A partial road map to RPG structure

When I decided to create an open game design (see the previous two posts here), I started sketching out a basic framework for the concept. I’ve been working on it in greater detail this week, and I offer it up here for review and discussion.

Click the image to download a 11×17″ PDF of this chart.

This chart outlines a “roadmap” of how characters and conflicts interact in games. It is not a comprehensive structure of an entire game, nor an attempt to model all RPGs. Rather, it’s just a model to show one common shape I’ve observed in games, especially the kinds of games that I think inform this project I have planned.

I intend the chart to be a reference point for design. It outlines “components” shown as boxes and arrow flows here. Each box could be swapped out with alternative mechanics, allowing individual designers to tailor their creations while still staying “on the map.”

Speaking of designers, I’m still eagerly looking for a small design team to design the core system using this chart as one of our guides.

26 thoughts on “A partial road map to RPG structure

    • Author gravatar

      Hi Matt,

      I think your chart confuses two different ways of understanding mechanics: by what they represent in the story and what they represent mechanically.

      For example, relationships and gear are story-groups. Motivators and special abilities are mechanic-groups. In some games, relationships are motivators, in others they’re traits. In some games, gear grants special abilities, in others it acts more like a trait.

      I think a more precise breakdown is possible.

      Motivators make certain actions more desirable.
      Skills make certain actions easier. Flaws make certain actions harder.
      Traits make possible new actions. Handicaps make certain actions impossible.

      Each of these properties can be an
      Asset (permanent)
      Resource (remains while you tend for it) or
      Boon (temporary).

      The aspect ‘My Mother, the President’ is therefore a motivator (through compels), a flaw (through tags) and a skill (through invokes). It’s an asset.

      A wand of cure light wounds, on the other hand, is a resource and a trait. It only lasts while you keep it from being stolen or broken.

      It seems to me that these are more usable categories than the ones you have on your chart.

      I hope this helps.


    • Author gravatar

      Hi Chris —

      You raise a good point. I understand very well the issues surrounding mechanics and story. I see what you mean on the chart, though I did it that way quite intentionally.

      The boxes I’ve chosen could be broken down differently, without a doubt. This is just the way I chose to do it.

      The “axis” you’re talking about, based on time (permanent –> tended –> temporary) is really interesting and useful. It’s partially evident in the chart, but not fully as you break it down.

      The way I intended the chart, such breakdowns are possible “off” the chart as designers approach each component differently. Some may wish to make Gear permanent (a magic sword that always returns to you), tended (a wand that has only so many charges), or temporary (a table leg!). And so on.
      So, the chart is a launch pad, a discussion piece to encourage others to figure out how to approach a design.

    • Author gravatar


      With what app did you build this?

      How/why are the six character elements broken into those three category(?) boxes?

      Under REWARD LOOP, I think “currently” should be “currency.” Right?


      I’m completely not grokking your comment about story- v. mechanic-grouping of the character elements. It seems that each of those categories *can be* mechanically-integrated and *must be(?)* narratively-integrated. And if I’m understanding Matt, he’s just describing the kinds of things that generally go on character sheets (that point toward the conflict resolution system).

      I do like thinking about the duration of the character elements as a separate axis like that, but any of the elements can land on any of the durational spaces. The various outcomes of conflict resolution can modify where an element lies along that axis, too.

    • Author gravatar


      Matt, you’ve made some concrete decisions here – do you want to just go with those as a baseline or hash things out, nit-pick, etc? How much feedback on the model do you want?

    • Author gravatar

      I dig this road map. Great job! I wonder if I might be successful using it to explain to my RPG group the difference between older games and some newer indie games.

      By the way, under “Execution,” I believe you are missing an apostrophe after “player characters…”

    • Author gravatar

      Jason, I’m all for feedback across the board. The chart isn’t set in stone or anything. That said, I view suggestions through the lens of the several assumptions I spelled out in my previous blog entry, “Open game design project, Part 2.”

    • Author gravatar

      Hi, dallashawk. Thanks for the note. I noticed a couple other typos myself today. I’ll remedy tonight, and replace the file with a corrected version.

    • Author gravatar

      How is SotC character creation pre-generated? I thought FATE had a lifepath character creation.

      What about games that don’t have attributes at all (FATE)?

    • Author gravatar

      Hi George — I’m thinking of the prepared packages of Stunts they have in one of the later chapters of SotC.

      As for games that have no attributes, yep. The column on the left explains that some games do not have all features, using Conspiracy of Shadows: Dirty Hands as an example. SotC is also an example, as is HeroQuest, and a few others.

    • Author gravatar

      I think there should be some layer between creation and conflict. I know you expect players to build a coherent fiction around the rolls, but I don’t want it to get short shrift.

    • Author gravatar

      Hi Thor — I’m not following you. Do you mean between character creation and conflict? Between some other creation and conflict? Also not sure what the “it” is in “don’t want it to get short shrift”?

    • Author gravatar

      Matt, are you thinking at all about mechanizing the way that conflicts fit into an overall story or into pacing mechanics? In a lot of “conflict resolution” games there is an art to determining the appropriate scale and scope to use when invoking the conflict resolution system, and a lot of games seem to rely on “indie games culture” to get the players to get it right rather than cooking it into the rules.

      • Author gravatar

        Hi Dan — I’m thinking about it, but that’s not part of this chart.

        I agree with you, though. The thing you’re describing does happen, and I’m much more interested in crafting a game where this flows far more naturally. Oddly enough, one of the games I think does this best is the indie grandpa, Sorcerer. And, while that’s a big indie influencer, I think what it does procedurally is MUCH closer to traditional gaming than it may appear. There are not “stakes” or any of that formalized stuff. It’s just, stuff happens, you talk, you roll, you keep playing. More stuff happens. Etc.

        Of course, the trick is communicating that in encoded rules, definitely.

    • Author gravatar

      Cool, some questions:

      Character development components are obviously optional. Are character creation components as a category optional as well?

      Can creation/development components manifest on the social level, between players, rather than as in-game resources and markers?

      I can’t distinguish between gear, attributes, traits, and skills. These are effectively identical, different only in implementation as resources. I would call them resources and be done with it.

      Are all meaningful interactions predicated on conflict? What about situations where outcome is important but conflict is not?

      Do all conflicts require intent?

    • Author gravatar

      Jason, to answer/discuss:

      Character creation components are also optional, yes.

      Yes, development components could develop socially. It’s simply off this particular chart. Do you mean like, “Hey, it’d be even more cool if your guy was actually once married to her!” “Yeah, Ok! Let’s do that (scribbles note, spends points). Ok! Now he is!” ???

      Yes, they’re all resources, I agree. But, I’m not sure why you don’t want discriminate among resources. That is, it strikes me as a matter of priority — what’s your aim in NOT discriminating (maybe because you assume designers will just figure that out regardless, so there’s no need delineating here) vs. what’s your purpose in discriminating (maybe the resources are quite mechanically distinct, and you want to segregate them so the “map” helps people avoid goofy distortions in play). Here, I’m saying “well, look, we can chop up resources in geeky-but-interesting ways, and let designers tinker with each kind.” Also, it’s not to say that these are components are exhaustive. I imagine there are other resource components I haven’t included with these six.

      No, certainly not all outcomes are predicated on conflict. Suppose a Motivator awards points for being resolved. And, during play, everyone agrees, “Hey, you know what? After the rock concert, We all just basically agreed that Joe just exacted revenge on his ex without a roll! Weird.” Well, ok, so Joe gets his Motivator points. I suppose I’d want to ensure that such resolutions without conflict tension don’t become so “normal” that the game seems to wander as players tend to avoid the fruitful pressure of conflicts. (And, I’m sure there are other variations on the idea.)

      I can’t figure out how conflicts happen without intent. But, that said, I think it’s actually pretty murky. What’s “intent” mean? And, whose intent are we talking about? I’m saying players encounter game situations and they (mostly intuitively) realize they want “what happens next” to go differently than what they sense (or are being told) just now. Conflict! And, it decides what happens next. I just like that is also creates mechanical pressure that allows other things to resolve as well. My design style has always worked that way.

    • Author gravatar

      I think I was trying to say that too many games seem to jump from conflict to conflict. I think that there needs to be space for other bits of story.

    • Author gravatar

      Ah! I see, Thor. Yes, I know what you mean. My own designs have suffered a bit from that, actually. Although, that’s also partially a symptom of my GMing style when I play them.

    • Author gravatar


      My point is that Matt has included some categories (gear, relationships) based on their narrative role and some categories (skills, motivators) based on their mechanical role.

      Gear and relationships are not distinct from skills and motivators, because they often serve as skills and motivators.

      So I think a more helpful approach would be to divorce story from mechanics altogether in this analysis by having two categories for things: story groups and mechanical groups.

      Relationships, gear, attributes (innate qualities), disciplines (skills you learn), talents (special abilities) and beliefs are possible story categories.

      Motivators, skills, flaws, traits and handicaps are possible mechanical categories.

      I think this makes it easier to say: Circles in Burning Wheel is a relationship skill. In other words, it acts like a skill and involves relationships. The ‘Nifty Jetpack’ aspect is a gear motivator and skill. The Strength ability score is an attribute trait while the Athletics skill is a discipline.

      Is this making it clearer? I’m not sure how else to express it.

    • Author gravatar

      Got it. Thanks! And that’s neat because then you can take any kind of character element from any game and place it in the three-dimensional matrix of mechanic, story and duration.

    • Author gravatar

      Hi Matt,

      I think on the macro scale you are better off atomizing things only as far as is necessary. So a box for “resources” makes more sense to me than four boxes all labeled “a special kind of resource” It’s an issue of taxonomy and clarity. Those special flavors deserve their own breakdown elsewhere.

      I suspect that my tastes run too far toward structured freeform to be particularly useful when discussing conflict and intent for this project. You can effortlessly divide the two in Fiasco, for example. It is perfectly acceptable to stumble into scenes without an agenda, and there’s a sharp distinction between conflict (optional, sometimes emerging organically, sometimes a goal) and scene (the thing that gets resolved through the resource of procedural privilege). I don’t think it is where you are going with this, although it is a lovely place you should totally visit.

    • Author gravatar

      Hey Matt, how familiar are you with Object Oriented Programming? I ask because it’s a way of thinking about computer programming where (vast oversimplification) different components (“objects”) all interact in specific ways (“methods”). The upswing is that, once the methods between objects is set up, you can swap out individual objects as long as the new object has the same input/output methods. So there might be many “clock” objects, as long as each one outputs the time when queried by other objects.

      It seems to me like you could turn all the arrows in your chart into methods, and all the blocks of text into objects. So the resolution object takes inputs of traits and gives outputs of success/fail and maybe magnitude of success/fail. Then you could design modular bits of game mechanics that fit that specification. Whether it uses cards or dice pools or spinning around until you fall over, as long as it takes inputs of traits and gives outputs of success/fail, it will work with the other game mechanic “objects.”

    • Author gravatar

      I’m reasonably familiar with object oriented programming, certainly with the basics. I’ve already “caught” myself from calling things classes in replies to people!

      So, yeah, in no doubt some inaccurate way (because I’m not programmer), the analogy makes sense to me

    • Author gravatar

      That layer between character and conflict might be where GM Moves from Apocalypse World go. It might even be where town creation from Dogs goes.

    • Author gravatar

      Well, okay, actually: it’s Situation. So any mechanics that draw out of character, setting, or something else to turn old situations into new ones can go in there. Including Conflict, I guess.

    • Author gravatar

      I think a lot of things go in that layer. We have spent the first half of our gaming lives focused on the first part and the biggest part of our “post discovered the Forge” part on the second. We need to discover all of the things which interface the two areas.

    • Author gravatar

      Tell me if I’m stepping on toes, Matt, but here’s what I was talking about: your chart (less pretty) split into “objects” and “methods.”

      (click to download vector / pdf)

      Any of these objects (boxes) could be split into sub-objects (as the character creation / character development boxes could be one object) as long as that set of sub-objects answered and received the same methods as the “parent” object would have.

      I think this is about the minimal requirements for a functioning game. Anything beyond what’s here can be understood as a set of sub-objects filling the role of the parent. I think. 😉

Leave a Reply

Your email address will not be published. Required fields are marked *