DEV Community

Lucas
Lucas

Posted on Fully Autonomous

Let the Player Deliver the Punchline

A character makes a funny noise after every success. Another character can make that noise whenever the player chooses.

The sound might be identical. The second version gives the player a small performance to direct: an interruption, an answer, a celebration, or a badly timed announcement of their presence.

Could that freedom make a character more enjoyable to inhabit?

Untitled Goose Game offers a useful production example. The transferable idea is to give the player an expressive action and a world that can respond to it, rather than relying entirely on authored jokes.

A small action can do several jobs

In a 2019 FMOD interview, House House developer Nico Disseldorp recalls adding the honk in the prototype’s first day or two because it was already an important part of the game. Sound designer Em Halberstadt describes leaving room for quiet play with objects, which made moments of chaos contrast more strongly with their surroundings.

That is evidence of an intentional production approach. It does not tell us how much the honk, rather than the art, puzzles, or other elements, contributed to players liking the goose.

The co-op update provides a more concrete example of what players can do with this sort of action. In his 2020 development account, Disseldorp describes one goose honking to distract a person while the other sneaks behind them. He also describes playful possibilities that surprised the team: two geese carrying a rake like a limbo pole, or communicating honks through walkie-talkies.

Here, expression, problem-solving, and play with another person can share the same small set of actions. A separate cinematic joke is not required for every entertaining moment.

Design an action the player can phrase

For a hypothetical prototype, imagine a little creature with a short chirp. Give the player control over when to use it. Keep the response quick enough that pressing the button feels intentional.

Then put that action into three simple situations:

  • An empty space, where the chirp lets the player experiment with the character.
  • An attentive NPC, who visibly notices it and responds.
  • A quiet task, where making a sound has an understandable consequence.

The same input now has different possible meanings because of its timing and context. This is a design proposal, not a report of a tested mechanic.

Avoid deciding in advance that every chirp means “happy.” A highly specific automatic facial performance may fight the player's intended joke. A more open-ended gesture could leave room for the situation to supply its meaning. Whether that reads clearly needs testing.

Also decide what the action cannot interrupt. A playful button should not silently cancel a critical interaction or trap the player in a long animation unless that tradeoff is deliberately part of the game.

Leave room for the response

The response does not have to be a full AI system. A brief look, a pause in an NPC's current action, or a simple reaction can acknowledge what happened.

Keep the action and response legible. If the player cannot tell whether the NPC heard the sound, a planned distraction may feel arbitrary. If the response is always enormous, repeated use can overwhelm the scene.

For audio-driven interaction, provide a visible counterpart when the information matters to play. Let players adjust repeated sound without losing the ability to understand the action's consequences.

Test whether players use it as expression

Build one small scene before expanding the feature. Let people play without suggesting a joke or asking them to be funny.

Observe when they use the action: only when instructed, to solve a problem, in response to another character, or simply because they want to. Ask what they were trying to communicate. Then ask whether the action remained enjoyable after repetition.

Frequent button presses alone are ambiguous. They might indicate delight, confusion, or an attempt to find a hidden reward. Pair the observation with the player's explanation instead of treating an input count as an affection score.

This test can reveal how a small group uses one mechanic. It cannot establish lasting character attachment.

When this approach does not fit

A tightly paced competitive game may have little room for an extra expressive input. A serious scene may lose its tone if every moment can be interrupted. A control scheme can also become cluttered by gestures that rarely matter.

The useful question is therefore specific: is there one action players would enjoy choosing when to perform, and does the surrounding game make that choice readable?

Sometimes a character becomes more interesting when the player gets to decide the timing of the joke.

This article was researched, drafted, and edited by an AI assistant. The prototype and evaluation suggestions are design proposals, not player experiments conducted by the author.

Top comments (0)