Fun-First Design (EAD's Kernel-and-Window-Dressing Method)
Find the fun mechanic first, then build only the minimum around it to bring it to life.
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 85%
When Shigeru Miyamoto was given his own R&D division inside Nintendo, he and his team rejected the era's default approach of designing games around a technical category or clone of an existing hit. Instead, the stated method was to search for an inspiring, replayable core mechanic first, prototype it stripped of narrative or art, confirm it holds up as genuinely fun and repeatable, and only then add the minimum story, characters, and visual world needed to frame that mechanic. Nothing gets added that doesn't serve the fun. This method produced enduring hits like Super Mario Bros. and The Legend of Zelda, games still replayable decades after release.
Origin
Formalized as Nintendo's EAD group manifesto after Miyamoto's Donkey Kong success proved that a non-engineer thinking about narrative and feel, not just circuitry, could outperform veteran engineer-designers.
Core principles
- 01Fun is a testable core mechanic, not a byproduct of genre or technology
- 02Add narrative and art only in service of an already-proven mechanic
- 03Replayability is the real success metric, not initial impressiveness
- 04Trust rare, proven designers with full creative authority rather than committee design
How to run it
- 1
Search for the core mechanic
Prototype small, isolated interactions (jumping, climbing, dodging) purely to find what feels satisfying to repeat, independent of any theme.
- 2
Test for replayability before anything else
Confirm the raw mechanic stays engaging over many repeated plays before investing in art, sound, or story.
- 3
Add the minimum narrative frame
Wrap the proven mechanic in the simplest story that gives players a reason to care, avoiding unnecessary complexity.
- 4
Cut anything that doesn't serve the core loop
Remove features, modes, or story beats that don't reinforce the central fun mechanic, even if they seem impressive.
- 5
Hand creative authority to the designer, not the engineers
Separate the role of 'game designer' from 'engineer' so creative decisions aren't constrained purely by what's technically convenient.
In the wild
Given only Radar Scope's leftover joystick-and-two-buttons hardware, Miyamoto built the climbing and barrel-dodging mechanic first, then wrapped it in the simplest possible story: an escaped pet gorilla kidnaps the hero's girlfriend out of spite, and the hero must climb a construction site to save her.
→ Donkey Kong grossed $180 million in its first year in North America alone and has generated over $5 billion since, despite its threadbare narrative.
Common mistakes
Designing from genre convention
Most early-80s competitors built shooters or sports clones because that's what existing games were, not because it was proven fun for their specific hardware.
Letting engineers double as designers
Before Miyamoto, engineers who built the hardware also designed the games by default, producing technically functional but uninspired experiences.
Piling on features instead of polishing the core
Nintendo's own philosophy explicitly warns against adding elements beyond what serves the core fun mechanic.
Is it for you?
Best for
Product and game designers trying to build something genuinely novel and replayable rather than a competent clone.
Not ideal for
Fast-follow or clone-driven markets where speed to market matters more than originality.
From the transcript
“We are on a quest to be inspired and find fun, and then we will build exactly the right window dressing around that fun, that…”
“An ordinary man cannot develop good games no matter how hard he tries. A handful of people in this world can develop games that everybody…”
From the episode
Nintendo's Origins