Miyamoto's Framework
On a forum post I mentioned that what most people consider a "game" today comes out of a system that Miyamoto was the first to systematize, and I thought I should expand it in an entry to work out the idea.

The idea is that virtually all games today fall into that framework, even though it's almost an arbitrary framework, the way that virtually all manga falls under Tezuka's framework even though his was almost arbitrary too.

The framework consists of these elements:

1) The player controls a character directly, and can move that character around. They press a direction, it moves in that direction This can be in 1D (platformers like Mario), 2D (overhead like Zelda), or full 3D (like flight sims). Sometimes the player isn't alive, they can be a block (as in Tetris), but there's something that you can "move" on the screen.

2) The player has to get somewhere, find something, or kill some enemy to win, sometimes with a time limit, but often without one. This is achieved by getting the player-moved thing to the goal through a long series of obstacles. Once done, it's the end of the game.

3) The player can die or lose by doing the wrong thing, such as falling into a hole or walking into an enemy or losing a battle. If they do so, they have to try again.

4) There are checkpoints or the game is segmented into spatial sections, stages, chapters, dungeons, or missions; it's not continuous, there are marked points along the way to the end of the game. If the player dies after completing one of these, they can continue from the start of the last one and try that section of the game over again. Everything in the game exists in space, with coordinates (in either 2D or 3D).

5) There's a gradual increase in the player's capabilities or items or weapons, the player gathers or purchases these along the way to the end of the game, such that at the beginning of the game the player's character is much weaker than that character is at the end of the game.

There are variations of this framework too. Sometimes the player controls a group of characters rather than one (as in a RPG). Sometimes the player does not control the player directly, but more indirectly (RTS and turn-based strategy). Sometimes the player can't die (adventure games).

These five principles are surprisingly consistent, even when one is broken, the rest are kept. Ten thousand great games have been made under this framework. They've become so prevalent and expected that anything not using these principles (or at least two or three of them) is generally not considered to be a game.

But, they're all almost arbitrary. There's no reason to use this framework and not some other framework. You probably could make ten thousand great games under five totally different principles, provided they worked together as well as these five do. Yet only a few truly great games don't use any of the five principles of the Miyamoto framework (some of Will Wright's games for instance), even though there's no reason why you couldn't have more.

Now I'm not saying we should have goalless games or games without any principles at all, just that the principles can be very different from the dominant framework. Other principles could just as easily work, I'll go through the five above and explore some alternatives:

Why do you have to control a character or a set of characters, why can't you control something else, like the level itself, or like the weather, or something indirect like that, rather than something which essentially follows your commands like a robot?

Why do you have to "get somewhere" or "find something" or "kill something" to win? One once-popular alternative was to get a "high score" instead. But there are other alternatives too: why not "arrange a system correctly" to win, as in Will Wright's games? Or why not "bring peace to things that fight each other" or "get things which are peaceful to fight each other" or any of a number of other goals?

Why does the player have to go back and try again if they succumb to an obstacle? Why doesn't losing turn you into an enemy, what you were just fighting? Why doesn't it make the screen gradually darker? There are a variety of ways to lose and a variety of obstacles, and "going back to the start to try again" is only one of them.

Why do games need to be sectioned up in distinct areas and dungeons, why not a continuous area? Why have spatial maps at all? You can create a great game without a map, you can set it in "dramatic space" rather than "spatial space", or any of a number of other ways to organize the game's content, such as a file system, or the way the internet is organized (there's no spatial map of the internet, just one-way hyperlink warps). Why do the objects in a game need to have a physical location, why can't they be more diffuse, like spirits which are everywhere but nowhere?

Why a gradual increase in the player's capabilities, why not a gradual decrease? Wouldn't that be more challenging (albeit somewhat depressing in theme)? Or why not make abilities and actions contextual rather than having a stock set of them that can increase? Or why not give the player all of them at the beginning, but not tell the player how to use them, so that they have to gradually discover their own latent powers, by experimentation?

It's clear that the framework Miyamoto set up works, and that it was an enormous achievement to find a set of principles like that which worked so systematically and wonderfully. And, as I said, ten thousand great games have since followed that achievement. But there are very likely other configurations of other principles which could work out just as well, and an additional ten thousand great games could be made under a newly discovered system which works as well as this one does. We just haven't found those configurations yet! But once they are systematized they'd be just as powerful as his.

I think the reason most "experimental" games fail is because they simply try to break the five principles of Miyamoto without forming a new system of new principles. They're basically like the "revolutionary" anarchist who is against everything but doesn't have anything to replace it with, they tear down but they don't build up anything new, and even when they do use a new principle, they don't use *many new principles systematically*, which is what would really be required.

The number of possible principles and especially the number of combinations of these is almost infinite, but most combinations will fail, and only a few like Miyamoto's combination will work. By work I mean be flexible enough to support a large enough gameplay variety to allow thousands of distinct games to be made using the new systemization.

If you take Miyamoto's framework and just reject principles at random and replace them with new principles at random, chances are you won't wind up with a new system that works, you'll wind up with a mess. You can't just "reject goals" and "reject the player dying" and make something without them and hope it'll be great, you need to form a new systematic set of principles which work as cohesively together as his do.

For instance, take the early efforts toward interactive storytelling, such as Facade or the Storytron. It's the beginning of a new system, it has one or two new principles, but not yet there, it's probably missing one or two other principles which it needs to truly compete with Miyamoto's framework. I can see it beginning to form, but it may take another ten years or so before it's systematized correctly. But I think that when it's systematized, its scope will be larger than Miyamoto's framework's scope.

Will Wright's "sim-system" framework has a small but interesting scope, because it allows potentially dozens or hundreds of distinctly great sim-system games to be made. But such games aren't for everyone, they're hard to pick up unless you are interested in the system being modeled, and they're inherently educationally oriented rather than dramatically or emotionally orientated, they're about the complexity of physical systems, which is a great thing to know about, but there are only so many complex physical systems to understand before you know them all, so I'd say the sim-system framework's scope is slightly smaller than the Miyamoto framework's scope, even though it might not seem so because it's less explored than his scope has been.

There are also more ancient game systemizations than Miyamoto's: puzzles, board games like chess and checkers and mah jong, card games like poker, and the like are one systemization first set up thousands of years ago involving tokens and objects with different values, and which was dominated over what we believed "games" to be for an incredibly long time. Sports, physical sports like baseball and soccer, are yet another systemization that existed in parallel to that one. Neither translates to computer gaming very well, although puzzles do so much better than sports, because they don't rely on physics, they rely on simple mathematics and geometry.

But those are just a few potential alternative game systemizations which have been discovered so far.