Ocarina’s Image: Freedom in Breath of the Wild


“When I was a child, I went hiking and found a lake. It was quite a surprise for me to stumble upon it. When I traveled around the country without a map, trying to find my way, stumbling on amazing things as I went, I realized how it felt to go on an adventure like this.” – Shigeru Miyamoto, 1993¹

Fifteen years after Ocarina of Time, the Zelda team finally retired its time-worn solutions to the problems of 3D. They shifted their philosophy, drew from the inspiration for the first Legend of Zelda, and rebuilt the beating heart of the franchise from first principles.

Previously we examined the challenges of building the first 3D Zelda, and the tradeoffs that Ocarina made to maintain approachability. These compromises attempted to balance two unsolved problems: decreased readability, and greatly increased complexity. In this final article, we’ll see how Breath of the Wild found new solutions to those problems.

Fujibayashi’s Vision

Hidemaro Fujibayashi first worked on the Zelda series at Capcom, where he was a director and writer for the two Oracle games and The Minish Cap. He eventually moved to Nintendo, worked on Phantom Hourglass, directed Skyward Sword, and then Breath of the Wild. At GDC 2017, he said:

“I wanted the user to be able to experience a new sense of adventure again and again, and to be able to freely navigate through it as they see fit. … A game where the user can think and decide, on their own, where they want to go and what they want to do.” – Hidemaro Fujibayashi, 2017²

The Wild leadership team was aware that something needed to change. Breaking conventions was a consistent theme in interviews around the game’s release.

“…you need to break conventions. You need to be creatively reckless. So perhaps it’s safe to say, in some ways, we game creators really enjoy violent work.” – Satoru Takizawa, 2017²

At that same GDC talk, Fujibayashi said that impassable walls and “a predetermined sequence of events” were conventions he wanted to break. “The first step,” he said, was to “put our sights on changing the structure of the game from a passive one, where you play within the confines of a pre-prepared mechanism, to one where the user can actively engage with the game.”

They achieved this goal by embracing emergence.

Physics and Chemistry Engines

“This would be a world where combining simple elements could produce complex results. Where you’d feel the thrill of imagining the things you wanted to try as you played.” – Takuhiro Dohta, 2017²

If the Zelda team was going to free players from predetermined sequences of events, they needed to rethink how players interacted with the game world. They couldn’t grant the freedom to problem-solve if they kept hand-scripting the solutions.

Fujibayashi refers to it as a shift from “additive design” to “multiplicative design”, where “objects react to the player’s action, and the objects themselves also influence each other”. In other words: emergent gameplay.

“This multiplicative gameplay is found in the space between situations and goals. We’ve filled every corner of the game world with these sets of situations and goals. …we want them to have fun thinking up and trying solutions on their own.” – Takuhiro Dohta , 2017²

First, they licensed the Havok physics engine. The physics engine governed movement, collision, and physical objects. Next, they built what Takuhiro Dohta calls a “chemistry engine”: a set of rules that governed “elements”, such as fire, ice, and wind. Finally, the two engines interact: Elements change the state of objects and other elements, such as ice freezing a moblin or water dousing fire. They can also exert physics force of their own, like wind pushing objects. The physics engine, in turn, controls how things move and come in contact with the elements. For example: throwing a flaming torch to ignite grass uses both. Filling the world with these objects and elements allows a rich and varied set of possibilities to emerge.

Natural Phenomena

Simple elements multiply to produce rich results. By basing their gameplay around simple rules, all they needed to fully arm their players was teach that small set of rules.

One strategy was to base their physics and chemistry rules on expectations from real life. Rocks roll downhill. Lightning is conducted by metal. Fire burns wood. Wind pushes things. Instead of having to teach these interactions with tutorials, they could trust players to learn them by exposure and experimentation. Once these interactions were learned they remained relevant throughout the game. And, crucially, there are very few exceptions to these rules.

“Players encounter in-game events they’ve observed in the real world and come to believe in the game’s particular set of rules. …these connections don’t simply occur by accident. Instead, they’re intentionally designed to have an overall consistency, and to feel instinctive.” – Takuhiro Dohta, 2017²

That consistency is what sets Wild’s emergent systems apart from previous Zelda titles. The Zelda team had based puzzles on real-life dynamics for years, but the rules weren’t fully embraced. For example: setting and dousing fires is a timeworn tradition, but previously we did it with specific objects meant to be part of a puzzle. In contrast, Wild laid a ground rule that all enemies, grass, and wooden weapons could burn.

Nothing could ever make 3D Zelda as simple to control as 2D, but easily learnable emergent play was the key to addressing the complexity problem. There were fewer rules that had to be learned, and many rules were easier to learn because they were consistent and based on real-life expectations. And experimentation was fun! The act of learning was player-driven, rather than scripted at key moments.

“We just give them the systems, give them the items, you get them and you think ‘well what can I do with this?’ And then you go out into the world and you encounter different things, and you try different things out. … you get your own hypotheses that you sort of test in the world.” – Eiji Aonuma, 2017³

In addition, the game could be played without learning all of the rules: most situations had many possible solutions, and one could play the entire game without learning how to capture a horse, or upgrade gear, or use advanced combat mechanics like shield reflection. Much of the complexity could be ignored or learned at the player’s own pace.

The Triangle Rule

“…we needed a certain level of realism, and we needed an information-dense, mature art style.” – Satoru Takizawa, 2017²

One of the major readability problems in 3D is that players can miss important information because they control their own camera. Wild addresses this with a carefully crafted landscape and a masterful visual language.

A “visual language” is a set of rules that developers follow to communicate to their players. Players pick up on these rules while they play. In Assassin’s Creed, eagles fly from vantage points to draw your attention. In Horizon: Zero Dawn or God of War (2018), climbable spots are highlighted with yellow detail.

Wild has one of the most consistent and readable visual languages I’ve seen. First, they made it easy to identify points of interest. The same orange light highlights unsolved shrines, magic puzzle orbs, unopened treasure chests, and more. They tried to avoid this shade of orange elsewhere. Similarly, any time you saw the alien curves of Sheikah architecture it stuck out, drawing attention.

A more subtle example of visual language was the “triangle rule”. The landscape was built mostly with triangular hills and mountains. Big triangles were landmarks, medium ones obstructed player view to create surprises when traveling, and small ones provided texture. Triangles with more irregular shapes, like extra peaks or divots, tended to hide Koroks. This would draw the player’s eye, but most players were not consciously aware of this pattern.

Visual language like this infused rich information into the environment. Care was taken to avoid using these cues at the wrong time. Using a cue incorrectly dilutes its meaning. Building a world that consistently adheres to rules like these requires immense care. If all goes well, the results are so subtle that players aren’t conscious of it. This was the magic that made Koroks feel so rewarding: we were drawn to places that felt different without always knowing why. Pulling that off requires a dev team that deeply understands and values the language. Nintendo’s teams have always been among the best at this, and Breath of the Wild shows singular mastery of this craft.

The High Ground

With its exceptional visual language, Wild ensured that players would understand gameplay information when they saw it. So how did it ensure they saw it in the first place?

The first solution was simple: By shifting away from scripted events and towards emergent play, they largely dodged the problem. In a traditional Zelda, there are thousands of points where players need to spot things or else be stuck forever. But in Wild, it was okay if players missed things! Almost nothing is required to progress.

“Link can climb up high and look down from cliffs. It’s kind of like 2D gameplay.” – Eiji Aonuma, 2017

The second solution was to elevate players and then train them to see. The game encourages players to reach high places. The high ground was valuable because you could shoot arrows further, or roll things onto enemies, and Koroks or other goals were often placed there. Most important of all, you could only paraglide from elevated positions.

Wild is a game about seeing the environment. Its visual language was designed to be readable from a great distance. Once the player is on high ground, the game encourages them to actively survey the environment through the use of strong landmarks, frequent elevation changes that break line of sight, and the ability to spot shrines and place markers at a distance. This active seeing, combined with the strong visual language, helps ensure that players will see everything they are meant to see.

“Zelda has an epic story and all, but the truth is, to me it’s all about hiking.” – Shigeru Miyamoto, 1998

Breath of the Wild

So let’s recap.

Nintendo really values approachability. They want many people to be able to enjoy their games. 3D Zelda had two major problems that hurt approachability: it was too complex, and vital information wasn’t readable enough. They compensated by reducing combat difficulty and spatial thinking in dungeon design, avoiding emergent gameplay, and protecting players from making mistakes. These underlying problems, and Ocarina’s compromises, were baked into Zelda’s design for the next 15 years because of Ocarina’s immense success.

The Breath of the Wild team wanted to break conventions and provide a high degree of player freedom. They looked to the first Zelda for inspiration and redesigned from first principles. In the process, they found new solutions to those two problems. High complexity was addressed by building the game out of emergent systems. This reduced the number of rules that needed to be learned, and made much of the complexity optional. Low readability was addressed by a masterful visual language and encouraging players to actively survey the environment.

With these problems eased, the Zelda team was finally unleashed on a game without Ocarina’s compromises. The world got dangerous again. The dungeons whole-heartedly embraced mind-bending spatial thinking. The entire game was built on a foundation of emergent, “multiplicative” gameplay. And they felt freed to trust the player: tutorials were minimized, barriers were removed, story-gating was abolished, and players were encouraged to make mistakes and express themselves through play.

Zelda was back.

“The spirit, the state of mind of a kid when he enters a cave alone must be realized in the game. Going in, he must feel the cold air around him. He must discover a branch off to one side and decide whether to explore it or not. Sometimes he loses his way.” – Shigeru Miyamoto, 1993¹