GMTK Jam 2021 Post Mortem

The Jam Overview

This is my second time entering the GMTK Game Jam. It's a 48 hour jam hosted by the GameMakersToolkit YouTube channel. Mark from GMTK plays the top 100 ranked games and settles on his own top 20 as the winners to be featured on a video on his channel.

The jam sets records each year for largest online jam and for good reason. It's very well run and typically has good theme choices. This year's theme was Joined Together. I am happy to have this as the "big" jam I join this year and set aside time for as if it were a vacation.

My Game Two Henries

This year I set out to have some collaborators from the start. I spoke with my friend LobsterBlues (an awesome illustrator) and managed to steal some of his time on Saturday to make some art for the game. He was busy for much of the jam but having a real artist really helped the game to stand out and have a ton of personality and charm that it otherwise would not have had. It also freed me up to spend more time on mechanics and notably level design that would have otherwise been spent pixeling. We also managed to snag a quick last minute tune from ScottyRich and one from Spaceboy. They were both awesome but we ended up using the one from ScottyRich as it fit our general feel a bit better.

Two Henries is a physics driven puzzle platformer where you control one of the henries to platform around and swing the other Henry, who is essentially a wrecking ball hanging from you, into or around obstacles to reach the end of each level. The idea was not incredibly unique. I think it was more unique in the way it was used, but there were many "wrecking ball" or "two characters joined by a rope" type games. The thing that set us apart in my opinion was really the art and charm. LobsterBlues really did a great job giving the Henries a real feeling in very few frames. The Henry you control has a whopping 4 frames while the wrecking ball has 3. We used some squash/stretch magic to make it feel a bit more alive than that but I'm very happy with the general look and feel of the game. It's even more incredible given the art was mostly made in a period of a few hours on Saturday.

Approach to Jamming

The jam started at 2pm my time on Friday. I worked until 5pm so the 3 hours between were spent half working, half brainstorming. Lobster was able to help me brainstorm through some ideas. We had ideas about a car game driven entirely by grappling hook, talked about doing this from the top down or more of a side view like a platformer. I think these were good ideas but I was a bit more concerned about our ability to get a car game done as that was not a genre I had ever touched before. Also because Lobster's time was limited and we needed to be able to get most of the game's art done in just a few hours. We ended up settling on one of our first ideas, which was a platformer where one of the characters was stuck swinging from the other. The actual initial idea was for the main character to be a mech but for its pilot to be swinging helplessly from a rope controlling it rather than being nested behind its armor.

I was lucky enough to have a project with some general setup for platforming so it didn't take too long to get a guy moving with the other character swinging below. After some more talk about the idea, playing with the prototype and the mech making it more of a shooter, we decided to change the way we would apply the idea of this character hanging from below. Rather than the main character being a mech blasting away enemies trying to protect this dangling pilot, we had more fun swinging the "pilot" around into stuff. Lobster also was intent on not using violence as the easy "verb" of the game. So given these two things we pretty quickly pivoted to a puzzley level based platformer where the main thing you did was intentionally swing that lower guy around. The rest of the day Friday was just playing in a test room with different ways to use the swinging guy as I needed to have most of the "items" laid out for Lobster's free time the next day to art this stuff up.

Saturday morning was mostly spent polishing, implementing art and sfx generally for the game. We had basically no real levels at this point. Just three test rooms connected by some logic I had whipped up to make actual level progression work. I also had been periodically sharing the web build of the game for some playtesting of the main mechanics.

Getting the art in really brought the game to life. It stopped being a fun test room prototype and starting having its own character and charm. This was incredibly motivating and led me to jam through about 15 levels over the rest of Saturday. A bunch of them needed some reworking and a few more mechanics were thrown in (most notably the buttons and the platforms driven by interaction with the buttons). But by the end of Saturday the game was basically complete and playable.

Sunday was spent reworking some mechanics/levels, adjusting volumes, working on the menu and sliding the tunes in at the last moment. I'm pretty happy with how the game turned out even if I'm not sure it will place well. We made the game we intended to make and executed on it pretty well. I didn't work past 11pm either night and didn't get back to working until around 10am the following day, zero feeling of crunch. The level design could still use some work and there were definitely new things learned but overall very pleased with how the game turned out.

Lessons Learned From Last Time

In the post mortem for my game last year I broke down a few things that went wrong. I wanted to right these things this time around and mostly did a good job on that. I had set aside the weekend long in advance and worked out of my actual work office (no one else was in on a weekend) to avoid random minutia of home life stuff. I roped Lobster in pre-jam to do art and worked the game schedule around his free time to get his great artstyle in. I worked in Unity for the first time in a very long time in order to get a web build out instead of the download I was getting with my Monogame based engine I would normally use. My suspicions about a web game getting way more play than a download were spot on. More plays in the first 12 hours of Two Henries than the entire week of Speclucky. I also did not feel like I had to push the game to the smaller streamers just to get people to play it. That just happened naturally!

What Went Right

Project Organization

No crunch outside getting the music in late & trying to get that build out with music included last minute. It wasn't overly stressful though as I had been pushing builds periodically and had a playable product from night 1.

General Gamefeel

The game mostly just feels good to play. There is some amount of wonkiness because of the physics driven nature but that actually contributes to the charm and feel, I think, more than it takes from it. The only gamefeel harming bug I had was a double jump caused by some bad coyote time code. Would be a one line fix but found it only just after the jam ended. Doesn't actually hurt gameplay but lets you break a few puzzles. Overall not a big deal.


I had not used Unity in forever but now am a convert after using XNA/Monogame for about 10 years now. The 2D workflow in Unity is waaaay better than it was last time I tried it and the inspector only bothered me mildly. The benefit of putting things together, ease of build/setup, numerous packages/tutorials, and general niceness of the system vastly outweighed the two crashes and slowish compile times. It helped that my Monogame engine setup is somewhat similar to Unity's general structure with the ECS. I only missed a few functions that I was used to having and because they both use C# it was easy enough to bring that stuff over. I will likely make my own utility package to help with some of the math functions and whatnot that I was missing and make Unity my main go to moving forward, particularly for jams.


I may also be a convert from my typical very low res pixel art style to using these large "illustration" style images and just manipulating them with some squash/stretch or shaders to give them more of an in motion effect. Previously I always went low res so that I could actually make decent animation in a short period but the wobbly squash/stretch and just a few frames worked so beautifully and gave the game a style that was not seen much in the jam.

What Went Wrong

The Concept Was Not Unique

One of the 3 things you are rated on is Originality and while the game was well executed and fit thematically in the jam, I don't think it caught people by surprise with the "what a cool idea" the same way Speclucky's tweet based level seed had. Next time around I hope to be able to combine the execution of this game with the conceptual uniqueness of Speclucky. I have a sneaking suspicion this will be our biggest issue when it comes to jam placement and making that top 100. We will see when the ratings come out I suppose.


As incredible as the art ended up there were a few mechanics that just didn't get Lobster art. The most notable was the button that my dumb self could not even get to look remotely decent. I wish I had been able to steal just a bit more of Lobster's time. Next time I will try to get him (or whichever artist) much earlier in the process and hope they have just a touch more free time to push more art into the game. It did restrict a bit on my desire to create new mechanics once I knew we likely would not be getting actual art for them as I did not want the new items to stand out as "ugly" or not fitting with the rest of the game. I want to clarify this is no slight to Lobster as he gave his free time on a weekend he was spending time with his family to give some art to the game, just selfishly want more of his time (sorry Lobster family in advance).

Wrapping Up

I am writing this before the ratings come out. I had been anxiously awaiting the final ratings but am feeling much less pressure about this as I write it now. Definitely curious how the ratings will land, but feeling no pressure about it at this point. Very happy with the game we made and even more happy with the lessons learned and with how well the overall process went. Looking forward to applying all these lessons for GMTK 2022!

My Winners So Far

Again this is pre-ratings so I may have missed some incredible games. A few stood out to me of the about 50 games I rated however:

  • Eject Bombing was my absolute favorite with incredible execution on a pretty good idea

  • One Gets It All was a very fun game though the art and presentation were somewhat lacking. Just a super fun concept worth playing.

  • Seamless is a great concept with a very very cool visual style to match. I wish the platforming felt a bit less floaty and the levels had a bit more to them but the presentation was so top notch it had to be included.

  • Slimes in Space is another game with unreal presentation/gamefeel for a 48 hour project. The game itself is fun as well and has a shocking number of levels especially considering the polish in a 48 hour period.

  • Glob is a game about making pictures out of letters. The cool thing about it is that you can make your own picture and upload it to the server for others to play. Another cool idea that is just very well executed. Its less of a game and more of a toy I suppose but its really cool either way.

Andrew Napierkowski

Professional Software Engineer | Hobbyist Game Developer