D&D 5E Sandbox Play

[steeldragons]

Lots a good suggestions here. I would concur that it is more heavily reliant on the GM to do (set up and maintain) well. But, at the same time, it is "light" set up since you are leaving the directions and outcomes to the players.

I would start with a "hex" worth of a local map. The immediate/starting location and the immediate (within 1-3 days' travel) worth of the surrounding environs. The space that a) if they are from there, the PCs will be well acquainted with and/or b) if they aren't from there can easily find out about (with reasonable accuracy and detail) from locals.

Someone else mentioned "3" plot hooks, and that has always been my practice as well. Start out with 3 possible locations/adventure plot hooks. Have a general idea of what the potential adventures, associated NPCs and monsters/problems, are for each of those 3 places.

Set up a list of of rumors if the PCs are going to be interacting/asking questions. [*SANDBOX TIP: Even if they don't, keep this list for other times/places. In short order, after say 3-5 different locales, you'll have notes of ready made rumors you can use in any location.] I would think 10 is sufficient (maybe even less if it's a small or insular locale). Half of those 10 are related somehow to what you've written. Half of those (1/4 of the total 10) should be inaccurate/false. The other quarter is worthwhile/helpful info. The other 1/2 of 10 are totally made up/false. Though, sometimes, turning something you'd intended to be untrue actually makes for a great adventure/story! If that's what the PCs want to follow, go for it!

I normally, to move the game along, will go with 2 "false starts". That is, if the PCs go off on 2 initially false leads, on the third one I just make that lead to somewhere I have prepared.

If the PCs take things in a different direction it's effortless to simply "move" the things you have notes on/prepared to accommodate their movements/locations and tie it in to whatever they are doing.*

[*SPECIAL NOTE/TIPS: and as a DM in a sandbox environment -every/any D&D except a flagrant railroad, I would say- you need to be prepared for this to happen at least 50-80% of the time. Accept that before you agree to sandbox play and you will avoid no end of disappointments.

SPECIAL NOTE II: moving where one of your prepared adventure hooks is is not to say "railroad the PCs"! They have no set outcome. They have no place they "have to go/be next". But use what you have worked out for your benefit, ease and speed of play/getting to the adventure part. I would not spend more than 30 mintues to an hour having the party chasing red herrings. In the beginning, it might get a few players trying to think/figure things out after a false lead or two. More than that and it just becomes aggrevation and boredom. Bad. Pacing and immersion are very important...again for any D&D I would say, but particularly in a sandbox.]

For example: You have set up the 3 hooks above. Don't force them to do anything. But there's no harm putting things in the party's way. For that reason in particular, I usually make -as part of my 3 "prepped" ideas- mobile: a wandering owlbear is in the area slaying livestock; a witch was seen flying over the Gobbler farm last night; a PC finds a platinum coin on the ground in the market place from the now fallen/former empire; ankhegs/bulette attack on day 3; etc...

I mean, obviously, if the PCs don't/won't leave town, you can't just plop a cairn in the town square [though sinkholes are a lovely way to introduce a dungeon/tomb/old forgotten city into an urban environment.] But the zombies from the cairn might start wandering around the homesteads outside of town at night. The necromancer seeking out the magical secret of the cairn is -turns out- the town mayor. Now you're having to deal with him and his constable (who doesn't know his true identity, just following orders) and their harrassment to get these "heavily armed trouble makers" out of town. The PCs are going to deal with them...or leave.

Point is, if Mondrago the Mage Mercenary won't go to the mountain, bring the mountain to Mondrago. In a sandbox, you can use what you have prepared, even [or especially] if it's not how you originally envisioned it being used.

Another tip, that kinda goes with the "players aren't going to always do what you think", do not plan out too far until the campaign is rolling. Keep notes as you go and don't be afraid to use them. Leads that were missed/not followed up on. Rumors that, maybe you didn't intend to use, but now have an idea of how to work in. Adventures that got started but then the party left alone/got sidetracked. Lends depth and immersion to the setting if things the PCs do actually are seen/remembered and have effect on the setting.

E.g.: The thief PC stole that sacred gem from the temple in Poortown and no one noticed. Now it's four months later and they're going back to Poortown -or even just the province that Poortown is in. "Hey, look! This poster kinda looks like us. That's Dashington and Madrigal. This bearded one's has a helmet like mine and Booster...Booster! What did you do!?"

As time goes on, and you expand out the setting in this direction and that, as you and the players become more familiar with their world and impact things in their world, making things up on the fly becomes easier. But notes are often necessary. Lots of hidden gems for fun and adventure and detail slip through the mental cracks easily.

SO, yes to pretty much everything you put in the OP. hahaha. Have a basic map. Have 3 possible plot hooks. Have 10 possible rumors. Take notes as you go. Be prepared to make things up and change things on the fly.

Sit back, let the players tell you what they're doing, and let the dice fall where they may.

Love, encourage, heavily recommend, and endorse sandbox play. My preferred mode of D&D.

 

[steeldragons]

ALSO! Important to keep in mind/note, the world keeps on spinnin'. That is, you can have recurring villains, you can have plots within plots, you can do all of those "big plans" you work out. If the PCs aren't following up on things/taking care of villains "once and for all", have their plans continue...and bring them back around to intersect with the PCs when appropriate (that is, when the plans come to fruition and effect the PCs or when the PCs get back to seeking them out).

If the party gets distracted, avoids or ignores that mayor-secret-necromancer when they're 1st level, and now have become 5th and are heading into some rumored haunted cairn, that then-4th level-necromancer is now 8th or 10th or has had the time to build up a fortress of [just say for example] 9 wraiths and dozens of zombies and ghouls.

He found the Skull mask of Undead Conqueror he was trying to excavate from that cairn back in Tavernville and is now out to claim the Duke's throne. Will they ever know/notice/recognize that, "Hey! That's that dick mayor from that Tavern-place where we met!" Maybe not. Will the players/PCs care? Maybe [even probably] not. Don't have to. That's what's been going on in the world while they were hunting the Treasure Tiger of Hidden Dragon.

So, "keep the world spinning" is another trick of the trade, again I would say applicable to any/all D&D, but particularly useful/important to sandbox play.

 

[Rabbitbait]

I'm a subscriber to the rule of three as well. Three clear choices always works well for me. I like starting out with a well detailed world - I'm currently setting up a campaign in Eberron based 30 years after the events caused by my last campaign. I picked a town I want the campaign to begin in and have asked the players to create well rounded characters with decent backstories. I'm happy for them to write new world history or add NPCs etc if it helps them to do so. I make sure that each of them has a number of NPCs that are important to them and that may provide opportunities to drive plot.

Then from there I will design an overarching plot for each of the characters based on either their background or their goals, as well as some world spanning plots. The first game will be the hook to get the characters together and from there there will be three clear themes they can pursue. Or they might choose to go in another direction altogether.

Once they have decided what they want to do I can design more detail in that direction. I think a good sandbox shouldn't have players scratching their heads around what to do as they have no real reason to get involved, it shouldhave them scratching their heads over which mission is the most important that they must do and how they should prioritise their competing demands. After each big mission concludes try to have another decision point where there are three main options.

I think the big secret with this is to be willing to scrap what you have planned if the players decide to go in another direction altogether. This is what happened in my last game which lead to the players overthrowing the church of of the Silver Flame in Thrane. Totally unplanned by me, but happened while other parts of the campaign I had designed were going on. Apologies for errors in spelling etc - I wrote this on a phone.

 

[aramis erak]

For me, when I do sandbox...

1) a map. Tell players the major players on the map
2) a bunch of rumors their characters have heard. Sometimes individualized to backgrounds and provided individually, not in session.
3) a bunch of things that following the rumors lead to.

So, if I put a ruined castle in a rumor, or on the map, i have a dungeon prepared for it.

I find a fully prepped sandbox to be a waste of my time, however, as my players tend not to be sandbox-play oriented. They are, however, the types who like to have choices to pick from.

 

[LostSoul]

I use some procedures for sandbox play.

1. I generate a hex map. Empty to start off with. I make sure that there's a nice variety of terrain types, and I usually put mountains in the middle (to separate the map).

I use 5 miles per hex and a map that's only 10 hexes by 10 hexes. That keeps it managable.

2. I stock the hexes. I draw cards to see what's in each hex: settlement, dungeon, terrain feature, gate, NPC/monster lair, or nothing (cards numbering 2 through 9). I go through a deck at a time, so I know that I'll have at least four of each (and two gates - the jokers).

3. For each hex I roll on some random tables I designed to add detail. Since I haven't finished all the tables this means writing up an entry for a hex. The hexes should all have something interesting about them, and ways in which interaction with hex features can change characters. They should also have connections to other hexes, so that the PC's actions in one hex reverberate throughout the setting.

Here's an example of one that the PCs recently explored in a game we're currently playing:

[sblock]RUINS FROM THE AGE OF GIANTS
A - Weird or interesting ruin, emphasizing the slumbering giant and its effect on the terrain
B - The giant in a torpor and its valuables
C - A table that personalizes the hex
D - Something that can change the character if the ruin is explored (generally extracting the valuables from the slumbering giant, risking rousing it)
E - A link to another hex

ROLL RESULT
A. A 20' mound or rocky outcropping, covered in old vegetation, that has the impression of a man sitting with his head held in his hands. A vein of salt crystal begins where the eyes of the man would be and trickle down the hill like dried tears. The entire hex gives off the smell of sulfur.
A dozen or so 16' tall mounds are scattered within a mile of the strange hill.
B. A hill giant rests here in a deep torpor since the Age of Giants. It has risen through the eons, building the mounds before settling back into its slumber. Slowly weeping over the ages, its tears have turned into valuable deposits of salt (about 4 tons; 1# of salt = 5 GP).
C. (I) Through the ages the hill giant has been dreaming with Tiamat and has come to regret its actions that led in part to her death. It still has a need to make hills but, like all giants, has no direction for this urge.
The hill giant will wake if any mining or digging is done on its hill. It will struggle to avoid the immediate and nearly-overpowering desire to build hills, giving those who have awoken it the chance to speak with it (it only understands words of power). It can be given a new purpose, though any task which it feels will harm the body of Tiamat will earn its immediate wrath (any major change to the landscape, basically - dams, changing hexes, that sort of thing; digging moats and raising berms is okay). Run this as a social conflict. If dialogue is not begun within 1d6x5 seconds its urges will overpower it and it will attempt to use all those who have awoken it as mortar for a new hill.
D. Characters who taste the salt gain the hill giant's ability to sense geological features for a week (allowing them to construct earthworks at twice their normal rate and detect precious metals and minerals in hills), but their body odour changes to that of sulfur permanently (-2 Reaction Rolls when it counts).
E. (I) An NPC in a nearby settlement or lair has studied this hill and has plans to try to control the hill giant for his or her own ends; however, he or she cannot speak the language of power but is trying to learn it.

(The (I) is part of a random table for this entry, so that each time this entry comes up it's a little different.)[/sblock]

3a. I also roll to see if there's anything else of interest in the hex - extra treasure, more monsters, or just something strange. I don't have any tables for these, I just add to the hex based on how the map is shaping up.

4. I pick a starting settlement for the PCs and present a package of rumours and a general blurb about the world. (We're playing in the Hyborian Age right now so that latter bit wasn't necessary.)

5. The PCs create characters.

6. The PCs pick a Quest from the list of rumours. Quests are the answer to "what do you want to do today?" and are the main source of XP. There are only a few types of Quests - exploring hexes, defeating NPCs, harassing a lair, increasing influence or reputation in a settlement, clearing hexes, and exploring dungeons. (Dungeon Quests are a little more involved but follow the same basic pattern.)

They also pick a Goal - a major achievement. A "Name Level" sort of thing - building castles, defeating powerful enemies, starting a settlement and seeing it grow, and some other PC-dependent stuff. This also grants XP, if the PCs do two of the following while taking a necessary step towards the completion of the Goal: put themselves in danger, spend resources, learn secrets, or explore new features.

7. The PCs adventure for a while, completing Quests.
7a. The terrain type determines the level of risk so that players can judge what they're getting themselves into. The level of risk determines the amount of reward (XP, GP, magic items).
7b. As they move overland or through dungeons, I roll for wandering monsters/random encounters. (And of course they encounter things that are there.)
7c. Encounters begin with Reaction Rolls.
7d. Combat drains resources and grants a few XP. Wandering monsters carry no extra treasure. NPCs placed in hexes have treasure, and some can be found lying around.

8. Eventually they'll run low on resources. They come back to town to replenish them. This takes time and requires them to interact with NPCs in different ways (based on class). They also spend money gained from adventuring in town on different things (hirelings, gear).

Taking time allows the setting to react to the changes the PC's have made. Interacting with the NPCs allows them to tell the PCs how things have changed.

9. After a few cycles of that, the PCs have enough XP to level up. This takes time, GP, and requires the PCs to interact with NPCs in different ways based on class.

Taking time allows the setting to react to the changes the PC's have made. Interacting with the NPCs allows them to tell the PCs how things have changed.

10. After a few cycles of that a month will have passed. I check to see how much GP the PCs have spent in town; if it's over a certain threshold, the town "levels up" - gets bigger. This grants more access to gear and hirelings. I make a roll to see what the PC's reputation and influence is, based on their actions over the past month. I make a roll to see if the NPC lairs have "leveled up" - gained new recruits, which allows them to expand their influence over the map.

NPC lairs can grow very powerful and become great threats over time. They'll also strip treasure from dungeons and other areas. This creates some time pressure for the PCs to get things done.

11. The PCs continue Questing until they've completed their Goals or have outgrown the area.

 

[the Jester]

Longtime sandbox DM here. I consider myself to be pretty far to the 'sandbox' side of the sandbox ---- storygame spectrum.

The main thing about running a good sandbox is you need the area where the pcs start to be dripping with adventures, where they can find an interesting site, scheme or npc around every corner. Maybe the local tavern has the stereotypical bulletin board with adventure options posted, maybe the city the pcs are in is rife with gangs and cultists, maybe the local temples have undead-plagued graveyards that need to be pacified. Whatever- the adventure opportunities should be both frequent and obvious, so that the pcs don't end up sitting around clueless and bored.

A sandbox doesn't mean that you don't drop hooks all over the place; part of a sandbox, IMHO, is dropping too many hooks for the pcs to ever follow up on. Not all at once- gradually, over time. But if you average two or three new hooks every few sessions, the pcs will know that there are things going on around them, that the world exists outside of their pcs.

Another thing is to offer a variety of risks and rewards from the beginning. Maybe everyone knows about the gigantic blue dragon up in the hills, and the pcs can go there (and be eaten) at the very beginning if they want to. Its hoard is legendary in scope, full of money and magic items; the rewards are commensurate with the risks. On the other hand, the pcs could be 12th level and decide to go fight goblins and reasonably expect that the goblins will be mostly easy to kill. But, by the same token, they'll have a few hundred sp and gp between them.

Sandboxes don't lack a story; they lack a pre-existing story. In a sandbox, the story emerges from play. It's what you talk about after the game is over, not what you plan ahead of time. At a certain point, a sandbox may start to strongly resemble a story-driven game, since the pcs have engaged a long-term enemy and are attempting to foil its scheme(s). But in a sandbox, the enemy isn't necessarily chosen in advance- the pcs might have ended up opposing the Ghoulish Council, the nobles of House Ilmixie, the fey Court of Lightning, the corrupt local mayor and his administration, the cult of Graz'zt, the agents of the aforementioned blue dragon, etc. In a story-based campaign, you know from the beginning (even if the players don't) that the final enemies are the vizier and court mage of the Court of Lightning.

In my own urban 5e sandbox campaign, the pcs have so far interacted with street gangs, gone into the local megadungeon, hunted bandits, sought out a killer who was ripping the hearts out of young lovers and other isolated people, tried to track down a master monk, tried to kill some evil things underwater, rented a fancy house, put on a rock and roll show using magic for special effects, rescued some enslaved people while slaying one of the local slaver bosses, worked security at a farmers' protest and fought amongst themselves several times. There is no overarching plot, but there are lots of threads and lots of pc choice.

 

[Sailor Moon]

Whenever I run a sandbox, it's either in an existing place such as FR, or my homebrew.

I run a mini synopsis on various places and I let the players mingle around in towns picking up rumours such as the dwarves in the north being assaulted by the frost giants for example.

Sometimes I may have something happen the town which can put them on a path for something someplace else.

 

[Staffan]

If the party gets distracted, avoids or ignores that mayor-secret-necromancer when they're 1st level, and now have become 5th and are heading into some rumored haunted cairn, that then-4th level-necromancer is now 8th or 10th or has had the time to build up a fortress of [just say for example] 9 wraiths and dozens of zombies and ghouls.

I would mostly use this in the cases where the PCs avoid something. However, if they defeat a foe and let that foe live, I would prefer to have that foe come back as an ally or at least a resource later - or if they come back still an enemy, one that hasn't grown overly much in power since.

Recurring NPCs are fun. I once had an Eberron game where one adventure had the PCs stop some plot of an Inspired telepath (can't recall exactly what that plot was, but it did require them to go to a ball to stop him). A few adventures, when the PCs had come across some nasty world-threatening thing, the Inspired guy showed up again - this time with an offer to help. I mean, if the world got taken over by gibbering things from the Far Realm, that's not exactly part of his plan either.

And the best thing was that I got to use a line from Babylon 5 when the PCs asked where he had gotten the information he gave them: "I'm a telepath. You do the math."

 

[Prism]

Lots of great ideas in this thread - its my favourite type of play. Just one thing to add that has worked really well in some of our campaigns. Just because there is no major plotline to follow there can certainly be a strong campaign theme. Two of our best campaigns have been based on the themes of exploration and another on survival. You can then help the players build a suitable character/backstory for such a sandbox.

The exploration theme worked well because we and our whole village migrated to a new, exciting and dangerous land. Our characters all had different story reasons to embrace such a theme. One wanted to make a name for himself, another find ancient relics from the previous fallen civilization, another dark magics. The group was given the first few missions to get them used to the new land but after that's it was up to the players to advance their stories. What is important here that the characters bond together over the first few levels so they are willing to help each other with their own quests down the line. We took this one to 14th level (3e)

The survival them created an instant bonding of the characters although most of them were evil alignment. They were stuck in a chaotic recently fallen city (Ched Nasad if you want to look it up) and we simply had to survive until we could find a way out and it was totally up to us how to do that. The DM threw all sorts of stuff at us but in general unless we needed to fight we mostly looked to hide or flee. As typical though of an evil group, once the party escaped, the bond weakened and the characters went their own way. It actually made a lot of sense as when a sandbox campaign is over the players tend to decide rather than a predetermined story. We made it to 8th level (3.5e)

Someone earlier in the thread mentioned Murder in Balders Gate. I found this to be a really good sandbox-ish style adventure. Not a totally open city adventure as you were often given 'jobs' to do but you could pick who you worked with and which jobs you took on. The more jobs you took, the darker they got so really it was about finding the line in the sand that the party was unwilling to go past. A test of character if you wish. Good stuff. We played this one in 4e

 

[Rhenny]

Oooh...I just thought of something else. Sandbox style works really well in gritty worlds where nearly everyone is more or less neutral (good or evil) and looking out for themselves. This shifts the focus away from good/vs. evil (which tends to create clear cut choices and quest options) to more survival type or personal gain type play (which opens the door for more ambiguity). I ran a long term 3.5e campaign using Sanctuary (Thieves World campaign) as a hub. It was the kind of campaign where everyone had questionable motives and nobody could be trusted.