(Long, sorry. Just getting back, so my writing style might be a bit rough for a while.)
A while back, I wrote this translation of a Terranova article.
Now, I'm going to tell you what I think about defining a game by its rules.
Some games are defined completely by their rules, but most are not. Most games are heavily influenced by non-rule things, such as what mood a given player is in, or how much they hate another player.
Most games that are defined completely by their rules aren't defined adequately by them. For example, the games of go and chess are defined completely by their rules. There is theoretically a "best play" at any given moment. But the rules produce such a stunningly complex situation that the player cannot actually find a best path. Therefore, the player substitutes mood, emotion, and gut instinct when skill fails.
Above that, all games are run by people, for people. The effect games have on players before, after, and during play isn't contained in the rules. People play the stupidest games - for example, drinking games. They don't play to explore a dynamic play space or judge a complex pattern or whatever. They play because it's fun to hang out with your friends and get drunk, and a drinking game provides a framework to focus all your friends in the same direction. Sharing fun has never been easier.
This idea of directed attention - usually directed enjoyment - is universal to games. If a game doesn't do this, it feels painfully dull, and simply isn't any fun. But the paths games take to direct and keep directing the player(s) are as widely varied as the games themselves.
An RPG uses moderate statistical and story foci to drive the player. Apples to Apples keeps all the players oriented on the same words, allowing them to share the laughs. Competitive FPS games immerse all the players in the same game space, in a "focus or die" situation.
The tricks they use to keep the players focused, to keep them immersed, are also widely varied. In a game like Apples to Apples, the game relies on the natural pull of friends sitting around together. A game like WoW uses that, too, but the pull of friends with only a chat window connecting them is significantly less, so WoW uses a lot of other tricks to keep its players immersed together.
"Speed" of these tricks is also important. Fast, widely variable games like Apples to Apples can play it fast and loose - because they are fast and loose. But if the turnaround isn't very fast, then the friends-sitting-around focus doesn't work as well. Think of it as each time a friend says something, it adds up. So, if they don't get as many chances to say something, it doesn't add up as fast.
Now, your friends could say things seperate from the game. But those things aren't focused by the game and aren't reliably a part of the game's pull. So they shouldn't be counted.
Every focus is like this, but usually with another form of time pressure: a given instance of a focus takes a specific amount of time for a player to comprehend. Moving instances too fast will confuse and frustrate the player, whereas moving them too slowly will bore her to tears.
Example: there are a million games where you fire little colored balls at other little colored balls to make them vanish. These games generally move pretty quick, because the games are not usually complex. On the other hand, Freecell is generally played pretty slowly, because the situation of the game is fairly complex and needs to be thought about in some depth.
Some games use a timed progression to force the players to progress. This generally is to make up for an otherwise weak play experience or to delay games so that slower players don't get left behind. There's nothing wrong with that. Most games use a "when you're ready" progression - allowing you to think about your play for a reasonable (or unreasonable) amount of time and allowing you to make your move whenever you want. You had better have a pretty good focus to allow that form of voluntary progression. Even Apples to Apples has rules to keep players from dawdling.
Where am I going with this meandering mess?
Well, what are the rules of the games?
They are simply attempts to make the play deep enough, complex enough, fast enough, and/or shared enough to keep all the players directed into an exciting shared experience. Even if there's only one player!
This means that talking about a game in terms of its rules is a bit like talking about a movie in terms of the wattage of stage lights it used and the brand of camera. But even the most diehard filmie doesn't talk about a movie solely in terms of its technical manufacture. Instead, even the most geeky filmist will say things like "exciting" or "moving" or "catastrophically pathetic". A light, a brand of camera - these things can and should be known and used, but they are used to focus the audience more carefully, and that focus is what everyone talks about.
When we make a game, we often think of something specific we want to do. A cornerstone. For example, in B1nary Her0, it's a goofy, geeky spinoff of a goofy hit game. Darius didn't go "it should make the players feel like rock stars", he said, "boy, we could do hex codes on this guitar!"
But according to what I've seen of the designers of Guitar Hero, they thought in terms of focus. They said, "our players should feel like rock stars". This served them well, because everything they put in made the game feel more drugs, sex, and rock-n-roll.
They didn't include any real multiplayer support, unfortunately, and I don't think the game is as much fun for a non-player audience as it could be. They thought about the right thing, but too tightly: the audience (and there often is one) should feel like a rock audience. It's the other half of the equation.
The purpose of a game is not to make a player feel a certain thing, or focus in a specific way. It's to make all players share a focus. Even if you design a one-player RPG, you should do things which will get your players talking to each other about that cool scene, or that neat trick.
Once you have that core idea, you can use rules to shape a game. Like cutting away all the stone that isn't David, you start with a blob that has a golden image inside, and cut away all that fat that isn't the image by using rules. Don't cut too much away!
I've always thought of rules as bricks and mortar, but now I think they are scalpels and chisels.
Make sense? Thoughts?