Ladino's Rogue Thoughts

What Makes a Good or Bad Structure?

Not all scenario structures are born the same.

Some are beautiful discoveries stumbled into by players who happened to find them while resolving some unexpected situation in the fiction. Some are meticulously built experiences made manifest after many iterations of game design refinement.

Others, however, are haphazard jumbles of good-enough practices put together to tenuously hold the game together until a better structure takes over. Others still are GMs desperately trying to fit a square peg to a round hole, unwittingly using a tool that not only doesn’t help with the situation, but ends up being worse than using no tool at all.

As a result, some structures are great at what they do: they can cover a lot of different situations and moods with grace, they present players with meaningful choices, they tell the GM what to prep, and give everyone a good starting off point from which to explore the fiction. Others, not so much: they are dry and overly procedural, they barely cover the situation they were made to resolve, they lead players to dead ends, they throw GMs into the improv waters wearing cement shoes.

Today I’m offering my thoughts on what makes for good and bad structures. Let’s build off of the features Justin Alexander identifies as defining complete and effective scenario structures in his series on the topic (these criteria are scattered in a couple different parts of the series, so I can’t give you more specific links):

  1. They have clear and specific triggers – so we know when to use them.
  2. They provide players with a default goal and a default action – so players are never left flailing without knowing what to do while it is in play.
  3. They should be able to loop through the default action – so the structure is “complete” and won’t routinely throw players into situations without a structure.
  4. They are easy for the GM to prep for and run – so GMs are never left flailing without knowing what to do before or while it’s in play.
  5. They are flexible for the GM – meaning GMs have a lot of freedom in what they can put into the structure.
  6. They are flexible for players – meaning players can make choices the structure hasn’t ordained or predicted.
  7. The decisions players make are meaningful – we could say this means the decisions should be informed and impactful, per the ICI doctrine.
  8. The decisions players make as part of that structure should be associated with the choices the characters make – i.e., not dissociated, per the Alexandrian’s understanding.
  9. They should tie into the reward structures of the game – so players have a reason to engage with the structure.
  10. They should be fun – so everyone actually enjoys playing them instead of just suffering through them to get to the good stuff.

These are a lot of requirements, and any one of them might be tricky to fulfill. I mean, ”being easy to prep for and run” alone is a beast of an issue, considering there are so many different preferences in how to do either of these things.

But then, I think some of these are not really universally applicable as a standard for good structures. Not all structures serve the same function, and as such they should emphasize some of these qualities more than others.

Let’s look again at Justin’s party planning structure as an example. Here’s a quick rundown of what that structure is comprised of:

Note that Justin doesn’t bother finding, or at least overtly expressing, a default goal and a default action for this structure. Is he a bad designer who can’t follow his own advice? That’s for each individual to judge, but for this specific instance I’d say he just intuitively realized that parties don’t benefit too much from these predefined goals and actions. What would those even be?

Default goal: Talk to everyone at the party? Have a good time? Get lucky?
Default action: Eat and drink? Dance? Mingle?

No need to bother with those – best to employ parties when the players can define or already have defined their own goals, and make their actions whatever they can think to do to achieve those. Which means that flexibility for the players is a particularly important quality for the party structure, whereas a default action isn’t.

Similarly, the party structure doesn’t try or need to form a closed loop – in fact, it even features a timeline of events, establishing a clear progression to its proceedings and a looming end point to the party as a result. There’s no pretense that you should be able to loop the party structure forever, the way you could with a hexcrawl or a dungeon-crawl. This is actually also true of combat structures. They generally have a clear end point in sight: the defeat of one of the opposing sides. You can’t (or at least shouldn’t) loop it forever. You can iterate combats forever, sure, always finding new things to fight, but this isn’t the same as looping the same hexcrawl forever into a West Marches campaign, for instance.

The reason why it’s fine that Justin’s party structure doesn’t feature these qualities, I think, is that it is not built to be a load-bearing structure. It isn’t meant to sustain an entire campaign of just parties. It is a tool to create a memorable sequence, as a change of pace to a normal campaign and as a believable and actionable model of how a party can be used to get new information, create drama or serve as a backdrop to some intrigue and action.

So, some structures can be good at what they do even if they lack some of those listed qualities, like being able to be looped. Then again, some qualities in that list are also a matter of preference. A group that has no issue playing with dissociated mechanics, for instance, probably won’t mind if the structure asks them to make some dissociated decisions here and there. Groups that are strongly intrinsically motivated are probably fine with a structure that doesn’t tie in to any mechanical progression rewards. Not all of these qualities Justin mentions seem to apply equally to all structures, then. I think the ones that really matter for any scenario structure, no matter whether they are a core feature of a campaign or a one-off gimmick to bring life to an unusual set of circumstances, are these ones here:

Being easy to prep and run just means the structure reduces or streamlines the amount of work needed for the game to happen, which is always a positive effect when it comes to allowing us to engage with the hobby in the first place. The value of flexibility, for GMs and players, is in allowing our creativity and imagination to shine when dealing with the game, keeping it surprising and engaging throughout. The value of meaningful decisions is in providing an interesting game to play.

And really, that last one, of being fun, is such a difficult quality to pin down that it really doesn’t serve as a guide for how to design our own scenario structures. It is more a reminder to diagnose the structures we are using to see if they’re clearing this bar, which will be different depending on what fun is supposed to look like for a given group and a given situation. So, we find that, as with so many things, a structure can only be good contextually, as compared to what it sets out to do and the effect it tries to achieve in what it does. If game structures are (a collection of) tools, we should understand them to be more like paintbrushes, rather than screwdrivers. That is to say, they’re tools that combine being adequate to their function with the flexibility of artistic expression, not tools where there is a single optimal solution that is right for the job.

What about bad structures?

It is tempting to imagine that bad structures are simply the reverse of good structures, then, which is to say that they would be:

Again, as with the definition of good structures, the last point is a bit tautological: a structure will be bad if it is boring, and it will be boring if it is bad, so this doesn’t tell us much.

As for the other points, though, I’m not sure failing to live up to the standards of ease-of-use and flexibility are really what makes a structure fundamentally bad.

If a structure demands a lot of work to prep and run, sure, it runs the risk of burning a GM out and maybe wastes more of their time than it should; but it might be worth all the trouble. A balanced, tactically complex and thematically strong combat encounter in a game like dungon dargon 5e takes some doing, but can also provide an incredibly exciting couple of hours in play and fond memories for a lifetime. (Not to mention that there are freaks out there who might actually enjoy the laborious process of making a complex structure sing! I used to be one of them!)

Similarly, make no mistake: flexibility, as an expression of the infinite possibility space of RPGs and a tool for allowing the fiction to take precedence over any mechanics, does make structures that allow for it better! But sometimes a structure is meant to cover a situation quickly and abstractly, giving just enough detail that the players can engage with the activity they cover, but not necessarily allowing players to hone in on the minute details of the thing, so as to not detract from the primary focus of the game.

A structure that just offers players a set of constrained choices, then, might not clear our definition of a “good” scenario structure, but it’d be too much to say that it will necessarily be bad. It could still be perfectly functional. Do you really want a complex subsystem that covers everything a PC might do with any given alchemical reagent they harvest, defining its magical properties and the various processes that can be used to extract them, or would you rather just let them pick between making one of two different potions with it and move on with the game? There are games for which the correct choice is, inequivocally, the second one!

So what would a bad structure look like?

Bad structures, in my mind, aren’t so much the ones that are difficult to use or limited in what they can do, but more so the ones that fail at doing what they ostensibly set out to do. In particular, bad structures will:

That is, bad structures will ask GMs to prepare and juggle useless information for the situation they are trying to run. They will ask players to make choices that are irrelevant and provide no feedback, leading to a frustrating game. Failing to provide any meaningful choices, then, is the essence of a bad structure.

This is why applying a structure that is good for a given situation can cause really bad results when used for a situation that doesn’t call for it – the information and decision points that are central to a particular scenario might be pointless in another. Resolving the players’ intent to go out shopping with a dungeon-crawl structure of exhaustively describing every location, object and NPC they stumble across while getting from where they are to the shop is usually bad because it is adding a lot of pointless noise to the proceedings. The rhythm of describing a location and asking players what they do in it or where they go next remains a flexible structure, but its flexibility is now worthless: the infinite choices players can make now lack impact to the goal they’re trying to accomplish.

Conclusions

The axes that push a structure toward being good and those that push a structure toward being bad aren’t diametrically opposed. Qualities that generally make structures better can exist in bad structures, and can be lacking in good ones. Problems that plague bad structures might be the right choices for another situation.

We can and should value structures highly if they are well-suited to covering many similar situations without flattening them to identical experiences and if they empower the players to approach it creatively, making it easy to turn their decisions into relevant consequences.

But we should only really deride structures if they waste our time with pointless busy-work or force us to make meaningless decisions.