[3.5] Sandbox Campaign Advice

[Thurbane]

So, just looking for some general advice and feedback.

Our group rarely plays sandbox-type games. We nearly always play pre-written adventures. Or when it's not a published adventure, it is usually a home-brew campaign with a fairly linear story.

And that's fine, that's what we are used to, and we enjoy it.

But our current DM may need a break soon, and I was thinking about tackling a sandbox type campaign.

Will be starting at ECL 3 (we nearly always start form 1st, but I want to skip those early levels). I'm throwing pretty much any official (non-Dragon) source open for players (most games at our table tend towards core-only).

We have three players: one is leaning towards Rogue, the other leaning towards Barbarian. The third (my wife) hasn't let me know yet, but she nearly always plays Druids, Rangers, or someone else with some kind of animal companion or familiar.

I like to have a 4 person party, so I will roll up an NPC to fill whatever gap is left in party roles. I always play them as a passive there for support, not a DMPC.

In terms of setting, I usually default to Greyhawk, but this time I'm going to try an original home-brew setting (which will mean a ton of work). I'm going to slot in as much setting-neutral stuff from 3.5/3.0 books as I can, like the cities from Cityscape, non-setting specific deities etc.

So, I guess what I'm looking for is suggestions or feedback on:

- Pitfalls of running a sandbox, to watch out for.
- How to be prepared as a DM for the party going off in an unexpected direction.
- Guidelines on homebrewing a campaign setting.
- Any other general advice.

Cheers - T

[Saintheart]

On homebrewing a campaign setting:
- Get the free version of Worlds Without Number, whose campaign generation guide is exactly what you're looking for. It covers homebrewing whole settings in a practical manner and works quite well for that. It is system-neutral and gives you a very good place to start. That said, when homebrewing a whole campaign setting, don't get too specific or do too much work. It gives you less room to move if you find something inspired out of what the characters do or say.

- Best creative advice I ever had for building a campaign setting: eliminate at least one common element that would exist in a generic D&D setting. In my own homebrew, I removed elves entirely. No elves, no half-elves, no PrCs or classes that require elves, nothing. You then immediately have a point of distinction in your setting and a plot hook for the adventurers: where did [generic race] go? (I, for one, made the elves disappear having been driven out after being evil overlords and slavemasters to every other race, left elven ruins all over the place as adventure sites, and made the study or interest in elven lore or knowledge a taboo subject in-universe. This immediately guaranteed at least one player would beg to take elven as one of his languages :D )

- Similar creative advice: eliminate at least one common character class. Similar reasons as taking out elves: it immediately gives the setting some texture and invites some creativity about why that is. Remove druids, for example, or make them strictly NPCs ... maybe even evil NPCs, similar to how the Midnight setting denies PCs access to the cleric class, makes it an evil NPC-only class, and allows clerics full sourcebook access while PCs only get Core for their spells.

Pitfalls of running a sandbox:
- Biggest issue is what I call the Buridan's Ass problem. The party doesn't know what to do because it has too much to choose from and therefore stands still and doesn't choose anything. This is particularly so when the party is used to being put on plot railroads, er, published adventures :) Therefore, give them the illusion of choice: start them as being on a minor quest of some kind, and then, as they're going on with that quest, throw them something else they can look into - e.g. while they're on the escort mission, have them notice a ruined castle on the next hill which they can choose to look into or choose not to.

- When going into towns, make sure you tell the players what their characters would normally do when they get into a town - e.g. clerics first arriving would normally go to a temple to pay their respects ... and then have something happen while they're on their way there. You've got to teach people how to play in the sandbox, moreso if it's an unfamiliar setting, and that means holding their hands at least for a while.

- "Sandbox" is a misnomer. The players can think of it like that, I strongly suggest at least to begin with you - as DM - think of it only as a standard dungeon delving mechanism with equipment purchase minigame outside, connected together with the wilderness survival minigame.

- By the way, if your group is ECL 3 and they're in a sandbox, and you want to make wilderness travel or ruin-delving meaningful -- then forbid all spells and items that allow infinite food and water supplies. No Everfull Mugs, no Everlasting Rations, no Rings of Sustenance, and if they really want to have Create Food and Water, it only covers them for one meal. This brings back logistics as part of the game, which is a handy thing in sandboxes.

[pabelfly]

My go-to for homebrewing a campaign setting is using Microscope. It lets you generate a historical timeline with equal input from all players, and you can set your campaign in the middle, or at the end of it. It's great in getting players invested in the setting created and finding out what players want in an organic way.

[Crake]

- Biggest issue is what I call the Buridan's Ass problem. The party doesn't know what to do because it has too much to choose from and therefore stands still and doesn't choose anything. This is particularly so when the party is used to being put on plot railroads, er, published adventures :) Therefore, give them the illusion of choice: start them as being on a minor quest of some kind, and then, as they're going on with that quest, throw them something else they can look into - e.g. while they're on the escort mission, have them notice a ruined castle on the next hill which they can choose to look into or choose not to.

To build upon this: sandbox doesnt NEED to mean without direction. You can have a sandbox with a story in it, but the difference is how players approach the resolution to the story, which is to say, they drive the story, they build and devise the solutions, and they deal with the consequences. You can still have set pieces, and villains, but you need to be much more fluid with running them, they need to be able to adapt to what the players do, and be agents in the sandbox themselves.

I think thats one of the greatest pitfalls of a sandbox, thinking you can just throw the players in and magic will happen on it’s own. Sandbox vs rails, I think, is more about how you get to your destination, not about what or where or if the destination itself even exists.

[Palanan]

Originally Posted by Saintheart
- Best creative advice I ever had for building a campaign setting: eliminate at least one common element that would exist in a generic D&D setting…..

- Similar creative advice: eliminate at least one common character class.

Not sure if I’d agree with this, at least not as a sweeping principle. This approach may work if you have a specific world in mind that needs this change to function; but removing X race and X class simply for the sake of Being DifferentTM can end up seeming contrived.

With regards to the specific homebrew setting that Saintheart mentions, I’ve seen it up close and I love it. The absence of the elves is well-thought-out, it presents a series of mysteries to the players and it works organically in the timeline as a whole, to the point that it’s a defining axis of the setting.

But I would say this is primarily a reflection of the depth and quality of thought that Saintheart put into his setting, rather than a consequence of simply removing one of the core races—and removing another race from another setting won’t necessarily result in the same quality of world-building, nor the same degree of distinction from other game worlds. In my primary setting gnomes are virtually nonexistent, but after Session Zero it never came up again. (And nobody's missed them since.) That's one race removed, but I wouldn't consider it a defining feature of my setting and it's changed absolutely nothing else about the setting.

I’d be even more wary about removing one of the classes. From a player’s perspective, it could feel too much like an arbitrary restriction that limits options—and in terms of world-building, eliminating a single class may not change the setting enough to make the effects more than cosmetic. Wizards and clerics would be exceptions, since they’re T1 full casters and baked into expectations of how the game functions; but a setting without paladins or monks will probably work much like a setting with them—apart from limiting those options for the players.

[AvatarVecna]

Reply to a previous thread on similar topic:


My own attempts to run sandboxes, even ones as sandboxy as "an official adventure with a big sandbox in chapter 3 for players to explore until you feel like moving onto chapter 4 of the railroad", has taught me a very important DMing lesson:

If the players knew what kind of story they wanted to play through, they'd be the one behind the screen instead of you.

Players don't want the freedom to go wherever they want, because if given the opportunity, they bicker amongst themselves and ultimately decide to wait until they gain more information to base a decision on. And either the plot advances without them and they fall behind (which is your fault), or you put the plot on pause until they figure out what they want and the game stagnates for 6 months (which is your fault). When players say they want the freedom to choose, they mean they want the freedom to opt out of dumb railroads, or to go on tours around town before hopping back on the train, instead of being bound to the rail for all eternity.

But if they really, truly didn't wanna be riding someone else's railroad? If they had a genuine thing they wanted to chase down and they weren't gonna let a silly thing like "democracy" stand in their way? They'd be building their own railroad. And probably a DMPC.

I ran the sandboxy adventure for a group that's been roleplaying longer than I've been alive, most of them having been DMs previously. And the second things started grinding to a halt, at least one of them would notice and would make a snap decision on what to do next, and the others rolled with it because they understood the issue. A railroad locks you into a very particular path through the stations of canon for whatever game you're running, but it's also supplying all the momentum. All the players have to do is not actively refuse the Call To Adventure as they're dragged along for the ride. And that kind of refusal still happens an awful lot and slows down games that aren't sandboxes. When the players get to pick the direction, they also have to supply a lot of the momentum. This is something that needs to be communicated to them during Session 0 in some fashion, because it's a conflict that's baked into the core of being "not a railroad".

If the DM presents 1 option of where the plot could lead next, that's a railroad. If the DM presents 10 options of where the plot could lead next, but 9 of them suck, that's still a railroad. If the DM presents 10 options where the plot could lead next, and all of them have pros and cons, that's not a railroad. But also it means that "where do we go next" becomes a difficult question for the players to have an objectively correct answer on, and so they're gonna debate it a lot. You're going to need some kind of OOC agreement to keep those kinds of debates limited, because otherwise any one of them could last multiple sessions all on its own. What shape that OOC agreement takes is up to you and your players, but IMO you really need to have one.

[Harrow]

I would give two pieces of advice for trying a sandbox. First, unless your players are unusually self-directed, you'll need to give them a push. They need something to hook them into the world that makes them want to start pulling levers of interaction around them. The second would be, play to your strengths. This is a homebrew sandbox, that means you don't need to rely on having a dungeon made of force walls because an adamantine pick ruins your plot. Just let the players solve things how they think is effective. If the players get clever and bypass something, just move on to the next thing. If something ends up being too effective, and starts to become repetitive and boring, then start using counter-play against it. This is a living world, people will react to how the players do stuff. Scry-and-die on the king of the invading army will work, charitably speaking, once. Word gets around. Premade adventure modules cannot predict every player action nor respond to them. You can't do the former, but absolutely can do the latter.

These pieces of advice can be combined. I would (and, in fact am planning to, as I'm about to start a homebrew sandbox after a string of premade adventures myself) give the PCs some sort of long-term, open ended problem to solve. "You're in debt to the gnome mafia! Do you take on dangerous or illegal jobs to pay them back? Skip town and hope for the best? Try to ally with the competing, elf dominant Thieves' Guild and take the gnomes on directly?" You can also give your PCs nice things, like a home base, then use it against them. "You return to the city from your latest caper to find it under siege! Do you try to hit the invaders supply lines and starve them out? Assassinate their general? Travel to their home and set stuff on fire until they turn back to stop you?" This method relies on them actually caring about the things you're threatening, though, so make sure to build up some attachment to it before your threaten to take it away.

[Paragon]

All those insights are really interesting.

I'll add mine to the bunch since I've been running a sandbox campaign for over a year now (from level 2 to 12).
Funnily enough, I get what Saintheart meant by taking something out and making it a plot point : I took out full casters.

I agree with the railroading needing some train stops to breathe before going back on the rails. Also, letting players decide what to do is a real issue as AvatarVecna and Saintheart said ; I learnt it the hard way.

On what I can contribute, I found that creating a web of different political powers and how they hold each others as is really good for immersion. I made a simple 1 slide power point with all the links between them (endgoals, alliances, ...) and it is a strong basis when you need to improvise something without setting up paradoxes later (my players keep reaaaallly good notes sometimes...)

Another tip, is really go in depth in the few majors NPCs your players will encounter and interact with whatever path they choose.
One of such NPC became there BFF which is a good set of rails when the need arises.

Finally, this goes without saying but using your players' backstory to weave the story around it makes for epic plot twists.

You'll make a good campaign I'm sure :)

[Twurps]

My 2cts after having a run a sandbox lvl1 through 16:

1) Slow down the pace. Railroads often come in the form of: 'Do X before Y happens', introducing a lot of time pressure. My sandbox didn't have any of that pressure, but after quite a few years of railroading, my players were still always in a hurry. A sandbox works best if the player actions actually influence the world they're in, but most things in the world take more time than '4 level appropriate encounters a day, every day' will allow. 'you saved the village from starvation' is one thing. But 'you now see the village thrive' really takes some time. So don't be shy with downtime, and make sure your players understand downtime is going to be common.

2) Consider what types of travel you want to allow, and where. If you're players start teleporting to random continents / planes of existence, they will encounter the limits of your preparation very fast. In my world, planar travel wasn't a thing (and why planar travel failed was actually one of my plot hooks). Most players will understand this, but it's still good to include in session 0 as it might influence build choices.

3) speaking of 'the limits of your preparation': Have 'fillers'. You cannot have every detail of every part of your world prepared beforehand. Most sessions, you'll be able to predict where your PC's might end up, and prepare the required details, but every now and then, they'll blind-sight you with a choice/direction you hadn't anticipated, and they are approaching the limit of your prep halfway the session.
That's a perfect time for that band of thugs to ambush the party, or that strange sound at night, luring them to a local graveyard, or.. You get the drill. Have at least one of these ready for each type of environment/location they could possible end up in.

[Bonzai]

3) speaking of 'the limits of your preparation': Have 'fillers'. You cannot have every detail of every part of your world prepared beforehand. Most sessions, you'll be able to predict where your PC's might end up, and prepare the required details, but every now and then, they'll blind-sight you with a choice/direction you hadn't anticipated, and they are approaching the limit of your prep halfway the session.
That's a perfect time for that band of thugs to ambush the party, or that strange sound at night, luring them to a local graveyard, or.. You get the drill. Have at least one of these ready for each type of environment/location they could possible end up in.

I can't emphasize this enough. Players can and will throw you curve balls. You may spend weeks crafting an amazing dungeon that should keep the group busy for months, and they may decide to skip it entirely. Then they may come back when all the encounters are long since trivialize and need to be completely rebalanced. You may have intended for a long series of quests to be able to challenge the BBEG, and they come up with some crazed, absurd plan actually works and bypasses it all in one session.

Being flexible and able to improvise is a crucial skill for a sandbox, and not every DM can handle it.

I also agree that sandbox does not mean that there can't be a story. I think of it as leaving bread crumbs instead of using rails. The party should have a common goal and reason for being together. Throw out the crumbs. How they get to them, and the order is up to them, but once they pick one up, it adds to the story and should bring them a little closer to their goal. They don't have to grab them all, but eventually they will wander into end game by chasing them.

[Harrow]

One thing to keep in mind is that, when players complain about railroads, they generally aren't complaining about there being a plot or even having predetermined goals. IME, when someone is complaining about railroading, what they're saying is that they came up with a clever workaround, something they may even be convinced is the "intended" path forward, only to learn that they aren't allowed to do what they want because, even though it makes sense, the writers of the plot either didn't think of doing what they're trying or they did think of it and contrived an obvious patch-job to prevent it. As long as you can roll with the punches and let your players feel clever, don't take a sandbox to mean the players get a blank sheet of paper that they can draw whatever they want on. You still can, and in most cases should, have some amount of structure in a sandbox. Unless, of course, you've talked to them and ultimate freedom is what they're after. In that case, just make sure they have some kind of plan of action to prevent choice paralysis from kicking in.

[Zanos]

Tell the players to come with characters who want things. A big problem with adventures where people get to decide what they're doing is that most characters don't really seem to want anything, and unless poked repeatedly, will choose to do nothing. Playing characters with some ambition is a good point to start in a campaign where the characters are choosing what they want to do. A character I played for example was looking for arcane lore that would allow him to resurrect his murdered son, partially because he actually wanted to do that, and partially because his pride prevented him from asking or paying someone to do it for him. It helps if the goals are something that progress can be made on in the short term, and to have other goals(or just character development) such that the character doesn't immediately retire when their immediate goals are achieved. I had another character that just wanted to claim a plot of land to rule, kingmaker style, and the party defending and improving the land to establish a power base made up a good portion of that game.

With that said, you need to provide a setting. Expect questions from your players like (i'll use myself as an example) "who are big experts on necromancy in this area? are there any names I would know as a result of my backstory studying these topics? do I know anything about their personalities? how they treat their subordinates?" or "how are new arcane discoveries typically made in this setting? are there any places where experts on these topics frequently gather? if there are, how would someone usually gain access to that knowledge? do i know anything about what kind of defenses they have about theft?" There's a couple of pretty easy adventure hooks in there; meeting with some important mage to become an apprentice, gaining admission to a mages college, perhaps heisting from that same place. If you aren't using a published setting you will need very detailed notes on persons and locations of interest in your setting, since the knowledge you provide to the characters will be the main source of direction for the campaign.

And with that in mind, the party needs a reason to care about eachother. There will be times where the party doesn't all agree on what to do and they may wind up on a mission that only one person is really invested in. This can be a "favor for a favor" type relationship or even in character debate, but generally there needs to be an understanding that while the characters collectively decide what to do, there will be times when the current goal isn't something that immediately forwards everyone's desires.

I will counter-signal the advice to just run a non-sandbox but pretend the players made decisions. I don't know why you would pretend to run a sandbox while actually just shifting everything around so players see all the things you want them to. This is transparent to people that are even remotely clever. While you can certainly reuse maps and maybe even encounters in areas where they make sense, you generally shouldn't have to do this because if you run for a group of characters that are consistently motivated, their decision making won't be throwing you for a loop too much. Just prep stuff 1 or 2 sessions in advance, and when you aren't sure what the players are up to...just ask them? Nothing wrong with just asking folks what they're trying to get done in the next couple of sessions. Not saying you'll never have to throw anything out, you will have to sometimes and that's okay, but you can minimize it by just talking to your players about what they're interested in.

I would also generally advise against trying to keep a coherent overarching narrative from 1-20 or even 1-10 or 3-10. The story in a sandbox game is defined by what the characters are choosing to do, not by who their opponents are. Characters inherently outgrow certain conflicts, and that's fine. I'm not saying you need to make everything slapdash, but if the player characters decided that the orc clans they're fighting are too entrenched for them to deal with, they should be able to abandon that conflict. There should be narrative consequences to deciding to do so, but I wouldn't make it so the conflict chases them no matter where they go because it was the story and needs to be concluded. I would generally plan for a major conflict to last 1-2 levels at most. Frankly, I find modules or adventure paths that neatly sort related villains into a progressive series of more difficult encounters to be a little ridiculous, and it is totally fine for the guys you fought at level 3 to be unrelated to the guys you're fighting at level 8. The great thing about running a sandbox is that you don't have to figure out how the threats are all neatly organized to filter in a line to give the party XP, the characters will generally decide on objectives they can realistically accomplish and will face appropriate threats as a matter of natural selection.

And last but not least, playing a sandbox generally requires a full party of players that are actively and enthusiastically engaged for the duration of the session, who put a lot of thought into what their character wants, how they fit into the setting, and how their goals can realistically be achieved with the means at their disposal(or how their means can be increased). It's not the type of game that everyone thrives in, and that's okay.

[rel]

- Pitfalls of running a sandbox, to watch out for.

The biggest danger in shifting to sandbox play especially after running published adventures is the players not having anything to do in the world.
So make sure you include lots of interesting things for a party of adventurers much like yours to do. And make sure the players will find out about them.

For each piece of prepared content in a location consider how you can organically have the PC's stumble upon it.
And whenever the PC's seem unsure of what to do next, start dropping those plot hooks.

- How to be prepared as a DM for the party going off in an unexpected direction.
world building helps with this. If you know in general terms what the PC's will find in each location then you can improvise more easily.
If the PC's ask whether there's anything for adventurers to do in the random village they just gotto and I happen to know there are owl bears in the nearby woods and a powerful wizard in retirement on the outskirts I can rapidly come up with 3 or 4 promising quests.

Further, if you make the world dangerous and include barriers like high level areas, you can keep the PC's somewhat constrained and thus focus your building efforts.

Finally, if the PC's do something really unexpected, don't be afraid to ask for a 20 minute break to prep. Chances are it was time to order the pizza anyway, and the PC's will enjoy knowing they are operating way off the map.

- Guidelines on homebrewing a campaign setting.
Have fun with it. GMing and world building are not chores. You're another one of the players, and if you aren't having fun, that's a problem.

Focus on those aspects of world building that you enjoy and handwave the things that don't interest you.

Try and focus on things the PC's will experience. the 100 year history of the kingdom will be less
important to the PC's than the quirky blacksmith NPC in the one-horse town they happen to be in.

- Any other general advice.
Make sure you all know what the game is about before you start playing. If half the group is sitting down to play a game about hardcore resource management and lethal combat and the other half is there for flashy encounters and improv acting everyone will end up disappointed.
A sandbox game can easily accommodate either style of play, but not both.

And have fun, keep the game moving, downplay the flaws and inconsistencies in your game and setting and don't sweat the details.

[Thurbane]

Thanks all for the feedback an suggestions, very much appreciated.

Update on the party - here's what looks likely:

Human (?) Barbarian 3
Human (?) Rogue 3
Tibbit Dragonfire Adept 3
Human Cloistered Cleric 3 (NPC)

[AMFV]

- Pitfalls of running a sandbox, to watch out for.
- How to be prepared as a DM for the party going off in an unexpected direction.
- Guidelines on homebrewing a campaign setting.
- Any other general advice.

Cheers - T

I don't run that many sandbox campaigns, but I've got at least some good advice (this is good advice for any campaign). Write out what will happen if your players do nothing. If your players don't interact with anything going on... what happens. Then you'll know what's happening with each part of the world they're largely ignoring. Now you might have to modify things as things change, but it'll give you a baseline. This is also good for prewritten campaigns, as well. Because then you can figure out how to progress a pre-written campaign if the players fail a step and you don't have to force a success.

Other general advice... A sandbox set in Europe in 1931 is pretty good, a Sandbox set in Europe in 1939 is not. Basically the point being that you want lots of tension and under the surface stuff, if there's a major "plot" thread going on that the players can't possibly avoid it's not a very sandboxy setting. So you'll want like interwar drama or cold war drama over like an overt hot war that consumes the world, and that holds true on a smaller scale as well.

Dungeon Masterpiece has some of the best advice for how to run a sandbox, he's a YouTuber, he does a lot of cool stuff about the politics of D&D settings, but he gives a lot of really good advice about running a Sandbox. Also the Alexandrian has some. The big thing is to think about the motivations of factions and major players in the setting, once you know those then even if the players decide to go off in a random direction you can figure out Who is there and What are they doing and What do they want. Like if you've thought about the Geopolitics in the area you can even come up with an answer fast as to the details.

Another piece of advice, while you can't "script" major plot stuff, you can script interludes, the random encounters on the road players usually run into. You can use these to dangle various pieces of things going on in the world in front of their players. Unlike a conventional hook you're not necessarily trying to get them to bite on the hook, just showing them it's there. Think of it more like a net than a hook, they aren't enticed to bite it, but it's there if they want to.

Edit: Also if your players aren't used to a sandbox, maybe do like a transition period where they're in somebody else's service or something at the start, something where they have a wide latitude in what they do, but they still have to kind of do a specific thing, then eventually sever that connection.