Sandbox is a term we use in games to indicate that the players can pretty much go anywhere and do anything they like – they set the agenda, they choose the direction, and they go. This is a little bit ingenuous, though; they may get to go anywhere, but it’s anywhere on a list of places that exist in the game world. They can do anything they like, but in-game situations and out-of-game rules constrain those actions to a degree. Sandboxes aren’t completely player-driven, much as we may like to think they are. They’re a menu of options that the players can choose from.
That said, it’s important to have a wide range of options available if you don’t already have an idea for where the campaign is going. This will allow a broad spectrum of experiences for the players to choose from, and let you experiment with different tones, moods, themes, and techniques to find what works for the players1, and what works for you. There’s an added advantage (to my mind) of having a wide array of things in your Sandbox – it makes the world feel bigger than the characters, and more alive.
So, how do you build a Sandbox game?
The Nature of the Sandbox
Two games that I’ve run that have had great success with Sandbox-style play are The Armitage Files and my Fearful Symmetries DFRPG campaign. The type of Sandbox in each campaign is structured differently, and you need to decide up front which kind of structure you’re going to use. In The Armitage Files, the Sandbox consists of a set of documents liberally sprinkled with references to people, items, events, and places that are not explained, but sound mysterious and intriguing. In Fearful Symmetries2, the Sandbox consists of a list of locations, threats, people, and themes.
Really, the type you choose is going to depend on – and determine – what type of campaign you’re playing. The default assumption in Cthulhu-based games is that the PCs are investigators seeking out mysteries, so it makes sense for the Sandbox to be constructed of rumours, clues, and hints. In DFRPG games, the default assumption is that the game takes place in a given city3 that the characters know fairly well, so it makes sense for the Sandbox to be constructed from people, places, and groups in that city.
The two approaches are not mutually exclusive, of course. In fact, they can’t really be exclusive. You need some concrete things in a informational Sandbox to give the characters something to grab hold of in order to kick off an adventure. And you need informational things in a geographical Sandbox to let the characters know where the cool stuff is happening. But the balance between the two is important to consider.
Let’s take a look at a video game example to illustrate one end of the spectrum. Oblivion is very much a Sandbox game, with a strong geographic4 focus. Yeah, you get information to follow the main story along, but you also spend a fair bit of time just wandering around the countryside, stumbling across random dungeons5, and cleaning them out.
The Armitage Files highlights the opposite end. Aside from the assumption that the game is set in Arkham, Massachusetts, there is very little concrete or geographical structure to the Sandbox. The occasional reference to a specific place – Kingsport, Zurich – is still just an informational cue for the game. Most of the clues could lead the characters anywhere.
Most campaigns deal with going somewhere and doing something. If your Sandbox is primarily geographic, then characters will be going somewhere to see what’s there. They will look at the map6, and say, “We go to Bitter Creek. What do we find there?”
If it’s primarily informational, then characters will be going somewhere to do something specific. They will look at their information and say, “We head to Bitter Creek to find the missing prospectors.”
In either case, it could wind up with the same adventure – searching for missing prospectors – but the hook in is different.
Determining how much of the Sandbox is informational versus geographic will shape the ways the players interface with the game fiction, and say certain things about the campaign. So devote some thought to where you want to set that slider.
Putting in the Toys
When you have decided on the nature of Sandbox, you need to fill it. There are a couple of ways to do this:
- Solo. This is the traditional way to design a campaign. You sit down with your blank Sandbox, and think up all the stuff that goes in it. Pros: You get to put in exactly what you want, the players don’t know any of the secrets. Cons: You are limited by your own creativity, you have to do all the work.
- Collaboratively. This is the default in DFRPG. You and the players sit down and populate the Sandbox together. Pros: Less work for you, you get the advantage of everyone else’s creativity, players get invested in the game. Cons: You don’t have complete control of what goes in, the players know secrets, requires the players to agree to participate.
I have to say, I’ve become a huge fan of collaborative setting building, mainly because it gets the players excited about the world and it puts in things that I never would have considered7. That said, it does require that the players be good about separating player and character knowledge.
Whichever way you do it, it is vitally important not to do too much detail work8. You never know what is going to be important at this point, so you may wind up wasting hours – days, weeks – fleshing out things that never get touched in play9. Not only will this frustrate you, spending the time will delay the start of the game. And if you’ve taken the collaborative approach, every day you spend tweaking the things the players helped you come up with, their attention and enthusiasm will wane just a little bit more.
So, go high-level. Add a city to the Sandbox? Write two or three sentences about what the city is and what it means to the game. If you’re using a Secrets Deck, make sure you come up with at least one secret for the thing. For example:
Belys is a prosperous city-state ruled by a collection of genasi noble houses. It evokes the Thousand and One Nights Baghdad feel crossed with Renaissance Venice, with wondrous magical devices for sale and convoluted politics and scheming behind the scenes. This is the foreign city that becomes the characters’ home base in the Paragon Tier.
Secret: The mystic power of Belys is based on an arcane machine that imprisons a legion of djinn and efreet, harnessing their energy for the use of the noble houses.
That’s more than enough to go on with. Now I know enough about Belys to seed some hints in the rest of the game, and to improvise if the players suddenly decide that they really need to go there now!
If you’re putting lots of elements in your Sandbox – and that’s really kinda the point, after all – coming up with just this much for everything is going to be more than enough work. I recommend tossing in a few evocative references with nothing attached to them for developing later – the ruined tower of Asterys, Kraken Bay, the Rookery, whatever sounds cool and fits in the campaign. That way, if you have a good idea after the game starts, you have something to attach it to.
The nature of your Sandbox – it’s place on the geographic-informational continuum I made up in the topic above – will determine what sorts of elements you put in it. If the structure is primarily geographic, the elements are mainly going to be places, with some people and rumours thrown in. If it’s primarily informational, then you’re going to have a lot of clues, rumours, hints, and people with information, with a few places and items thrown in. Mix and match as required for your vision of the campaign.
Showing the Sandbox to the Players
Once the Sandbox has toys in it, you have to show it to the players. How you do that is going to depend on what sort of Sandbox it is, and the forms the toys take. If the game is primarily geographic, you may want to hand them a map with the various locations labeled on it. If it’s informational, you might, for example, hand them a mysterious document with a number of unexplained but intriguing references.
If you’ve done setting creation collaboratively, the players will already know a fair bit about the Sandbox. In these cases, I often just type up and flesh out the notes we came up with at the setting creation session and distribute that to the players10 as the setting bible. It’s important at that point to have a discussion with the players about segregating player knowledge from character knowledge, but so far I’ve found with my players that their involvement has made the setting cool enough to them that they will happily ignore anything their characters shouldn’t know so as to have the fun of finding it out in play.
The point is, of course, to let the players see what options they have. You don’t have to give them a look at all the elements in the Sandbox, but they do need to see where a few things are, and get an idea of the scope and nature of the setting so that they can start making decisions. I mean, yeah, you can plop them down in the middle of nowhere with no map and say, “Where do you go from here?” But that initial decision, being pretty much totally random, is meaningless to the players, and to the characters11. You need to give them some context and structure to complete the buy-in and make the game matter. You need to give them some sort of map, even if it’s just a blank sheet with a dot that says You Are Here, two dots marked Sweetwater Gulch and New Zion, and a line connecting the three points marked Road. Now they’ve got real, meaningful options.
Setting the Agenda in Play
Okay, so you’ve got your Sandbox all set up, and you’ve shown it to the players. What next?
Now, you have to start structuring the actual adventures. In a broad range of choices, it’s easy for the characters to become paralyzed with indecision about what to do next, so you have to point them subtly12 towards the adventure. The best way to do this at the beginning is to constrain their choices.
Yeah, that sounds like a bit of hypocrisy after the whole bit about building in choices and making sure the choices are meaningful, but hear me out. Traditionally, RPG adventures initially place the characters in a reactive role: something happens, so the players have to respond. A stranger in a bar needs help, so he asks the PCs to go into the dungeon. A socialite is murdered, so the PCs have to find the killer. The supervillain is robbing a bank, so the PCs have to stop her.
It can take some training before players will actively set their own agendas and seek out adventures. They need to see that they have the power – the agency – to set the agenda, and you may need to lead them to that realization gradually. So, start small, dropping pointed opportunities rather than outright adventure hooks: instead of the bartender telling them that some punks have stolen the bar sign and the PCs have to get it back, just have the whole bar be surly and upset, and let the characters figure out why that is and decide for themselves what to do about it. It’s a small step, but it will eventually lead to PCs telling you what they want to do in the next adventure13.
Even if your characters are used to setting their own agenda in games, you still want the choices to be a little limited at the beginning, just to help them get into playing their characters and interfacing properly with the campaign and setting. Leave the big choices for later in the game.
Now, once the players start really taking the lead in setting the agenda, you will sometimes find they have a tendency to deliberately try to surprise you, or put one over on you, or fake you out. This is an artifact of the adversarial-GM fiction that I’ve talked about before – the players “know” the GM is out to get them, so they have to trick the GM in order to win, whatever winning means. How do you deal with that?
Easy. Ask them not to. Tell them that, while you’re totally cool with them setting the agenda, you need a little prep time to make sure you have interesting things for them to do. If you talk to them about it reasonably, and play fair with them14, they will be more than willing to be honest and upfront with their plans, so you can make plans of your own.
Which brings me to…
The beauty of the way Sandbox games are structured is that you don’t have to build in a lot of depth before you need it. You don’t have to have thirty fully prepared scenarios ready to go at a moment’s notice, just one. As long as it’s the right one. That’s no more than you need to prepare if you’re running a more traditional campaign, where you as GM set the agenda and dictate the adventures, but it has the added bonus of being something you know the players and characters are interested in because they chose it. They have choice, you get to flesh out that choice to make it cool, then they get to play through it and make the whole thing cooler.
Now, making sure you have the right adventure ready is very much a matter of communication with the players. For the first adventure, I talked about constraining the choices available to the characters, and I gave a couple of reasons. There are other reasons, having to do with preparation: if you limit their choices, you need to prepare less for that first adventure. I recommend building just one adventure, but have a couple of different ways into it. Yeah, this is a bit of a cheat15, but it gets you playing and pulls the characters into the game. Then, at the end of that adventure, ask them what they want to do next.
Couple of important points about that:
- Ask them at the END of the adventure. This gives you time to prep the next adventure based on what they want to do.
- Force them to a decision. Don’t bully, but make them choose something specific so you have a starting point for your prep work.
- Get them to commit. If you’ve put in a month’s work on an adventure that they’ve said they want, and they show up at the session, and say they’ve changed their minds, I think it’s allowable to strike them in the head with something heavy16. Make it clear that their choice is binding, and if they come up with a better idea at the start of the session, defer it for a later adventure.
- Make sure you accept their decision. Point out options, offer opinions, but don’t try and make them choose something they don’t want. Once they’ve chosen, don’t try and weasel the adventure around to something else. Don’t use the adventure to punish them for not going with your idea. Basically, don’t be a dick.
When they’ve told you what their plans are for the next step, prepare the adventure based on that. I don’t pretend to know what kind of prep work you need to do for your game – that varies from system to system, and from campaign to campaign, and from GM to GM – but spend that time trying to make the characters glad they chose the option they did. Pour coolness on the idea, throw in some neat twists and surprises17, make the opposition interesting and engaging, and do what you need to do to make the adventure rock.
A crucial part of preparing for a Sandbox game is keeping track of what happens so it can inform the rest of the game. In a linear game, this is pretty straightforward, but it’s a little more complicated for Sandbox games. Take notes, and leaf through them when you’re prepping adventures to see if there’s anything interesting that you can call back up to add some continuity. Make sure you don’t lose the name of the NPC that you made up off the top of your head but has now become important. Keep track of any surviving villains and not-quite-extinct plots and conspiracies, and any extinct NPCs or cities or helpful organizations. This becomes invaluable when it comes time to start pulling the threads together for the emergent storyline that the campaign generates.
No matter how much you prepare, though, you’re going to wind up having to wing it from time to time, so make sure that you’re ready for that, too. Keep an encounter or two salted away for when you need to send in the ninjas18, and try and tie these encounters into other aspects of the game. Using a system that is easy to improvise in – GUMSHOE and FATE, for example – means that you can get a lot of mileage out of a single encounter, while systems that aren’t quite as easy to improvise in – D&D, for example – may mean you need to have a couple encounters ready just in case. If nothing else, sending in the ninjas gives you a little bit of breathing room to cope with the unexpected player choice that prompted your little panic attack.
When you do improvise, it’s even more important to take notes to keep things straight. If you haven’t tied the improvised section into the main plot before, take a good postmortem look at the notes, and figure out how it’s related after the fact. It helps build verisimilitude. Not that it has to be tied to the current main plot; sometimes, it can be fun to throw in an alternate storyline to see if the characters19 can figure out that there’s actually two different things going on at once. Or, it can be a hook into a new adventure, showing up a little early.
Lather, rinse, repeat.
That’s about all I have to say about Sandboxes. The next emergent storylines post will show up within the next couple of weeks, and I’ll talk about Watching Their Eyes.
- I’ll talk about this in much more depth when we get to the Watch Their Eyes post. [↩]
- And any DFRPG game that goes through the setting-building section of the game. [↩]
- Though this is not necessary, and is addressed in the setting construction chapter. [↩]
- Most CRPGS work on a geographic slate. It’s certainly easiest to conceptualize in that medium, and allows the programmers to scale the difficulty of the opposition by geographic area. [↩]
- Seriously, what’s with all the dungeons littering the countryside? Why does every little cleft in the rock open into a vast underground network of caves filled with monsters? Shouldn’t someone be doing something about that? [↩]
- Or whatever equivalent you have in the game. [↩]
- Baba Yaga in the sewers of Dublin, for example. [↩]
- That way lies madness. No, really. Madness. [↩]
- I’ll talk about this more when I get to the Discard Liberally post. [↩]
- Less any secrets I’ve thought up in the meantime. I mean, the players need some surprises, right? [↩]
- Not to mention that it undermines the notion that they have free choice in the campaign, because they can’t see that their choice makes any difference. [↩]
- And sometimes not-so-subtly. [↩]
- While this may feel odd to you as a GM at first, embrace it. It lets you focus your creativity on what happens in the adventure, rather than on what the adventure is. Relax and let them boss you around. [↩]
- This is, of course, an important point. Don’t screw them when they do what you ask them to. [↩]
- I don’t like inflexible rules in anything, including running games. I will use any tool I need to in order to build a play experience the players enjoy, even if it means I have to lie, trick them, and cheat behind the scenes. I will do what I need to do to bring the cool. [↩]
- Disclaimer: Maybe I don’t really think this. But when it happens, I want to. [↩]
- But, as said before, don’t weasel it around to a different adventure. [↩]
- Whatever the ninjas happen to be in your campaign. [↩]
- And players – this can be very confusing for players, so be cautious. [↩]