I was running a game this past week and the session, which was otherwise entirely either narrative or conversation-based, ended with a combat situation. The opponents were some capable but ultimately nameless guards in a place the PCs needed to infiltrate, and sneaking by had proved impossible. The PCs are pretty badass, so in short order there was just one gravely injured guard remaining. I was about to tell the group that they finished him off with ease when one of the players said they attacked him and started rolling dice. Okay, I thought, I’ll let the mechanics say what I was about to. Except they didn’t. The player rolled very poorly and failed.
Again, I was about to say that despite that character’s bad luck the group puts a quick end to the fight, when another player also started figuring out the specifics of their attack. I nodded and let them make their roll. That character also failed.
At this point, the dice mechanics had diverted expectations—not just player expectations, but GM expectations. The game took the story where no one was expecting. Some people like to describe this as being “emergent.” No one was telling a story, the story emerged from play. Both GM and players discovered the story as play progressed through a mixture of character actions and dice rolls.
Other people would say that the mechanics got in the way of the story that was being told. This turn of events offends a sort of narrative sensibility. When the mechanics allow for something to happen that feels wrong to the GM or the players, many people respond by questioning whether the rules are facilitating the style of play and emulating the kind of stories they’re interested in.
In this specific case, the situation was in between these two viewpoints. I wanted the story to say, “you guys are so capable that you win easily.” The players wanted that too, but they wanted to express it through the mechanics. They were eagerly grabbing dice so the dice could tell them what they wanted to hear. But did they do that because dice are the only means at their disposal? Would they have felt less capable if they didn’t roll the dice and I just said they succeeded?
I think for some players, the answer is yes. I think that many players feel like if you don’t roll dice, you’re not really doing the thing. It cheapens it. While this is sort of patently ridiculous if you examine it too closely1, it’s also understandable, because there’s a sort of social contract at work in many games that says “the mechanics are the way in which you interact with the world.” The rules are the laws of reality.
For other players, however, the answer is no. Doing 3 points of damage versus 4 points of damage isn’t the important thing to them. It’s participating in the telling of an exciting or interesting story. In the same campaign, just one session later, the players made the decision to head to a location that wasn’t at all important but was full of enemies. As the GM, I had three choices.
1. Play it as it lay, and spend the rest of the session on the PCs fighting their way through a bunch of enemies, which would likely not advance the story at all.
2. Rearrange things behind the curtain so that the place they were headed was indeed important to the matter at hand.
3. Narrate the PCs’ success in a number of pitched battles in a few sentences, and describe that unfortunately what they were looking for was not there.
In this case, I chose option 3, because the PCs (and therefore the players) are after something, and the fights aren’t the point. Just to be sure, I checked in with them to make sure I wasn’t stealing away anyone’s fun, but I read the room correctly.
The point is, the players (and the GM) didn’t want the story to emerge from the mechanics in this case, because the mechanics would have taken a long time to tell what was likely to be an unsatisfying story. (The counterpoint, of course, is that maybe a completely different but just as interesting story would have emerged. And perhaps it would have. So it was a calculated risk.)
From a designer’s point of view, you need to think about these things. Back when TTRPGs were new, the designers really didn’t, and players could pick how to handle these issues as they wanted. It’s worth noting this because the adventure I was running was originally designed in the 1980s.
Today, however, I think a designer needs to pick an end goal and design toward it. You can absolutely have a game where the story is entirely emergent and even the GM is surprised by the twists and turns. Such a game should have a lot of random tables for determining events, NPC actions, and maybe even the physical layout of the world and its locations. The designer might even want to consider using swingier dice (like a d20 or percentile dice) to increase the randomness. GMs in such a game would be encouraged to not fudge dice rolls and roll where everyone can see.
The downside is that such a game could be punishing to unlucky players, or those who make poor choices. The story that emerges might be an unsatisfying one.
Alternatively, you can design toward the goal of having the narrative (as desired by the players and GM) trump any mechanics—or better yet, integrate such things into the mechanics, granting players the ability to engineer auto-successes at points where it seems appropriate, and then giving the GM the ability to do something similar. Any dice mechanics should involve a normal distribution, like a bell curve or certain kinds of dice pools, so that events are mostly predictable and extraordinary outcomes happen very rarely.
The downside for this kind of game could be that the mechanics themselves don’t engage the players much, which often makes advancement less interesting. Also, predictable outcomes are, of course, predictable, and some players will find that dull.
And there is a spectrum of options between these two choices. But a designer needs to approach the issue deliberately. It might very well be one of the first decisions you should make when you sit down to design a brand new game.
The distinction is ridiculous because in neither case is the archer actually shooting a real arrow. Whether it’s through GM fiat or the result of a die roll, ultimately the GM is stating success or failure.
Still More on Narrative Versus Mechanics
Even though the discussion may feel rather circular, I think that it is still a challenge for any GM to make these choices. The way we have figured this out is by providing a better system for Story Design. The thing about game systems is that they provide rules for GMs to run the game and provide advice when there is story to be designed or "decided/created". So when they provide rules for those decisions or creations it may feel as if creative freedom is reduced. Our Story design system establishes a scope frontier, so GMs make sure the story will flow as they planned in the broader picture, while at the same time leaving the mechanism to provide the micro narrative (like combat results, success on overcoming an encounter and the like). GM designs only the stuff behind those challenges and makes sure they are linked to the characters enough so they don't derail and they are pushing towards the story. It is unlikely that mechanics will clash with the narrative this way, and instead they are the spice of each session, the unexpected for all (including the GM).
There's so much to discuss on this particular topic, and I think it would be worth writing a column where you can expand this part:
"...you can design toward the goal of having the narrative [...] trump any mechanics—or better yet, integrate such things into the mechanics, granting players the ability to engineer auto-successes at points where it seems appropriate [...] Any dice mechanics should involve a normal distribution, like a bell curve or certain kinds of dice pools, so that events are mostly predictable and extraordinary outcomes happen very rarely [...] The downside for this kind of game [is] predictable outcomes [...]".
That first part (making the mechanics control the narrative) is basically what you did in the Cypher System (via Effort, Skills, and Assets). It would have been interesting to make the dice mechanics more predictable with a non-linear result (a 2d10 pyramid curve instead of the d20, for example).
But I think the aesthetics also matter when it comes to the design process, don't they? I read once (I do not remember where) that you liked how the d20 rolls, and that it was part of the designing process when it came to Numenera.
And what about the "swinginess" of a linear result (single dice roll, like the d20 system)? Was that part of the consideration since it makes what happens in the Ninth World more... "unpredictable"?
Such a fascinating topic. Thanks, Monte!