8. A First-Rate Link, Even by Nintendo Standards

Initially, Zelda: Twilight Princess was being developed solely for the GameCube, but you then adapted it for Wii. I’m sure there were many vexing problems along the way, but could you tell me how you found your way through the various complications?

Well, I felt confident that applying Wii’s unique control system to Zelda would heighten the game’s charm. The pointer function, for example, would make the use of in-game items in the first-person mode both easier and more intuitive. I told Aonuma-san that for the Wii version we would just have to change the controls without having to change the game's content, but there were some concerns about this.

At first, the staff and I were at a total loss as to what to do. We’d been shown a new device, but weren’t happy to just brush it off with: “We don’t have enough time for that!” (laughs) So we started trying out various ideas. One of which was dropping the 3D stick altogether and instead just using the Wii Remote only for its pointer function in a similar way to the Nintendo DS stylus.

You mean, you wanted to control Link by using the Remote to point to a certain destination?

Right. In the end it didn't really work. But even though we felt that it probably wouldn't suit a Zelda game, it was worth giving it a try. We played around with a lot of ideas for the controls and camera work, including those that we thought may have been a little too revolutionary. Then, when we exhibited the game at this year’s E3, we had a version that we felt was just about in a finished state. But there were still things I wasn’t completely satisfied with, and I was concerned about what kind of reception it would receive from the crowds there. As I’d feared, the feedback was quite negative.

Oh, really? I was under the impression that the people who were able to try it out at the show weren’t that critical.

Well, the ones who only played it for a while seemed to enjoy themselves, but the hardcore fans who had gotten used to the GameCube’s control method, and those people who were having a quiet word in Miyamoto-san's ear (laughs) seemed to be less than impressed.

I did tell him, with perhaps the hint of a threat, that people thought it was terrible (laughs)

(Wry smile)

Getting this kind of chance to sound out people's reactions before a piece of software is released is actually something I welcome. But personally the thing which most affected me was hearing how smooth the controls on Mario Galaxy were! (laughs)

(laughter)

Well, you know, I did tell myself at the time that Mario Galaxy was developed for Wii from the start, but I knew the customers wouldn't have any sympathy for that kind of excuse. In any case, it was clear that some serious changes had to be made. Actually, it was the time after E3 that gave us the most headaches with the controls for the Wii version.

So, which parts did you begin by changing?

Firstly, we tried to take out all the aspects that had a negative reception during the showing at E3. Then from there, we took things back to the drawing board, going back to the original concept: “What are the best things about using the Wii Remote in this version?” We heard a lot of things like: “We want to use the controller like a sword” at E3. Of course we had already done some experiments involving swinging the Remote like a sword and seeing Link’s action mimic this on-screen. During these early experiments however, we were faced with the fact that players doing this repeatedly would become tired, and so in the end we chose to remove it. Especially at the beginning, we made the game respond to various movements of the Remote. For example, if you swung the Remote vertically, Link would swing his sword vertically. But we felt this would actually become restricting to the player, and would tire the player out if they kept playing this way.

But the crowds at E3 seemed intent on being able to wield the Remote like a sword. And so you found yourselves stuck between a rock and a hard place.

Perhaps the most persuasive point for us was that the players who tried the game at E3 instinctively swung the Wii Remote around like a sword. It wasn’t just the sword either, we also noticed that during the fishing sections, players were manipulating the Remote like a fishing rod and reel, even though it was controlled by the buttons. Seeing this, we realised that this must be an intuitive movement. We knew we had to make some kind of adjustments to the game to incorporate this. When people first pick up the Wii Remote, they are expecting the game to respond if they swing it. That's why we knew we had to integrate this functionality into the game. We worried a lot about the actual implementation of these features, but decided to simplify the system so that swinging the Remote did in fact create a sword-swinging motion in the game, but didn’t cause people any stress when trying to do so. Luckily, Zelda isn’t just about slashing away at enemies from start to finish, and I was glad when we were able to recreate a spinning attack with a simple flick of the wrist. Getting to that stage though was very time-consuming, and full of adjustments.

It certainly seemed that once E3 had ended, there were a lot of issues concerning the Wii control method for Zelda, but that you really knocked it into shape. In retrospect, even though it was a struggle, are you glad you had a playable version of the game in time for E3?

In the end, I'm glad we managed it.

I was rather surprised by the reception the game got at E3 as well. I’d expected it to fare a lot better and, in all honesty, I wasn’t prepared for the number of people that found it difficult to use the directional pad while gripping the Wii Remote. Since we’re all very used to using the Remote and very dexterous when it comes to using all the buttons while holding it, seeing the way first-time players gripped the Remote really tightly was surprising! (laughs) It made me realise that it's not easy for most people to move their fingers, as if to utilise a directional pad, when holding something rod-shaped.

One thing we know for certain is that we didn’t perfect the control system with a single idea: making the Remote work by swinging it, or by minimising the need to use the directional pad, for instance. Even once we’d decided to proceed in a certain way, we still conducted countless tests in an attempt to iron out the small imperfections and give the system a complete overhaul. It was Miyamoto-san and I, through endless discussions of: “You mean like this?” “Yes, like that”, and plenty of tweaking that got it done.

Yes, it was just like he said. For software in the development cycle, it’s relatively easy to assign functions to a certain button, but the problem was how to make the players really feel they have total control over Link’s movements. If we didn’t go through all the possible control functions thoroughly, and make sure that things were absolutely perfect to prevent minor control errors, we couldn’t really say we’d “tested it”, could we? We honestly did a huge amount of fine-tuning!

We certainly did!

Even though we managed to find our way through that particularly gruelling stage, probably the biggest obstacle we had to face on the long journey to completion was at the beginning when the team leaders were under the impression that, even though the game was going to be released on Wii, they didn’t think they’d have anything to do with the development of that version.

I see. Back when you first decided to produce it for Wii?

Precisely. The staff only thought they were to be involved in the GameCube version, which would be compatible with Wii. They believed that the Wii version wasn’t for them to worry about. So it was only myself, Aonuma-san and the lead programmer who had fully devoted ourselves to doing this version! Once the revelation that “we’re making a Wii version of Zelda!" had sunk in, everybody began putting their all into the project.

Ah…

And so at the very end, when we entered the debugging phase and I announced to the staff that they were to begin debugging the Wii version. I was worried that everyone would say: "But we did the most fine-tuning on the GameCube version!" But luckily it didn’t take long before everyone adapted to the Wii version and subsequently found it tough going back to the GameCube one! (laughs)

In interviews with the other developers, the general feeling is that they feel confident about both versions.

I’m glad to hear it! (laughs) A lot of the staff became really passionate about the GameCube version, but they found that because of the differences in the aspect ratio on the Wii version, the player's field of vision is bigger than on the GameCube version. For that reason, a lot of them felt that the Wii version was easier to play. I think that in the end, everything worked out just fine.

The fact that on the GameCube version left and right are reversed also adds a new enjoyable dimension to the game.

I see. Next, I'd like to talk about graphics. In Twilight Princess, the style of the graphics is more real, but your aim wasn't to make them photo-realistic, was it? Aonuma-san, you originally come from a design background, so I'd like to know how you came to settle on the final style?

When I first found out that this Zelda was going to feature more realistic graphics, my initial concern was that we'd just be making a lot of extra work for ourselves. In particular, when a realistically proportioned Link jumped in the game, it wasn't a good representation of how someone would jump in real life. If I tried to make it too realistic, the game would suffer as a result. And so, even though there are parts that lacked total realism, I focused on the adjustments needed to make Link move smoothly. This was very difficult, but we didn't have to make it ultra-realistic. To quote Miyamoto-san: "What's happening in the game world should feel true to life." And so our ultimate goal was to to ensure that the character's design and movements weren't laboured and were at least close to reality, without us expending too much time and energy on it. We didn't have to do this for everything of course, just those certain aspects that required it. Finding those aspects however, was the most challenging part.

In other words, it's like setting the scene for a play rather than recreating the world as it is. If you don't tell people they should be making a stage, they go ahead and try to make an entire world. There is an art to how you properly set a stage and that's what I had to carefully explain.

Is there anything you could tell us specifically?

Well for example, imagine a scene with lots of small stones. On the one hand, you could make it so that the player can move every single one, but if you can successfully communicate the premise that there are also stones that don't move, they will accept this. Making them move isn't the problem, making them move realistically is much more demanding. When and where to do this while constructing the game's world is something that clearly wasn't going well when I stepped into the development of this project, and I found myself having to drill the details into the staff.

Did those problems have anything to do with the sheer size of the project?

Yes, I think they did. These days development is much more fragmented with separate people working on tasks such as graphics, movements and item placement. When creating realistic movement within the game's environment everything is essentially connected; whether it's the way the graphics are drawn, how the program is written or even where the objects are placed, even the designs have an impact on this. And when one aspect is out of place in the game world, it loses its seamless connection to the other parts. When I find one of these problems and attempt to locate the cause, people tend to point the finger at someone else and have me going round in circles! (laughs) And just like I've mentioned before, I'll then get angry: "Who put this stone here?" Somebody must've put it there! When I try to track the culprit down, it always comes back to the director. "The designers and programmers didn't do it, so it must've been the director! Who's in charge here!" This happened quite a lot.

But there really are times when a stone is placed somewhere for absolutely no reason whatsoever!

And when I find one of them, I'll always ask: "Why was it put here?"

And the answer is usually: "I just felt like it." (laughs)

I know, or "I just thought I'd put it there" is another one. (laughs)

I'd actually be happier if someone said: "It looks good, don't you think?" (laughs)

"I just felt like it" really is the worst possible answer you could get!

I mean, everyone's so busy every day dealing with other problems that they rarely have time to give the stones a lot of thought. They often just stay exactly where they're put.

I see. Now, let's turn again to the game's graphics. When we see a picture of Link, we don't merely see a CG model, or a comic book character, we see a fully-developed character who feels realistic when he moves and has a really unique sense of balance. I personally felt that this style was unique, but how do you feel Aonuma-san, speaking as the person who realised this style?

But it's not just the individual pictures, I believe the amalgamation of everything produces a real power which makes Link unique. There are so many people working on the various designs that there is inevitably a degree of inconsistency. This is where the senior directorial staff, beginning with Takizawa-san, have to step in, and as we were saying just now, decide the points that are essential, focus on retaining them and thereby maintain the game's coherence. But this time round, this was especially challenging due to the sheer amount of these tasks. But thanks to a lot of trial and error they conducted right down to the lighting and the very atmosphere of the game, the whole world really came together well. And the result of all these elements coming together is that Link now moves exceptionally well. The atmosphere in this game reminds me a lot of Ocarina of Time. That game's atmosphere really was in a class of its own and I'm sure fans of the series will agree. The staff here held that as one of their aims and believe that, by using Ocarina as a basis, they've been able to revive that atmosphere in a new form.

The game certainly does have that feeling about it. What do you think of the new Link, Miyamoto-san?

The new Link is truly wonderful, isn't he? It won't be easy to make something as good as this again. Even by Nintendo standards, this is first-rate.

Wow! (laughs)

No really, I think this is something we should all be proud of. The person in charge, Nishimori-san, may be one of our younger members of staff but he's been working closely with an experienced programmer. Aonuma-san and myself have been discussing things with him since the planning stage and have checked everything exhaustively. We would discuss the best methods of programming right down to the nearest millisecond.

So, Miyamoto-san, you've no more complaints about Link?

None at all. But there's plenty I'd like to say about the movement of the animals! I'd be saying: "You call that an animal?" (laughs) "You call that a horse? Go and see what a horse really looks like!"

Yes, that has come up in the interviews before! (laughs)

Ah, they told you about that? (laughs) But back to Link, I really think that for a player character the quality of his movement is unsurpassed. Even right near the end when I spotted something and said that it would be a shame if it wasn't perfect and that we should fix it, everybody was happy to put the work in and get the job done. I said a bit hesitantly: "At least let's fix this", but the team responded with: "If we're going to fix it, let's fix it properly!"

You were hesitant? (laughs)

(laughter)

Also, this Link not only moves very well, but he also has real charm.

Even though he's the player's character, he comes with his own personality, which sounds a bit strange, but I think that his uniqueness probably became apparent right after Miyamoto-san had him throw that goat! (laughs)

This was one time I just had to speak out. With this project I was supervising, and entrusted the planning side of things to other people. But in exchange I said: "Let me throw that goat!" (laughs)

Another act of violence! (laughs)

Well, that was one big goat! Up to that point Link looked a bit doll-like, but once he threw the goat we started to think: "Maybe he could do this" or "Maybe he could throw that!" The things Link could do grew and grew…

Throwing the goat, throwing the mid-level boss…

Let's try not to give too much away! (laughs)

Oh, yes… (laughs)

Whatever the case, this Link feels very responsive.

Yes, and there were plenty of other things I wanted to add! But if I mentioned this you'd probably say: "What? You want even more time!" (laughs)

I probably would! (laughs)

But even so, in all honesty there were other things I wanted to do.

I think what we did was enough, but even that didn't satisfy my desire to do more. If only I could change it just a little bit more... (laughs)

I suppose that's what makes it Zelda.