Janna at Dungeonmastering recently wrote up an interesting article about running non-linear games . They’re a neat and slightly under-explored concept, so I thought I’d bring in other people’s discussions and see what we can figure out. They are commonly discussed using a variety of names… sandbox play is another common example.
In one sentence: In a sandbox game, you wander around and do what you want instead of following a plot.
The world exists independently of the PCs, who wander around and pick their adventures and plots. Events tend to follow as consequences of previous decisions and are usually logically inferred from NPC reactions to the actions of the PCs. “Destiny” and strong plots are rare.
Rather than creating an adventure that feels like a story, with opponents carefully gauged to the PC’s powers, thematic elements tying to the PC’s backgrounds, and the like– instead the GM creates a world. The world can be (and usually is) complex– a lot of deeply detailed homebrew worlds are created for this style of play. Often the world is created long before the PCs are generated.
Structure of a Sandbox Game
Unlike the short campaign notes Phil talks about in Lessons From The Long Campaign– Prep Only What You Need , one ideal sandbox game is built on extensive notes, a world history, detailed NPCs, etc. (More like the failed Mutants & Masterminds game from his example.) Ben Robbin’s successful sandbox game was called Westmarch. (Read all five articles to see how it developed and played out, or just the linked post for a sweeping overview.)
Sandbox games often involve more detailed creation of a smaller area than typical GMing. The same area will be tread across again and again, as the PCs investigate one story, switch to politicking with the city council, then infiltrate the black market, etc. (Or, as in Westmarch, when different groups wander through the same wild area.)
Plot lines are often less boldly presented and the game is likely to have several plots and events running at the same time. Often players are given a chance to choose which plot of several they intend to address– whether they’re clearly identified as plots or not. Many “plots” are just NPCs acting to accomplish their goals. Sandbox games can even lack plot altogether– though that’s more often leveled as a criticism of the style.
Modules in the Sandbox
Typical module play shares some elements of a sandbox game– the plot and setting exist independent of the PC’s actions and any group of PCs can get roped into solving the module. However, modules usually feature a narrower plot and try to keep PCs moving from one plot point to another– in part to spare the GM from having to invent a lot of material during free exploration. Some modules try a hybrid approach– providing a clear setting [often one structure, like a haunted castle] and a timetable of events.
A different approach to module design is to set up a complex situation and see what the PCs do with it. The Well of Souls  is an interesting Glorantha scenario setup in just that way. A few background events are laid out, but what happens once the game begins depends entirely on who the PCs befriend and ally with. (Character selection and history in such an intimate setting strongly skews the list of options your PC can even imagine.) Sandbox play often results. NPCs want things and will ask, demand, or negotiate for them– and it’s up to the PCs to decide who they trust, if and how they want to investigate the death, and even what solution they’re willing to try, if any.
Advantages and Disadvantages of Sandbox Play
I’ve played in both good and bad sandbox style games. The best games remain directed– PCs with strong drives and motivations can provide this, as can an oppressive situation that inspires PCs to work to change the status quo. For example, Sorcerer requires PCs be built with a kicker (a dramatic event that kicks the PC out of their routine) to provide the initial drive.
Bad sandbox games can feel like plotted games, but with a lot of wandering around and boredom between plots. Players who are used to plotted games will sometimes wander around and interact with NPCs mechanically, just trying to identify the next plot. PCs who quickly achieve their goals (and have nothing left to motivate them) can leave the player rudderless. Standoffish and loner characters may find it difficult to get invited (or drug into) fun plots.
Janna’s advice about using plot hooks to tantalize the players instead of smashing them with plot hammers is excellent. While consequences are an excellent driver of play, you need to make sure that the consequences enforce the feel of the game you’re looking for. If you want dramatic heroics, suing the PCs for property damage after their first fight will chill what you’re trying to encourage.
Above all, if a PC repeatedly stays home to play computer games, something needs to change. The player may need the structure of a plot, or the contents of the sandbox may need tweaking to engage the PC more thoroughly.
What are your experiences with Sandbox style play? Have you built a sandbox game for your players? How did they respond? Did it take any effort to orient them to the new playstyle?