Sandbox Play is a term that comes from computer RPGS, though it is said to refer back to when war games were played on sandboxes where the wet sand could be shaped into any kind of terrain. It refers to CRPGs where there is no plot that the game is taking the player through, but instead the player is free to roam around the world and interact with it. The player might encounter NPCs who will offer missions, but is free to ignore them or abandon the mission in the middle if it’s too hard or boring; the player can also just instigate things on his own. Grand Theft Auto: Vice City is offered as an example of this style of play (I’ve never played it myself).
By extension in tabletop RPGs Sandbox Play refers to the same sort of qualities in a campaign: no particular overarching plot, with the players free to pursue their own plans and interact with the world as they see fit. There’s no built-in supposition that the players will go on any particular quest or be employed by any particular NPC or organization. There’s some dispute among people who use the term as to whether such things as plot hooks or adventure seeds even exist in Sandbox worlds, but the general consensus seems to be that as long as the players are free to ignore the cries for help, or the reports of trouble brewing in the coastal provinces, it doesn’t diminish the Sandboxness of the setting that it contains such things. In some sense, it’s the essence of the style that such things go on whether or not there are PCs there to interact with them.
Things that are anathema to Sandbox play include: “level appropriate encounters” such that wherever the party goes they only encounter things that they can defeat, but not too easily; Schrodinger’s NPCs, who always are just about to get mugged no matter which alley the PCs are walking by; GMs applying meta-game pressure to the PCs to go on a particular adventure because “this is what I’ve got prepared for this week”; travel that moves at “The Speed of Plot” so that however the PCs choose to travel, and whatever stops they make on the way, they always arrive at the villain’s castle just as the final evil ritual is about to commence. In other words, anything that disturbs the illusion that the game world exists in all its detail and with events proceeding along their course whether or not the PCs are looking in that direction.
Pretty much anybody who’s played in one of my campaigns will recognize this as my default style of GMing (one-offs are a completely different kettle of fish, since they are usually defined by the particular story the PCs are caught up in and end as soon as it’s resolved). Certainly I’ve done things at various times that have violated the spirit of the Sandbox, out of laziness or because the players have started to flounder and asked to be gaffed with a plot hook, but basically Sandbox Play is what I like to GM for anything reasonably long-term.
I think part of it is that I don’t seem to be very good at the overarching plot style of campaign. While I certainly admire the aesthetic of the Sandbox, I’m not married to it. But when I’ve tried the plot-driven campaign (because my players have asked for one)….it doesn’t seem to work out very well. Frankly, their characters usually take one look at the set-up and run away. I don’t know…maybe I’m not very good at conveying the scope of the problem they’re facing versus the extent of the character’s powers, so they think it’s overwhelming or insoluble. My players will sometimes say it’s because “we suuuuuck” but they don’t really believe that, and neither do I. It’s true that my games often feature large-scale unintended consequences, but that’s a reason for caution, not outright paralysis. I mean, it certainly doesn’t paralyse them in the Sandbox campaigns…they just think twice about doing anything major and sometimes prepare contingency plans, then scramble to pick up the pieces if something unforseen occurs. But somehow when it comes to a campaign with a built-in plot from on high, they chicken out.
Now, one thing I haven’t yet tried is making such a campaign set-up do-or-die. The Sandboxer is strong enough in me that when the characters have fled their destiny, I’ve shrugged and let them run, coming up with new stuff in the setting in whichever direction they’ve bugged out. Their destiny has never come and dragged them back kicking and screaming. Maybe it should. It could be that I’ve been confusing character reluctance to buckle down and follow the course of duty with player reluctance to embark on the campaign. I certainly don’t want to force the players down the path of a long and complex adventure if it doesn’t seem fun to them, but maybe once in a while they want to be railroaded–at least to kick off the campaign.
Tales of the Rambling Bumblers 
This goes against all my instincts as a GM. It sounds great in theory, but in practical terms, how can you have material to keep the players interested the whole session if you don’t know at all which path they are going to take? If you have the whole Parlainth region mapped in your head, full of hooks and NPCs, but the players decide to go to Kratas to sell some stolen loot, what do you do? Improvise the whole session? Because you simply cannot have something prepared for absolutely anywhere they may decide to go.
I guess you can improvise and still come out on top, but not every session. And as a roleplayer who prefers solid, detailed and surprising plots, I assume a little nudging the players in the “right” way is a small sacrifice to enjoy the game. As long as it does not degenerate into railroading, of course.
It’s not for every GM or game group. It’s best for the type of GM who likes to work out lots of fiddly little details of the setting in advance just for fun, or else has improvisational skills and memory for consistency that is top notch. If you’re comfortable using published settings, there are some that go into sufficient detail or just have been around long enough that they can be a big help with Sandbox Play.
Another approach that some people use is to have lots and lots of stuff be driven by rolling on tables. It’s actually pretty easy to scroll the world in a direction that hasn’t been visited before if you’ve got charts for generating it on the fly. The key there is that you need to have good charts that vary by region (and perhaps by things like season and time of day) so that the world doesn’t seem wacky and random. Just a set of monster encounters won’t do. It also helps to be willing and able to guide or supplement the chart on the fly to improve the consistency and creativity.
It’s a bad approach for GMs that aren’t into world-building as an activity in itself, and who will be bothered if none of the players ever run into all this lovely work they’ve put into it.
It’s also emphatically not for every player. Sandbox-style games pretty much come with a guarantee that the PCs aren’t the focus of the world, and no guarantee that an avenue that the PCs choose to pursue over many game sessions won’t peter out in anti-climax or end in failure. If your players want their PCs to be the axis around which the world turns–which is a perfectly reasonable thing for them to want, since it’s the default assumption of almost any story-oriented game–Sandbox Play seems like it’s designed to deliberately thwart their fun and destroy their reason for playing. It’s also not a good match for players who don’t make characters that have goals of their own that they’ll pursue without Fate nudging them (or railroading them) along. Characters who have something specific they want to accomplish, whether it’s explore the world, climb to the top of the thieves’ guild hierarchy, or settle a particular region, are much easier to accommodate with Sandbox Play than characters who just want to wander around until adventures overtake them.
Finally, just because the point is that the characters have the illusion of a seamless world at a consistent high level of detail doesn’t necessarily mean that the players have to share the illusion. It’s perfectly ok (IMO anyway) to tell them that you’ll have to call the session until next week so you can prepare if they suddenly decide to take it in a surprising direction. It’s even ok, though not perfect, if that causes them to decide to hang around the current area for a couple days of game time so they’ll have stuff to do that evening. Naturally, this can’t happen too often or it’ll become frustrating for the GM and the players, but it’s not impossible to have a Sandbox-style campaign just because you can’t have something prepared for absolutely anywhere they may decide to go. It helps that in Sandbox Play you almost never just wave your hands and say “three weeks later, you arrive in Region X”; it’s expected that if they set out in the direction of Region X they have to play it out…part of the point of a Sandbox campaign is that interesting stuff may happen on the way.
This is great, Josh. I thought I had a good understanding of sandbox play, but your article here gave me some cool new ideas.
@amatriain: In sandbox play, the scenario you described (forsaking Parlainth for Kratas) comes with a whole host of challenges. Simply getting there is an adventure unto itself…as Josh noted, PCs don’t move at the “speed of plot,” so a journey from Parlainth to Kratas can quickly become an expedition lasting a session or more.
This, to me, seems to be the essence of sandbox play: No story element should be overlooked. GMs shouldn’t gloss over days-long travel stints; rather, these are opportunities to explore the world and react to new discoveries.
What games that we’ve both played in would you classify as “sandbox”? The one that comes to my mind is John’s Water Uphill World.
If I’m understanding your use of the
phrase, to the extent that I ever
ran sandbox style, i.e., dungeoneering parties without motivation other than loot and power, I largely ditched it because it didn’t appeal to you, and began adding “plot”. So I am somewhat surprised to find you so passionate about sandbox play. (Of course, maybe I just ran a really sucky sandbox game…)
@PatrickWR: I know a lot has been written in the blogosphere recently about “speed of plot”, but I wanted to say that what you describe is not exclusive to sandbox-style play. My games are always plot-oriented, and I’ve often turned a trip between two points into a session onto itself. It’s a great opportunity to run an “isolated” story between two sessions relevant to the overarching campaign plot. And I find it’s a good idea to put as much local color into this kind of sessions as possible, in order to flesh out the setting and give them a feeling of the world. In sessions related to major plot points, there is often not much room to introduce local color.
@Russell – Water Uphill World is probably the only one where we were both players. I think of most of of the Neng campaigns as being Sandbox, and Neng itself designed for Sandbox play. Despite there being big shiny campaign seeds lying around, nobody has ever really been under any obligation to pick them up, and I think all of the groups but one have completely shifted focus at least once based on the characters’ changing interests and opportunities. E.g. The Friends of the Fox campaign was intended to be this big overarching story, but the party went in a completely different direction almost from the git-go…to the extent that I’m not sure any of the players even remember having met the Fox back in session 1.
And Dungeoneering games don’t, IMO, tend to make for very interesting Sandbox settings, because the Dungeons are generally sui generis, with no particular conceptual connection to the outer world or to each other. They have Sandbox-like features in that the players tend to control where they go next in the Dungeon and at what pace, but the games tend to resist the characters taking up any activity except Dungeoneering. If you want to stay in town, there’s often literally nothing for you to do except shop. It would just be bizarre to decide that you wanted to run for mayor, or open a trade-route through the mountains to make a fortune. To the extent that you would ever get involved in doing something like that it would be because some NPC hired you to do it.
Someone, I forget where, probably one of the big RPG forums, once said that Sandbox games don’t have Plot, they have Plans, and that resonates with me. In a game with a Plot, the GM can often block out roughly what the scenes will be before the players even roll up characters. They will interrupt the mugging, be appealed to for help, discover the murder, go to the abbey to investigate the manuscript, be set upon by footpads, discover the plot to blow up the bridge, and confront the wicked uncle. Playing the game is just filling in the details. A more complex scenario might cover what happens if they go investigate the old mansion before they investigate the abbey, or if the footpads win, or they fail to prevent the destruction of the bridge. It doesn’t have to be railroading, and it doesn’t have to be destined for one particular ending, and it can present the characters with a big honking moral dilemma to wrestle with, but the shape of the story is all laid out in advance. It is plotted.
A Sandbox game might have all the same elements, but they wouldn’t be slated to occur in any particular order, and the PCs wouldn’t be penciled in for any particular role in the happenings. Instead the NPCs would have certain plans: the wicked uncle scheming to blow up the bridge to prevent the garrison from the fort from relieving the town when the uncle’s paymasters raid it; the virtuous niece looking for someone to help her against the mysterious men bent on kidnapping her; the false abbot and his brigand/monks planning on looting the town and recovering the Orb of Phantasmagoria, etc. The PCs would be presumed to have their own plans, perhaps to set up a trade route through the mountains to get rich off trade between the town and the Dwarves. The GM would set it up and then it would just unwind. At certain points, depending on whether any of the Planners have interfered with each other, some of the events mentioned before might happen (if unchecked, the uncle’s men will blow up the bridge, if the PCs won’t help the niece, maybe the dashing young cavalry commander of the fort will). It could even turn out that when it was all over, the recap of the Sandbox game matched the recap of the Plotted game in every particular. But the process by which it came about would have been completely different.
Well, there’s a lot of room between a sandbox game and a plotted (or pre-plotted game). In my hero-cities game, there was certainly no predefined plot, and there were many player initiated agendas. They set up a merchant company, became double agents,
conquered and ruled a village of humanoids, etc., without my knowing more
than a session in advance. But it wasn’t a sandbox game, because I basically guaranteed that whatever they did, there would be A plot. I didn’t know what the plot would be, but it would exist in retrospect. And there were NPC plans that succeeded or failed, but I really didn’t make up these plans until I at
least knew which direction the group was
heading. If they had an agenda, I let them pursue it, but if they seemed to be at a loss, NPC’s with problems would keep turning up until they accepted a mission.
I certainly don’t remember meeting the fox! But we seemed to be given a mission
(rescue the “princess”, prevent eternal winter) that didn’t leave many alternatives.
@Rusell- I’d agree that there can be degrees and degrees of plotting vs. Sandboxery.
Actually, though, the ones you mention were two separate adventures, and I think offer a reasonable example of Sandbox Play.
In the first case, iirc, nobody approached you about recovering the Princess…but you had met her and even though you didn’t much care for her, you didn’t particularly trust her guards (with good reason as it turns out) and went off to look for her on your own account. When you found where you thought she was, you recruited help when you decided the opposition was too tough for you–as it was, since I made no particular attempt to scale it to your “level.” More than that, though, on your first pass you rescued the doppelganger despite the Princess’ being there all along for you to find if you’d looked harder, and returned “her”…advancing the villain’s schemes in a way that totally surprised me. Then, when you could have wandered off with your reward and the feeling of a job well-done, leaving the doppelganger in place for me to make hay with later…you decided to go back to the cave complex and take another look. This time you discovered the real Princess, and the moment of realization and the look on your faces is one of my cherished memories of the campaign. Even better because I never engineered it that way, and probably would have discarded it as too unwieldy if it had occurred to me. Finally, the choice to smuggle her back to the Palace and swap her for the doppelganger, but rescue the doppelganger and let her choose her own path was pretty much all yours (within certain constraints that I had set up about how magic worked in the setting that made simply killing the doppelganger risky).
In the second adventure, it’s true that preventing the everlasting winter was a pretty clear-cut task and your characters probably felt they didn’t have a lot of choice at that point (though if they had fled, it was inevitable that some heroes would have been found to take up the task). But the plan to remove the dangerous artifact from the world and prevent anybody from attempting it again by giving it to the God of Death, that was all you guys. To this day, I’m not even sure it was such a great idea, though it sure made for some great adventures.