GMTK Game Jam 2022 Post Mortem (Patreon)

Hey Patrons!

So, if you somehow missed it… July was GMTK Game Jam month!

Every year I hold a game-making marathon, where people try to design and develop a game in 48 hours. And even though the jam itself takes a weekend, the whole event - with planning, making videos, doing streams, etc - essentially eats up an entire month!

You can see the results in the video above. And play the games for yourself on Itch.

Now you’re probably sick of hearing me say this, but it’s true: this was our biggest jam yet. 

Though, it was very close! It looked like this one was going to be a little smaller than 2021, but just before the jam shut, we hit 22,000 sign-ups…

And more than 6,000 games. That’s a 27% follow-through rate (accounting for people in teams, and those who didn’t finish). The same percentage as last year!

Now, in terms of how the jam was run… well, the GMTK Game Jam is mostly pretty steady at this point, and it doesn’t change a great deal from year to year.

Over the last half-decade (what?!) we’ve figured out rules, timing, and systems that work for the majority of people. And so it’s easy to just re-use them, perhaps with minimal tweaking (like, changing the “fun” rating criteria to “enjoyment”).

But, of the stuff that changed, here’s what went right, and went wrong…

What went right: 50 hour jam

Last year, I talked about how the stress of dealing with late submissions led to panic and anxiety. Some people just get really frustrated and angry if they miss the deadline or find a bug in their game, and end up taking it out on me.

It also leads to loads of extra work - the days after the jam are spent dealing with support issues and back-and-forth emails. It’s really not fun.

So this year… I changed it. I made it abundantly clear that there would be no late submissions, and to account for this, I made a change to the jam. Instead of being 48 hours long, the jam was 50 hours. Those last two were supposed to be dedicated to uploading the game, looking for bugs, and catching last-minute issues.

I think overall this went well - it alleviated some stress on Itch (though, the servers did still go down as people uploaded their games in the last few seconds), there were fewer people who were upset, and it allowed us to tell people who came in late “hey, you had two hours, sorry!”.

I did bottle it at one point. Too many people managed to delete their game or remove the game from the jam. So I let people email me to sort this out. Even this was a tremendous pain, so I’ll avoid that in future.

So I’m pleased with the change, but next year I need to work on the messaging and communication a bit more as there was some some confusion about how it all works.

I think I’ll let the jam clock be 48 hours, but then add 2 more just as it closes. That will really sell that this is bonus time, not game-making time. And it’s always better to suddenly be given more time, as a surprise, rather than less.

What went right: even clearer rules

Last year I put the rules on a website. This was great, as it gave us a one-stop-shop for all the jam info. But because it was outside the jam, a lot of people missed it.

So this year I put them directly on the Itch page. I found a handy bit of HTML that lets you show and hide big blocks of text, which stops the page from being overwhelming, and helps people find the answers they need.

I think this worked well - it felt like more people knew the rules, and we had fewer simple and basic questions.

But I have an idea for how to improve it next year. As the jam was coming to an end, I added a notice section at the top of the jam page - just letting people know that time is almost up.

Next time I’m going to use that notice section throughout the entire jam, and put the stuff people need to know, in that moment. How to submit, while the jam is open. How to know their game is definitely in the jam, as it comes to a close. How the voting period works, in the voting week.

The goal is to train people to know that when they need answers, there’s a place to check!

What went wrong: a tiring theme

This year’s theme was “Roll of the Dice”. I was really curious to see games that played with random chance and probability - but I hoped the theme would also allow for other types of games, like ones about rolling.

Like last year, the theme ended up being quite polarising.

In terms of using the theme as a design starting point, I think it was pretty successful and many people enjoyed the challenge.

Plenty of others found it too difficult, though. We saw a lot of games that were ultimately spoiled by adding randomness - because if you don’t use it carefully and judiciously, you end up with a game that just feels unfair, unpredictable, and tedious.

Perhaps this theme didn’t really suit a 48 hour jam. With only one weekend to make a game, you need to just pick an idea and go - but tricky design challenges like this requires time to think, prototype, playtest, change course, and iterate on ideas. There’s just not enough time for that.

Also, some people found it restrictive, as they felt it forced them to make games about luck and randomness. That’s definitely not the case, however - many of the games in the top 100 (like Berb Hike and Rolling World) aren’t random at all, and just use the shape, design, and look of a die as inspiration!

An alternative thumbnail, featuring DIEmunition

But I think the main issue with the theme became obvious when people started to play, rate, and stream the games: dice fatigue.

Almost every game featured a die in some way - throw a die, be a die, fight a die, yeet a die into a zombie’s face, date a die - and so it was quite easy to get sick and tired of dice-based games.

My evidence for this (other than my own dice fatigue while streaming) is the number of ratings. it’s the only number that went down from last year - from 140,000 to 130,000.

So, next year I’ll be going for a slightly more open theme - and no nouns next time! I may also check the theme with some non-jamming helpers, just to get a second opinion.

Some more stats

131,589 ratings were given out during the voting period, and 99.7% of games received at least one vote. Most had many more - the average number of ratings per game was 21.4 and the median was 14.

The most rated game of the jam was Roll N’ Skate with 1029 ratings. It also received the most comments - over 400 of them, at jam close. It didn’t make it into the Top 100, though.

2687 games had the word “Dice” in their title, and 463 had the word “Die”. That’s about 50% of the games! Other popular terms were “roll” (710), “snake eyes” (26), and “loaded dice” (20).

As always, I asked the developers if they made their own art for the game, or used pre-existing assets. 79% of developers made their own artwork. I also asked jammers if they made their own music. 51% of them did.

Also, each year I ask participants to say which engine they used.

The most popular engine, as always, was Unity with more than 61% of the games made in that engine. Godot came next with 16%, then Game Maker with 6%, and Unreal Engine with 5%. 6% of jammers used an engine that wasn’t on the poll, and 3% used no game engine at all!

That’s pretty much the same as previous years, but Godot has increased its share again. From 12% in 2020, to 13% in 2021, and to 16% in 2022. Go Godot!

Final Thoughts

Finally, I just wanted to share some thoughts and feelings on running the jam.

This was the sixth year that I’ve done this thing. And when I started the process of hosting the jam, I was a little bit… demotivated. It felt more like a chore - a big ol’ thing I have to do each year. I started to wonder if the jam should perhaps wind down…

But when the jam actually started, I was reminded why I do this. The enthusiasm and the excitement from all involved. The sheer passion and creativity on show from the games. The hype in the Discord. The love shown to creators in the comments. People really enjoy the jam - and once it gets going, I do too! (And it was even better this year because the post-jam experience was not ruined by support issues!)

So, it makes me really proud to be able to bring this super fun annual experience to the game dev scene, and I vow to continue doing it until, at the very least, our 10 year birthday.

I can’t do this jam alone, of course. So thank you to… everyone who signed up, submitted a game, or rated games. My Discord mods (Slam, Adam, Sandra, Amit, Hobusu, Teslobo, Lawn, Merus, theChief, Brioche, Walrus). Everyone who worked on the Team Finder website (Dotwo, Meanwhile, Slam, Alexandre, Brysen, Glorfindel). The people who run Itch, especially creator Leaf.

And, of course, all of you Patrons! GMTK Game Jam has no corporate sponsors, so this is possible thanks to your donations. I haven’t forgotten about my offer to play your jam games - that video will be coming soon!

Back to normal videos now. Speak soon.

Mark