What is a roleplaying game?

At the start of this year, I made a post wherein I attempted to work out my philosophy of game design. (Philosophy meaning "the critical study of the basic principles and concepts of a particular branch of knowledge, esp. with a view to improving or reconstituting them") Though it helped to advance my thought process, I can't say I succeeded in reaching a clear thesis at the time. The other day, while explaining roleplaying games to a non-geek, I realized I now had a clearer idea of just what (in my mind) makes something a roleplaying game:

A roleplaying game is a form of group storytelling with rules that dictate the mechanics of player success and failure. These rules enhance the experience in two ways: First, they make the story more compelling by providing concrete gauges of a situation. Second, they provide the player with a series of interesting strategic choices, turning the experience into game.

Read on for a more detailed description of what the different parts of this definition mean, as well as an explanation of how they relate to GNS theory.

"A roleplaying game is a form of group storytelling. . ."
I could get into the use of a GM here, but that's a common aspect, not an essential/defining one. Same reason that I don't mention, say, the use of random chance.

". . .with rules that dictate the mechanics of player success and failure. . ."
I had my doubts about this phrasing- I was imagining the possibility of an existing or future RPG which "killed the sacred cow" of rules dealing with player success. But then I realized that without the possibility of varying degrees of success, you can't have a game- it would just be a roleplaying exercise.

". . .These rules enhance the experience in two ways. . ."
If you want to get technical there certainly other ways that one can use rules to hypothetically improve the player's experience in one way or another. But the following two are the only ones I consider significant and defining.

". . .First, they make the story more compelling by providing concrete gauges of a situation. . ."
For example, your current level of injury. It's not actually necessary for the player to have access to a given gauge- for example, in Unknown Armies the GM keeps track of everyone's hp and just gives players rough descriptions of how their characters feel. The point is that the players know there's a gauge measuring their character's health- that there are procedures in place which will dictate what happens, even if that procedure is sometimes just "it's the GM's call".

I'm actually tempted to split this bit into two subcategories- rules as a source of consistency and rules as a frame of reference to help us understand a situation.

". . .Second, they provide the player with a series of interesting strategic choices, turning the experience into game."
I was going to say that it makes the experience fun. But fun, defined broadly, isn't essential. A roleplaying game doesn't have to involve a "fun time". It's easy to envision a well-done roleplaying game where the player's primary emotions consist of sadness and anger, and the PCs strive to achieve their goal but still ultimately fail. "Fun", in the sense of pleasant sensations, wouldn't seem like it apply to such an experience- but a different sense of the word would still be quite appropriate.

In his book on game design, Raph Koster says that games are fun because we derive satisfaction from learning how to overcome the challenges they present us with. Even faced with inevitable defeat at the hands of endless enemies, we enjoy trying to kill as many as we can. In this sense, a game can be simultaneously fun and depressing- imagine playing a game where the goal is for your character to kick as many defenseless puppies as possible within a set time limit.

How this relates to GNS Theory: As I've said before, I think Jon Edward's concept of creative agendas (as seen in both his old threefold model and his current theories) has merit because it identifies three key ways a roleplaying game can be enjoyable, but that his view of these "agendas" as separate audiences that an rpg designer must choose between borders on a violation of common sense-there is no reason why a player can't enjoy both the strategic and dramatic aspects of a given situation. What he views as different tastes, I view as different aspects of what makes a good roleplaying game. I would say that the view he describes as "Narrativist" focuses on the group storytelling aspect of a roleplaying game; his concept of a "Simulationist" view would focus on the ability of the rules to provide consistent, concrete gauges of a situation; and his conception of a "Gamist" view would focus on the game that the rules create, obviously.

I'm not quite done writing about this manner. The above is a definition of roleplaying games; but it doesn't touch on what I think makes roleplaying games special. That'll be another post. . .