How The Legend of Zelda: A Link Between Worlds Was Made
On April 17, 2013, at the very end of a Nintendo 3DS Direct presentation, Satoru Shibata introduced a new game in The Legend of Zelda series. Viewers immediately recognized the Hyrule of A Link to the Past: the same top-down view, familiar roads, trees, and ledges. It was easy to mistake the game for another Nintendo 3DS remake.
Then Link walked up to a wall, turned into a drawing, and slid along its surface to the other side of a chasm. That single move made one thing clear: Nintendo was not simply bringing an old game into stereoscopic 3D. The developers intended to use a familiar world as the foundation for a very different Zelda.
The game came to E3 2013 in June. Nintendo gave it a relatively small corner of the 3DS area, yet visitors often had to wait several hours to play. The demo brought together the project's main features: a multi-story tower, spring-loaded platforms, and movement along walls.

Visitors play an early demo of the new Nintendo 3DS Zelda at E3 2013; this was the first public opportunity to try Link's transformation into a wall drawing and the game's multi-level puzzles.
Katsuo Ishida / 〖E3 2013〗今度のリンクは壁画に変身!「ゼルダの伝説 神々のトライフォース2」 / GAME Watch
By then, the game's central idea had already existed for almost three years. During that time, the first proposal had been rejected, a prototype had been approved, development had been halted, and the project had later resumed with a different group of people before being revised several more times. It also did not begin as a sequel to A Link to the Past. To understand how that happened, we need to go back to late 2009.
Three People Remained After Spirit Tracks
Work on The Legend of Zelda: Spirit Tracks for Nintendo DS ended in late 2009. Most of its creators immediately moved to Skyward Sword for Wii, while only three people took charge of the next handheld Zelda: Hiromasa Shikata, Shiro Mouri, and another programmer whose name Nintendo never disclosed.
Shikata had worked on the series since Ocarina of Time and was to serve as director for the first time. Mouri had been the lead programmer on Phantom Hourglass and Spirit Tracks. They were later joined by planner Kentaro Tominaga and lead designer Koji Takahashi. As with other Zelda games of the period, Eiji Aonuma was the producer.

The five main development leads after the game was completed: Eiji Aonuma, Hiromasa Shikata, Shiro Mouri, Kentaro Tominaga, and Koji Takahashi. Their responsibilities and the points at which they joined the project differed considerably.
Famitsu.com / 開発秘話&トリビア満載でお届けする『ゼルダの伝説 神々のトライフォース2』開発者ロングインタビュー / Famitsu.com
This was a familiar process at Nintendo. In a 1992 interview, Shigeru Miyamoto explained that a new Zelda would first be assigned to a small group that experimented and searched for the foundation of the future game. More people joined later, once it became clear what exactly needed to be made. The process in 2009 began in much the same way, except that the three developers still had no suitable idea.
They spent about a year exploring different possibilities. Roughly six months in, Shikata brought Miyamoto a proposal for a Zelda built around communication between players. Nintendo never published a detailed description of the concept, so it can be judged only from Miyamoto's reaction. He said it looked as though it had been invented twenty years earlier. The developers abandoned that direction and started over.
At that point, A Link to the Past was not part of the discussion at all. Shikata later told Satoru Iwata directly that the original proposal contained neither a sequel to the Super Famicom game nor even the ability to enter walls.
A Prototype in One Day
The new idea appeared almost by accident. During one meeting, Shikata suggested allowing Link to enter a wall. Mouri and the second programmer were immediately interested, even though Shikata himself could not properly explain what the ability would be good for.
The second programmer—the one Nintendo never named—was especially persistent. He was normally calm, but this time he began arguing passionately. He believed the idea was extremely strong and did not want it discarded without a practical test. Shikata decided that the first thing to check was whether Link could move around the corner of a wall. Mouri said he would build that section in a week, but out of sheer stubbornness finished it in a single day. This was only the corner-turning test, not the entire prototype, which was expanded gradually afterward.
When Shikata saw a flat Link in motion, the point of the mechanic finally became clear to him as well. An ordinary three-dimensional character turned into a two-dimensional image, moved around a corner, and reached a place that had previously been inaccessible. One action immediately provided both a new form of movement and the basis for puzzles.

The early prototype still used a wide-eyed Link resembling the hero of Spirit Tracks and an angled camera; handwritten notes mark the places where wall movement was tested.
Nintendo / 社長が訊く『ゼルダの伝説 神々のトライフォース2』, part 2 / nintendo.co.jp
This prototype looked little like the future A Link Between Worlds. The camera viewed the hero at an angle, as in Spirit Tracks, and Link retained the cartoon-like appearance of the handheld Zelda games for DS. The developers simply continued using familiar solutions so they could test the new mechanic more quickly.
Miyamoto reviewed the experimental version and allowed work to continue. About two weeks later, however, all three developers were reassigned to games intended to launch with Wii U. This happened around October 2010, nearly two years before the console's release. Shikata began working on The Legend of Zelda: Battle Quest attraction for Nintendo Land, while Mouri moved to New Super Mario Bros. U.
The project effectively lost all of its principal creators. Shikata did not believe they would ever return to it after such a halt. Before leaving, the developers installed the prototype on three 3DS systems, attached labels bearing its internal title, and handed the devices to Aonuma, Miyamoto, and Takashi Tezuka. Nintendo never disclosed the title itself. The systems were meant to remind the executives that a working idea already existed.
What Nintendo Said in 2011
While Shikata and Mouri were working on Wii U games, Nintendo released Ocarina of Time 3D. After its success, journalists increasingly asked Aonuma when the 3DS would receive a new Zelda of its own. His replies were evasive: he discussed individual ideas for a future game but said nothing about its story or how far development had progressed.
In September 2011, Aonuma merely hinted to Game Informer that stereoscopic depth in the new Zelda would be more than a visual effect and would become part of the gameplay. Two months later, he told GamesRadar that Miyamoto wanted to return to a top-down Zelda. A Link to the Past particularly appealed to him because the 3DS screen could show its multi-level world with genuine depth.
In an interview with the German magazine GamePro, Aonuma was slightly more forthcoming. He said discussions about the next Zelda for 3DS had begun during work on Ocarina of Time 3D, but the project was later put aside because the developers were busy with Skyward Sword. He also promised that the new game would continue the story of an earlier entry in the series.
However, Aonuma's statement that discussion of the new Zelda began during work on Ocarina of Time 3D and was later paused because of Skyward Sword does not match the detailed timeline Nintendo published after the game's release. According to the developers' recollections, the project began immediately after Spirit Tracks. About a year later, all three original developers were transferred to Wii U launch titles, and Miyamoto did not suggest using A Link to the Past as a foundation until May 2012.
Nintendo never explained the discrepancy. Aonuma may have simplified the sequence of events in a short interview, or he may have been referring to a direction that had not yet been formally approved. The only firm conclusion that can be drawn from the 2011 interviews is that Nintendo was already considering a top-down Zelda and ways to use the 3DS stereoscopic screen. Those remarks do not prove that a sequel to A Link to the Past was already in development.
Outside the company, players argued about remakes of Majora’s Mask and A Link to the Past. Inside Nintendo, meanwhile, people were trying to save a very different project—the prototype built around movement along walls.
A Return Without the Original Developers
Skyward Sword was completed in the fall of 2011, allowing Aonuma to return to the handheld game. He did not wait for Shikata and Mouri: both were still occupied with Wii U titles, while Nintendo wanted to release the new Zelda for 3DS by the end of 2013. Starting only after their return would have made that deadline impossible.
Kentaro Tominaga joined the project around November. He had previously worked on The Wind Waker, Twilight Princess, and Skyward Sword, but this was his first handheld Zelda. Together with another programmer, Tominaga continued developing the abandoned prototype and built several small dungeons around movement along walls.
In May 2012, Tominaga proposed making fifty such dungeons. Miyamoto rejected the plan but suggested using A Link to the Past as the foundation. Aonuma expanded on the idea: he proposed placing the camera directly above the hero and combining that viewpoint with Link's ability to become flat.
That was the moment when A Link to the Past first became part of the project. Before then, Nintendo had been looking for a new handheld Zelda with an unusual mechanic, not deliberately making a sequel to an older game.
The camera change made the new ability far more expressive. In the early prototype, players always saw Link at an angle, both while he ran across the ground and while he became flat and entered a wall. The transition between those two states was therefore not especially noticeable. From a strict top-down view, the difference was immediate: on the ground, the player saw Link from above, but on a wall, from the side. The transformation changed not only the hero's appearance but the viewpoint itself.
The overhead view led Aonuma to another thought: if the new Zelda already resembled A Link to the Past, it could also use the familiar map of Hyrule as its foundation. The others worried that this would make the game look even more like a remake. To demonstrate how the old world could work in a new game, the producer opened an editor himself, recreated one area of the map in three dimensions, and placed Link inside it.
Aonuma later recalled completing the work in about three days, although Tominaga immediately objected that it had taken longer. They never settled on the exact duration, but the more important result was clear: after seeing a familiar Hyrule in three dimensions, the developers and Miyamoto agreed that the old map would not necessarily turn the new game into a conventional remake.
Around July 2012, the developers showed the project to Miyamoto for the fourth time. About two months had passed since the unsuccessful presentation built around fifty dungeons. This time Miyamoto approved the new concept, and Nintendo moved from small experiments to development of the complete game.
After the project was approved, Nintendo began assigning more people to it. Shikata and Mouri were still busy with Wii U launch games, although they occasionally joined discussions about the new Zelda. They could not switch over completely until late 2012. Koji Takahashi joined at about the same time after finishing his work on Animal Crossing: New Leaf.
An Old Hyrule for a New Game
The developers now had to decide what to preserve from the old game and what to create from scratch. A Link to the Past was released in Japan in November 1991. Twenty-two years later, Nintendo retained the familiar arrangement of forests, lakes, mountains, and settlements, but set the story in the distant future. Hyrule had different inhabitants, and every dungeon was newly created.
The familiar map offered two immediate advantages. Newcomers received a compact world, while anyone who remembered the original constantly compared the old route with the new one. An obstacle that once had to be walked around could now become a shortcut: Link turned into a drawing, crossed a vertical surface, or slipped through a narrow crack.
Before full production began, all the main developers played through A Link to the Past, but their familiarity with it varied. Shikata had played it in his youth, yet reached the ending for the first time only because of the new project. Tominaga had previously skipped that installment entirely. Memories of the classic game were therefore not treated as an untouchable standard.
Old enemies and puzzles returned only when they remained interesting on 3DS. Sometimes the familiar solution was more difficult than the new one. Where an experienced player might try to fight with a sword or shield, turning into a drawing could be safer. The developers did not want to punish people for knowing the original, but they used the habits of longtime players as material for new puzzles.
Why Link and the Rupees Had to Be Tilted
The familiar map could not simply be transferred into a three-dimensional world. Hyrule also had to remain readable from the chosen viewpoint. A camera directly above the hero emphasized Link's transformation into a drawing, but ordinary models looked wrong from that angle. If a three-dimensional Link stood upright and was viewed from directly overhead, the screen would show little more than his green cap. The same problem affected other characters' faces, rupees, and many small objects.
The developers chose to trick the eye. Characters and objects were tilted slightly backward toward the camera. They look correct from the normal gameplay angle, but from the side it becomes obvious that Hyrule's inhabitants stand at a strange slant and the rupees appear to be falling over.

The normal gameplay angle is shown on the left and the same scene from the side on the right; tilting Link, the enemy, and the rupees made them fully visible beneath an almost vertical camera.
Nintendo / 社長が訊く『ゼルダの伝説 神々のトライフォース2』, part 4 / nintendo.co.jp
Mouri proposed another technical decision. He insisted that the game update its image sixty times per second rather than thirty. With stereoscopic mode enabled, this made it easier to maintain a clear picture, while movement and controls became noticeably smoother. Aonuma initially believed that thirty frames per second would be sufficient for Zelda, but a comparison of the two versions convinced him.
This was a difficult target for the 3DS because the two images required by the stereoscopic screen had to be rendered at high speed. Mouri confidently said it could be done if the entire game was designed for sixty frames per second from the beginning. The complexity of the models, environments, and effects could then be adjusted in advance to meet the target; rebuilding completed scenes later would have been far more difficult.
When Link Became a Wall Mural
The high frame rate made it possible to show the transformation smoothly, but it did not answer another question: what should flat Link actually look like? In the early prototype, his three-dimensional model was simply squashed against the wall. After the switch to an overhead view, Takahashi decided that this was not enough. Entering a wall changed both the viewpoint and the controls, so Link's appearance needed to make it immediately clear that the hero was in a different state.
The designers first tried an almost Cubist drawing with a deliberately distorted body and one enormous eye. Mouri strongly objected. He had become attached to his own prototype, and the drawn Link kept moving his arms and legs independently. This weakened the player's sense of controlling the hero. After lengthy arguments, Takahashi proposed not merely a “drawing,” but a mural that would look natural on a stone wall and remain consistent with the Zelda world.

Link's final wall form, created after experiments with an ordinary flattened model and much more avant-garde drawings.
Nintendo / 社長が訊く『ゼルダの伝説 神々のトライフォース2』, part 5 / nintendo.co.jp
This artistic decision also shaped the story. The game now needed to explain why Link could turn specifically into an image. That led to Yuga, a vain artist and sorcerer who transforms people into paintings. A hero's appearance usually follows the story; here the process was reversed, and an effective way of presenting the mechanic forced the writers to invent a suitable enemy.
On a wall, Link could move only left and right, go around corners, slip between the bars of a gate, and cling to the side of moving blocks. For a time, the developers also allowed him to jump, almost like Mario, but they removed the ability. The more actions that were possible in the flat state, the harder it became for players to understand its rules. The transformation was kept primarily as a way to find a path.

Link crosses a narrow side surface in mural form, where an ordinary three-dimensional character could not stand; the vertical meter on the left limits how long he can move this way.
Famitsu.com / 開発秘話&トリビア満載でお届けする『ゼルダの伝説 神々のトライフォース2』開発者ロングインタビュー / Famitsu.com
The time limit was introduced because otherwise Link could travel endlessly along walls and bypass too many obstacles. The transformation was given an energy meter that gradually refilled. Mouri later suggested using the same meter for the bow, bombs, and other items instead of keeping separate supplies of arrows and ammunition.
There was another reason for this decision. In earlier Zelda games, arrows or bombs placed before a puzzle often revealed the answer in advance. A shared meter removed that hint and also spared inexperienced players from worrying about using their final projectile before a boss. Energy regenerated automatically, but players still could not use every item without restraint.
Depth as Part of the Puzzle
Turning into a drawing was not enough on its own. The developers also needed to show why the game required the 3DS stereoscopic screen at all. They found the answer in the Tower of Hera, a location from A Link to the Past that already featured more changes in elevation than most other dungeons.
Tominaga built the new tower in late 2012. Link struck a spring-loaded platform with the hammer, stepped onto it, and was launched directly to the next floor. The lower levels remained visible beneath him, so the player did not merely know which floor Link was on but could see the distance between levels. Link could also become a mural on the tower's outer wall and travel around the outside of the entire structure.

The multi-story dungeon is designed so that lower rooms remain visible through openings; on the 3DS screen, the stereoscopic effect further emphasized the distance between levels.
Famitsu.com / 開発秘話&トリビア満載でお届けする『ゼルダの伝説 神々のトライフォース2』開発者ロングインタビュー / Famitsu.com
As new people joined development, the Tower of Hera became the example used to explain what the new Zelda should be. The stereoscopic effect did more than decorate familiar rooms: the arrangement of the floors helped players understand where they could rise and how to reach a destination. The tower also became the basis for the E3 demo. In just a few minutes, visitors could see the depth between floors, the smooth movement, and travel along walls.
This brought together two ideas that had existed separately for a long time. Miyamoto wanted to see a classic Zelda with genuine depth, while Shikata and Mouri had conceived a hero who moved from three dimensions into two. A Link to the Past became useful to the project for more than nostalgia: its overhead viewpoint made that transition especially easy to understand.
How the Dungeons Stopped Following a Fixed Order
Alongside the technical challenges, the developers reconsidered a long-standing rule of the series. Shikata said that some people he knew had abandoned Zelda after getting stuck in a single dungeon for too long. Until the player found the one correct route, the rest of the adventure remained inaccessible.
He wanted to let players complete the seven dungeons in the second half of the game in different orders. The traditional Zelda structure made this difficult. Link would normally find a new item inside each dungeon, use it during the second half, and then open the way to the next one. Changing the order caused the entire chain to fall apart.
The first solution was a shop containing all the main items, but this created another problem. If they were sold cheaply, players would immediately take everything and rupees would quickly lose their purpose. If prices were high, the freedom would be largely artificial because players would still have to spend a long time collecting money first.
Aonuma remembered one of his hobbies, which required expensive equipment. He never specified what the hobby was. Beginners were advised not to buy everything at once, but to rent the equipment first and purchase the items they liked later. Ravio's shop was built on the same principle: a required item could be rented cheaply, then purchased permanently and upgraded afterward.

Nearly every main item is available in Ravio's shop before the player enters the dungeons; renting made it possible to choose a route while preserving the value of rupees and the later purchase of equipment.
Famitsu.com / 開発秘話&トリビア満載でお届けする『ゼルダの伝説 神々のトライフォース2』開発者ロングインタビュー / Famitsu.com
The developers considered several return conditions: a fixed rental period, the following morning, and even a late fee. In the end, they decided that Ravio would reclaim the items when Link died. The risk of losing equipment encouraged cautious play, but the developers did not want to force players to repeat a route they already knew. Items could therefore be rented again immediately, warp points were added inside dungeons, and fast travel was made available across Hyrule.
The internal structure of the dungeons changed as well. Since the required item was already in the player's possession, the familiar pattern—find the item first, then use it to complete the second half—disappeared. Shikata believed a handheld game should not keep someone inside one dungeon for several hours, so the dungeons became more compact and some puzzles were moved into the overworld.
Dungeon entrances were marked on the map. Players knew in advance where they needed to go, but still had to determine how to reach the location, which item to bring, and where to use the wall. Renting changed not only the order of the dungeons but the distribution of challenges throughout Hyrule.
Aonuma later said that the developers had reconsidered the established rules of Zelda in this game. Shikata, however, remembered that this had not been their goal at the beginning. Each change solved a separate problem. A player stuck in one dungeon could go somewhere else; item rental prevented rupees from losing their purpose; shorter dungeons were better suited to a handheld system. Together, those decisions changed the familiar structure of Zelda.
The Title, Release, and Early Results
Despite all the changes, the game remained a sequel to A Link to the Past. Its title therefore had to emphasize the connection with the original without suggesting that Nintendo had made a remake. In Japan, the game was called The Legend of Zelda: Kamigami no Triforce 2: the number 2 directly identified it as a sequel to the original. Some staff worried that buyers would still mistake it for a remake because of the number; even the joking title New Zelda, modeled on New Super Mario Bros., was discussed. Yet the game used the same Hyrule and overhead viewpoint, so Nintendo kept the 2.
Outside Japan, the title omitted the number. By E3 2013, the Western version had received the name A Link Between Worlds—a link between worlds instead of a link to the past. The new title referred to travel between Hyrule and its parallel reflection, Lorule. The game itself was set in the distant future after the original.
The title had been selected by E3, but development continued almost until release. In late September 2013, Aonuma told 4Gamer that the game was still being made and that he had been unable to take a summer vacation. Fewer than two months remained before the Western launch.
The game was released in Europe and North America on November 22, 2013, and in Japan on December 26. According to Nintendo's report, worldwide sales had reached 2.18 million copies by December 31, including 1.85 million outside Japan. In other words, most of that result was achieved during roughly five and a half weeks of sales in Western markets, while the Japanese version had been available for only a few days.
In 2014, A Link Between Worlds won the Game Developers Choice Award for Best Handheld/Mobile Game. At the D.I.C.E. Awards, it was named Handheld Game of the Year and also received nominations for Adventure Game of the Year and Game of the Year.
After the release of Breath of the Wild, it is easy to see A Link Between Worlds as an early step toward a more open Zelda. The developers themselves, however, did not draw that connection. They were addressing a more specific problem: they wanted players to choose their own path through a small, familiar Hyrule while preserving traditional dungeons, items, and puzzles. Even so, the result proved important for the entire series. A Link Between Worlds showed that the established progression order could be changed without losing the qualities players valued in Zelda.
The sequel to A Link to the Past grew from a project that originally had nothing to do with the older game. Shikata proposed letting Link enter walls, Mouri demonstrated that the idea could become a game, and the familiar Hyrule appeared only two and a half years after work began. Nintendo did not merely return to an old map: the walls that blocked Link's path in 1991 had themselves become paths.
