RPG Errata: Darkest Days, and Random Difficulty

Darkest Days is a dark fantasy RPG designed by Bell Moon Games. A bit souls-like, the game sees its players as former legendary heroes returned to life by ancient ritual, in a desperate attempt to save the world from demons and monsters. It has an interesting “pip” system for skills, which gives you bonuses from both your stats and your equipment.

One of the more interesting parts of the Darkest Days ruleset is how it handles skill checks. In most modern RPGs, when a player wishes to undertake a task, they are given some measure of difficulty. Most systems call this a “difficulty rating” or a “target number.” You have to roll “better” than this number to succeed at your task.

Okay, but where does this number come from?

Again, most systems rely on the GM to divine the number, based on their judgment of the situation, setting, and the character’s abilities. Some systems provide a chart of difficulties and their corresponding numbers as a guideline. Some recent systems have a set TN that applies for every situation.

Darkest Days goes in a different direction; it chooses the TN randomly.

In the Skill Checks section of the rulebook, Darkest Days tells the GM to roll a die or two every time a player wants to undertake a task. If the players are level 1-5, for example, the GM rolls 1d20-1d6. whatever the result is, that’s the TN for the task.

This might seem odd to those of a more deterministic view of GMing. After all, the average of 1d20-1d6 is 7, but it’s still a spread of 1 to 19. what if the player is trying to do something particularly hard? What if the dice decide a run-down wall is harder to climb than the steel gate is to pick?

Give it a moment of thought, however, and you’ll hear the echoes of my post on dice narratives. What if the dice say an easy wall is too hard for the thief who rolled a 1? What if the nearly impossible puzzle is solved by a lucky roll from the barbarian? Dice already wrest control of the narrative away from us, so why not subject the GM to the same forces?

Is that railroading?

Looked at that way, the randomized TN is just another kind of Oracle. With Oracles, you roll the dice to see what challenges are next, and with this TN method, you also roll to see how difficult the challenge is. It’s a way of affirming the RPG theory that the GM isn’t here to guide or shape the meta-narrative, but to simply be a referee.

Ordinarily, while the dice decide whether a player succeeds or fails at a challenge, the GM is the one who decides the probability of success. By virtue of this power, the GM shapes and guides the direction of the story. Players become more or less likely to take certain actions, which can then be prepared for. If the cliff is smooth and shear with a 25 TN, but the elevator gate locked with a 10 TN lock, the players will likely slip through the gate rather than risk falling and dying. That lets the GM plan for a broken-down elevator, or a griffin attack, or archers and oil from above, or even just an ambush of soldiers while the players are trapped in a small brass cage.

With random TN, the GM is forced to play in the fair unfairness of probability. If the players want to scale the cliff instead of put themselves in a slowly-rising cage, then it’s the dice that decide how viable a plan it is, not the GM. If the GM wants this door to be particularly difficult to break down, they don’t have the power to force the players down a different path.

At the most balanced, this method creates a natural bell-curve of challenges, some very easy with others quite difficult. At its worst, its no worse than the tyranny of the dice from the PC’s perspective.

Fundamentally, this design pushes the GM of their lofty perch as something beyond the other players. They aren’t in control of the story any more than the rest of the table; the dice decide both the challenge and the outcome. The GM now has to come up with reasons why this door is easier or harder to get through, much as Players have to come up with reasons why their characters do or don’t manage to succeed.

But that phrase: “fair unfairness.” There are other alternatives, right? The OSR movement doesn’t aim for “fairness” in its methods, it aims for player agency. If a wall almost impossible to climb, it should have a high TN; if the players decide to take the nearly impossible route then they accept the consequences. TNs aren’t just arbitrary obstacles, they’re ways of representing the narrative world in the system.

Using this framework, the die-roll has more narrative agency than one might expect. It’s a roll that could be made by the GM, if we wanted to be thematic about it. A lucky or unlucky roll shapes the story, allowing for fantastic and surprising moments of heroism. This framework requires a lot of creativity and flexibility to work well, but when it does, the dice shape a story within an established setting, rather than shapes the world around it.

As always, there’s value to all methods, but I am surprised that it took until now for such a simple design change to be experimented with.1 Either way, it’s a system well worth trying, and worthy of a place in the toolkit of RPG design.


  1. Assuming, of course, that it wasn’t. I’m not going through every RPG from the 70s til now to be factually accurate on this point. ↩︎