Skip Navigation

Top Menu

Home Archives About

Blog Post

Twist Endings: Good vs. Bad

Last night I read a book, or at least the first 25% of a book and the last few pages. The writing was poetic. The setting and characters were interesting, if a bit cliched. I was just starting to hunker down and enjoy myself when, around page fifty...

The heroine got amnesia.

Yes, amnesia. Selective amnesia, of course, around some shadowy "accident" that left her a sickly, paranoid mess. She'd be going about her business and suddenly drop to the floor, clutching her head and crying. But she couldn't remember what happened "that night," no matter how hard she tried.

Right away, I knew I was going to hate this book.

I skipped to the end and, sure enough, the heroine knew little more at page 200 than she had at page 50. I skimmed through to the Big Twist: it turns out that "that night," all of her friends had died. The whole time she'd been hanging out with them, they were either ghosts or medication-induced hallucinations.

The reaction the author was presumably going for: "My God, I totally didn't see that coming! This book is amazing!"

My reaction: "My God, I'm glad I got this book from the library and didn't waste my time and money on it."

Here's the crucial difference between a good twist ending and a bad one. In a good twist ending, the author was honest. In a bad twist ending, the author lied.

Giving your heroine amnesia and portraying dead people as if they're alive is lying to the audience. It's withholding crucial information that you could give your readers upfront, but you don't.

Why don't you? Because at some level you know that if you don't lie, your story will be boring. Withholding the truth is the only way you can come up with to artificially draw out the tension, to keep people reading for three hours until you finally admit, "Actually, what really happened isn't all that interesting. Sorry."

Let me refine what I mean by "withholding the truth." In every story you write, you will withhold the truth to some extent, or reveal it gradually.

If you're writing a mystery, as the book about the amnesiac was purported to be, you don't name the killer on page one. But you do put clues on page 30, page 82, and page 135. If readers are paying close attention, the identity of the villain will be as plain as day. But with some clever misdirection (the clues are mixed up with red herrings, the detective misinterprets the clues, other people are hiding secrets that make them look guilty, etc.), most readers won't put it all together until the very end. And then they'll say, "It all makes sense now! Why didn't I see it before?!"

But you don't lie to your readers by omission. You don't have your brilliant detective, during his Unmask the Murderer Party, allude to clues that weren't in the story—clues that he conveniently kept to himself so the reader wouldn't catch on. That's cheating. It's like giving a child a puzzle box that can't be opened, no matter what the kid tries, because you hid the key in your pocket.

Even if you're not writing a mystery, the same rules apply. If readers come to the wrong conclusion, it should be because they were dazzled by your sleight of hand, not because the truth wasn't there.

Criteria number two for a good twist: In a good twist ending, the truth is more interesting than the false conclusion. In a bad one, the twist is a letdown.

Last week Sweetie bought a video game by an independent developer for a dollar. It was a narrative game, a sort of interactive visual novel, so he encouraged me to play it first.

The game starts out with you, the player, arriving at your family's new house in the woods after a year-long trip abroad. It's midnight. A thunderstorm rages outside. The lights are flickering ominously. And your family is gone.

You proceed through the dark, creepy house, finding clues about what happened over the year you were away. You discover that your father inherited the house from his uncle, an eccentric shut-in whom the locals affectionately called The Psycho. There are signs that your father, a failed writer, had been deteriorating mentally; that your teenage sister had been dallying with witchcraft to conjure up Uncle Psycho's spirit; that your mother had been getting very friendly with a hunky coworker, and Dad might have gone all The Shining on her and Little Sis. And then...

LOL, JK. The house is empty because Mom and Dad went to a couples' retreat and Little Sis ran away with her girlfriend. The End.

Writing an ending like this is playing a prank on the reader. It's like showing someone a treasure chest and saying, "If you give me a buck, I'll show you what's in this chest. It's really amazing, I promise." When they give you their money, you open the chest...and there's nothing inside but a few dead bugs. And you feel clever and chortle, "Ha ha! You totally fell for it! Sucker!"

This is why I get very, very upset when a book ends with one of the following.

  • The protagonist wakes up. It was all a dream.
  • The protagonist turns out to have a split personality. He or she is the killer, the villainous mastermind, the monster who's been terrorizing the town at night, etc.
  • The protagonist turns out to be dead.
  • A character readers empathized with turns out to be a run-of-the-mill sociopath with no redeeming qualities. AKA Ready-Made Villain in a Box: Just add psychosis and shake.
  • Everything your hero did was part of the villain's carefully orchestrated plan, and/or the villain is actually a good guy who was only "testing" the protagonists.
  • Aliens. Just...aliens.

All of these endings are not only overdone, they're incredibly lame. You can write the best book/movie/game in the world, but in one swift stroke of "Aliens!" you've killed it.

Some people, oddly, like being punked. They'll think you're brilliant for pulling the rug out from under them. But the rest will hate you. They will fling your work into fireplaces. They will never trust anything you create again.

For most stories, there's more or less a consensus about the quality, barring the petulant one-star ratings from a few disgruntled outliers. There will always, always be that handful of people who want to punish the author because their Internet connections were spotty and they couldn't download the book from Amazon after they'd paid for it. But if you look at a graph of ratings, most will concentrate around one number, with a curve down on either side.

But a bad twist ending is very polarizing. Some people will give the work five stars and say it's the most awesome, mind-blowing experience ever, and others will give it one and say they feel like they were conned.

Readers should never feel conned. They can be disappointed, they can be angry that their favorite couplings didn't work out or they didn't get the happily ever after they wanted, but they should never feel like you lied to them.


Alicia Butcher Ehrhardt (September 30, 2014, 3:36 pm)

Each one of those now-bad twists might have worked once, the first time it was used. They are all lazy writing jobs.

I think the people who give something like that a five-star rating are encountering it for the first time. I also think you can't count on that kind of reader for a career.

I always read at least a few of the negative reviews - it saves a lot of time if someone has already mentioned your pet peeve.

Nice analysis. Looking forward to reading more of your posts, past and future.


(Will not be shown)

What is the first letter of "Pennsylvania"?