In a role-playing game, you customize your character’s abilities, refining that customization as the game progresses. Your character will typically have a class (fighter, mage, priest, rogue, etc.), which you pick at the start of the game and generally doesn’t change; this sets broad limits of your character’s abilities. You also typically have cross-class attributes (strength, dexterity, intelligence, wisdom, charisma, constitution, etc.); these increase as you level up, letting you focus on areas that are more important to you. As you level up, new class-specific abilities also appear: for example, a mage may learn new spells as she levels up. Sometimes these new abilities are determined for you; sometimes you have a range of potential abilities to learn that you can choose from. And sometimes there’s further branching: for example, in the most recent role-playing game I played, Dragon Age, each class has four specializations (of which you can eventually choose two), and each time you level up you’re able to pick one of a dozen or so abilities to learn, some of which are tied to a specialization, some of which are tied to your class, and some of which are available across classes. And that’s not all: some games allow you to change classes if you level up far enough, and some dispense with the concept of class entirely, allowing you to mold your character as you choose based on the attributes and abilities that you select.

In general, I’m drawn to these more flexible games: I like the idea of shaping a character according to my playing style, according to the way I like to behave, even according to my idea of the sort of person that I am. Even in these more flexible games, though, classes are latent: not all combinations of choices are equally powerful, so if you search forums or wikis or gamefaqs, you’ll find recommended builds, giving you combinations of attributes that work best to, for example, allow you to do the most damage in the game, or absorb the most attacks, or heal your party members the most. (The presence of multiple members of your party encourages this sort of specialization.) You can ignore these, but you do so at your peril: you’ll have a hard time making it through the game on hard difficulty settings if you’re haphazard about the choices you make.

This focus on character construction is one of the ways in which role-playing games are adolescent games: they’re all about figuring out whom you want to be when you grow up. Which, like all great themes, actually spans all ages. Or maybe I’m stuck in adolescence, because I’m still trying to figure whom I want to be when I grow up!

And, just as I prefer more flexible approaches to building your character in role-playing games, so too do I prefer that when constructing my life. I’ve already changed classes once: while I don’t talk about it much here, I leveled up for years as a mathematician, eventually spending five years as a postdoc at Stanford. At which point I changed classes, becoming a programmer. (Fortunately, I’d leveled up that skill while growing up and during my undergraduate years.) One debate that I’ve had over the last few half-decade was whether I wanted to change classes again, and become a manager; I dabbled with that for a few years, though these days my general feeling is that I’m a better fit as a programmer. (Actually, my real feeling is that I’m a better fit as a member of an anarchist collective, where either nobody or everybody is a manager, but never mind that.)

But, of course, not all programmers are alike: and life is definitely the sort of game where you have to choose where to improve your abilities, you don’t have preset skills that you gain as you advance. (Outside of college curricula, at least! Which is one of the reasons why I don’t like them, I suppose.) I have a habit of going deep enough into some area to be able to do pretty well at it, and then switching over to something else rather than going all-in. (The exception, I guess, is agile: it’s both deep and broad enough that I haven’t yet gone all-in with agile, but I also haven’t stopped exploring further.) I switch domain areas fairly frequently: programmer tools, then streaming video, then video games, and now computer security. And, as a bonus, I don’t even go all-in on programming as a whole: yes, there are 208 posts on this blog tagged with ‘programming’, but there are 327 tagged with ‘video games’. (Similarly, I read a lot more novels than math books when I was a grad student. And, going back to my undergrad years, there’s the whole ‘Sanskrit major’ thing.)

That’s my life; what would games suggest will be the outcome? When I make a build like that in a role-playing game, it’s fun enough to play, but never really optimal for progressing in the game. So I’m pretty sure that, if you looked at the great strategy wiki of life, you wouldn’t find my particular career trajectory on the ‘recommended builds’ list. Which is something that I’ve been thinking about with this latest job change: in fact, that’s as good a way as any to thinking about the last parts of my recent “job search and narrative” post. As a result, I’m now focusing on a build that will let me do well with startups, that will present me as an agile server-side programmer who can deal with large amounts of data, and whose checkered past will make him ready to face with equanimity the sorts of surprises that happen frequently in startups. I’m still pretty far away from the recommended startup build—I’m 40 years old rather than 24, I’m not going to be putting in 60–80-hour work weeks—but there’s a certain strength and coherence to it.

The truth is, though, I’m better off not following a recommended build. A recommendation isn’t made in a void: and recommended builds are designed to let you progressing through the game as smoothly as possible. Games have metrics that you can use to measure that progress; life has metrics as well. You can choose which ones to pay attention to, which ones to ignore; and even if you are paying attention to a metric, you can choose how to respond to it. Returning to Dragon Age, that game has per-party member approval meters: you can try to raise them all as high as possible, you can try to raise a few specific ones as high as possible, you can even try to lower them all if you’re feeling particularly perverse! Or you can step away from the idea of controlling them, and instead see them as gauges reflecting your personality, reflecting your nature, reflecting your character. My build may have made certain battles tougher than they would have been otherwise (though, to be honest, it’s probably really my RPG-playing skills that are to blame for that), but I’m fine with that: I’m more interested in seeing my choice of actions reflected in the reactions of those around me and in the tales that unfold.

Which is a pretty good way to think about life as well. GDC just ended, and I can’t imagine a better way to have spent a week: stunning talks, I feel privileged to have been able to spend time with people as I did, and honored to have been able to play a part in facilitating one such meeting. And today I’ve been sitting around the house, working on my pro keys skills, having Miranda show off the underwater house she’s built in Minecraft, picking out tunes on my new guitar, hanging out with Liesl, being glad that Zippy can still make his way around, and I’ll read a few chapters of a Scala book when I go to bed. It’s a good life: if my character build has somehow led me here, it has more than done its job.

Post Revisions:

This post has not been revised since publication.