It is hard to tell, though, which emotion Disney Pixar’s latest animation-wonder, The Good Dinosaur, is targeting. As Catmull unveiled the never-before-seen trailer for the film, which is due to hit screens on Thanksgiving, I was temporarily overcome by a scrum of emotions while watching the star character, a kid dinosaur, in action. (Hey, like you didn’t cry during Frozen ….)

While the act that The Good Dinosaur has to follow, Inside Out, is a tough one—that film has grossed an astounding $436 million worldwide since its release on June 19, according to Box Office Mojo—the new movie isn’t likely to embarrass at the ticket booth. That’s not simply because of the reception to the trailer, raucous as it was, but rather because of Pixar’s breathtaking run of hits—a string of movies that include Toy Story (1995), Finding Nemo (2003), The Incredibles (2004), and Up (2009), among other blockbusters.

So how exactly does the studio manage to pull off one monster winner after the next, Fortune‘s Michal Lev-Ram asked Catmull? The answer, he said, is to not keep doing what you’ve done before, even if it’s phenomenally successful: “If something works, you shouldn’t do it again. We want to do something that is new, original—something where there’s a good chance of failure” each time at bat.

And failure there is. The rest of us just don’t get the opportunity to see it. Unlike with software, there are no patches or updates or 2.0s for a movie, once it’s out. So Catmull and his team rework stories and scripts intensely—often going through the expensive process of restarting the development process from the very beginning—rather than have a flawed project end up on screens. “We’d rather face the failure internally rather than release it that way,” he said.

Among such “complete restarts,” he listed are such critical hits as Toy Story 2 and Ratatouille. And, notably, The Good Dinosaur.

In that rebirthing process, Catmull said the key issue—perhaps the only issue that he can assess as a boss—is how well that team works together. When any bunch of Pixar creatives begins a new project, he said, “it always sucks—and I don’t mean that in a self-effacing way. I mean, it always sucks.”

To get beyond that, Catmull relies on what he calls “the brain trust,” a notion that he says he arrived at by accident. The brain trust isn’t one specific cluster of individuals, but rather any group that follows these four golden rules:

1. Nobody can override the director. “Basically we remove the power structure from the room,” he says. “If they know that John [Lasseter, chief creative officer of Walt Disney and Pixar Animation Studios] or I can override the director, then they will enter the room in a defensive way.”

2. Peer to peer. The conversation has to be “filmmaker talking to filmmaker,” said Catmull. “It’s not boss talking to filmmaker or boss talking to employee.”

3. All team members share in one another’s success.

4. They give and take honest notes.

Every once in the while, typically when those rules are followed, magic happens. “And what I mean by magic is the loss of ego in the room,” he says. “It’s an amazing event when it happens.”

One example of that magic? Frozen. The project had been around in one form or another for years, said Catmull. And it wasn’t clear what the story was. “We had an offsite where all of a sudden everything clicked. We had the realization of what this movie was really about. It turned on a dime and we knew we had a movie.”