You spend years reading fantasy and then realize maybe five or six ideas actually changed how you write it. Here are the ones that stuck with me.
The Rules You Invent Constrain You More Than the Ones You Inherited
When you write a story set in the real world, the constraints come free. Gravity works. Governments exist. People die when they're stabbed. You don't have to think about any of it.
Fantasy gives you none of that. Which sounds like freedom until you realize you now have to build every wall your characters will run into. And the walls have to hold.
Naomi Novik understood this. The Wood in Uprooted operates on rules Agnieszka doesn't understand for most of the book. She can't fight it because she doesn't know what it wants. She can't negotiate because she doesn't know what it values. The rules of that corruption, the way it spreads, the way it takes people and gives them back wrong, those rules ARE the central conflict. Novik didn't need a villain with a speech. She needed a system with internal logic that her protagonist couldn't see all of.
The moment you say "magic requires a price," that price will come to collect. And "it takes a toll" isn't specific enough. What toll? On whom? How much? The freedom of fantasy is also its trap. You have to invent the constraints, and vague ones will buckle the second your plot leans on them. If you're building a fantasy world from scratch, the constraints deserve as much thought as the wonder. Maybe more.
Fantasy Has to Earn Every Moment the Reader Spends Away from Reality
There's a test I keep coming back to: why does this have to be fantasy?
A real-world setting is cheaper. Easier. More immediately relatable. When you ask a reader to learn a new geography, a new political system, a new set of physical laws, you're asking for cognitive investment. The world has to be worth it, and "worth it" means the world is doing work in the story, carrying weight, generating conflict, making the plot impossible outside this specific setting.
Joe Abercrombie's The First Law passes this test in an interesting way. It couldn't happen in the real world because it's actually about the fantasy genre itself, about what happens when the barbarian hero, the noble quest, and the wise mentor are all running on corrupted software. The world is the argument. Take it out of fantasy and you lose the thing the book is actually saying.
I think about this whenever I see a fantasy novel that's basically a contemporary drama with swords added. Cool worldbuilding alone doesn't pass the test. The world has to carry weight.
The Chosen One Problem Is a Symptom, Not the Disease
The disease is a protagonist who's special because the author says so.
Tolkien's Frodo is worth mentioning here, briefly, because he's the anti-Chosen One. He's chosen because he's too small to be corrupted by power. Not because he's the strongest or the bravest or the most talented. Because he's the least likely to be ruined by the thing he's carrying.
Patrick Rothfuss takes the opposite approach and still makes it work. Kvothe IS supposed to be extraordinary. But Rothfuss earns it through specificity. The sympathy binding system in The Name of the Wind takes pages of math and physics. It has rules that can be tested and broken. When Kvothe does something impressive, you understand the mechanics well enough to feel the difficulty. Three pages of binding logic, not three words of incantation.
What matters is whether you've earned the specialness on the page, in scenes, through friction and failure and specificity. I don't know what to make of the fact that both approaches, Frodo's smallness and Kvothe's brilliance, work equally well. Maybe the trope was never the problem.