Building a Sandbox

[Jorunkun]

Before you read any further, I suggest you check out an excellent series of posts over on the Ars Ludi blog. . It describes in great detail the design theory of and learnings from the author’s “West Marches” campaign, a sandbox setting that ran for over two years, with some 20 (!) players.

Inspired by this, I’m now in the process of building my own sandbox campaign, for when 4e launches. My core design tenets are these:

1. “Wilderness as a dungeon”
The most basic sandbox campaign is an outdoor map and a random encounter table – it’s all you need for the PCs to freely move, explore and get into trouble. I figure that the more detailed this map and the more region-appropriate and varied the encounter tables, the more interesting the game will be at its heart. This is why I have built out a huge 80 x 120 mile map, at a scale of 1mile per square.
Essentially, I’m treating every micro-region as I would treat a room in a dungeon, setting down some specific detail for every part. It’s quite a bit of up-front work, but I think it is time well spent, as players will go back and forth on the map, so these “rooms” will see lots of re-use.
Also, since the wilderness will be the main setting of this campaign, I intend to really play out the mechanics of overland travel, including seasons and weather effects, climbing, tracking, boating and riding, finding food and shelter etc.

2. Top down, inward-out encounter planning
Next, I am dividing up the map between the most powerful entities of the world - individual monsters or groups. I see these like radio towers, broadcasting their presence into further-away squares in proportion to their power. I am dividing up the related encounters into three types: Random-encounter/information, set piece, dungeon.
For example, the most powerful entity on the map is going to be an evil red dragon, who lives in a volcano and is worshipped as a god by a clan of dragonborn. The closer players get to its lair, the more likely it is that they will see signs of the dragon and his followers (scorched earth etc) and run into related random encounters (ie. Dragonborn patrols). If they proceed, they will find (pre-placed) outliers; small dungeons or scripted encounters like in the Book of Lairs. Further in still, and there will be a full fledged dungeon; a substantial map with the dragon and its treasure at the centre.
As the players move about the map, they will be exposed to overlapping, different-strength signals, allowing them to make informed decisions about where (not) to go.

3. Parallel, emergent plots
The biggest difference between my previous campaigns and the sandbox design will be the absence of a scripted plot. Having run plot-heavy, custom-tailored episodic campaigns for the last twenty years, I’m really excited about trying out a more simulationist/game-ist approach next.
Essentially, there will be half a dozen or so competing factions (cities or organizations) on the map, that the players can work with or against. The more powerful they become, the more influential their support will be for their chosen faction and the state of the world. I am hoping to come up with a sort of meta-game for this that gives rise to a plot, much like a game of Civilization does.

Obviously, I would love to compare notes or collaborate here, if anyone is interested.

J.

 

[S'mon]

Another thing you can do when your players are wandering where you haven't planned for...

Run them through the planned adventure, but as something they're imagining...

This seems pretty heavy-handed to me, unless it grows naturally from previous campaign events (eg the PCs know the Dreamweavers are hunting them). An alternative is to have a few mini-adventures prepped that can be run instantly, wherever the PCs go. An old ruined tower or keep with some monsters, enough for a single session, is ideal. The Book of Lairs I & II are a good source for this, as mentioned, but there are plenty of published scenarios that work - I like Goodman Games' The Mysterious Tower, the C&C version needs minimal conversion for 2e.

 

[S'mon]

Jorunkun, your approach sounds good. I'd advise you to leave plenty of room for later expansion, and personally I wouldn't want to map the whole campaign area at 1 mile per hex/square; even 2 miles/hex reduces by 4 the amount of keying you need to do, and I've found 5 miles per hex is good for an overview map that's detailed enough to show every village (in a sparsely settled frontier type setting) while large enough to include tons of adventures. If you have a giant piece of hex paper though, go for it.

I agree plots in sandbox should be emergent, based on the plans of the competing factions, not pre-scripted. You may want to pre-script a few possible things on a rough timeline, but be ready to change it as the game develops. If you have an interesting event in city X planned, it may be a waste to have it occur off-stage when it can wait until the PCs are actually in X.

Edit: The classic example of Wilderness-as-dungeon is the Wilderlands of High Fantasy. The WoHF box set from Necromancer is awesome and well worth getting just for inspiration. It also seems like it should work perfectly with 4e and its 'Points of Light' ethos.

 

[Dannyalcatraz]

This seems pretty heavy-handed to me,

It is, it is! It really should only be used when you're in extremis.

I've only used that method a few times, myself. The last time I did, I was pressed for time before the session, and the game went "off script" within 5 minutes! In the end, the "dream adventure" provided mental clues to the real adventure- so it was kind of a shared premonition, "sent by the gods."

 

[Jorunkun]

I'd advise you to leave plenty of room for later expansion, ..

In terms of scope, this campaign is supposed to cover the epic (lv 1-10) range, and a very dense area - I'm thinking more in terms of a mega-module than a nation or world sized setting. I'm expecting that the characters will stay in the region I outlined until they are kings of that particular hill, and either go on adventuring in neighbouring planes or retire after that. Like many posters upthread, I find it hard to create a believable setting that spans a broader range of levels, and don't care much for high-level play anyhow.

... and personally I wouldn't want to map the whole campaign area at 1 mile per hex/square (...) If you have a giant piece of hex paper though, go for it.

Well, I must admit I actually enjoy making outdoor maps; I find it's very inspiring. I use the Civ4 map-editor, snapshot the end result and piece it together inside Maptool. This is because my group plays online exclusively, where maps are key. For a standard ftf game, I'd agree this may be overkill.

 

[Raven Crowking]

I haven't had a chance to read through this thread yet, but I will do so later. Great idea for a thread!

Some things you might want to consider:

* Descriptions. A sandbox lives or dies by how "real" it seems to the players, so have some descriptions available that make different bits of forest or stream seem to actually be different from each other. AEG's Toolbox is great for this.

* Real Life Relics. From the stone foundations of a burnt-out farmhouse, to a lost wooden comb, to a broken child's doll, have some relics of past lives around that aren't adventure sites and aren't real "treasure". Not only do these add to the verisimilitude of the setting, but they make some real treasures (and threats, and adventure sites) less obvious.

* Normal Animals. Especially in a sandbox setting, players should run into normal animals and plants. Encounters with deer in the forest, sighting lizards and toads, etc....even mention of things like mice and dragonflies....Not only do they help the setting seem more real, but they provide food for PCs, things for druids to command and question, cover for spying familiars, and potentially clues as to the nature of an area (as in, "There must be open water around here somewhere, with all these dragonflies.")

* Religion. Sites for pilgirmages, including shrines, remote temples, etc. Roadside cairns. Burial mounds. Not every holy site needs attending clerics; a stone near a bridge where folks leave odd coins is just as good. More than just having a list of gods, you should consider at least a basic philosophy for each major religion in the campaign area. This is the sort of thing lots of players have fun with, so you should consider recruiting them to do some of the work!

* Calendar. You should have a blank campaign calendar to mark out when things happen and where. Plan out major events ahead of time (the 1e Oriental Adventures has a usable system for this). Think about major holidays, what they look/feel like, what happens during them, and how the PCs can get involved. Include the phases of the moon(s) for lycanthropes and strange magics. Consider the impact of seasons. Use a weather generator (there are several good ones online) to generate weather a year in advance. That way, when the players decide to hide out in Wildgate for six months, you won't be caught unprepared!

* Travellers On The Road. Who travels on the road, and why? Encounters should include more than merchants, patrols, and monsters. There was a great Dragon article on this a long time back (sorry I don't know the issue # off the top of my head). Meeting other folks on the road, in ways that make sense, is a lot of fun, though!

* Interconnect Adventure Sites. Especially if using pre-written adventures, take some time to ponder how various adventure sites might be related to (and point to) each other.

* Cull Names. Ever hear that common complaint that some adventure has a great background but the players will never get to know it? With a sandbox game, the DM can go over that background information well beforehand, and seed it throughout the setting. By the time the PCs encounter the Lost Tomb of Ballyho, they not only know who Baron Ballyho was, but who he was betrayed by, and what happened after. Suddenly, what seemed like meaningless detail gains a poinancy that adds to the experience of exploring the Tomb.

I'll try to add more when I have more time/have read through the thread.

Best of luck with it, and keep us posted!


RC

 

[S'mon]

In terms of scope, this campaign is supposed to cover the epic (lv 1-10) range, and a very dense area - I'm thinking more in terms of a mega-module than a nation or world sized setting. I'm expecting that the characters will stay in the region I outlined until they are kings of that particular hill, and either go on adventuring in neighbouring planes or retire after that. Like many posters upthread, I find it hard to create a believable setting that spans a broader range of levels, and don't care much for high-level play anyhow.

Well, I must admit I actually enjoy making outdoor maps; I find it's very inspiring. I use the Civ4 map-editor, snapshot the end result and piece it together inside Maptool. This is because my group plays online exclusively, where maps are key. For a standard ftf game, I'd agree this may be overkill.

I'm pretty similar - I think a 10 level range is about all a sandbox can usually handle, but I normally make them at 5 or 10 miles/hex to fit on a single A4 page - 5-mile is a particularly good scale, I use it for my Duskmoon Hills PBEM, which is an NC-17 rated take on the old Basic D&D modules (B1, B2, B3, B7 etc).

 

[evilgamer13]

One way to resolve the level scaling for the sandbox is the Pandora's box like ancient evil that the PC's inevitably let out. So your general adventure range is between 1-10 until the PC's go after some super KEWL treasure or are sent to collect some mcGuffen or other (by someone who is actually was duped by an evil cultist), of course in the proses the PC's break the ancient seals (or the first of the seals). Then the big bad gets stronger as it wakes up or reclaims sundered portions of it's old power.

 

[DestroyYouAlot]

I'm going to ask you for some DMing advice.

What do you do when, in the middle of the session, the PCs start heading for Whatzisnehm's lair but you don't have any encounters statted out?

Several years ago I was running a sandboxy FR campaign, and the PCs would occasionally do something unexpected like that, and I would find myself having to call a 15-minute smoke break while I rushed to stat out, say, a gang of bandits, or prepare an encounter, or whatever.

How do you deal with this?

Do that. You got it right the first time.

In a pinch, you can always try and distract them - dangle something shiny and pre-planned that you've been keeping in your back pocket in front of them, hope they'll bite. Sometimes they'll be stubborn (*snort* - "sometimes"), and you'll still have to either a) run whatever it is on the fly (TAKE NOTES WHEN YOU DO THIS), and if it's not as super-awesome-and-cool as you envisioned it, well, the players don't know that, or b) figure out why they can't go there. Sometimes the door's just locked, whatever. *shrug*







My advice on sandbox campaigns: True, there is a LOT of work involved at first - to steal a 3e-speak term, it's very "front-loaded." Unless you're Super GM, there's no way you can detail everything to any appreciable level, so do a little work on a LOT of stuff. If you have access to any old Judges Guild stuff, take heed - it's a bazillion quick blurbs on NPCs, businesses, locations, groups, whatever. When it comes up, you'll be able to take the minimal prep work you've done on [X] and spin it into something more substantial. Moreover, if you have a shallow understanding of a large area (town, dungeon, wilderness area, whatever), you'll be able to "see" how the whole thing works as a whole, and then "zoom in" mentally on any one point when you need to.

I mentioned it already above, but TAKE NOTES. LOTS OF THEM. ALL THE TIME. Without doing an inordinate amount of work before ever actually getting to play the damn thing, you'll be able to get maybe one town pretty fleshed out as far as major shopkeeps, authority figures, guildmasters, etc., but when you have to go down the road for a minute, you'll end up spinning some out your ass. WRITE IT DOWN. Preferably as it spills off your tongue. The next time the PCs run across whatever it is, you'll at least be able to sort back through your session notes to find out that the third farmer on the left is named Sven and has one arm.

Quests. Jobs. Hooks. Whatever; the idea is, have a lot of them hanging around. Don't bother developing much for them; just spin up a sentence or two to dangle in front of the party, and a sentence or two more that they'll find out if they look. The rest, you work out as needed. The thing you want to avoid at all costs is a bored party - bored parties go around looking for things you haven't touched yet, and poke around behind the scenery. They have an amazing knack for that. You'd rather they go for something you've at least considered for five minutes on your lunch break. And, while - from one point of view - I'm all for "it's there whether you look at it or not," the actual details of what "it" is can be freely changed depending on the needs of your campaign. You're getting off the railroad - that's GREAT - so this is where you get to inject your story bits. If the party knows that there's bugaboos to hunt in the Bugaboo forest, they'll probably go there at some point. And they should find bugaboos no matter when it is (or what the relative challenge is - if bugaboos are too tough for them, well, hey, you warned them). However, if there's an NPC you want to meet, guess what? He's now an avid bugaboo hunter. If there's a nifty little dungeon crawl you dreamed up the other day? The entrance is now in bugaboo forest. The idea is that the framework for the world is what it is - the details, you simply can't fill out ahead of time. So you fit things where they fit. (I know this seems counterintuitive when we're going for sandbox play, but this is how it can actually work, and you still get to hold down a job and see your family on off days. )

It's also been mentioned that you can either make certain wilderness areas "safe" for low-level parties, or take the 1e approach and have everything be totally burly and crazy outside the town walls. I'm in favor of a mixture of these techniques, but there is something to be said for the latter technique - at first, the party can only make short expeditions into the wild, so you have them on a short leash, and you can detail just ahead of them. As their "operational range" increases, so does the area you've got detailed and - eventually - memorized. YMMV - like I said, I tend to mix the two approaches - but don't discount the latter approach entirely (as I understand it, Wizards has taken quite a shine to it, lately ).

The beauty of sandbox play - whether it's a whole region, or a hardcore old-school megadungeon - is that it's a changing environment that takes a LOT of initial investment from you, but that you can use for YEARS, and it takes less and less effort from you to maintain the environment (as you come to know it intimately and be able to run it by reflex), so you have more time to concentrate on the big picture - your focus can gradually pull back as the PCs grow in power, until they're doing the stronghold thing, you're moving political groups around, and they're thinking about maybe taking Loki down a peg or two. Good luck.




Dannyalcatraz mentioned regional encounter tables - +1.

Products (most of which have already gotten plugs, but these are my picks): As mentioned, AEG's toolbox. The 1e DMG - how anyone can run a FRPG without this at the table is beyond me. Book of Lairs, I + II. Anything Judges Guild (Wilderlands stuff especially, but it's all minable for material.) Dungeon mags (old ones especially, they're more site-based). WSG, DSG, 2e Campaign Sourcebook, 2e Worldbuilder's Guide.

 

[Jorunkun]

I'm with Destroyyoualot in that half the fun in sandbox play is that it empowers, nay, requires the DM to improvise within a preset framwork. I'm also all for letting the dice fall where they may wrt encounters and combat, provided you gave the players enough opportunities for getting information to know what they were getting into. I mean, for once you don't have to worry about your precious plot progressions being derailed because of PC death.

<ramble>What I am wondering though is how much plot a sandbox should have. Or rather: How much plot is too much?

I am assuming that one of the defining criteria of this kind of campaign is that there is no pre-scripted plot beyond setting up a balance of powers that the player can relate to, and see what happens. The story results from what the players do, avalanche style.

How far do you take this framework? Is it acceptable to fudge encounters when things get boring? Do you set up encounters that will happen to the players regardless of where they go? Should there be accidents waiting to happen, triggered by player choice or by them reaching a certain level etc? Is there an end state to the campaign, like "kill foozle", or are you rolling dice to see what faction will win?
</ramble>

What kind of plot generating preparations are right for sandbox play?