1998’s Ocarina of Time is a pivotal game in the Legend of Zelda series. For one, it’s where the series’ timeline splits into three different branches. It’s also when Nintendo’s long-running adventure game franchise split into two major styles: 2D and 3D. The 3D games have come to be viewed as the “main event,” the big-budget, 50-hour affairs that come around once every 5 or 6 years. Meanwhile, the 2D games have taken on a bit more of a “filler” role, in that they typically enjoy a shorter turnaround time and help keep the series active while the next 3D game is in development.
Given the shorter development times and lower budgets of the 2D games it makes sense that Nintendo is more prone to experimenting with different ideas while creating them. Whether that experimentation comes in the form of touch controls, or multiplayer, or non-linear structures, very few 2D Zeldas are alike. This works out in everybody’s favor because it helps keep things interesting, and also because the constant cycle of experimentation with the 2D games has saved the series’ bacon on several occasions.
Let’s take a look back at the history of the 2D, top-down Zelda game, and see how each one represented a unique sort of experiment.
The Legend of Zelda (NES, 1986)
“We started to work with Legend of Zelda at the same time as Super Mario Bros., and since the same people did both games we tried to separate the different ideas,” Shigeru Miyamoto once said in an interview with the Swedish magazine Superplay. “Super Mario Bros. should be linear. The next step in SMB should be obvious. Zelda should be Mario’s total opposite.”
Looking back, the decision to create the first Zelda can seem like an awfully brave one. Going from straightforward arcade action games like Donkey Kong and Mario Bros. to a game about free-form exploration across an entire world strikes one as an immense leap in ambition, planning, and execution. But it didn’t start out that way.
It took the 6-person development team a fair bit of experimentation to settle upon the final design. Zelda 1 actually began as a much simpler concept that would involve the player selecting a dungeon from a menu, entering it, and finding their way around. There was no world map, and no design that called for the player to feel their way around an enormous world. And by the time a world map existed, there was considerable apprehension among the members of the development team about whether players would even understand what they were meant to do, given that open-ended games weren’t common at the time.
“I remember that we were very nervous, because The Legend Of Zelda was our first game that forced the players to think about what they should do next,” Miyamoto recalled in the same interview. “We were afraid that gamers would become bored and stressed by the new concept. Luckily, they reacted totally opposite. It was these elements that made the game so popular, and today gamers tell us how fun the Zelda riddles are, and how happy they become when they solved a task and proceed with the adventure. It makes me a happy game producer.”
Zelda II: The Adventure of Link (NES, 1987)
Shigeru Miyamoto likes the games he’s deeply involved with just fine. Every so often, though, Miyamoto will take one of Nintendo’s other games and throw it under the bus. Zelda II is one of those games.
“It was my idea, but the actual game was developed by another team, different people to those that made the first game,” Miyamoto would say to Superplay magazine in 2003. “Compared to Legend of Zelda, Zelda II went exactly [as] we expected. All games I make usually get better in the development process, since good ideas keep coming, but Zelda II was sort of a failure,” he said. Ten years later, he’d bring Zelda II up again as an example of a game he “could have done more with,” although he didn’t use as harsh language this time around.
Failure or not, it doesn’t change the fact that Zelda II is one of the more interesting Zelda games. It was another attempt by Nintendo to create something completely different—which is commendable, given what an incredible success the first game was.
Unlike the other 2D games in the series, Zelda II for the most part does not take place from a top-down perspective but from the side. Development began with Miyamoto telling the team—not the Zelda 1 team, but an entirely different team—that he wanted to make a side-scrolling action game where the player had to attack with, and defend against, high and low attacks.
“Rather than being a continuation of the series, it started as a new sword and shield type of action game,” director Tadashi Sugiyama recalled in a 2016 interview. “We were experimenting while producing the game so we didn’t really have the first game’s systems in mind while developing it. As for it being unique within the series, we were searching for new ways to play so you could say it’s like a spin-off.”
Because Sugiyama and his team weren’t really concerned with creating a game that was based off Zelda 1, The Adventure of Link couldn’t be more different. In addition to its side-scrolling action stages, the game also featured an RPG-like experience system, where you could level up your life, magic, and attack stats. Beyond these traits, Zelda II is also remembered for its quirky towns that Link can visit, and its downward stab—both of which directly inspired indie hit Shovel Knight.
“Failure” is a harsh word to describe Zelda II. Especially given that plans to remake the game for the Nintendo 64 eventually resulted in Ocarina of Time, one is reminded yet again that even so-called failures have their place.
A Link to the Past (SNES, 1991)
A Link to the Past is one of the less experimental Zeldas. After the vastly different Zelda II, Miyamoto wanted to create a “true sequel” to the original Legend of Zelda for the new Super NES platform.
The final version of A Link to the Past is significantly different from what early design documents called for, which would have been much more experimental. Originally, the game was meant to use a multi-world structure, where you would operate out of a hub world and travel to other, connected worlds. Decisions in the hub world would have implications in these other worlds.
Interestingly, one of these worlds was probably going to feature a sci-fi setting, if this piece of concept art from Hyrule Historia is any indication. Unfortunately, the idea was ultimately scrapped and the three planned worlds were reduced to just two, the Light and Dark worlds that were sort of mirrors of each other. The sci-fi themes were jettisoned entirely.
“At first there were three worlds, but players would’ve gotten confused,” Miyamoto explained in a 1991 interview published in a Japanese strategy guide for A Link to the Past. “That’s why we had to fix things up. It’s difficult to plant a new concept like that in an action game, you see.”
Link’s Awakening (Game Boy, 1993)
The first Zelda developed for a portable system, Link’s Awakening is a bizarre, otherworldly game with a heavy focus on characters. The only rule of its development was that there were no rules. “I really had free rein to do what I wanted, so long as I didn’t make Miyamoto angry,” said the game’s script writer Yoshiaki Koizumi in 2007. Director Takashi Tezuka said the environment was like the “free spirit of an afterschool club.” Miyamoto was busy with other projects, and so the development team did what they wanted. And what they wanted was something like the American television show Twin Peaks.
“At the time, Twin Peaks was rather popular. The drama was all about a small number of characters in a small town,” Tezuka said in a 2009 interview. “When it came to The Legend of Zelda: Link’s Awakening, I wanted to make something that, while it would be small enough in scope to easily understand, it would have deep and distinctive characteristics. After that, in The Legend of Zelda: Ocarina of Time and The Legend of Zelda: Majora’s Mask, all kinds of suspicious characters appeared. I didn’t tell them to do it that way, but personally, I did find it considerably appealing.”
Without Link’s Awakening, we may never have gotten the character-focused intrigue of Ocarina or Majora, which in a sense makes it the most important 2D Zelda game in the series. As further proof of its long-lasting influence, Link’s Awakening was also the first Zelda to feature a fishing minigame, something that has been present in nearly every Zelda game since.
Later this year, Nintendo plans to release a completely overhauled remake of Link’s Awakening for Switch. Given how different Link’s Awakening is from its contemporaries, it’s the perfect Zelda to remake.
Oracle of Ages and Oracle of Seasons (Game Boy Color, 2001)
The first and, so far, the only time that Nintendo has released two Zelda games simultaneously, Oracle of Ages and Oracle of Seasons actually started out as three games. Maybe even four games, depending on how you look at it.
The story behind the Oracle games begins in 1999, when Capcom’s Yoshiki Okamoto, responsible for games such as Final Fight and Street Fighter II, approached Shigeru Miyamoto with the prospect of working on a Zelda game. Before long, an arrangement was struck, and Capcom were hired to work on not one, but a series of Zelda titles for the Game Boy. The first of these would be a remake of Zelda 1, with the idea being that Capcom and Nintendo would introduce the game to a new generation of players.
This didn’t quite sit right with Okamoto’s team, who wanted to skip the remake and get right to work on an original game. Their solution was to remake Zelda 1 as per their arrangement with Nintendo, but have it be part of a three-game set where all three games would be able to link up via the Game Boy link cable. The things you did in one game would affect the game after it.
“I wanted, for example, that if you missed an enemy in the first game, you would encounter it in the next one,” Capcom’s Hidemaro Fujibayashi recalled in a 2001 interview with Nintendo Online magazine. Fujibayashi was the young designer that had thought up this idea, and was eventually chosen to direct the three games, which were to be called The Tale of Power, The Tale of Courage, and The Tale of Wisdom.
Unfortunately, this was easier said than done, and Capcom were forced to reduce the number of games to two. The resulting games are Oracle of Ages, with a greater emphasis on puzzles, and Oracle of Seasons, which has a greater emphasis on action. Linking the two and playing both changes certain events in the games, eventually culminating in a “true ending” to the story.
Fujibayashi would eventually leave Capcom and join Nintendo; most recently, he directed Breath of the Wild.
Four Swords Adventures (GameCube, 2004)
Following the release of The Wind Waker, the Zelda franchise found itself in a slump. It had succumbed to “gamer drift”—a term Nintendo used to describe fewer and fewer people being interested in playing videogames on the whole. The development team felt that some sort of reimagination was necessary to push things forward.
Around the same time, Zelda co-director Eiji Aonuma told Miyamoto, his boss, that he was tired of working on Zelda. Miyamoto suggested that Aonuma step into the role of producer instead, where he could manage the series from a more macro level, and re-examine what it needed to be, to combat gamer drift. To help with this, Aonuma once more enlisted the aid of Capcom.
One of the ideas Miyamoto had to shake things up was to create a multiplayer Zelda that would make use of four Game Boy Advance devices connected to a GameCube. In the overworld, players would see the game on the television display, but when entering a dungeon they would each have their individual view on their GBAs. It would be based on the original Four Swords, which was a bonus multiplayer mode in the GBA port of A Link to the Past.
If that sounds needlessly complex, it’s because it was. Four Swords Adventures flopped.
“I believe this result stemmed from the need for each player to have a Game Boy Advance and the need for each player to also have a cable to connect that Game Boy Advance to a GameCube,” Aonuma would later say in a Game Developers Conference keynote. “I felt requirements like these prevented it from doing as well as we’d hoped, but there was another problem. I think you might have noticed this as I was explaining the game, but it suffered from seeming too complicated. It was too difficult to convince the consumer they wanted to play the game.”
The Minish Cap (Game Boy Advance, 2004)
Work on The Minish Cap had begun as early as 2001 at Capcom, but was put on hold to free up resources for Four Swords Adventures instead. The goal was to create a 2D Zelda game that was just as impressive as the 3D games. 3D Zeldas were considered the “event” games, but the 2D ones were easier for players to understand and simpler to play. To try to create something that would be as impressive as the 3D games, Nintendo wanted to use Capcom’s talent for excellent sprite-based graphics to convey a sense of shifting perspectives, based around the idea of “big and small.”
“When I was making Oracle and Four Swords, I had my own personal image of Hyrule, and I was trying to figure out a way to keep that intact while still making a bonafide Zelda game,” director Hidemaro Fujibayashi said in an interview. “I was also thinking about the antonyms (or maybe ‘symmetries’ is a better word) that have defined Zelda games previously, like light and darkness. As I thought of that, the concept of ‘big and small’ came to me.”
The idea for Ezlo, the talking cap that accompanies Link on his quest, came from the Gnat Hat in Four Swords. While brainstorming items for Link to wear, including masks and caps, the team settled upon a talking hat, which led to Ezlo.
Phantom Hourglass (2007) and Spirit Tracks (Nintendo DS, 2009)
From 2001 onwards, Zelda was in a constant struggle to find its identity. The tastes of Japanese and Western players were so wildly different that it seemed Zelda couldn’t possibly hope to appeal to both markets with the same game. 2006’s Twilight Princess, with its more “realistic” feel, had the U.S. and European markets covered. But in Japan, the videogame market was in rapid decline with only the Nintendo DS holding it up.
The touch controls that were reviving and expanding the Japanese game market with games like Brain Age, Eiji Aonuma felt, were the solution to getting the more casual fan interested in picking up a Zelda game. Aonuma put a small team of developers to work on prototyping a system of touch-controlled movement on the DS. In addition to potentially exposing more casual players to Zelda, this would also allow for some interesting creative choices with puzzle-solving.
“We thought Brain Age was our rival,” Aonuma revealed in an interview in 2007. “Brain Age’s like that smart transfer student. The Zelda team’s not in the top places, but it studies hard. And then comes this transfer student and easily gets the first place without studying. That’s very frustrating. After three long years, we finally finished Twilight Princess and the transfer student’s the one that’s smart and cool and gets the first place? Damn it!”
Aonuma’s hunch paid off. Phantom Hourglass sold over 900,000 units in Japan alone—the highest any Zelda game had sold in its home region in many years. It also contains some of the most clever puzzles in the series, a number of which wouldn’t have been possible without touch input. Spirit Tracks, which launched two years later with similar controls, wasn’t quite as successful, but also wasn’t as good.
A Link Between Worlds (Nintendo 3DS, 2013)
By 2012, Zelda was in a slump again, still in that struggle to figure out who exactly these games were for, and what they needed to be. The Legend of Zelda: Skyward Sword on Wii hadn’t performed to expectations, and there was a very real risk of the brand losing its relevance. The word on the street was that Nintendo had forgotten what people liked about Zelda in the first place: exploration.
A Link Between Worlds was created as part of an effort by Nintendo to “rethink the conventions of Zelda,” as it often said at the time. The starting point for this was to allow the player to tackle dungeons in any order that they liked. To accommodate this idea, the team came up with an inventive item-rental system. The rental system would allow players to borrow any of the game’s items at any point, if they had the cash. That would let them tackle the dungeons in any order.
This meant that Rupees, not items, were now the only thing standing between you and wherever it was you wanted to go. And making money was never too difficult. It was the perfect solution, one that allowed for a more open-ended style of exploration similar to the original Legend of Zelda.
“When it was time to make this game, I had the vague idea that in A Link to the Past, you could clear multiple dungeons in parallel,” director Hiromasa Shikata recalled in Iwata Asks. “But when I played the game again, that wasn’t very true. So I thought it would be good to do that for this game and we made it so that when it comes to the seven dungeons in the latter half, you can go to any of them as you like.”
Tri Force Heroes (Nintendo 3DS, 2015)
Tri Force Heroes isn’t particularly “innovative” by Nintendo’s own admission, but it is interesting. Series producer Eiji Aonuma modestly describes it as another attempt to make multiplayer Zelda work, but naturally it isn’t quite as straightforward as that.
In Tri Force Heroes, three players can partner up online to solve puzzles. In order to communicate with one another, they use emoji on the Nintendo 3DS touch screen. The development team did actually experiment with including voice chat, but found that being able to communicate via voice meant that even one player figuring out the solution to a puzzle meant it was spoiled for everyone, and this reduced the need for players to coordinate with one another.
“It’s almost like the ‘Like’ button on Facebook,” Aonuma said in an interview. “When someone presses ‘Like,’ no one really knows what that ‘Like’ implies. Going off that idea, with these icons, it’s not just a direct meaning—it hides another meaning behind it. Figuring that out ended up being pretty interesting and a good feature, so we went with that idea.”
So, what’s next for 2D Zelda? It’s a little too early to say. What’s coming in the immediate future is a completely overhauled remake of Link’s Awakening, but that isn’t necessarily representative of where Nintendo might take the 2D Zelda series next.
What we do know is that Eiji Aonuma and Nintendo are more willing than ever to experiment with new Zelda ideas going forward. For instance, Aonuma recently mentioned to Kotaku that he would even be open to the idea of a “Zelda Maker,” similar to Nintendo’s Mario Maker series of games, should fans find the new dungeon creator aspect of the Link’s Awakening remake enjoyable.
Beyond this are even crazier ideas that have been contemplated in the past. At one point, Nintendo contracted Japanese studio Vanpool to create a horror game starring Tingle. More recently, Nintendo lent the Zelda license to the indie studio Brace Yourself Games for Cadence of Hyrule, a Crypt of the NecroDancer spin-off set in the world of Zelda, featuring its music and characters.
At this point, Aonuma and his team appear to be treating Zelda the way Nintendo treats its Mario franchise: Anything goes, as long as it’s fun, interesting, and there’s an audience for it.
Ishaan Sahdev (@ishaansahdev) specializes in the coverage of Japanese video games, with a focus on trend/design analysis. In addition to Kotaku, he also writes for Gamesindustry.biz.