23. Action Rolls - Player Goals
Action rolls tend to be the heart-and-soul of Blades, and like Clocks before us, they’re at their best when they’re formalising existing play structures. Negotiating position and effect, for example, is at its best when it’s not a correspondence course, but rather a flowing back-and-forth between the table.
However, action rolls have a pretty significant flaw in the fact that they are inelastic. And I’m going to leave rules elasticity on the table for a while until we get to players rolling the dice.
For better or worse, this is likely to cover the 6 steps of action rolls across 3 or 4 days.
“1. The player states their goal for the action.
2. The player chooses the action rating.
3. The GM sets the position for the roll.
4. The GM sets the effect level for the action.
5. Add bonus dice.
6. The player rolls the dice and we judge the result.”
I have personally criticised Blades for being “overly procedural in its approach” and I think this 6 step process shows why. It feels less like negotiation and more like renegotiation. It feels like we constantly have to dip back into previously “settled” parts of play, which is ultimately one of the reasons it’s so inelastic.
We talk about goals, then dice, then possible consequences, then dice, then judging and resolving. The nature of formulaic resolution isn’t inherently bad (after all, it’s what PbtA has been about for years), but those formulas need to be smooth and natural. There’s a old approach in Systems Thinking from aviation that your systems need to have a low usage cost. You can’t give aircraft maintainers really difficult procedures and expect them to follow it to the letter. This is the Idiot Proofing approach that led to Lockout Tagout procedures. The secret is: People are lazy and you make your system easier to use than it is to not use. It’s easier to leave a padlock on the machine then it is to remove the padlock to use it. In medicine, if you have a problem with over- or up-dosing analgesia, you can reduce it by literally making the ampules smaller, meaning lazy administrators have to physically crack a second jar in order to exceed a regular dose. Not a problem if it’s necessary, but can avoid “well, I don’t want to have to write up this half-syringe, so I’ll give it all.”
Action Rolls kind of miss that natural flow, in part because they add tasks without creating the procedure to check for it, or make it easier. For example, position and effect requires both parties to be aware of the stakes and benefits before the roll, but the game doesn’t check for it in any meaningful way. This is what makes it inelastic and what drives players to us it as a reference text.
I just want to start with the first two steps. Because it’s so fascinating to me. One of the core elements of game design is “what is your game asking of the players?” This could be a clear question, like “Can you X before Y happens?” “Can you kill the bugs before they kill you?” (Arrowhead, 2024) These are usually a test of skill or timing.
Or a game could ask you about your values: “Who will you become when your status quo is rocked?” (Northway, 2022)
Or a game could ask you about your methodology: “How do you go about tending to multiple competing needs?” (Ludeon, 2013).
But games have unsolved questions. The players answer them through play. That’s the nature of consequential interactivity. And these can be big macro questions like the Play To Find Out question, or they can be micro questions. These questions are asked through the flow of play, so changes in mechanics will change what that question is and how it’s asked. Blades, for example, asks you How You Do Things. I know it says that it starts with players stating the goal, but it doesn’t actually need that. The first decision that the game “checks for” is action rating, which is not a function of goal but a function of approach. And often the goal is self-evident, or will be checked for later (this is that elasticity that we’re gonna talking about at some point). It is easier to assume the Player Character’s goal than it is to ask for it. Especially in Pawn Stance play where we tend to talk about actions and not context. To build on the book’s example: “I punch him” not “I apply violence to remove him as a threat.”
The game recognises that “Usually the character’s goal is pretty obvious in context, but it’s the GM’s job to ask and clarify the goal when necessary” (Action Roll p18). But the question is “When” is it necessary for the GM to clarify the goal? When setting position and effect, right? So goal is actually a function of position and effect “object” and not this opening salvo. This procedure tells you to gather information from the table, then do nothing with it for a while. And that’s important because that’s the kind of design burden that undermines a whole system. One question worth asking of a tabletop game is where do players race ahead or lag behind your intended structure. And in this case because the action rating generates dice, and dice are security, players tend toward deciding on their approach rather than their goal.
If you wanted players to have to consider their goal, you would have to mechanically tie goals to the system, which is what we say of PbtA. Instead this kind of unmoored “what’s your goal” question (not asked by the game, but mandated to the GM, who won’t use it for a while) is endemic of a mechanical framework (the Action Roll) that wants you to think about a lot of things, but doesn’t provide concrete ways for that consideration to be brought into the middle part of this resolution flow. This is the difference between telling people to use IIEE (as Blades does), and having mechanical systems that reinforce those questions.
For example, and hugely in its favour, the overlap of Blades’ Action Ratings means that there’s usually a few interesting questions to “I fight”. Do you skirmish, hunt, or wreck? Oh you talk the guard down: Do you Command, Carouse, or Sway? Oh you get the lock off the door? Do you finesse, tinker, or wreck? By creating a system that means the answer to those questions matters, it forces us to ask them. A player CANNOT open a locked door under risk without telling the MC that they want to do it quietly, or quick but loudly. And that’s great! That’s the point of a good role playing game system.
Our old friend Baker-Boss says “System is consensus” (Baker & Boss, year unknown), which I usually see interpreted as “System is how we come to consensus”, but I want to assert that just as much, if not perhaps more, system is what we come to consensus about”. For example, when threatening someone in Apocalypse World 2nd edition (Baker and Baker, 2016), the system absolutely cares if you’re willing to pull the trigger or not. You are forced to come to consensus about your character’s willingness to shoot. That’s because if you are, you Go Aggro, and if not, you do something else (Manipulate, perhaps). The move is predicated on the willingness to engage with all of its results (or at least enough that you can make the move function). In D&D 5e, there is no such requirement to come to that consensus. Players are free to leave their willingness to deliver on a threat up to the whims of the universe, because in both cases we roll an Intimidate check.
This is why “letting the player choose their action rating” (#2 on our list) is important. It’s not because players have to have authority or anything as formalistic as that. It’s about forcing the player to generate the consensus of what their character is doing. From How to Play:
“When an action roll is called for, the character is usually already in motion, doing something in the fiction. This thing they’re doing will almost always determine which action rating to roll. However, a player is free to revise their character’s action in order to use a different rating, as long as the character performs the new action in the fiction.”
Declaring an action rating is the first sign that we need to build consensus, and the first question we need to build consensus on is “what’s happening in the fiction?” The best way to do that is to ask “exactly how are you going about this” and a key part of that question is the Action Rating. We don’t need goals to answer that question, and in fact, they’re not important to the game at all at this point. Not until we get to effect.
I’ve been talking a lot recently about “what does your game ask players when they engage with it”, and I think it’s fertile design space. About what facts are you forced to build consensus. And then, elasticity, how far into play can get you get before the game alerts you that you’ve missed a step?
And when it comes to setting goals, that’s quite far. But of course, we’ll talk about that tomorrow.
