So, this is one of those idea that illustrates where narrative and games do a strange dance. When you create a scene in a story, there are lots of elements in play, but one of the most critical is how it’s going to end. In the crudest sense, if the scene is a fight, then it’s generally going to end with one side or the other winning. Usually the protagonist wins and we move on to the next scene.
This is a terrible example, of course, for a couple of reasons (most of which John Rogers articulates incredibly well). The most obvious is that in most cases you can be confident that the hero won’t lose because that would interrupt the story. There are fight scenes which have uncertainty, but they usually require significant framing to create that uncertainty.
To make these scenes more interesting, there is usually at least one other uncertainty in play. Let’s say, for example, we have the classic pulpy showdown with an evil henchman while an NPC is in a deathtrap that’s going to trigger soon (a candle is burning through the rope that holds up the deadly spike!). Now there are 4 possible outcome:
- Hero Wins, NPC Rescued
- Hero Wins, NPC not rescued
- Hero Loses, NPC Rescued
- Hero Loses, NPC Not Rescued
Now, if you’re creating a story, this is suddenly a little bit more interesting – you’ve got a number of different ways they scene could close. This is, of course, a greatly simplified case – if you’re actually writing a story the options may be less binary, but I’ll be using this more simplified model for illustration.
When I talk about a drama in an RPG session, I am thinking in terms of those outcomes – how is the scene going to break in an interesting way. This is why, from a dramatic perspective, a 10×10 room with some goblins in it is super dull – there are only two outcomes. And if I see the problem in this way, then my instinct is to add something to the fight to increase the number of outcomes, so there is a “beat the goblins, but…” element
This is a good instinct. It generally means the fight with the goblins will be more interesting than simply dropping them in a room. But it is also only half of the picture.
Where RPGs differentiate from stories is that what may seem like a simple outcome in a story can be finely graded in an RPG. The dramatic outcomes of facing the goblins are “Win and proceed” or “lose and perish (or something). The RPG outcomes are:
- Win handily
- Win but we’re a little banged up
- Win but we’re hurt badly
- Win but we lost a party member
- Win but we had to burn through a lot of potions
- Win but we had to burn through a lot of spells
- Losing, so we made an orderly retreat.
- Lost half the party, the rest ran away.
- Lost badly
And this list is still abbreviated because many of the options combine. It was an Orderly retreat but we burned a lot of spells. We were hurt badly and had to go through most of our potions. In short, that dramatically simple outcome is much more nuanced from a game perspective.
One practical upshot of this is that dramatic thinking can make the fight scene better (by making it more than just goblins in a room) but by itself it may not produce a satisfying outcome.
In a vacuum, the game approach seems sort of silly and full of unnecessary detail. And it kind of is. If you’ve ever played D&D where you are having a “clean” fight (One where you start at full resources and will have a chance to recover to full afterwards) that feels very different than working your way through a dungeon.
The key here is one of continuity. If you are thinking dramatically, the things which transfer from one scene to the next are only the things which are dramatically interesting. If you’re hurt, then that might carry forward, but not in any kind of fine-grained sort of way. Depending upon the kind of story, it might not be carried forward at all.
Anyway, one of the interesting design challenges of RPGs of a more dramatic bent is to keep that sense of continuity while still staying relatively dramatic. Torchbearer does an amazing job with this by effectively maintaining a list of prices that are easy to spend but hard to recover. Fate uses Consequences to reflect important things carrying forwards.
But the key is that even these more abstract systems are more information that someone reading a story would want. And, honestly, for the story of the game that you tell once you’re done, they are details that largely get swept under the rug. Even in D&D, we are excited that we landed a crit at just the right time, and we describe the attack – rarely do we include the exact damage number because that’s not part of the story.
But what it is is part of the experience. The continuity of those resources can tie together the game in a meaningful way. To say that they are at odds with story is like saying that a muffin tin is at odds with muffins because it’s inedible.
Now, here’s a qualifier. There’s a lot of fun play to be had sticking to purely dramatic rules. There’s also a lot of fun to be had by discarding drama in favor of awesome rules and emergence. I don’t pretend to know the right way to go about this, and I know my own play is an unstable mix of these priorities, but that is how I like it. On some level, I don’t believe that gaming is simple enough for any one solution to fit every game, and that the pursuit of fun tends to get a little muddy.
Not that I would have it any other way.
And it’s important to note the independence of the axes. The NPC in danger is not the “stakes” of the scene – if that were so, there would only be two outcomes – win and rescue the NPC or lose and the NPC perishes. At first glance, this may seem like you’ve made the simple conflict more meaningful, but you’ve actually doubled down on predictability – you have raised the stakes enough that the hero can’t lose if the story is still going to progress. ↩
This is one of the things that made 4e such an oddball. Recovery of resources between fights was not terribly onerous, so very little really carried from one fight to that next , so the fight outcomes we closer to the dramatic model (because a cost that is immediately reimbursed is not really a cost). ↩