Jump Jousts Jam

What you need to know before starting jump jousts jam

I signed up for the jump jousts jam because the prompt was simple enough to not cause scope creep. Jump mechanics and vertical combat in a small window. Most people overcomplicate it on day one. They start building a full menu system, a tutorial, particle effects, and then realize at hour thirty-six that they haven't touched the core loop. The core loop is everything here. If the jump feels good and the joust interaction is readable, you're already ahead of seventy percent of entries. The rest is polish.

jump jousts jam how to make something people actually finish

Here's the practical approach. Pick a single mechanic first. Not two. Not three. One. I chose a double-jump with a downward slash attack in a previous jam. That decision took about ten minutes and defined every other system. Everything else became a question of whether it served that one mechanic. Build the jump in a gray box with zero art. Just a square that goes up and comes down. Get the gravity, the jump height, the air control to feel right. For platformer-style jump jousts jam entries, the gravity should be higher than you think. Something around twenty-five to thirty units per second squared usually works. Lower gravity makes everything floaty and kills the feel of landing after a joust exchange. Test it with just a flat floor and a few platforms. Spend two to four hours here depending on your engine familiarity.

Once the jump feels solid, add the joust. This is where most jams fall apart. The joust needs to be readable at a glance. I've seen too many entries where players can't tell if they hit their opponent or got hit themselves. Use clear visual feedback. A flash, a knockback direction, a sound cue. Three signals minimum. If you can only provide one, make it knockback because that communicates both offense and defense simultaneously. One thing nobody warns you about: the collision detection on joust mechanics. Box colliders will fight you. I ran into this exact problem during my last jam entry. The moment I used simple AABB collision for the joust hitboxes, the player would occasionally phase through an opponent's attack zone and the hit would register on the wrong frame. The fix was switching to a one-frame snapshot system. Instead of continuous collision checking, I captured the state at the start of each jump arc and resolved collisions based on that snapshot plus trajectory prediction. It cut the bug rate by about eighty percent and took maybe twenty minutes to implement once I stopped trying to make the physics engine do the work.

Scope management during the jam window

Most jump jousts jam submissions die from feature bloat. You will want to add a health bar. You will want to add power-ups. You will want to add a second character class. Don't. Ship a game with one move set and one enemy type instead of a game with five move sets and no finishing touches. Players judge completion more than content count. Time budget breakdown that actually works for a forty-eight hour jam: twenty percent on the core loop prototype, thirty percent on the Joust mechanic and feedback, twenty percent on level design or a simple arena, fifteen percent on UI and menus, ten percent on audio, and five percent for actual bug fixing. Yeah, five percent. You won't have more. Accept it and write your code so bugs are obvious rather than latent.

The mistake I keep seeing beginners make with jump jousts jam entries is building the game first and testing it last. Reverse that. Playtest your gray box prototype on day one. Not day three. Day one. If the jump doesn't feel fun with square sprites, it won't feel fun with artwork either. Art doesn't fix bad mechanics.

Technical considerations specific to this theme

Camera movement matters more than people expect. When the player launches into the air for a joust, the camera needs to follow without inducing motion sickness. I use a simple lerp-based camera that tracks the player vertically with a slight delay and clamps the horizontal movement. The delay gives a sense of weight to the jump. The clamp prevents the camera from going where it shouldn't. Input buffering is another thing that separates finished entries from abandoned ones. Allow jump input up to 0.15 seconds before landing. Allow attack input up to 0.1 seconds after landing. These small windows make the game feel responsive without requiring split-second timing that frustrates players. I've seen people discard entirely functional jump jousts jam prototypes because the timing felt too punishing. Ten milliseconds of input buffering usually solves it.

Sound design for the jump impact alone can take half a day if you go down the rabbit hole. Use free assets or simple synthesized hits. A short low-frequency thud for landing, a sharper noise for the joust contact. That's two sounds and you're done. Don't spend more than forty-five minutes on audio unless you're specifically good at it.

What to submit and how to present it

Build a web build if your engine supports it. It removes the friction of download for judges and players. If you're using Unity, WebGL export. Godot has decent web support. If you must ship a desktop build, make it under 200MB. Larger files get ignored. The description matters. Write three sentences. What the game is, what the core mechanic is, and one honest note about limitations. "This is a compact jump-combat game made for jump jousts jam. You control a character who defeats opponents mid-air. Only one enemy type is implemented due to time constraints." That kind of thing. Honesty reads better than boasting about scope you barely scratched.

If you're asking where to find the jam itself, check the major game jam platforms. It typically runs on itch.io jam listings or Ludum Dare archives depending on the year and hosting platform. The community page for jump jousts jam entries is usually linked from the main jam announcement. There's no single official download of the game template because it's a theme prompt, not a pre-made engine pack. You build from scratch each time. The biggest pitfall I can think of: don't start your jump jousts jam entry at midnight on Friday expecting to pull an all-nighter. I've done it three times. The fourth time I started at noon on Saturday instead. The entry was cleaner, the code was less buggy, and I actually slept. Cognitive decline after hour thirty-six of a jam is real and it shows in the final product.