February 17, 2004

An Economy of Rules (part 3) – [musings]

Part 1
Part 2

Rules of structure

Proposed structure rule: Nonlinear plots are better.

In the old Scott Adams Adventure game, there was a dragon sleeping in the woods. If you wanted to get rid of the dragon, one possible solution was to take the wine bladder and fill it with swamp gas, then use the bladder as a bomb to blow up the dragon.

It turns out that you don't want to do this, because the resulting explosion will also destroy the dragon eggs, a treasure you want to collect. But the game lets you do it anyway, with the result that you can't win the game.

The full solution to Adventure is less than a page long. Each puzzle has one correct solution, but the puzzles can be completed in a flexible sequence, and so from a close perspective, there are a number of possible ways to win the game, with each solution being some permutation of the same elements. Because of those permutations, you can't really say that the game has a linear plot. Taking each puzzle to be an episode in a story, where each puzzle has only one solution, it is possible to interpret the episodes as linear episodes. Although the plot is non-linear from a close perspective, it isn't a particularly interesting kind of non-linear; and since the episodes are all linear, most people would refer to the game as having a linear structure — where linear has come to mean uninterestingly non-linear.

(I am bulldozing over most of the interesting nuance of interactive fiction here. Nick Montfort's "Toward a Theory of Interactive Fiction" should provide a good counterpoint.)

In 1997, Westwood Studios released the Blade Runner game, which was based on the 1982 film (in turn based on a novel by Philip K. Dick). In the game, which had some stunning as well as some not-so-stunning visuals, the final outcome of the game was determined by how you resolved the conflict with the characters. You could play the loyal blade runner, a kind of mercenary cop who makes his living by "retiring" replicants who are on Earth illegally, or you could empathize with the replicants and consequently be forced into exile. The Blade Runner game has a truly non-linear plot. It was also, in many ways, an un-loseable game: so long as you didn't get arrested or get yourself killed in a fight, you couldn't make any incorrect choices when advancing the plot. Play the game long enough and you would always get to some conclusion.

(Note: The default configuration for the game doesn't allow the user a choice of topics when talking to other characters. When you click on a character, the game chooses the topic based on some hidden logic. This makes it very difficult to steer the story along certain paths without figuring out which actions trigger which conversations.

With the "advanced" conversation option, it is still difficult to tell how the chosen topic will affect the story, e.g. selecting the "Voigt-Kampff" topic might lead you to a Voigt-Kampff test with that character, or it might not. The result is a game that was too confusing to appeal to a large audience, in spite of its non-linear storyline.)

The overall plot in Blade Runner was still limited to one of twelve(?) possible endings, six(?) of which were very similar. The plot was non-linear by way of branching. Is a braching plot tree interesting enough to be called non-linear, when the permutational plot of Adventureland isn't?

What strikes me odd about this stucture rule, after looking at the above two games, is that there isn't much of a history behind the preference for non-linear plots. In the history of computer games the vast majority have linear storylines.

Where did this idea about non-linear stories come from? I think it is based on a poor syllogism:

  1. novels have linear plots
  2. games are not novels
  3. games have non-linear plots

Looked at this way, the appeal for non-linearity is an appeal to maintain a certain distinction between novels and games. Competition between novels and games is a conflict on the level of form. It is likely this isn't just a conflict with the novel, but a conflict with film, with comics, and possibly with music. There is an expectation that games, by the nature of their form, can provide an escape from linear structures, structures which so seriously limit other forms.

There is some trouble here, because I've just described games as having a natural form, yet the structure rule under discussion is a rule of what games should ideally be. Few games meet the criteria of the rule, yet there are undoubtedly a number of things out there which we call games, which are considered to be very good games. What, then, is the function of this rule? I think it serves to both prevent games from devolving into non-games, and to push games to evolve into something else. The former is an economic role: maintain the quality of games. The latter is a more commercial role: keep pushing games into new areas, to create new markets.

(continued ...)

Posted by B Rickman at February 17, 2004 07:18 PM | TrackBack

"Proposed structure rule: Nonlinear plots are better."

This rule is for designers who would rather throw in with a pat, sweeping absolute than acknowledge that games are an art medium where every creative choice can be made to work depending on the artist's intentions and the context of the piece. I say that not to establish some sort of intellectual happy land where nobody is wrong, but to point out that the linear / non-linear spectrum (a very complex and hotly debated issue you have chosen for your discussion of rules - did you really intend to dive into this issue in earnest or keep the focus on the meta-issue of rules?) is one of preference rather than empirical truth. For this reason I'd say your rule is more of a craft rule than a structure rule (game design IS a craft, though it is often ignored or subsumed into programming and art content creation), and a flimsy one at that, tantamount to a Rule of Visual Design that reads "Too much red is bad".

Intellectual curiosity might motivate us to say that strongly non-linear game structures are "better" (as strongly linear design is a very well-explored field while the most interesting frontiers of open-ended games are, most likely, still to come) but intellectual maturity would require that we view it in terms of costs and benefits. Anyone who advocates either perspective as the "proper" way to design a game is limiting their options and attempting to reduce an art to a craft - this happens all the time in the mainstream game industry though, where any creative decision that leads to fewer sales is a bad one and is rigorously codified as such.

Also, why the use of the word "plot"? Does that concept even apply here? Games only become stories when you talk about them in past tense, a game session that has already been played.

Posted by: JP at February 23, 2004 08:11 AM

Here I'm using "plot" as a convenient synonym for "story", and using neither to talk about the user experience but rather the story embedded into the game. I know this is all part of the big messy debate about games and stories, but for the moment, on this blog, I'm ignoring the positions of that debate.

And I chose this rule largely because it is such a messy one. I think it is a structure rule because it deals with relationships between elements; the construction of those elements requires craft, but where those elements go in a larger framework is a structural process. I would include level design and game balance in structure as well, because although they require some crafting, they require more in the way of structuring, in looking at different arrangements and choosing the most interesting ones.

I have a larger argument which I am trying, hopefully, to outline with these essays. The argument is that the purported rules of game design act to keep games, as a form of expression, from turning into some other form. The rules are conservative, they act to maintain games in a certain state. But different people have different perceptions about what games are, so while everyone is trying to preserve the game form, they are actually pulling the form into several directions. When someone insists that this is what a game should be, that plots should be non-linear, they may end up with something that doesn't classify well as a game.

Posted by: Brandon at February 23, 2004 12:49 PM

I like some of the analysis here, but with regards to your larger project, I have to ask: who is actually proposing such a rule? Or at any rate, a rule as simple as "Nonlinear plots are better"? Iirc, Ed Fries was instrumental in installing greater nonlinearity of a sort in Crimson Skies, and I think other games, although it's unclear to me that this was really motivated by an essentialist idea about what games are. And in general I'm not sure that this is the case, either. As JP noted, highly linear games are well explored, quite prolific, while the gaming masses are starting to feel the constraints of such design. It seems to me that market forces, in attempting to accomodate both a rising demand as well as differentiate product, would adopt a *mantra* that nonlinear plots are better, but not as a general rule of design.

Posted by: Walter at February 23, 2004 01:35 PM

"The argument is that the purported rules of game design act to keep games, as a form of expression, from turning into some other form. The rules are conservative, they act to maintain games in a certain state. But different people have different perceptions about what games are, so while everyone is trying to preserve the game form, they are actually pulling the form into several directions."

I see your point. Definitions and rules are indeed potentially dangerous things and can be created and wielded in the service of an agenda. That bothers me as a creator because I feel the only "correct" direction is Every Possible direction. Rules and axioms like that limit what is possible. Definitions are merely lenses through which we can view a subject, but those can also be proscriptive in a way that disregards valid avenues of study.

I'm coming to believe more and more that the central conflict in the study of games and game design is not the spectral "narratology VS ludology" but rather a constant and low-level lack of understanding between those who are studying games as they *could be* (academics, perhaps, and designers who want to push the boundaries of what is possible) and those who are studying games as they *should be* - as dictated by market forces. Like I said, some of the fondest advocates of "game design rules" just want a fool-proof formula with which they can make endless piles of blockbuster games.

Posted by: JP at February 23, 2004 03:44 PM

Yes, and if the movie industry is any indication, there will simply be no way to drive out rule-based ideologies of production.

Posted by: Walter at February 23, 2004 04:07 PM