Many, if not most, TTRPGs use dice as a randomizer to introduce a bit of chaos and uncertainty into their games. Dice most often determine chances of success, but they can also prompt encounters, environments, NPC reactions, and much, much more. Randomness keeps things exciting, tense, and unexpected. Some games even go so far as to make character creation completely random.

I’m a big fan of games that let dice take over the story. Powered by the Apocalypse style games, for example, don’t just give you pass/fail on a simple roll of a d20 + modifier. Instead, every time you roll, you do so because your actions trigger some sort of move, and each specific move has a set list of outcomes–questions you should answer to determine how the story of your actions unfold based on the result of the roll. Instead of just seeing whether you success or fail. you instead roll some variation of a “mixed” result, introducing some consequences as it fits in the narrative. Drama is built into these type of systems.

But I think dice can do more. In Ironsworn (which is pretty much a PbtA style game with its moves and 3-tiers of success/failure), is meant to be played without a GM. To that end, whenever you have to make a move, you roll a d6 + modifier to determine how well you do. That’s your Action Score. But you also roll 2d10, which are the challenge dice. You’re trying to roll your action score higher than either or both of these challenge dice. If you roll higher than both d10, you get a strong hit–you succeed without consequence. If you only beat one of the d10, it’s a weak hit–you do it, but at a cost. Otherwise, it’s a failure–you fail to do it and there’s a major cost.

And it’s these 2d10 that really intrigue me, and what make me writhe this post. Since there is no GM in Ironsworn, there is no one to determine a base difficulty for a given task or action. There are no systems for advantage and disadvantage, for positive or negative modifiers outside of what your character brings to it. No arbiter of how hard a challenge should be. Instead, that’s all handled by these 2d10. If they both roll high, then so is the difficulty. If they’re low, the challenge is less complicated. And everything in between.

I just love this kind of stuff. And I think this design ethos could and should work even in GM-led games. In fact, I’m of the mind that the GM’s job should be less adjudicating things like difficulty and consequences, and more helping the players to tie it all together into a fun bit of narrative. Rolling dice to set the difficulty of a challenge adds this extra layer of narrative chaos, which is exactly what I want from a narrative system. (Obviously, for a tactical game, it could get quite annoying.)

Like, let’s say you’re trying to get past a guard into a restricted area of the city. Usually, the GM would determine how difficult it is to convince the guard to let you pass, or to sneak around them, or whatever else you want to do. Instead, though, the GM rolls some sort of dice to determine the difficulty of the getting past the guard. If the result of that challenge is low, maybe the guard is asleep, or easy to bribe, or just disgruntled with their job. If the challenge rolls high, the guard might be vigilant, or there are more guards than you thought, or some other thing is going on that makes it harder to deal with it.

See, instead of the GM having to think about these things to set difficulties, the dice inform the GM about these things instead. A simple roll can be a jumping off point for the fiction of this guard post, which the GM can interpret as it makes sense within the narrative (or interpret it together with the players in a more collaborative game). It’s just such a simple and elegant way to take some of the work away from the GM, and to let a bit of chaos inform the fiction.

I’m currently drafting some ideas for a skill-based, narrative focused game that uses dice pools. Right now, I’m thinking of letting the players roll 2 or more dice of different sizes (d4-d12, maybe). These dice come from “attributes” and “skills” and probably gear, perks, and other sources. You always roll at least 2 dice, and have lots of options to add more. Maybe at a cost, or through limited uses.

And on the GM side, I want to use something like challenge dice to introduce the chaos I just rambled on about. I think the GM rolls 1 or more d6, based on how difficult the general challenge is. Yes, the GM still sets an overall difficulty, but rolls the dice to find out how difficult the challenge is in this exact moment for this exact approach. To return to the guard example: getting into the restricted area is always a bit of a challenge, so let’s say 3d6. Maybe there is some “default” value the GM can reference, like everything within the restricted area is a “Tier 3” challenge, with each tier being a d6.

Either way, players and GM roll their dice, and lay out their results. Players then can use their die result to cancel out challenge dice. A player die of equal or higher value than a GM die cancels that die. If they cancel at least one challenge die, they succeed, but at a cost. If they cancel all challenge dice, they succeed fully, no consequences.

Players can also push themselves, canceling a challenge die of their choice, but taking “stress” equal to what the canceled die rolled. (No idea what “stress” means at this point).

There is probably more and interesting interactions to dream up for players to manipulate the dice pools, challenge dice, results–rerolls, cancellations, modifiers, and so on.

As an example, you are trying to talk your way past the guard.

• You have a social attribute of a d8, some sort of social skill of a d6, and maybe a forged paper that claims you’re invited to a noble’s estate in the restricted district, which has a quality of also a d6.
• The area is a Tier 3 area, so the GM determines that the challenge to talk your way past the guard is 3d6–no other considerations made to increase or decrease that challenge dice pool.
• You roll your dice (d8, d6, d6) and land on something like 6, 2, 5. The GM rolls their 3d6 and gets 3, 4, 6.
• You can cancel the 6 with your 6, and the 4 with your 5. Since you canceled at least one GM die, you will succeed at getting past, but there’s a consequence. Maybe you arise suspicion, or the guard sticks an escort on your to reach the noble’s house.
• You could now push yourself to cancel the remaining challenge die of 3, meaning to take 3 stress. Doing so would remove all challenge dice and give you a full success, no consequences or setbacks, but at the cost of taking on stress, which probably comes with its own issues later, but at least you succeed in the short term.
• If you had rolled no dice that could cancel out the challenge dice, you’d fail or you’d have to push yourself to cancel at least one die for a success with consequence

I think there’s something here. Design space could include things to remove challenge dice for players. Like a “Fast Talk” perk that lets you reduce a challenge by 1 Tier when lying. Likewise, your character could be wanted for some reason, so the GM would increase the tier for this challenge by 1 for a total of 4d6. Maybe the number of remaining challenge dice on a “mixed” result determines the severity of the consequences, incentivizing you to push a roll, cancel a challenge die, even if it won’t be enough to cancel them all.

What this does is exactly what I was talking about. The challenge is somewhat fixed in the sense that the number of d6 rolled give you an indication on how difficulty it might get. But all those d6 could roll a 1, making it a cake walk for you, as the guards are not that observant, or don’t care, or know exactly who you are but let you through anyways for some other reasons. Or the dice could all roll 5-6, meaning the guards are good at their job, vigilant, or the checkpoint is better guarded than you thought.

I just like the idea that you can set a challenge is really high or really low, and the dice might have a different idea in the end. And this particular version is just me thinking about my own, unique way of approach this. Challenge dice, in this early draft of a system, would allow the GM to easily assess a situation without needing to reference NPC statblocks or other systems. Since this is a narrative game, a simple Tier system to determine how many d6 a challenge is worth, will take a lot of a cognitive load off the GM, keep the focus on the characters and their choices, and allow for a smooth flow between dice and fiction.

At least I hope so.