-recorded trucks and got them together pretty quickly, my current major project works a lot with tile-maps so I had a lot of experience setting up and creating 2-D paths so I did not have to waste much time learning how to use them.

Moving the player was as simple as holding WASD and moving a little square around and area collisions would be attached to trucks to report when a player got too comfortable with being alive.

Within the day, I had set myself up with the prototype, I sent a copy of the game to a few friends and some glaring issues came up. Initial play-testers discovered a bit of an exploit.

The "Slip Lane" Problem

As playtesters got deeper and deeper into the game an optimum strategy was born colloquially named the "Slip Lane." Players avoided driving trucks over a straight path to keep it free as an easy way of running to the exit.

This behaviour trivialized the second half of the game, The planned mad-dash crescendo moments were being reduced to an easy stroll to the exit. As Soren Johnson mentions on his blog...

"Given the opportunity, players will optimize the fun out of a game" - Soren Johnson

I didn't see this as I was designing the game, and picked up on it was because I was doing the exact same thing without realising it. This needed to be addressed because reaching the finish line needed to feel like an achievement.

To correct this, I settled on making the entire map covered in a mud which would slow down the player, then driving a truck over the mud would create a walk-able path that would not slow down the player.

This would surely correct the problem as it would stop players from avoiding driving trucks over areas because they would need walk-able paths. Right? Well, sort of...

Testers wised up to the fact that they could just drive one truck to create a path and then finally dodge that single truck. I was in the same boat again... the slip lane strategy reigned supreme.

Back to the drawing board, New strategy. I implemented a new system which would apply oil slicks across the map. These would return movement speed back to the same as mud.

I would query which tiles were driven over the least and apply them to those tiles. To remove the hazard, they would have to route a new truck over the least travelled tiles. This solved the issue of the slip lane, but this in turn caused a new problem.

Players grew to resent oil slicks, players were running out of moves trying to take out of the way paths to rid themselves of the oil slicks. What I needed was a way to reward players for engaging with them, finally an idea came to mind.

The solution, Oil slicks would give you an additional move.

This solved two issues as more and more oil slicks came on the map. This gave the trucks more and more freedom to make moves and the less driven paths were starting to get more and more trucks on them.

The Final Polish Stretch

A lot of the issues I was running into was simply because players were not understanding how the game worked so I had to add extra GUI and tutorial messages to keep the game flowing.

I made a win state and spent the final day and a half polishing the game. I'm proud of what I came up with and I was ready for the world to see. I published the game four hours before submission closing time and could not wait for the world to see my creation.

I was too tired to stay up however the last 72 hours of development had taken the wind out of my sails, I was too tired to do anything else and my friends in my discord call can attest that I was starting to get more and more loopy.

The Review Period

The next day, I began to start reviewing some of my competitors games. I checked on the status of my game...

One review, Sure, that is disappointing however I am sure more and more will play soon. I experienced a bit of revulsion as I opened the first game, I had to download an executable? As the thought hit me I opened as my eyes and my jaw dropped. My game was exactly the same...

This heavily limited engagement to my game as not only did they have to click into my game to begin with they had to download something to their machine as well. The extra resistance I believe could kill a lot of initial reviewers. Why did I not think of this, the System Administrator side of me was screaming!

Even without a web version, I also should of compressed the file into a 7z or zip. Although small for me, 99 MB could still be rather big of an ask for some users.

With that realisation and the burn out from the last couple of days work I slid back into my chair afraid of what the future would hold. In the end my game would amass only seven reviews over the voting period.

In addition, when searching through the projects and option was present to sort projects by reviews. Following the "Matthew Principle" the review rich just kept getting richer.

I was dejected, Tired and sad with the initial engagement, I decided to lick my wounds instead of trying to figure out how to boost exposure. Looking back I know I made a major error.

My description page was as interesting as tax law, in fact I am pretty sure people would prefer to read through tax law than the page I created. Even while writing this it sits on an uninteresting white page.

Screenshots, Gifs, YouTube Trailers on how the game plays will increase the chances of someone engaging with your players. You need to give them an excuse to join in.

The Results

In the end the results were as follows, As well as links to each of the ranking category...

Seeing the raw score for the majority of my ratings hover around 3.5 stars was a nice treat. The final score was scaled down due to the lack of reviews but on reflection I am glad the seven that did review seemed to like it.

In the end it was a good break from my major project and I don't regret it. The initial constrictions of a small time period really did trigger my ADHD hyperfocus and I really did make a game that I am proud of.

Back to work...
Have a lovely day dearest reader!

-Nathaniel