I’m thinking about a new game idea, which affords me the opportunity to tinker with new game system ideas. One such gadget that popped into my head was the idea of completely integrating the ideas of classes and actions. One of the inspirations with which I’ve been rather taken as of late is the way Leverage‘s iteration of Cortex+ boils character actions down to the role that they fulfill: anyone on a Leverage team can hack, or con, or plan, but you may not be as good at it as the guy or girl you pick as your hacker or con artist or mastermind. In short, you have a range of dice to hand out to the various roles, and you put your good ones in the slots you want your character to do well in. An interesting takeaway from that is that, if you think about it, none of the characters actually have classes so much as they have classes that they are most often the highest-qualified team member to perform.
We could potentially extrapolate from this that there are only a set number of actions a character can take in a game: in most games, this is finely parsed out into a skill list, and Leverage uses role + skill to create its panoply of different possible actions, but what if we just folded those into role or class divisions? What you’d end up with, then, seems at first blush like it would be rather constricting, but only if you adhere to the framework of only letting the most qualified member do the thing they’re qualified at. Let’s look at this from a more generic fantasy angle: instead of approaching it as “only the fighter can hit, only the wizard can cast spells”, characters become a multi-channel blend of different capabilities. In a pinch, the fighter can use his terrible “wizard die” to throw out a spell, even though that’s not his bag*. The door opens to create some of those interesting cross-class character concepts, like the magical thief or the martial mage, by allocating the dice and prioritizing the actions you want to take.
To ground this in what I’m working on, I look at what characters are going to be doing in the game’s premise and decide that there are about six or seven things that they’ll be called upon to do. These range from fighting to communication, so I play with the idea of mapping them to the color spectrum in order from most active (red) to most passive (violet). Each character role gets a signature color and the corresponding action as a “freebie” each round: anyone can fight, but the hitter gets an additional swing each go, for instance. The mapping doesn’t quite work as perfectly as I’d like – where does movement go on the scale, and should there be a class devoted to taking movement actions? – but it’ll do for now. Aside from your free in-role preferred action, you might get two or three other skills to choose from to specialize in, getting an increased ability to succeed in them. This does not have to be your preferred role action, though, which raises some interesting possibilities: someone who can perform an action very often due to familiarity, but isn’t quite as good at it as someone else.
The other complication that becomes apparent is that such a system assumes that actions are intransitive: once you start introducing objects or targets for the actions, you start to see why the stat+skill construction is a useful syntax. To say “I attack” normally assumes there is only one thing to attack. Sure, you can take a swing at someone’s head… but what about attacking someone’s reputation? Their latest repair job? Their confidence? Their spirit? If you’re working in a design paradigm that allows free-form descriptors (which I may or may not use), the application of Action to Target relies on a lot of adjudication, and breaks up the granular nature of character roles. The guy who’s good at Fixing computers is not assumed to be good at Fixing people, and a system that lumps those together is going to have to justify it or require a healthy suspension of disbelief.
To be clear, I haven’t thrown out the idea of [Role Actions] applied to [Tags], but it’s probably going to require a bit more scaffolding around it to prop up the system. We’re still early in the process.
* The earliest example of this that I’m aware of is Deadlands, where your stats are rated by the various dice you want to put in them. Anyone could be a gunslinger or a magician, say, but characters were defined more practically by their abilities than their class.
PS: Somewhat appropriate tangent: I ponder some similar thoughts in my response to this game design thread at RPG.net.