Kind of. All of these three points can be encouraged by the game rules themselves and they can all be encouraged by the group. Drama points are an actual rule in Buffy that encourages Narrativist play. Hit-location tables are actual rules that encourage simulationist play.
GNS Theory is a little outdated, and it has its flaws. But it does provide a pretty useful foundation for discussion.
My own little pet theory uses GNS as a part of it, but I’m not convinced that GNS is functional enough to really work by itself (the Big Model Theory agrees in this sense, as it grew out of GNS). Rather, a useful way of breaking down an RPG has to do with, well, what an RPG actually does, and what its components are:
1. Agents (Characters) with varying levels of competence in different areas of expertise
2. Challenges (creation and resolution) for those characters to face
3. Participants (generally at least two) who will create characters to face the challenges
GNS can be useful in describing these different aspects on an RPG, but they really aren’t sufficient to describe an entire game itself (or at least they can’t do it in a useful way). It’s more helpful to identify a particular mechanic or gameplay practice as Gamist, Narrativist, or Simulationist to try to understand what a given game is trying to do or to help understand what you want your game to do.
I’ll give you an example from my game, d20^. There is a rules subsystem in the game called “Dramatic Interactions” which is intended to be used for non-combat situations that really need additional focus, and are run as a series of skill-checks. Say there’s a chase. At the start of the chase, the hunter gets five dice and the target gets five dice. The two sides then take turns naming skill challenges. The Target might decide that he wants to try to hop a fence and lose the hunter that way, and the Target will do so by swinging over the fence on a low-hanging fire-escape (making an Acrobatics check). The Hunter then says how he’ll respond to this (in this case, maybe he’ll just climb over the fence with Athletics). Each side wagers dice (up to three… one is a small risk, two is a moderate risk, and three is a major risk) and rolls their wager (you don’t need to match the other guy’s wager), selecting the highest result from your pool. Whoever wins the opposed check pockets all of the dice, and can use them on later checks in the encounter.
Surveying this with GNS, the overall layout of the rule is pretty Gamist. Wagering and betting tends to be something that people enjoy, as well as rolling a lot of dice, and going head-to-head with this wagering and trying to outwit an enemy are all pretty heavily gamist. However, the actual structure of each individual check is strongly narrativist: the players are encouraged to essentially “paint in” details of the location for the chase on the fly, adding fences to hop, ladders to scramble up, etc. The GM doesn’t have the burden of mapping out the whole chase-route, and can instead focus only on what is particularly appropriate for a given moment during the chase itself. The mechanic as a whole is weakly simulationist, as the pool of edge dice which the hunter and the target hold are meant to provide a rough abstraction of how close the target is to escaping or how close the hunter is to catching the target. As you start to lose ground, you start to run out of options, and it’s harder for you to escape or catch up. However, simulation definitely isn’t the goal here, as abstraction is more favored in order to encourage narrativist and gamist play.
And something I missed in earlier posts: no one part of GNS Theory is “right” or “wrong” or “good” or “bad”. In gaming more than other areas, different strokes for different folks should be a fundamental of the community.
