Deepen Before You Widen
Three things done well beat thirty done shallowly — expand what you already have before adding anything new.
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 90%
Sanderson's third law reads: expand what is already part of the magic system before adding something entirely new, since new additions change how the system fits into the world. Its origin is a near-disaster. Mistborn shipped three distinct magic systems across three books, and readers responded by asking how many The Stormlight Archive would have. His answer was thirty. Sitting down to build it, he recognised the trap: bigger is better. His correction is the core of the framework: 'three well done magic isms are way better than 30 non-well done magic systems.' The input is the urge to widen. The mechanism is a forced return to the question of what the book is actually about, followed by a search for the existing element that could be deepened instead — very often a character who is thin rather than a system that is missing. His analogy is early Elder Scrolls, which advertised a thousand procedurally generated dungeons built from about thirty recombined elements; players were bored after the second, and the fix was fewer, hand-crafted dungeons. The output is work with fewer moving parts and more depth per part. He pairs it with a deliberate override, the zeroth law: always err on the side of what is awesome. If something breaks the outline or the system but is genuinely too cool to omit, try it — you can always revise the foreshadowing in.
Origin
The law came out of Sanderson's own scope spiral: after three magic systems in Mistborn he planned thirty for The Stormlight Archive, and recognised while building it that the ambition was killing the book. He notes this is also part of what killed the original 2002 version of The Way of Kings.
Core principles
- 01Expand what is already part of the system before adding something entirely new, because new elements change how the whole thing sits in the world.
- 02Bigger is better is a trap that grows out of the success of the previous project.
- 03Three well-executed systems are worth more than thirty half-executed ones.
- 04Procedural recombination produces the sensation of vastness and the experience of boredom — an ocean an inch deep.
- 05When you feel the urge to widen, the underlying need is usually depth in an existing element, often a character.
- 06The rule is a check on yourself, not a law of art — a zeroth law overrides it when something is simply too cool to leave out.
How to run it
- 1
Restate what the project is about
Before adding anything, answer the question Sanderson forced on himself: what is my book actually about, and what worldbuilding genuinely enhances that story? Everything downstream is judged against that sentence.
Pro tip Write the sentence before you look at your idea list, so the list does not bias the answer.
- 2
Inventory current scope
List every system, subplot, faction or feature already in the project. Most scope spirals happen because nobody has looked at the full list in one place.
Pro tip Count them. The number itself is usually the argument.
Watch out An inventory taken mid-draft always finds more than you expected.
- 3
Convert every addition into a deepening
For each proposed new element, name an existing element it could enrich instead. Sanderson's own reframing: instead of adding a new character to widen the story, dig deeper into who an existing character is.
Pro tip The best version of a new idea is often a second layer on something already loved.
Watch out Adding rather than deepening makes the story wider but more shallow.
- 4
Kill recombination masquerading as content
Identify anything that is existing parts rearranged. That is the thousand-dungeons pattern: it produces apparent scale and actual tedium.
Pro tip Fewer, hand-crafted units make everyone happier — including the maker.
Watch out Procedural or template-driven expansion is the easiest kind to ship and the fastest to bore.
- 5
Re-execute the survivors to full depth
Take the reduced set and do a genuinely good, thoughtful job on each. Three, well done, is the target shape.
Pro tip Depth is what readers describe as immersive; breadth is what they describe as impressive and then forget.
- 6
Keep a deliberate override
Apply the zeroth law: always err on the side of what is awesome. If an idea breaks the outline or the system but is too cool to leave out, try it and then engineer the foreshadowing to support it.
Pro tip Sanderson's reasoning: if it genuinely makes the story better, you will figure out how to make it work in revision.
Watch out The override is for exceptional ideas only — used habitually it simply reinstates the scope spiral.
In the wild
After Mistborn ran three distinct magic systems across three books, fans asked how many The Stormlight Archive would have, and Sanderson's instinct was thirty. Building it, he caught the trap — bigger is better — and stopped. He rebuilt around a smaller set executed properly, asking instead what the book was actually about and what worldbuilding would enhance that story.
→ The Stormlight Archive shipped as a deep, coherent series rather than a catalogue, and the original sprawling 2002 draft stayed shelved.
Sanderson uses early Elder Scrolls games as the cautionary case. They advertised a thousand procedurally generated dungeons, but those dungeons were assembled from roughly thirty recombined elements, so players were bored after the second one. The later fix was to hand-craft fewer, better dungeons.
→ Players were measurably happier with less content of higher craft — the phrase that stuck was 'an ocean an inch deep.'
Common mistakes
Letting the previous project's success set the scope
The pressure to widen usually comes from outside — three systems worked, so how about thirty. Sanderson notes he fell into this despite having already been burned by it once with The Way of Kings Prime.
Mistaking a character problem for a content shortage
The felt need for something new is often an underwritten character. Adding a new one makes the cast wider and every member shallower, and readers report the whole thing as thin without being able to say why.
Using the zeroth law as a general permission slip
'Err on the side of awesome' is an override for exceptional ideas. Invoked routinely it becomes the justification for every addition, and the third law stops functioning as a check at all.
Is it for you?
Best for
Creators and builders working on a sequel, second album, second product or expanded universe where expectations have escalated.
Not ideal for
First projects and early exploration, where breadth is how you find out what the thing is.
From the transcript
“I had started to fall into a trap and the trap is bigger is better”
“three well done magic isms are way better than 30 non-well done magic systems”
“it's it's an ocean an inch deep you want to avoid that in your storytelling”
“sanderson's zeroth law is always air on the side of what's awesome”
From the episode
#794: Brandon Sanderson on Building a Fiction Empire, Creating $40M+ Kickstarter Campaigns, Unbreakable Habits, The Art of World-Building, and The Science of Magic Systems