Juice: The Small Animations That Make a Game Feel Good

Take a well-made match-three game and strip out everything not strictly necessary. Tiles still swap, matches still clear, the counter still goes up. Every rule survives intact, every input works, and the thing becomes almost unbearable to play.

What was removed was not function. It was the tile lifting slightly before it moves, the cleared pieces scattering rather than vanishing, the counter overshooting and settling back, the board nudging when something big happens. None of that changes the state of the game. All of it changes whether anyone wants to keep touching the screen.

Game developers have called this juice for about twenty years, and the word stuck because no more dignified term captures it. It is the layer of motion sitting on top of a system that already works — and in a format where a stranger decides within seconds whether this is worth their attention, it is not a finishing touch. It is a substantial part of what the ad is selling.

These get conflated constantly, and keeping them apart is the thing that makes juice possible to reason about rather than just pile on.

Feedback is informational. Its job is to tell the player that an input was received and what it did. A press state, a highlight, a piece snapping to a slot — each of those answers a question the player was actually asking. Remove it and they are genuinely confused, because a fact they needed has been withheld.

Juice answers no question. When a cleared tile bursts into fragments that arc outward and fade, the player learns nothing they did not already know a frame earlier. The tile was already gone. The burst exists to make its going feel like something.

That sounds like a case for cutting it, and it is the reason juice is always the first casualty of a size budget. But it inverts under a different question. Feedback determines whether the ad is *usable*. Juice determines whether it is *enjoyable* — and enjoyment is not a bonus in an ad a person did not ask to see. It is the entire basis on which they keep watching instead of hunting for the close button.

Juice looks like a grab bag of tricks until you notice that nearly all of it is doing one thing: pretending the objects on screen have mass.

Nothing on a screen has weight. A rectangle that moves from one place to another at a constant speed and stops dead is what a computer does naturally, and it looks exactly like what it is. Every technique below is a way of faking the behaviour of a physical object, and they come more or less directly from traditional character animation, where the same problem was solved decades earlier with pencils.

That is also why they compose well rather than competing. They are not separate effects; they are parts of one imitation.

It is worth taking a single interaction apart frame by frame, because juice is invisible at normal speed and that is exactly the point. Here is a player tapping a block that clears — roughly half a second of screen time, with each layer labelled by what it is doing.

Note what is not in that list: a pause. Nothing in the sequence blocks the next input. The player can tap again at 100 ms and the new interaction begins while the previous one is still resolving, layered on top.

That is a hard constraint rather than a nicety. Juice that gates the next action has stopped being juice and become a loading screen with better art, and in a format measured in seconds, a half-second of enforced watching is expensive. Effects must be interruptible, overlappable, and entirely optional to the flow of play.

Most juice budgets go to success, which is intuitive and slightly backwards. A player who succeeds is already having a good time; the animation decorates a moment that was going to feel fine regardless. A player who fails is at the point where they decide whether to try again or close the ad, and that decision is made on feel rather than reasoning.

This is the mechanism behind a pattern anyone who plays games recognises: some games are enjoyable to lose at. The block that does not fit tilts, teeters and topples with visible physics. The stack collapses elaborately, taking a satisfying second to fall. None of that softens the failure — it is still a failure — but it was interesting to watch, and interesting-to-watch is what converts a loss into another attempt.

The inverse is grim and common. A player misses, the piece vanishes, a red flash fires, the board resets instantly. Every fact has been communicated accurately, and the experience is of being told off. In an ad nobody asked for, being told off is the last beat before the close button.

Juice is unusual among design principles in that applying it enthusiastically fails about as badly as ignoring it, and does so more often in work by people who have just read about it.

The board is information the player is reading in order to decide what to do next, and every effect occupying the screen competes with that reading. One tile bursting is a celebration; six tiles bursting while the view shakes, three counters punch and a banner flies in is a screen where nothing can be located. The player is not delighted — they are waiting for it to end so they can see the board again.

One test resolves most of these arguments. Add the effect, then ask whether a first-time player can still tell, at any moment during it, what the state of the board is and what they can do. If not, the effect is not too flashy in the abstract; it is hiding the game, which is the one thing it can never be worth.

The second limit is budget. The good news is that the expensive-looking techniques are mostly not the expensive ones: squash, overshoot, stagger and scale punch are transformations of assets already in the file and cost almost nothing. Elaborate particle systems consume the budget, and they are what you can most afford to lose.

A playable ad makes an implicit claim: this is what the product is like. The mechanic carries most of that claim, but feel carries a surprising amount of it, because feel is the one thing a player can evaluate in two seconds without knowing anything about the game.

So an under-juiced ad for a well-made game is not a neutral omission. It is an inaccurate description of the product — it says adequate about something better than that, in the first two seconds, to somebody deciding whether to continue. That is the defensible answer to why any of this deserves budget. It is not decoration competing with the mechanic for space; it is the part of the ad showing that the thing on the other end of the install is worth the download.