When speedrunner Vytox clocked Donkey Kong Bananza in just under an hour, he wasn't following Nintendo's intended path. He was soaring through the sky using movement techniques the developers never explicitly designed for, skipping entire sections, and generally making a mockery of the game's carefully crafted progression.
The Bananza team knows this is happening. More importantly, they seem to have accepted it.
At the 2026 Game Developers Conference, producer Kenta Motokura and technical programme lead Tatsuya Kurihara shared a talk titled Constructive Destruction: Fusing Voxel Tech and 3D Action Platforming in Donkey Kong Bananza. During this discussion, they were candid about a problem every modern platformer faces: the more freedom you give players, the more ways they'll find to break your game.
The Constraint Problem
Motokura initially explained their approach to stopping players from going where they shouldn't. The core gameplay style of Donkey Kong Bananza centres on movement and interaction within 3D space, relying on the player's ability to navigate complex environments while using the character's strength to alter the stage. So how do you prevent sequence breaks in a world you've deliberately made destructible and interactive?
Engineering solutions help. The team designed surfaces Donkey Kong cannot climb and created barriers that are simply impossible to bypass without solving required puzzles. Yet when pressed on whether it was becoming harder to contain players, Motokura's answer was essentially: yes, we're losing this battle. "It is getting very hard to keep players from going all over the place," he admitted.
Learning to Embrace the Chaos
But here's where the developers show genuine wisdom. Rather than getting frustrated, they've started designing around sequence breaks. The developers acknowledged that the game includes skips intentionally, as shown in a reply by game director Kazuya Takahashi in an interview before the game's launch.
The most telling example comes from the Racing Layer. Players who manage to skip the intended Rambi ride and reach the end anyway are rewarded with special dialogue acknowledging their trick. Motokura explained the thinking: "Sometimes there are sequence breaks in game that you can, once you learn about them, design around so that there is a gameplay experience on the other side of that sequence break."
This mirrors what the Super Mario Odyssey team did, placing coins in hard-to-reach places that savvy players would inevitably discover. It's a form of game design that trusts the community and rewards curiosity rather than punishing it.
Where This Comes From
The philosophy runs deeper than just accommodation. When Motokura and his team were making the 2017 Switch game Super Mario Odyssey, they were considering the implementation of a destroyable rock. That experience shaped how Bananza approaches player freedom.
Attendees were shown a section of the original Super Mario Bros., specifically World 1-2, which offered multiple ways to proceed, including one in which Mario could smash blocks to create a path through. A game from 1985 provided the template for a game from 2025: give players tools and trust them to experiment.
According to software engineer Tatsuya Kurihara, the game features an average of 347 million destructible voxels per level, allowing players to tear apart terrain, throw pieces of the environment, and interact with levels in dynamic ways. That scale of destruction is not accidental. It's permission.
What Surprised Them
Kurihara revealed something that caught the team off guard: how many players attempt to destroy every single voxel on each layer. The intent was there; the enthusiasm exceeded expectations. But the bigger surprise was what Motokura observed in the speedrun community. "The surprising ways that people are using voxels for movement, not just double jump, but other movement techniques entirely that they discovered on their own to get to some very interesting places," he said.
That's the dynamic at play now. Speedrunners are discovering exploit techniques the developers didn't anticipate. The developers watch it happen, sometimes laugh about it, and design future games knowing that player creativity will always outpace design intentions. Rather than seeing this as failure, they're framing it as collaboration between developers and community.
The real question is whether this approach scales. Right now, four weeks after its release, Donkey Kong Bananza is one of the most active games on speedrun.com and strategies are still being developed regularly. As new tricks emerge, will the team continue to design around them, or will some techniques cross a line from clever to broken?
For now, the developers seem genuinely delighted by what players are doing. They've stopped trying to win the containment game and started celebrating the fact that they made something players want to break. In an industry obsessed with control, that's refreshingly honest.