Learning from games: eight virtues of good design

So I was watching this talk by game designers Jonathan Blow (Braid) and Marc Ten Bosch (Miegakure) and I thought, "This is something we can apply to writing," because that's what I do. There are a number of interesting little discussions here but the centerpiece is a list of eight virtues of good design according to a certain aesthetic both Blow and Bosch advocate as one good option among many -- one way to shrink the space of possible games such that you are more likely to create a good one. You won't have to watch it to understand this post, but here's the talk:

So now I'll list the eight virtues and describe how I think they relate to the problem of writing a novel. These rules wouldn't be shocking to good game designers and their applications in writing won't have shocking consequences either, but they can serve as a useful and memorable heuristic for making decisions (again, reducing the infinite space of possible works to something more manageable).

1. Richness. This one seems like it would require the least explanation. They describe "richness" as the virtue of selecting a space rich with potential consequences for you to explore. But the key here is selecting a space. The richness of consequences will emerge from rules -- in a game, these rules are called game mechanics. (Mario can jump this high, but he can jump this high if he has a running start.) I think a lot of writers find rules frightening because they limit possibility or because we are trying to write toward "the truth" and any limitation will perhaps stand between us and that truth. I think it's common to underestimate how many constraints any good piece of writing necessarily contains, and how these limitations ultimately create a rich space for exploration. Your characters, your choice of tone, your method of structuring scene and chapter and etc., all constitute rules (or even game mechanics: we are playing a game with the reader). In short, while it might seem that richness suggests excess and maximal inclusion, we actually need to be selective about the elements we include, or the novel will not be rich so much as an incomprehensible blur, a smear of language. Think about the very real limitations of Pynchon as a novelist: many complain about his flat characters and slapstick humor, but without those elements to manage the text and simplify it, his already dangerously complex fiction would become unreadable.

2. Completeness. This actually sounds like the opposite of what it is. You're not trying to put everything in, you're trying to use everything you put in as completely as possible. They describe it as "completeness of exploration." Jonathan Blow says that he will do this in a game even to the point of removing fun -- presumably because other, perhaps more complicated pleasures become available . Think here about chess. You don't really get anything out of playing one game of chess, and you don't get much out of playing five. At first, when you don't understand the possible permutations of the game, your moves seem meaningless. After several dozen games, you begin to understand the space of possibilities enough that many moves are meaningful. The better you get, the more you know what can happen in the game, the more rare it becomes for you to take a move just for the sake of taking a move. You can't afford to waste anything. The trouble with Chess is that it has too many moves. This is why I can only intermittently enjoy it: I know that mastery would require a lifetime. (This is also the genius of chess.) And of course Chess has evolved to reach its present state over a very long time, passing through the hands of countless people. Most of us don't have that long. We need fewer mechanics, fewer rules, a smaller space; once we have that, we can begin to explore the consequences of what we have.

When I wrote my novel Fat Man and Little Boy (forthcoming from no one at no time, as of yet) I limited myself to two primary characters who would explore four locations over the course of the novel: Fat Man, Little Boy; Japan (especially Nagasaki (after the bomb)), France (unnamed city, south of Paris), France (the concentration camp Gurs, remade as a hotel), and Hollywood, in that order. I also limited my important secondary characters (two women, one man). I made rules about how long a given chapter could be and how many sections could be in a chapter. There were half-length chapters (about four pages), full-length chapters (about eight pages), and double-length chapters (about sixteen pages). There were also a number of restrictions on the language, in terms of tone, style, and syntax, many of them deeply idiosyncratic. So here is how I wrote the book: I combined these elements in various combinations until I had exhausted their interesting consequences. I asked myself, "Have I checked in with Little Boy recently?" If the answer was no, then I wrote a chapter about Little Boy. I asked myself, "Has he spent significant time with this secondary character yet?" If the answer was no, then I paired him with that secondary character. Sometimes I still didn't know what to do, so I reminded myself of the underlying mechanic for each character: Fat Man was defined by gluttony and guilt (if in doubt, I made him eat something) and Little Boy was defined by shame and a desire for silence (if in doubt, I made him stay quiet against the wishes of other characters). When I had used all the best combinations in a given environment, I moved them to the next one. Do this enough times, put the results in a sequence, you have yourself a plot.

The primary method of advancing story in character driven narratives is to put two characters together who have not been together before. That's really what this virtue of "completeness" is all about.

3. Surprise. Blow describes the desire to make a game surprising as a counterbalancing force to the desire for completeness. You don't have to show everything because we know what results a lot of combinations would yield. By focusing on surprising results, we focus our attention and the reader's on things that will bring pleasure and new information. In a game, this means that Mario doesn't have to stomp a goomba in every possible situation. In your novel, it might mean that we spend very little time with characters who get along and agree on everything: they're not going to surprise us. My character Little Boy only really interacted with my character Rosie, a potential mother figure, when it was too late for her to mother him; because he was too old, and because there were other demands on her as a mother, she had a reason to actively try not to serve as his surrogate mother. Before that, they would have gotten along too well, so I mostly kept them away from each other: a scene where the motherly character mothers the childish orphan character wouldn't tell us anything we didn't know about the characters or their world.

4. Lightest contrivance. When we feel the author's hand too much in the text, we usually lose interest. We like it best when the mechanics of the text guide its outcomes in a way that feels organic. But, as Blow points out, relative levels of contrivance within a game (or, for our purposes, a novel) can matter a lot: if one mechanic is very contrived and another is not contrived at all, that looks weird and ugly. If the amount of contrivance is roughly even throughout a text, that bothers us less. As much as possible, though, we want to let our various rules interact with each other as cleanly as possible, and accept the results as the truth of the novel, even if we ourselves don't appreciate it very much. Unlike a game, the best outcomes in a novel are often (even usually) those that come about because of the rules and in spite of the desires of author and reader. Flannery O'Connor's worst stories are consistently those where she imposes her will on the story in the climax in order to avoid an outcome she dislikes. "Everything that Rises Must Converge" is my favorite example. The old lady's death has nothing to do with the reality of the story and everything to do with our desire to see her punished.

5. Strength of boundary. This is really about knowing the identity of what you're working on. They discuss eliminating unnecessary mechanics. I would say eliminate unnecessary characters, settings, chapters, paragraphs, sentences -- all with an eye toward clearly establishing a voice, style, and identity for your novel. Sometimes you cut things because they aren't you, not because they aren't good. This is another way of saying: revise.

6. Compatibility. This is pretty much the same principle as the last one, for our purposes. Don't add a new element if it doesn't interact with the elements you have in ways that reveal new and interesting things.

7. Orthogonality. Marc Ben Tosch argues that game designers should make sure their game's mechanics are orthogonal -- that they overlap as little as possible. You don't usually want to have two or three mechanics that do the same thing in the same way. In a novel, this often means creating characters that contrast with each other as much as possible. Characters can reveal more about their world and its rules if they have different desires and capacities. For this reason, my last three novel manuscripts have starred comic pairs. The first pair (both cops) was black and white, clean and dirty, careful and careless, empirical and fanciful, respectively. The second pair (Fat Man and Little Boy) had the obvious contrasts (size, apparent age, maturity, hunger, power) and some others (facility with language, virility, "softness"). The pair I'm working with now, two brothers, are contrasted in terms of intelligence, self-control, strength, attractiveness, moral clarity, and fear. Of course this is a very old technique.

You don't necessarily need pairs, though, you just need people that see things differently, or different modes of narration, or different objects for one character or one mode of narration to react with. Language-driven novels without much in the way of character tend to operate by applying one mechanic (one style of language) to a variety of situations and environments: the variety of objects refracts the language and twists it into new forms. (A lot of the writing we talk about here does this.)

One thing Blow and Tosch don't discuss -- probably because it's much more pertinent to fiction than to games, where this issue will often take care of itself -- is that contrasts become most effective when the things being contrasted have a lot in common. The characters, scenes, or situations in a given novel are usually variations on a theme, more alike than different in key respects. (Or often there are several interpenetrating groups of like characters, as in Pynchon's V, where certain tendencies repeat themselves across time and space.) In House of Leaves, the characters are defined by their reactions to the overwhelming reality of the house.

8. Generosity. In Braid, one of the key mechanics is that you can reverse time. You don't have a limited number of opportunities to do this: you can do it again and again. Why not let the player explore the consequences of your mechanics fully? In terms of a novel, this has fewer obvious applications, but I do think it's worth thinking about who is being generous to whom in the case of a novel. In a game, you create an environment and explicit rules for interacting with that environment. You're building a space for the player's agency. In writing a novel, the relationship is much less clear. You're still creating a space for the reader's agency, but you're not sure how that agency will operate, because you're creating the only concrete object that will definitely exist in each writer/reader interaction: they might write a review, they might blog about the book, they might write on the page or tear it up, but they might just read the book and think about it for a little while. So here the generosity has to extend to the person with the more concrete forms of agency -- the writer. You have to trust that if you explore what's interesting to you about the space of your novel, your readers will be generous enough to allow you the time and language that you need to do that. At the same time, you want to remember to allow your readers as much space as possible to experience the book in the way that they want while still maintaining the integrity of your story (such as it is).

Conclusion: Blow discusses how much pressure these rules have taken off of him as a game designer. If the game isn't fun and it doesn't make him rich, as long as it's guided by these virtues it will probably have something valuable in it. I feel the same way about the rules by which I write. So far I haven't published any of my novels, so I haven't been successful in a lot of key respects. I don't know if my books are good or fun or whatever. But I can have some confidence that by following my own rules, I did something that was potentially beautiful according to its own measures. For now, that's enough. As a general rule, I don't know what to do if someone tells me to say something smart. But if they tell me to say a sentence with seven words in it, I can probably accomodate that. These sorts of rules can't guarantee you a great game or a great novel -- but they can limit the space of creation such that creation becomes possible, and provide heuristics by which to judge your decisions.