There are a lot of people that want to run a sandbox, player-driven games, but it always ends up being a struggle. It takes a lot of skill to be able to improvise any kind of scene on the spot and create interesting narratives when players tend to go wild on the world, and on the flip side being able to do anything means you don't know what you want to do since you don't know which direction to head towards. But I think we recently found a system with some good foundations for tackling such mode of gameplay - the Chuubo's Marvelous Wish-Granting Engine.
Our group's GM has recently gone deep into playing I Was A Teenage Exocolonist, a life sim about growing up on an alien planet, and he wanted to run a game in the setting. Something more pastoral and laid back, so we settled on Chuubos. While the system is a lot to get into (we bounced off of it about three times before) once you get into the flow of it it does a lot of neat things.
The most important part of that system for our discussion is its Arcs and Quest system. Basically, every character picks a narrative they are a part of - the journey their character is taking through the story. If you are fed up with your mundane life and want to become someone you go on a Knight Arc. If you want to understake a spiritual journey and connect with a higher power, that's a Mystic Arc, while if you are training to overcome some personal limitation, that would be an Aspect Arc, etc. Every character's Arc will be developing in parallel as they play more and more of the game. It's basically your Hero's journey.
But a character Arc is a big structure, something that will take you an equivalent of an entire season of a TV show or more to go through. You need something smaller and more concrete, and those are the Quests - the steps on the Hero's journey. Something you engage with every session for a few sessions with concrete scenes in mind. For example, if you just so happen to want to blackmail your dog into learning martial arts, it might look like this:
Each of the Major Goals is something you would focus your part of a session on, at least a full scene devoted to it. The Quest Flavor on the other hand is just that - a bit of flavouring you sprinkle in whenever you want to remind everyone that you are indeed still blackmailing your dog into learning martial arts.
And with this structure we have our key for running low prep sandbox games - giving players a structure for their character's story to follow and key events they want to happen. Then each session all you need to ask the players is what scenes they'd like to happen to their characters and you have a good chunk of the session already accounted for. Similarly, since the characters are on an Arc that keeps the story directed towards a coherent resolution - maybe you start off the story by being defeated by a rival dog dojo, then you blackmail your dog into learning martial arts, and you finish it with an epic martial arts showdown... in which you realise that the real treasure is the time you spent with your dog and you don't need to perpetuate this bloody fued between your dog dojos... Or something.
Of course, this works only as well as the Quests themselves. Chuubos is a chonky book with a lot of them, but there is certainly room for improvements and eventually you find yourself wanting to make your own Quests, which just means you are actually doing prep for your low prep game. But on the plus side - that can be split amongst all of the players!
Another way you can engage with the system is by creating Group Quests - bigger things the player characters participate in and events that unfold around them. These work similarly - you have Major Goals that dictate bigger scenes that happen, and Quest Flavor that happens way more often. Here is where you can put the more traditional narrative of what the group is struggling against - perhaps dealing with a famine, or establishing themselves as a group of adventurers, etc.
So far, we have about 25 sessions of Chuubos under our belt and the structure has worked pretty well for us. Before each session we talk about what kind of scenes we want to happen based on what Major Goals we want to accomplish and that accounts for most of the session prep. Sometimes we go with the flow based on how the various scenes play out sometimes derailing our plans, but often it still flows naturally from one scene to the other. So far in our Exocolonist game we haven't had much of GM-led plot happen besides a few key events from the videogame being mirrored. Of course you still need to put in the legwork of preparing the cast of NPCs for players to interact with and the world for them to inhabit, but that's much easier with an established setting like this one.And this kind of mirrors what we have experienced running our other sandbox games, like our epic Exalted campaign. A lot of that game was figuring what are the big goals the players want to accomplish and giving them a neat structure to follow along that journey. We had entire sessions driven by players like me coming in with a set of tasks to do and banging them out as a baseline of the story. On the flip side the harder part was trying to motivate or help guide the players that didn't know what to do next sicne they didn't have a handy checklist of things to point to...
So it seems the key to running a lower prep game is to offload the prep onto game designers and the players to shape the narrative their characters will be on and then enabling them to follow that path into a hopefully satisfying journey, Chuubos style!
