Start with AI, then read the rest
The fastest way from “I have no idea what to make” to building is to shape your idea with an AI first, using a guide built for exactly that. It knows the theme, knows what makes a jam game work, and pushes back on a weak idea instead of just agreeing.
Building in MSU Space? It’s already there.
You don’t have to set anything up. The guide is built into MSU Space as a skill, so you just start talking.
- To brainstorm a game, say something like “What kind of game should I make for Vibe Camp Season 2?” and it starts shaping ideas with you right away.
- To get the jam details (dates, rules, prizes), say something like “Tell me about Vibe Camp Season 2.”
Building with your own AI tool? Give it these two first.
Same content the MSU Space skills run on, as two links you hand to your AI before you start.
- Vibe Camp Season 2 Persona is the brainstorming guide. Start your session with this and ask it to help you shape a game.
- Vibe Camp Season 2 Information has the jam facts: dates, rules, prizes, and links.
It isn’t one click, and that’s the point
Brainstorming with an AI gets you started. It doesn’t get you finished.
If you just ask it to recommend a game, you’ll get something perfectly reasonable, and so will everyone else who asks the same way. AI converges on the average, and average doesn’t win anything. A game that reads like the first thing the AI suggested is going to look like a dozen other entries.
So treat the first answer as a starting point, not an answer. Push back on it. Tell it what you actually find funny, which MSU character or moment you’d want to build around, what kind of game you personally can’t put down. Ask for five wilder variations and throw four away. Take one branch and go deeper. That back and forth is where your game stops being generic and starts being yours.
The AI can build fast. What it can’t do is decide what’s worth building. That part is still on you.
The rest of this page is the thinking behind the theme, written to be read. If you want to understand why Short, Simple, Tense actually works, so you can steer the AI with intent instead of guessing, keep going.
Here’s a question worth sitting with before you build anything. Why can a game with one button and no story keep you replaying for twenty minutes, while a game with ten times the content gets closed after a single run?
It isn’t the amount of stuff in it. It’s tension.
Season 2’s theme, Short, Simple, Tense, looks like three separate asks, but it’s really one design idea wearing three words.
This post is about that idea: what actually makes these games grip you, and how to build it on purpose instead of hoping you stumble into it. It’s what tends to work in this style, so take what’s useful and argue with the rest.
TL;DR
- Fun in this theme is tension. It all runs on one loop: see the move, commit, keep a little or lose it all, and you’re instantly back in.
- Short. One or two minutes a run, instant restart, nothing between attempts. Cheap failure is what lets players risk a hard move.
- Simple. Understood in three seconds, no tutorial. Build on instincts players already have. One or two inputs, but the depth lives inside that one input.
- Tense (the part most entries get wrong). The move has to be hard to pull off, not hard to figure out. And one slip has to cost the whole run, right then, not a little worse later. Keep it fair, never lag or luck.
- What to avoid. Difficulty that’s all thinking, a loop with nothing at risk, a commitment with no on-the-spot cost, or a game that plays itself while you wait.
- Make it feel good. Instant, punchy feedback is a real part of the score. Ask the AI for the juice and tweak until it lands.
- Test by playing. Finish a full run yourself, then watch one person play with no instructions. If they get it fast and say “one more,” it’s working.
The rest of this post is just the why behind each of these.
First, what “fun” even means here
Every game in this theme runs on the same tiny loop, whether it’s *Flappy Bird, Crossy Road, or Getting Over It.*
You see the move → you commit → it plays out for real → you keep a little or lose it all → you’re instantly back in.
That loop is the whole engine, and three things make it feel good.
- You’re in control, and the control has weight. The satisfaction comes from you pulling the move off. Not watching it happen, not picking the obviously-correct option, but landing something your hands had to get good at.
- The game answers you instantly. Every tap lands with a punch, every crash you can almost feel. When the game reacts in a blink, your brain reads the action as real and rewarding.
- There’s real risk, and it resolves right now. The move can fail, and when it does, you feel it immediately. No consequence-arrives-quietly-later. Right then.
Get those three working together and you get what every good jam game has, that “one more run” pull. Lose any one of them and the loop goes flat. The rest of this post is really just how to protect each one.
Short is about respecting the player’s minute
Aim for a minute or two per attempt, and make the restart instant. One button to jump back in, or no button at all.
This isn’t only about attention spans. Short runs are what make the whole loop possible. When a failure costs you ten seconds instead of ten minutes, you’re happy to risk everything on a hard move, and that willingness to gamble is where the tension lives. Long runs make players cautious, and cautious is the opposite of what you want.
So cut anything between attempts. No menus, no “are you sure,” no loading. The faster a player can fail and be back in, the more “one more” you’ll pull out of them.
Simple is understandable in three seconds, not shallow
A new player should get how your game works in about three seconds, with no tutorial and no wall of text. That’s the bar.
The way you hit it is not by explaining well. It’s by needing no explanation.
- Build on what people already know. Real-world physics is the biggest shortcut there is. Nobody needs to be told that things fall, roll, bounce, and slide. Drop a slime and it rolls, pull something back and it launches. When your rules match a player’s instincts, the game explains itself.
- Let the game teach by playing, not by telling. Don’t open with instructions. Drop the player straight into an easy first moment where the one thing they can do is obvious, and let them learn it by doing it. Trust their instincts. People figure out far more than tutorials give them credit for.
- One or two inputs, no more. One tap, one drag, one held button. That constraint is a feature, not a limitation.
Here’s the part people miss. Simple to pick up is not the same as easy to master. All the depth lives inside that one input. The timing of the tap, the angle of the launch, how much power you put behind it, that’s what players spend the entire game getting good at. Shallow is when the input can’t really fail. Simple-but-deep is when the input is easy to understand and hard to perfect.
Tense is the whole game, so get this one right
This is the heart of the theme and the part most entries get wrong. Tense has two requirements, and it needs both.
1. The challenge is in doing the move, not in figuring out what to do.
The player should always be able to see what needs doing. The hard part is pulling it off. A character with real weight and momentum that keeps getting away from you. A single tap that has to land at exactly the right instant. Aiming, timing, judging the power, reacting in the moment. The second you commit, it plays out for real and you live with it.
When the thinking is the hard part instead, tension breaks. Puzzles don’t get tense under a clock, they just get stressful, because a timer on a thinking problem only makes you scramble to think faster. That’s not the feeling you’re after.
2. One wrong move takes back what you built, right then.
This is the single most important sentence in this post. A miss has to cost the run itself, immediately. Off the edge. Back to the start. The height you’d climbed, gone. Not a little worse a few moves later. Gone, now.
And the stakes should grow the further in you get. Start gentle, then let every slip cost more the deeper you are, because the fear of losing a long run is exactly what makes your palms sweat near the end. Getting Over It is the pure version of this: one careless swing and everything you climbed drops all the way to the bottom.
One caution on fairness. The difficulty has to live in the doing, never in lag or luck. Every death should feel like yours to fix. The moment a player blames the game instead of themselves, they quit. Hard is fun. Unfair is the exit button.
One useful trick when you’re building with AI. Don’t ask for “harder.” Tell it exactly what a single slip costs. “If they miss the timing, the run resets.” “If they overshoot, they fall and lose all their height.” Describing the cost gets you tension. Asking for “harder” usually just gets you faster, and speed alone isn’t tension.
What quietly kills tension
It helps to see the failure modes directly, because most of them look fine until you play them.
- All the difficulty is in the thinking. 2048, Sudoku, match-three. Once you’ve spotted the move, making it is a swipe that can’t fail, so your hands never get good at anything.
- Nothing is ever at risk. Whack-a-mole is one note. React, hit, score, repeat, with nothing to lose. When there’s nothing to lose, the only knob left to turn is raw speed, and speed by itself is not tension.
- The commitment is free and the cost is late. Suika Game lets you aim the drop as long as you like, and the drop itself can’t miss, so your hands are never asked to do anything hard. And a bad drop doesn’t cost you on the spot, it costs you a little room a dozen moves later. The pressure seeps in instead of landing in one moment.
- The game plays itself while you wait. Idle games keep going without you, and the moments that want your hands are islands in between. If a player can look away and nothing is lost, there’s no tension to lose.
Two things that feel like tension but aren’t: weight and momentum by themselves, and reacting fast on its own. You need the weight and the on-the-spot cost. Either one alone falls flat.
Making it feel good is a real part of the score
You can nail the loop and still have it feel dead, and that difference is what the judges call Quality, or “juice.” It’s the instant, satisfying feedback that makes a simple action land hard.
- A tap that hits with a visible, audible punch.
- A crash, a fall, a fail that you can almost feel through the screen. Screen shake, a sound, a flash.
- Controls that respond the instant you touch them, with zero delay.
- Visuals clean enough to read at a glance, so the player always knows what’s happening.
The good news is you don’t build any of this by hand. You ask the AI for it and tweak until it feels right. “Add a screen shake and a heavy thud when they crash.” “Make the launch feel snappier.” Iterate on feel the same way you iterate on rules.
Since your game is built from MSU resources anyway, use them for the juice. A monster that squashes on impact, a familiar effect on a good hit, an MSU sound on the fail. The assets are right there.
Optional: a serious challenge in a ridiculous world
You don’t need this, but it’s one of the most shareable things you can make. A game’s challenge can be dead serious while everything around it is completely goofy, and that contrast is a joy on its own. Getting Over It is a shirtless man in a pot. QWOP is just trying to run. Goat Simulator and I Am Bread lean all the way into the absurd.
Their secret is that the jank is on purpose. The game underneath actually works and the challenge is real. The chaos is a coat of paint, not a broken game hoping it loops back around to funny. A wild ragdoll flop turns a loss into a laugh, and a laugh into “one more.”
MSU is full of raw material for this, from its characters to its memes. If that’s the game you make and you commit to it fully, that’s exactly what the Golden Jank award is looking for.
Test it the only way that matters: play it
Before you publish, play a full run yourself, start to finish. If you can make it all the way through, it’s ready, and anything that trips you up is just one more thing to tweak.
Then, if you can, watch one other person play with no instructions from you. This tells you two things fast. If they understand it in a few seconds, your Simple is working. If they say “one more” without being asked, your Tense is working. If they look confused or bored, you’ve found exactly what to fix, and it’s better to find it now than at judging.
That’s the idea
Short, Simple, Tense isn’t three boxes to check. It’s one feeling, the “I can see exactly what to do and doing it is killing me and I cannot stop trying,” built out of a tight loop, a single deep input, and a cost that lands the instant you slip.
This isn’t the whole map, just what tends to work in this style, so build one, play it, and trust what the playing tells you over anything written here. Share your progress in Showcase as you go.
Now go make something you can’t put down.