This 1989 discussion between Shigeru Miyamoto and Yuji Horii reveals never-before-seen details concerning the development of Dragon Quest IV and A Link to the Past. It also touches on each icon’s respective views on the future of game design as well as their predictions for what gamers might see in the future. It was retrieved from Game Staff List Association Japan, a Japanese website that, among other things, aims to summarize, transcribe, and categorize interviews with video game developers. Let’s introduce the participants!

Shigeru Miyamoto

The creator of a number of world-renowned video game franchises such as Mario, Zelda, and Donkey Kong joined Nintendo after graduating from the Kanazawa College of Art.

Yuji Horii

The man behind Dragon Quest got his start after placing in an Enix software contest he was covering as a freelance writer after graduating with a degree in literature from Waseda university.

The Discussion Begins

Horii: (Dragon Quest IV producer) Chida’s really on my case right now. *laughing*

Miyamoto: This must be your busiest time.

Horii: I’m thinking I’ll try to wrap things up within the next 2 months, though.

Miyamoto: So, that must mean the game structure and everything has been decided on.

Horii: Yeah, that part’s all done.

Miyamoto: From what I’ve heard, you’ve been really messing with the structure in order to make IV.

Horii: I actually think that the game structure was perfected around the time DQIII came out. The series gradually made the shift from the single protagonist of Dragon Quest I to the party of Dragon Quest III. We incorporated a variety of other elements as well. We could have just changed up the game’s scenario, but that would have been boring. I wanted to make a leap forward with the game structure itself, rather than just having a different story. If the changes we make are too drastic, though, some players might reject them. In order to avoid that, I did multiple scenarios and made the game beatable due to the simplicity of its structure. I think the best course of action is to supplement what you have and add a few new things as well.

Miyamoto: Is it true that the battles represent the most drastic divergence from DQIII?

Horii: The scenario and the battles, I’d say.

Miyamoto: Do you go around collecting party members? Or are they all part of your team right from the beginning?

Horii: Haha, good question… There are 10 party members in all, however.

Miyamoto: And how do they behave during battles? Is there a character who will just run away whenever he feels like it, for example?

Horii: They won’t run away on you. There might be someone like that in the game, but your main battling force won’t make a run for it. If they did, it wouldn’t make for much of a story.

Miyamoto: Does that mean the 2 supporting characters might run away? And if they do, they’ll be gone forever?

Horii: Nah, I don’t really want there to be anything in the game that can’t be undone. For example, if there was a character that ran away and never came back, I have a feeling you’d be left regretting whatever you did to make that happen. “Oh, man! I made a huge mistake and then I accidentally saved my game!” That’s not what I want. I think that, no matter what happens, you have to be allowed to undo it.

Miyamoto: The protagonist is a hero, right?

Horii: That’s right.

Miyamoto: Does that mean that during the battles and stuff, players won’t be able to do anything that’s not heroic?

The 1991 NES board game Itadaki Street Special

Horii: I’m keeping that a secret. *laughing* We’re developing a board game alongside DQIV that’s going to fall into a completely different genre than anything you’ve seen before, so it’s worthwhile for us to do something different.

Miyamoto: I always have a backlog of 4 or 5 games, all being developed at the same time. I’m currently working on 4 Super Nintendo games, from relatively simplistic titles to huge productions. And that doesn’t even cover all of it.

Horii: I heard that Zelda 3’s going to be for the Super Nintendo. Have you already started working on it?

Miyamoto: Yes, we’re making progress, little by little.

Horii: What’s it going to be like?

Miyamoto: Basically, I intend to make a return to Zelda 1‘s style. This is something I’ve had in mind since even before we began making Adventure of Link.

Horii: That’s awesome.

Miyamoto: Ever since I started making the first game in the series, I’ve been saying that the 3rd Zelda will feature a party, one that consists of the protagonist, who’s a mix between an elf and a fighter, a magic user, and a girl. The fairy that appeared in Adventure of Link was actually a party member designed for Zelda 3. A girl who looked a little like a fairy and whose role consisted of reconnaissance. Like the characters in action games that don’t engage enemies in combat but rather go and scout out the surroundings and return to you safely. It’s also fun when an action adventure game lets you choose who to send out. That’s the sort of thing I’m thinking I’d like to put in Zelda 3.

Concept art for Adventure of Link

I’ve never been too particular about the story in the games I’ve made in conjunction with (Earthbound creator) Shigesato Itoi. The stories of Mario and Zelda titles have always been supplemental to the actual gameplay. Action games only have stories attached to make the experience more interesting. Itoi is the one that writes the story, and I just help out a bit.

Horii: So, development is moving along?

Miyamoto: We’re more or less finished already. Now we just have to wait until release. But the factories won’t make any copies for us.

Horii: Really? Even though you’re producing it in-house?

Miyamoto: They have to make a ton of copies of Dragon Quest. I’d like to wait until that one comes out before releasing the game. Once Dragon Quest fever sets in, this year is done for. Curiously, staggering the releases affects the demand for the Super Nintendo.

Horii: How has it been, making your first RPG?

Miyamoto: We started out developing it as an action game, primarily. We’ve been careful to maintain a high degree of originality, noting where we’ve copied something from another game (albeit not substantially) and where we’ve done something completely new. We’re proud of ourselves for developing the game structure. The game structure of RPG titles is already more or less settled upon, and an RPG overworld is something anyone can make. But that’s all the more reason to ask yourself whether it’s good enough to use the same template as everyone else and simply expand the story on top of that. That’s where the challenge comes in. These days, there’s a gap between players who prefer a solid story to having new features and players who prefer having new features to a solid story.

Horii: That’s right, it’s difficult to balance the two. Do you think you’d like to continue making RPGs in the future?

Miyamoto: I hope to make a new game structure for people who can write scenarios like Shigesato Itoi can. Current game structures are being improved through the combination of elements from different game genres. I want to break away from that a bit. I think it would be ideal if I were able to create a game structure that represented a turning point, much like the Famicom RPGs of old, for example.

But more importantly, Mr. Horii… won’t you consider developing for the Super Nintendo?

Horii: Well… I don’t know…

Miyamoto: You really ought to give it a shot!

Horii: Hmm, maybe. I don’t really have too many ideas right now…

Miyamoto: With the way things are now, you can’t really express yourself, right? The Super Nintendo makes it easy. These days, it’s far more straightforward to realize all the things you’ve wanted to do but haven’t been able to. Game development has gotten much easier, in more ways than one.

Dragon Quest’s password entry screen

Horii: I definitely feel that we’re approaching the limits of the Famicom. There are certainly a lot of times during the creation process that I find myself worrying about what will happen if we do. If you were to ask me what I had trouble with while making Dragon Quest I, it would have to be the Spell of Restoration. The amount of information we made the players memorize was relatively small, to an extent. The password system was annoying, however, and I wasn’t in favor of using the idea in Dragon Quest II, since it employed a party system. I said that we should at least have shorter passwords when there wasn’t a lot of information to remember. *laughing* Dragon Quest III made it possible to save your data, amongst other things. We were able to create a variety of characters who were able to change their occupations, for example.

Miyamoto: I see what you mean. The Zelda games use disks, meaning players have always been able to save their progress. Zelda I and Dragon Quest I hit the market around the same time. It’s too bad that you had to resort to using a password back then.

Horii: I also struggled with memory limitations.

Miyamoto: I was amazed by how much data you packed into Portopia.

Horii: Yeah, we were cutting excess data right up until the final stages of development, but there was still quite a bit left. We were 2 kilobytes over, around 2000 characters. We had exactly 1000 messages, so we started by cutting out all the suffixes. “This sentence doesn’t need an object marker,” we’d say. “This emphatic particle can go too!” *laughing*

Miyamoto: I’ve sometimes had to do the opposite. We had 2 bytes left over in Super Mario Bros., so I decided to put something else in the game in order not to waste them. *laughing*

Horii: I’ve had times where I’ve found that ideas I’ve come up with have already been used in other games. There are even things I’ve been wanting to use in Dragon Quest IV that have already been done. Nothing to do but cut them.

Miyamoto: You take something out if it’s similar to something that’s already been done?

Horii: You got it.

Miyamoto: So, what you’re saying is that you wouldn’t hesitate to cut something that resembled what somebody else did, even if you’d thought of it before they had? I have to keep in mind the amount of time I have before my deadlines, meaning that I release things the way they are, regardless of whether they’ve been done before or not. I just think to myself “Oh, so someone else had the same idea.”

Horii: It also depends on the scale of the thing and the ideas that comprise it. If it’s something that doesn’t really matter, I just get rid of it.

Miyamoto: Ghosts’n Goblins debuted in arcades while we were making Super Mario Bros. You don’t get killed in one hit in that game, either. “They’ve gone and done the same thing,” I thought to myself, but I couldn’t get rid of that part of the game structure. It would break the entire game.

Horii: DQIII has a world map. We were pretty far in development when Mirai Shinwa Jarvas came out, which boasted a world map itself. I cursed my luck and called the staff together, but we couldn’t change anything at such a late point. If we’d altered the map system, we’d have had to change everything, so we were forced to leave it as-is. At our company, there’s always a huge discussion whenever something comes up that goes on until we reach an agreement.

Miyamoto: Talk about dependable.

Horii: Each staff member puts themselves in the players’ position and speaks from their perspective. They’ll go so far as to say things such as “If that’s what the game was like, I’d throw it at the wall!” But because they’re speaking from the player’s standpoint, we can only try to persuade them. You’ll hear comments like “The players will hate that!” and “Can’t we just do this?” Everyone gets a say.

Miyamoto: The way we make games is a little different. I can’t comment on the tiny details of the game program or make suggestions for other ways someone could go about what they’re doing. I just tell people what I want to do, and the programmers tell me whether or not they can do it. We try to reach a compromise. 2 or 3 days later they’ll come to me and say “Well, we can do this much,” and then, 2 or 3 days after that, “Well, we can accomplish that if we go about it like this.” It often happens that we end up achieving what we set out to do in the process. *laughing* I have my teams to thank for that. The Mario games have all involved the same group of people. The Zelda team is mostly the same as well, only the director has changed. We start out with 3 or 4 people, and then when we run into trouble we add around 20 more. If we were to start out with dozens of people, there wouldn’t be any work for them to do before we’d decided on the direction we were going to go with the game.

Horii: Our opening meetings are always comprised of 3 people: Me, Nakamura, and Chida. Once we’ve begun establishing the details, the number of people present at these meetings gradually increases.

Miyamoto: When we run into a big problem, we stop the people who are making progress in their work and ask for their help. This is mostly during the last 2 or 3 months of development, however.

Horii: Personally, I fix things I’m not happy with right away. I owe a thank you to the Chunsoft employees for taking care of that for me. They completely redid the map when we ran into trouble there.

Miyamoto: They do it because of their desire to make something they can be proud of, I imagine.

Horii: Adjusting the difficulty of the game is tough as well. Whenever I draw out a dungeon on paper, it always ends up being really hard. I test play it thinking it’s going to be easy, only to find it’s outrageously difficult.

Miyamoto: We always endeavor to reduce the difficulty of the dungeons by 20% once we finish making the game.

Horii: You decide that in advance?

Miyamoto: That’s right. I’m exaggerating a little when I say we always plan to reduce it by 20%, though. Our criteria isn’t that concrete. *laughing* When we’re doing an action game, we make the second level first. We begin making level 1 once everything else is completed.

Horii: I see. We’re actually collaborating with Famibou Tofuya (the pen name of former Famitsu editor Yoshimitsu Shiozaki) to make a board game. We made level 1 before doing anything else, and it ended up being really difficult. So, we decided to make a practice stage. It was still too tough. In the end, the first level we made was used as the fourth stage instead. *laughing*

Miyamoto: Do you think that RPGs and adventure games will become a substitute for novels?

Horii: Nah, I think that novels still have their place. Games are more active. If you were to write a novelesque story for a video game, players would feel that it dragged on and on. The sense that you were the one driving the story would disappear. I think the most important aspect of game design is to immerse the player in the game’s universe and make them feel like they’re actively driving the plot. That’s the reason I won’t risk having the protagonist speak, even though it would make writing the story much easier.

Miyamoto: That’s a common feature of RPGs these days.

Horii: Oh, yes? Generally speaking, I think having the protagonist speak alienates the player. He’s playing as though the character is an extension of himself, so why is his avatar suddenly speaking of its own accord? He’ll be struck with the realization that the character he’s been thinking of as himself up until now is actually someone else entirely. Having the protagonist speak for himself and decide own his own which way the story goes would make players uncomfortable. To tell you the truth, I actually did take that approach once. In Dragon Quest III, you rescue the pepper sellers from Kandar’s cave and they run into Kandar while they’re trying to escape, right? The protagonist, if he’s the head of your party, says “Leave him to us! Run! Quick!” I took that route because I couldn’t see another way around it, but there were a lot of people who were uncomfortable about the fact that the protagonist, who’d been silent up till then, suddenly spoke. It doesn’t matter how much talking the supporting characters do, only the protagonist’s lines will stick in the players’ heads.

Miyamoto: Cutscenes in action games are the same in that regard. There are scenes that make you feel as though you’re the one doing everything, and scenes that make you feel like you’re being pulled along against your will. I actually really dislike taking control away from the player. I want to do everything I can to ensure they feel like they’re in control. Mario grabs onto the flagpole, slides down to the bottom, and enters the castle on his own, right? I don’t like that at all. I want to let players enter the castle themselves, if possible.

Horii: So, action games run into that problem as well.

Miyamoto: R-Type’s cutscenes are really good. It’s like you’re sitting in the dock the whole time. It’s easy to grasp when you’re able to move on your own, and you can fire bullets whenever you want. Just between you and I, I don’t think Adventure of Link had very good cutscenes. You feel like you’re watching things happen rather than achieving them yourself. Exactly what you were saying, pretty much.

Horii: That’s what makes games different from movies or novels. If you make it work, you won’t alienate the players, and it’s possible to make them feel like they’re actually there.

Miyamoto: There’s also a big difference between feeling like you’ve figured out something on your own and feeling like you came upon the solution by chance. It’s really difficult to give players the impression that they’ve solved a puzzle themselves in an adventure game that consists of choosing commands from a menu.

Horii: You can’t let them solve it by accident, and it’s no good if they still can’t figure it out after trying everything. It’s a huge bother to go through a list of commands.

Miyamoto: Games that have you choosing icons or options from a menu allow players to stumble upon the solution. Did you worry about that when you were making Okhotsk?

Horii: Because you’re guided through the game and not given a lot of choices, I don’t think players felt like they were being pulled along. I actually decided to make a catch for the PC release. The PC version of Okhotsk was my first multiple choice adventure game. If you try out all the commands, you’ll eventually come upon the solution. So, what would happen if you weren’t able to do that? I was thinking about penalizing the player if they did something wrong, but I eventually gave up on the idea. If someone is really going to go to the trouble of trying everything, I might as well hand them the solution.

Miyamoto: So, you thought it was fair if players like that got a reward for their trouble.

Horii: When you think about it, there are always going to be some people who want to take the shortest path and some people who want to meet everyone and do everything. That’s not a bad thing, is it?

Miyamoto: Right now, I’m working on the sequel to Onigashima. It has a command system and I can’t help but worry.

Horii: Personally, I want to make titles that have a story that draws the player into the game’s universe rather than titles that fit nicely into a genre like “RPG game” or “adventure”. Today’s RPGs are incredible, in a sense. They get a lot of publicity and suck you into the story. I wonder, however, if we can’t take them a little further. I think one example would be an RPG game that made use of a network connection.

Miyamoto: There’s been talk at Nintendo about making a Famicom network. However, speaking as a player, it’s not going to happen as long as the biggest problem of the telephone age remains unsolved. People will go crazy for a game with a network connection once it’s released. Then they’ll see their monthly telephone bill and realize that they can’t afford to play anymore. NTT is going to have to change the way they look at phone bills and figure out how they’re going to separate the cost of connection fees from regular phone use. As long as phone bills remain fundamentally unimproved, it’s going to be a hard thing to pull off. Until NTT stops thinking that connection fees fall under the same umbrella as general phone calls, the potential networks have to reach a wider audience will be limited.

Horii: I don’t think people will start making games that utilize networks until they extend their scope. There’s no use in making something only a small number of people will play. Once networks can be accessed by a bigger audience, that’s when game worlds start getting interesting.

Miyamoto: It’s boring if only a certain subset of people can play.

Horii: Yeah, I agree. I feel like that’s the trouble with today’s computer networks.

Miyamoto: Anyhow, once the day comes that it’s easy for the general public to access different networks, I think we’ll be able to pull off something interesting.

Horii: I think there are a lot of ways to go about making games on the Famicom. It’s interesting to think about what rules you’re going to have when making a board game, for example. If people get tired of RPG battles, maybe new games won’t have any. If everyone makes fantasy games, people might get sick of those as well. Then another type of game world will come onto the scene. Instead of each genre stagnating, people will release games that push their boundaries. RPG games, action games, simulations… I feel that RPGs might be split into 2 types, for example. In one type, the story will take precedence, and people will take it very seriously. In another, only the world will be established, and you’ll get to eke out your own existence. Wouldn’t that be interesting? To have an RPG that allowed you live your own life?

Miyamoto: Sounds like fun!

Horii: We could even take RPGs a little further. Up until now, the computer has been the master of the game world. I think it would be neat to have an RPG where the player was a god. You’d be able to plant seeds by giving information to one of the world’s many inhabitants. Then you could have fun by observing what changes occurred the next day as a result.

Information on the 80s-era Puppy Love is scant.

Miyamoto: I wanted to do a game that revolved around raising a child. I might be ripping something off by saying this, but your kid would start off not knowing anything and not being able to speak and you’d teach them everything. If you taught them something contradictory, it would cause a disruption and you’d get to see their reaction. They’d keep getting smarter. Just as I was thinking this, though, a game called Puppy Love came out in the States.

Horii: A long time ago, before I made Okhotsk, I had the idea for a game where your partner was a robot that gradually gained new memories. You’d raise him RPG-style. I imagined it would be interesting to have a game that was two-sided. If you gave it an order it didn’t understand, it would ask you what you meant, and you’d tell it what you wanted it to do. Then, next time you gave the command, it would do it, growing smarter and smarter.

Miyamoto: How about a game where you get to be a mother-in-law who bully’s your son’s young wife? It’d be like in Star of the Giants where the wife wouldn’t submit to you and you’d have to compete with her by trying to throw her out of the house within a certain number of months.

Horii: That’s one kind of RPG, alright. You play a certain role. I think it would be neat to have a really tragic RPG as well. Because it isn’t real life, everything you do goes wrong, and you get to marvel at how bad the situation becomes.

Miyamoto: It would be fun to see just how far you could go with it.

Horii: “My wife walked out on me! Where did she go?! What should I do?” Talk about funny.

Miyamoto: I think people would accuse us game designers of being cruel if we were to do that.

Horii: I just want to relax once DQIV comes out.

Miyamoto: I’d like to do a game for the Super Nintendo that even fathers can enjoy. Something that makes people criticize me, wondering why on earth I’d make a game like that in modern times. Something that appeals to us dads.

Shigesato Itoi’s Bass Fishing No. 1