If you roll too high on some of these moves, you get an overkill: a success with a consequence because you were too powerful. This is a wonderfully flavourful mechanic that prompts you to consider whether you want all three statements to be true all the time. Maybe it’s safer to hold back so you don’t hurt someone close to you. The randomness becomes a big factor in establishing your character’s position in a scene. Especially when the odds of rolling 10+ on 2d6+3 are just over 58%!

Another important factor to consider when designing for randomness is the dynamic between the randomness and player choice. With the previous example of a move in Godkiller, the player’s action triggers the move (“When you inflict violence on someone...”) and then the player rolls to determine the outcome. The decision comes before the randomness. This is called output randomness. Output randomness can sometimes feel very chaotic, since players are forced to react to the consequences of the randomness as they appear, rather than be able to plan based on the results they know they could get.

The opposite of output randomness is input randomness, where the randomness occurs before the player makes a decision. A classic example of this is a hand of cards in Poker. The player gets their random hand at the beginning, then chooses how to bet based on it. In the comments of the Skeleton Code Machine post about Input and Output randomness, Exeunt Press describes input randomness in TTRPGs as “inherently limiting,” using a hypothetical example of a game that gives you a hand of random actions to choose from. But I don’t think this is the only thing that input randomness can mean in RPGs. For a different mechanical perspective, I want to propose Spark Tables, a mechanic that I love from the works of Chris McDowall, especially Mythic Bastionland. See the original post about spark tables on Chris’ blog here.

Spark tables are a GM’s best friend. I’ve been obsessed with them in my recent designs. They’ve shown up in my previous and upcoming solo games: Visitations (which came out last year) and the upcoming A Witch’s Pilgrimage (WIPs are available to paid subscribers!). They are a great example of input randomness: you are given some random words, then “react” by interpreting the words into a more fleshed out and meaningful narrative.

For a while I believed that input randomness was inherently better and lead to more player agency because of the greater information, but I don’t buy that any more. I think it depends on the range of the randomness in your game, something I have also previously written about. If the range of randomness in a game includes situations where the next decision is obvious, that will always have lower player agency than a game where the randomness always leads to interesting decisions. This can be okay! Especially if those non-decisions feel thematically appropriate. But they can also be frustrating if used in the wrong place.

I think this is why I find spark tables so compelling. The entries (at least with Chris’ tables) feel like there are almost always multiple ways to interpret them. Then, when you begin combining multiple prompts, the people, places and things you create feel completely bespoke to your game.

It’s like GM prep magic, I’m obsessed.

---

Thank you for reading! I really wanted to do a full rundown of Uncertainty in Games, but like I said I didn't have the time. All my blog posts are available for free, so join at the free tier to get them in your inbox!

Read more at the Crackerjackalope Games Patreon: