Shopping Cart Theory

tim-cain narrative-design morality npc-design player-behavior quest-design fallout

Abstract

Problem: How should narrative designers model morality and moral choices in games, when real-world human behavior shows that many people only act "good" when rewarded or punished?

Approach: Tim Cain uses Shopping Cart Theory — the observation that returning a shopping cart is a small, anonymous good deed with no reward and no punishment for skipping it — as a lens for understanding human morality and its implications for game design.

Findings: People split into those who do the right thing intrinsically and those who only act well when incentivized. This maps directly to NPC writing (some NPCs are just good; some are just bad without elaborate justification) and quest design (occasionally withholding rewards for good choices reveals player nature).

Key insight: Occasionally make the good path unrewarded or less rewarded than the bad path — then call players out for their choices at the end. Shopping Cart Theory predicts many will choose badly, and that tension is powerful narrative design.

Source: https://www.youtube.com/watch?v=PfiUhrG86Es

The Theory

Shopping Cart Theory is a simple philosophical thought experiment. After shopping, you can return your cart to the corral or leave it in the parking lot. Returning it is:

  • Not legally required — no fine, no arrest
  • Minimal effort — corrals are scattered throughout the lot
  • An anonymous good deed — nobody sees you, nobody thanks you
  • Helpful to strangers — the next person gets a tidy parking lot

Yet many people don't return them, and they produce elaborate excuses: rain, kids, sore feet, "the store pays someone for this." Cain notes the irony: these same people had no trouble getting the cart under identical conditions. Reversing the process is somehow impossible.

Rewards and Punishment

Some stores combat this with a quarter-deposit system — you pay to unlock a cart and get the quarter back on return. There's no actual reward (you leave with the same money), yet return rates jump. People respond more to the threat of loss than the absence of gain.

This reveals something fundamental: a subset of people will not do even the smallest good act unless rewarded — or will only act right when punished for not doing so. They may see themselves as good people, but Shopping Cart Theory exposes the gap.

Implications for NPC Design

Cain argues this maps cleanly onto writing NPCs:

  • Genuinely good NPCs don't need a written-in reward for their behavior. Some people just try to do right. It's realistic.
  • Casually bad NPCs don't need elaborate villain motivations. People are mean because they can get away with it. Realistic responses when confronted: "Other people do it," "I don't care," "Someone else will fix it," "My knee hurts."
  • Best villains still believe they're justified — their evil serves a greater good or lesser evil. But rank-and-file bad actors just... act badly.

Implications for Quest Design

The most interesting application is quest design:

  • Occasionally offer no reward for the good path — or a better reward for the bad path
  • Watch what players do — Shopping Cart Theory predicts many will choose badly and then rationalize it online
  • Call them out at the end — Cain loves end slides that confront players with consequences

The Fallout Example

In the original Fallout, the fastest way to solve the water chip quest is to take it from Necropolis and leave without fixing their water pump. The ghouls warn you what will happen, but you get your reward faster by ignoring them. At the end of the game, the slides reveal the ghouls all died — and you're called out for it.

This is Shopping Cart Theory in its purest game design form: make the selfish path easy and rewarded, then hold up a mirror.

References