Juice is how a hit explains itself
8 min read
By Eduardo Orellana ·
Two games can share a move and feel nothing alike. The difference is the instant of feedback that tells your hands the move happened.
The move is not the feeling
Some games feel alive the moment you move, and others feel dead before they have shown you a level. The dead one can be competent. A run-and-gun that registers a shot only as a number going down has, technically, shot something. Your hands do not believe it.
Juice is the explanation the hit gives your senses. A freeze of a few frames, a sound, a spray, a camera that nods. None of that is the design of the weapon. All of it is how the weapon becomes real. Overdo it and the screen is a cartoon of impacts with no readable combat underneath. Underdo it and a good combat system plays like a spreadsheet. The test is rude. Watch the player hands, not the health bar. If they do not flinch when they land the hit, the hit did not land for them.
More on this: Godot FPS templates, and why the feel is most of the work
Two games, one obvious difference
Some games feel electric the moment you move, and others feel dead, long before either has shown you a story or a level. Random Heroes is the flat example: a competent run-and-gun platformer that does very little in motion.
Put Super Time Force next to it and you can tell them apart instantly. It snaps and crackles, like a current running up through the controller. Neither game has told you anything about plot or progression yet, and one of them is already exciting.
That gap has a name, roughly: game feel. It is abstract, invisible on a screenshot, and central to action games and platformers. Players register it within seconds of pushing a stick and pressing a button.
Feel is decided in the basic action
Game feel lives in the simplest verbs a game gives you. It runs underneath everything else, in the second-to-second business of moving, jumping, shooting, landing, colliding, braking, turning.
To find it, take the game apart. Delete the score, the story, the art, the soundtrack, the smart level design. What is left is raw interaction, and the question is whether that is still nice to touch with nothing else to do.
The Mario games survive that treatment. Controlling Mario in Super Mario 64 is enjoyable on its own terms, and players will happily bounce around an empty room for a long stretch. The springy jump, wall kicks, triple jump, long jump and dive are fun before any level asks you to string them together.
Movement came first in Super Mario 64
Mario 64 was reportedly built in roughly that order. For the first months of development it was said to sit close to the blank-room state while Shigeru Miyamoto worked on how Mario moved. Given what the game turns out to be, that time is not indulgent. Movement is the product.
The feel sits in Mario's friction, momentum and weight. Everything else is service: the enemies and the level design exist so movement has somewhere to go, first letting you show off the move set, then asking you to master it.
Other platformers found the same thing. Super Meat Boy is satisfying at a purely kinetic level, before you care about any particular room, while plenty of others are dragged down by loose or stodgy controls no amount of design can rescue. Action games work the same way. The strong ones would still be worth playing if you were only shooting things in an empty box; the weak ones have no charge at all.
There is no recipe for the fundamentals
Nobody can hand you a universal formula for the core action, because the right feel is specific to the game you are making. "Make it feel like Mario" is true and close to useless as instruction.
What does generalize is the layer above it: a long list of small tricks that make a game noticeably better to play. Most of them get added late, during polish, and they usually travel under the name juice.
Juice will not save a core action that is wrong. What it does is take a good action and make it legible, physical, and worth repeating. It is the difference between registering an input and staging an event.
Shake the screen so impact is visible
Screen shake is the most familiar trick. Vlambeer is the studio most associated with it, and its games wobble on almost anything worth noticing: a gun going off, an enemy taking a hit, something ridiculous happening to a fish.
It works as feedback. You touched the world and the world answered back. A shot stops being a number coming off an enemy and becomes a disturbance that runs through everything you can see.
The trick is deciding what deserves it. Applied well, shake tells you which contact mattered. Applied to everything, it is noise, and the signal disappears. The goal was never to make the screen move.
More on this: Make an FPS: the weapon first, then everything else
Freeze the frame when something connects
A second trick is to stop the game dead for a fraction of a second when a hit or a kill lands. That tiny judder makes contact feel heavier than the animation on its own ever would.
Fighting games run on it. Street Fighter arrests the action briefly so punches and kicks land with force. God of War and many of the Zelda games use related pauses to put extra weight behind an attack.
Plenty of other signals do the same job. Flash the enemy white, knock it backward, spray blood, switch its animation, play a sound with some body to it. Each says the same thing: what you just did registered.
More on this: Make a fighting game with two characters before you make twenty
Tie the character to the world around them
Feedback is not only something you do to enemies. The character you control has to read as part of the same world. Small puffs of dust on landing, recoil through a gun, little physical reactions on contact all bind the avatar to the ground, the weapon, and the space around it.
Sound carries a lot of that. A gun that sounds thin and polite makes the shooting feel thin and polite. Give it bass and volume that matches the fantasy, and the same action starts to feel powerful.
Repetition is the enemy, so vary what comes back. Randomize sound effects and the hundredth one stops grating. Raise the pitch across a run of hits and repetition turns into a rising thrill instead of sameness. Mario and Peggle both climb the scale for exactly that reason.
The camera is doing feel work too
How the camera behaves is part of the feel, not a neutral window onto it. In Luftrausers it does not just track the plane. It frames the fight and pulls nearby threats into view, so speed stays readable and you never feel pinned to the center of the screen.
Hotline Miami takes another approach, letting the view push out ahead of where you are looking. That tells you where you are pointed and what is waiting there. In both cases the camera is deciding what you understand about the next second of play.
Match the camera to whatever the game cares about. A game about jumping needs a camera steady enough that a landing feels like your fault or your credit. A game about fast shooting can let the camera take the recoil, catch the threats, and sell the hit.
Oversize the things that matter
Scale is the bluntest source of energy available. Nuclear Throne makes the bullets enormous. Super Time Force makes explosions look like small atom bombs. Hotline Miami turns a blood spray into a geyser of red goop.
None of this is aimed at realism. It is aimed at readability and force. A big effect tells you instantly what just happened and how much of it you caused.
Hotline Miami adds something most action games throw away: permanence. Bodies and blood stay where they fell, so retracing your route walks you back through the mess you made. The fight reads as hard won because the room keeps the evidence.
Point every effect at the same verb
Polish is only worth anything when it reinforces what the game already is. Juice means doubling down on the thing at the center, not decorating around it.
A game about shooting wants guns that kick, a sharp fire rate, loud impacts, and a camera that shudders on every trigger pull. A game about jumping wants tuned friction, believable weight, landing feedback, and a camera calm enough to trust with a leap.
None of this arrives quickly. Feel can absorb months of tuning, and no single trick makes a game satisfying by itself. Nor can you settle it by reasoning about it: you get the build running, play it, and decide whether it feels right, because your own hands are the only real test. The principle is short enough. Get the fundamental action right, then make every piece of polish speak to it. When everything points at the same verb, the game comes alive.
Build it in Flockbay
In the Flockbay app, build one attack with no feedback and the same attack with a freeze, a sound, and a visible result. Play them back to back. Keep the quieter of the two only if you can still feel it. Add feedback until you can, then stop before the screen becomes the joke.
An AI platformer maker and a 2D platformer template make a single hit easy to repeat. One enemy is enough.
Keep reading
- YOU ARE THE MONSTER is built on one verb and refuses to add a second
- Cartel Pilots Wanted makes the cargo, not the flying, the co-op problem
- The jump is the whole contract
- The camera serves two masters
- The pad decides which games exist
- Keep It Up! makes not dropping the balloon the whole game
- Kick makes getting to school one long soccer trick
- Salt & Soul makes a salt shaker the whole kitchen
- MECCHA CHAMELEON makes your painting skill the stealth stat
- How to Fish is a fishing game where the physics is the joke
- Make the core action feel alive first, in a platformer you make with AI
Feel comes from the core action first. Every sound, pause, shake, camera move, and effect after that should be making that one action clearer and better to repeat in your own game.
More notes on how games play