The History of Nintendo EAD. Part 4: Tennis Without Buttons, Scales, and Ink

Wii Sports: every swing had to make sense without instructions

Wii Sports certainly demonstrated what the Wii Remote could do, but it was not just a technical showcase. Each of its five sports had to be understandable before the player opened the manual.

Tennis suited this goal better than most familiar genres. The player did not have to move the athlete toward the ball manually. They waited for the serve and swung with a familiar motion. The angle and timing of the swing affected the ball's flight, so even a simple gesture still required skill.

In baseball, the Wii Remote became a bat; in bowling, a ball; in golf, a club. The name of the sport already told the player how to move their arm.

The official Wii Sports gesture guide. It shows how one controller could stand in for a racket, bat, ball, club, or boxing glove.
Nintendo / Ｗｉｉ Sports - Ｗｉｉ / nintendo.co.jp

The set of five sports did not appear immediately. The developers tested many options and kept the games where the gesture was easy to recognize from the outside and produced a clear result right away. They also made sure that watching was at least as entertaining as playing.

In earlier EAD games, someone sitting nearby could do little more than suggest a solution in Zelda or wait for their turn in Mario Kart. Everything on the screen depended on the person holding the controller.

Wii Sports, by contrast, was meant to entertain the whole living room. A failed bowling throw, an overly powerful swing, or an awkward stance could make even the people waiting for their turn laugh. After a few misses like that, someone usually wanted to say, “Let me try.”

Wii Sports sales do need one qualification: the game was bundled with the Wii in North America and Europe, while in Japan it was sold separately. By September 30, 2016, Wii Sports had reached 82.79 million copies, including bundled units. Even with that caveat, the result is impressive.

Mii finally found something to do

But Wii Sports succeeded not only because its controls were easy to understand and the game was fun to watch. Its virtual Mii characters also played an important role. Their simple appearance took years to develop.

Early prototypes used simple human-shaped figures without arms or legs instead of well-known Nintendo characters. The developers tried replacing them with Mario, but then players felt they were controlling a familiar hero rather than a version of themselves. Wii Sports needed not a celebrity, but a neutral character in whom the player could see themselves.

It still needed a face. By that point, Nintendo had spent almost ten years trying to create an editor for virtual doubles.

Mario Artist: Talent Studio for the 64DD allowed users to transfer a photo from the Game Boy Camera onto a 3D character. Later, the team created Stage Debut for the GameCube, also known as Manebito. Both projects ran into the same question: once someone had made a character that looked like them, what were they supposed to do next?

Wii Sports finally answered that old question: the created character stepped onto the court, joined a baseball team, and met relatives there.

Satoru Iwata found a working portrait editor among the DS prototypes and showed it to Shigeru Miyamoto. At first, the team wanted to build the editor directly into Wii Sports, but it was later moved into the system-level Mii Channel. The same Mii could appear in different games.

Creating a Mii on Wii. A single, simple visual style worked better than earlier editors with many poorly matched options.
Nintendo / Ｗｉｉ Sports - Ｗｉｉ / nintendo.co.jp

Talent Studio let users assemble a face from individual parts, but it offered too many artistic styles. Stage Debut added hundreds of clothing items and accessories, yet players still asked what they were supposed to do with the finished character.

In the DS prototype, the faces became simpler and worked better together. Wii Sports gave the double a lasting role: the Mii returned to the court, appeared in other sports, and later moved from game to game.

Wii Fit turned scales into a controller

After Wii Sports, the department chose an activity that looked even less like a video game: weighing yourself every day. When Shigeru Miyamoto was in his early forties, he took up swimming and got into the habit of stepping on the scales daily. He enjoyed tracking changes in his weight and adding another point to the graph each day. This became the starting point for a project with the working title Health Pack.

The first assignment sounded unusual even by Nintendo's standards: create software that would make people want to weigh themselves every day. At that point, the project had neither a final form nor a finished board controller. Miyamoto imagined it as a family ritual: people would gather in the living room, look at the graphs, notice changes, and encourage or gently tease one another.

Sumo provided the idea for the controller. A large wrestler would be weighed on two scales at once, and the readings would then be added together. If a person shifted their body left or right, the numbers changed differently. The difference between the readings could reveal which way they were leaning.

Household scales measured pressure only four or five times per second. That response was too slow for a game, so the prototype increased the rate to 60 measurements per second.

The developers then argued about directions. The software team thought left and right weight shifts would be enough. The hardware specialists wanted to recognize forward and backward leaning as well, so the player could move with their whole body. They kept building prototypes—round, octagonal, and even one made with a saucepan lid.

In the end, sensors measuring metal deformation were built into the board's four feet. They detected weight shifts in every direction and weighed a person accurately, while the device itself had almost no moving parts. After testing, the developers adjusted the exercises while engineers refined the board itself—its shape, sensors, and housing. The Balance Board reached its final form at the same time as Wii Fit.

An official image of Wii Fit and the Balance Board. The device was conceived as a set of scales for daily measurements, but experiments turned it into a full controller for the feet.
Nintendo / Ｗｉｉ Ｆｉｔ / nintendo.co.jp

Wii Fit was released in Japan on December 1, 2007. The contrast was striking: Wii Sports encouraged players to hand the controller to someone else, while Wii Fit asked them to turn the console on again the next day, weigh themselves, and add another point to the graph.

A Wii Fit session could take no longer than an ordinary morning weigh-in, making the game easy to fit into daily life. The approach was successful: by September 30, 2016, worldwide sales had reached 22.67 million copies for Wii Fit and 21.12 million for Wii Fit Plus; both figures include bundles.

Super Mario Galaxy usually controlled the camera itself

While the Kyoto groups were learning to work with gestures and the Balance Board, the Tokyo team began a new 3D Mario game. In Super Mario Sunshine, players often had to track Mario and adjust the camera at the same time. The developers wanted to preserve free movement while demanding fewer manual camera corrections.

The idea of spherical surfaces came from the Mario 128 demonstration discussed in the previous part. Even the EAD Tokyo staff did not immediately believe that an entire game could be built on small planets. Yoshiaki Koizumi assembled a small group that spent about three months making a prototype with a spherical planet, where Mario could run freely thanks to gravity. The prototype convinced their colleagues that the idea could support a full game.

Mario runs across a small spherical planet. Such surfaces first appeared in Mario 128 and later became the foundation for Super Mario Galaxy's levels.
Nintendo / スーパーマリオギャラクシー：マリオ、宇宙を冒険する。 / nintendo.co.jp

Testers were given unfinished levels, and the team marked the places where they got lost or complained of motion sickness, then redesigned the camera. The controls were kept as simple as possible: only two buttons handled the main actions, while an attack was assigned to a swing of the Wii Remote. Mario spun after the gesture and struck nearby enemies. The move did not require the player to land precisely on their heads. A short pause between spins forced the player to choose the right moment instead of waving the remote continuously.

A second Wii Remote allowed a less experienced participant to collect Star Bits, point out directions, and briefly stop dangerous objects while the first player controlled Mario. The helper did not place a second character on the screen, so the camera still followed Mario alone. The controls were simple enough that even someone with no gaming experience could take the second role.

Super Mario Galaxy was only one of several major games EAD was developing for the Wii at the same time. The Kyoto teams were also expanding ideas about intuitive controls and cooperative play: Wii Sports Resort recognized movement more accurately with Wii MotionPlus, while New Super Mario Bros. Wii let four players complete the same level together. But the more projects ran in parallel, the harder it became to distribute experienced staff among them. By the time work began on Nintendo 3DS games, the shortage of people had become especially noticeable.

Mario Kart 7 started with only eight people

In early 2010, Hideki Konno's group launched two Nintendo 3DS projects at once—Nintendogs + Cats and Mario Kart 7. The first game was due earlier, so only eight people initially worked on Mario Kart 7. The plan was for colleagues from other teams to join them later.

The plan failed. EAD was working on several games at once, and development of The Legend of Zelda: Skyward Sword had also been extended. The people the Mario Kart team expected never became available. When the project entered full production, there simply were not enough developers, and new course designers could not be trained overnight.

Nintendo turned to Retro Studios in Texas. The American team had just completed Donkey Kong Country Returns and already knew how to work with Japanese leadership. Retro Studios was first asked to remake some classic courses, but its responsibilities gradually expanded. For the first time, a mainline Mario Kart game was being co-developed by teams in different countries.

Developers at Retro Studios in Austin and Nintendo EAD in Kyoto discuss Mario Kart 7 over a video call. Their collaboration helped solve EAD's internal staff shortage.
Nintendo / 社長が訊く『マリオカート７』レトロスタジオとの共同開発篇 / nintendo.co.jp

Course layouts, models, and feedback now moved constantly between Kyoto and Austin. Nintendo had to split one game between two teams and coordinate decisions by video call. The staff shortage was resolved without pulling employees away from other EAD projects.

Super Mario 3D Land became a “hamburger”

At the same time, EAD Tokyo began work on Super Mario 3D Land. The team's task was to make a 3D Mario game suitable for short sessions on a handheld system. Super Mario Galaxy 2 was a huge collection of ideas: director Koichi Hayashida compared it to a festive banquet serving many different dishes. He imagined the new game, by contrast, as a “hamburger”—small, straightforward, and quick to eat. Super Mario 3D Land had to be equally compact.

This was the team's first handheld project. Hayashida estimated the length of his commute to the office and suggested making the levels short enough to finish within a few train stops. When the train reached the right station, the game also had to be easy to pause.

Large areas with hidden stars were therefore replaced by individual courses with a flagpole at the end. A player familiar with the 2D Super Mario games could immediately see where to go, but now moved through three dimensions.

A selection of compact Super Mario 3D Land levels. The game kept 3D movement but brought back clear courses and an end goal familiar from the 2D entries.
Nintendo / スーパーマリオ ３Ｄランド : 「スーパーマリオ」の世界を3D映像で体験。 / nintendo.co.jp

The stereoscopic screen created another problem. During fast-paced play, people naturally moved not only their fingers but also their upper body. This shifted the 3DS away from the optimal viewing angle, and a strong 3D effect could make the image split in two.

Less than a month before development ended, the team changed the stereoscopic settings: the left-eye and right-eye images now converged at the point in space occupied by Mario. This more comfortable option became the default, while the mode with stronger depth remained available as an alternative.

The game also had to remain readable with 3D turned off. Clear shadows were therefore drawn beneath floating blocks, coins, and enemies. They showed the player where an object was positioned and made jumps easier to judge even on a flat screen.

Wii U: thirty prototypes without one clear pitch

While the Tokyo team adapted 3D Mario to the 3DS, the Kyoto groups searched for a way to explain the capabilities of Nintendo's next home console, the Wii U. Wii Sports had served that role for the Wii by immediately showing how to use the Wii Remote. Nintendo Land, developed almost alongside the new console itself, was expected to do the same. This time, the experiments centered on two screens—the television and the controller display.

The first mock-ups were assembled from whatever was available. The developers initially connected a Wii Zapper to a small display and configured the image to change as the device moved. The prototype was meant for Wii U, but Miyamoto liked the gyroscopic controls so much that he insisted on adding a gyroscope to the nearly finished 3DS as well.

The next mock-up already resembled the future Wii U GamePad: two Wii Remotes were attached to a small display, while two Wii consoles simulated the new hardware. The team tested about thirty game ideas on this setup. Many relied on the difference between the two screens: the player with the controller could see information unavailable to everyone else watching the television.

One prototype of the future Wii U GamePad: an ordinary screen with two Wii Remotes attached. EAD tested ideas on devices like this before the finished controller existed.
Nintendo / 社長が訊く『Wii U』 Nintendo Land篇 / nintendo.co.jp

Mario Chase was built around this idea. The GamePad owner could see the entire map and controlled the runaway, while four participants with Wii Remotes saw only the area around their own characters on the television and coordinated by voice. The runaway benefited from the complete view; the pursuers had the advantage of numbers and help from their teammates.

Four pursuer views in Mario Chase. On the television, each player saw only their part of the field, while the full map remained with the runaway on the GamePad.
Nintendo / Nintendo Land : アトラクション / nintendo.co.jp

But this two-screen system had one drawback. The idea behind Wii Sports was clear even to an observer: the player swung the Wii Remote like a racket, and the character on the screen copied the movement. Mario Chase was harder to understand. The television did not show what the GamePad owner could see, so a spectator could not tell how the players were interacting. Nintendo Land's creators themselves admitted that the game's appeal was difficult to explain in a single sentence. It became obvious only during play.

Of course, Nintendo Land alone cannot explain the Wii U's weak sales. But it worked less effectively than Wii Sports as a clear demonstration of the new console. The GamePad's main advantage—its separate screen, which gave its owner a special role—could only be appreciated by taking the controller and joining the game.

Splatoon: from tofu to squids

After the first Wii U projects were completed, producer Hisashi Nogami assembled a group of younger employees who had previously worked on the console's system menu, Nintendo Land, and New Super Mario Bros. U. They were asked to avoid relying on an existing Nintendo series and invent a new game.

Over six months, the participants proposed about seventy ideas. After several presentations, the team chose a working online prototype created by programmer Shintaro Sato. Two teams of four players shot paint and tried to cover most of the map. Instead of characters, white and black cubes with small noses moved through the arena—otherwise, it was impossible to tell which way the “tofu” was facing.

The first working Splatoon prototype, with white and black tofu cubes. Even here, two teams painted territory and could hide in paint of their own color.
Nintendo / 社長が訊く『Splatoon（スプラトゥーン）』 / nintendo.co.jp

The core mechanics already worked: the cubes painted the floor and hid from opponents in their team's color. But the cubes themselves were only temporary models, and the team still had no heroes capable of becoming the face of the new game.

When the cubes were replaced with humans, the idea failed: the characters' silhouettes remained visible even after they submerged in paint. Rabbits came next. Their long ears made their direction clear from above, and their white fur could be colored to match each team. But the developers could not explain why rabbits would shoot ink and hide inside it.

During another discussion, the developers decided to split the character's abilities between two forms. In humanoid form, the character held a weapon and wore clothes. In squid form, they dived into paint, hid, swam quickly, refilled their ink supply, and climbed painted walls. This ability gave wall painting a purpose: wall coverage did not affect the score, but painted walls opened routes to new places.

The finished Splatoon: a squid swims through and hides in its team's paint. Changing form connected movement, concealment, ink recovery, and climbing walls.
Nintendo / Splatoon（スプラトゥーン） | ナワバリバトル / nintendo.co.jp

Once the squids appeared, all the elements finally formed a clear concept: the team that paints the most territory wins. Reaching that simple formula took six months, about seventy ideas, and several unsuccessful character designs.

Splatoon was released in Japan on May 28, 2015. Satoru Iwata noted that fourteen years had passed since Pikmin before EAD launched another new series with new characters. By December 31, 2016, Splatoon had sold 4.76 million copies—a notable result for a new series created by EAD's younger generation.

The EAD name disappeared, but the teams remained

Splatoon made a fitting final chapter in EAD's history. It followed the department's long-standing tradition: first find the game mechanic, then find the right characters for it. Less than four months after Splatoon's release, the EAD name disappeared from Nintendo's organization.

On September 16, 2015, the company merged Entertainment Analysis & Development with Software Planning & Development. The new division was named Entertainment Planning & Development—EPD. Shigeru Miyamoto became a Creative Fellow, while Shinya Takahashi took charge of the combined organization.

On paper, much changed. Nintendo wanted to release software for its own consoles and smartphones more quickly and make broader use of its characters. EAD's internal teams and the SPD employees who worked with outside studios now reported to the same leadership. The department's structure and the meaning of the reorganization are examined in more detail in a separate article about EAD's structure.

The games from EAD's final decade were very different, but each was built around a simple and understandable action. Wii Sports asked players to swing an arm, Wii Fit to step on scales, Galaxy to run around a small planet, and Splatoon to paint more territory than the opposing team. Not every idea was as easy to explain: Nintendo Land showed how difficult it can be to communicate a game's concept to someone who has not yet held the controller.

After the reorganization, the developers continued building prototypes, observing players, and reworking anything that proved unclear. September 16, 2015, therefore marked the end of the EAD name, but not the end of its teams. Their story continued inside EPD—on Nintendo's next console.