Successful Sandbox?

[King of Nowhere]

Regarding power creep, gaining power is fun.
The party starts from nobodies doing odd jobs, progresses to second-rate guys who take care of secondary objectives while the real champions take the spotlight, to world heroes who have to pick up the pieces when the former champions are outclassed.

And 3.5 is slow because it has so many options. Sure, i can make a generic fighter in minutes, but i can put in the extra effort when i want to make someone more memorable. It's not a bug, it's a feature.
I don't know anything about superheroes systems, but from what i picked up in this forum they involve too much dm-may-i for my tastes. I want a system with objective rules that will let me engage the players in actual challenging combat. That system should then include a resurrection spell, because actual challenging combat will occasionally require it.

[Telok]

I don't know anything about superheroes systems, but from what i picked up in this forum ...

GitP is not a heavy supers games forum. I would not use the impressions here to make generalizations of the (extremely wide & varied) supers genera.

Basically there are supers systems that are straight ports of every D&D, narrative, light, heavy, GURPS, a Palladium/Rifts verson, VtM supers knock offs, "high school supers slice-of-life sitcom", and the assorted stuffvthe actual comics companies have put out. Its a bit insane. But with a bit of effort you can find something with your desired system attributes, the in-progress-pay-to-playtest-can't-tell-if-its-a-dumpster-fire-yet Marvel 616 thingy system seems likely (so far) to not rely much on DM fiat...yet.

[Stonehead]

This concept you outline is called a simulation radius. The 3x3 box can be perfectly literal if using a square grid map. Arguably it also serves as formalization of how human memory and information processing works anyway. There are, however, several easy methods to hack this. A global calendar, for example, allows for plotting events like "town [1][1] attacks town [7][7] after X game turns have passed" regardless of where players happen to be. Another way is to keep a separate chart for relationships between characters or political bodies (such as towns) that's updated separately, for example, using a Chess game to model how a war between two countries is going on in the background (moves in the Chess game are made at rate of 1/X turns).

Oh cool, I didn't know it had a formal name. If you aren't using a physical grid, it's pretty easy to only track things that are "conceptually adjacent" to the players. So even though they're several dozen miles away, Kingdom[7][7] is still worth tracking, because the Fighter upset their king so much that the kingdom is sending assassins to get revenge.

While I have run several successful sandbox games, I have never tried to do it in the D&D setting because of the big problems with sandboxes in the D&D rules.

For me, a good sandbox game needs to be in a rules environment where two things are true:

(1) The characters don't get significantly more powerful after completing an adventure or two.

This is crucial. For example, if you dangle various plot threads for the PCs to encounter, but they're all going to involve CR1 monsters, once the PCs reach level 5 in a D&D game, none of those plot threads are worth picking up. So, you're going to need more new plot threads all the time.

(2) NPC character sheets need to be able to be created in about five seconds (while still being as effective as desired).

One of the main ways to make a sandbox game work is to have a bunch of possible plot threads for the PCs to investigate, but you don't want to have to spend a couple of hours making NPC character sheets for characters that the PCs might never encounter. So, instead, you have only the broad strokes of what the characters are like planned in advance.

When the PCs do choose a particular plot thread, you need to have those NPC character sheets though, so you need to be able to create them in a hurry.

----------------------------------------------------------------------------------------------------

Of course, D&D fails badly at both of those criteria. So, I am surprised when someone can do a successful sandbox game using D&D rules. I much prefer using a superhero game system (even if not in a superhero setting) for a sandbox game.

I think these complaints are valid more in combat-grindy campaigns. The more combat-focused the campaign is, the more important it is to have balanced encounters to keep the combat intense, and unique npc stats to keep combat from getting stale. The more focused on intrigue, or city-building a campaign is though, the less important close combats and varied builds become.

If the pull of the "Get revenge on the man who killed my husband" quest is "solve the mystery of who killed my husband", then it's not super important if the players steam-roll the fight when they eventually confront the killer. If the main pull is "kill the murderer once you find him" though, then he should be a tough fight.

Same thing with npc stats, just using simple templates like villager, noble, thief, guard, etc, can be sufficient, provided combat isn't the main appeal. If the main appeal is manipulating political relationships, does it really matter if all the different kings have the same mechanical stats? I don't think so, as long as they have unique goals.

I know a lot of people feel super passionately about this kind of thing, so I should qualify everything by pointing out that a combat-focused campaign is just as good as a mystery-focused or politics-focused campaign.

[Quertus]

So, @Ixtellor, you've got 2 threads open that I've been reading. In one, you're asking for advice on a successful Sandbox; in the other, you ask if it isn't better to have a GM write a character's backstory in order to give the GM more control.

Putting those together, it seems incumbent on me to point out that giving the GM less control, and removing the GM's impetus for control (generally from their psyche, often with a clue-by-four) is a key to a successful sandbox. You can't really run a successful sandbox if you're still trying to run it like a linear adventure. Seeing as how those are, you know, kinda opposed virtues.

On a related note, a key to running a successful Sandbox is having content that is worth exploring. As a really bad example, on my homebrew world of Placia, "Deity" (ie, those who can crowdsource their magic) is a race, not a position. There is no mechanic for elevating beings to godhood. This one simple fact has then trickled down into a large number of eccentricities about Placia's deities (they way they create religions and interact with mortals, the way their spheres of influence don't follow traditional D&D lines, the path of religions throughout history, the way the religions interact). If anyone had ever wanted to investigate those little things, or leverage them as tools, the firm underlying structure of *why* things were this way had already been created.

Dragons, OTOH, are bounded by belief. So long as belief remains low, they sleep. When the party started putting together clues, like how the Elven Martial Arts were all about giving penalties to foes, which they evaluated as horrible for fighting more populous races like humans or goblins, and only made sense if the Elves were the ones ganging up on a singular powerful foe, and started trying to figure out what could possibly fill that ecological niche compared to the sparsely populated Elves, the Dragons stirred in their slumber.

IMO, creating a world of depth and consistency, where you start with root causes, and work through to the elements visible to the inhabitants of the world, is key to getting the right kind of players to engage with and explore the content of the sandbox.

Of course, if you don't have that kind of player, and you are *only* creating the world for the game (as opposed to me, who creates it as the home for my PCs, and as something to do other than suffer Ennui), then there's little point creating such thought-provoking content, if your players aren't the type to think about the content. Also, little point if you lack the skills to present the content in a way that your players can grasp. If they would care about it, but can't find any handholds into the world from your descriptions, then there's little value added to the game by you adding such details in.

I, personally, very much prefer having players who, when I present seemingly incongruous (or even simply anomalous) data, will (break to OOC and ask "are you sure?" before they) break out the magnifying glasses and start investigating, at least some of the time. Thankfully, I've had several groups with players like this, because that's what my GMing style is geared for: creating worlds to explore.

So, more generally, the key to running any game, not just a Sandbox, successfully, is trying to align the styles of the participants. Or, to switch to advice, "don't try to run a multi-year Sandbox, especially as 'probably the last game we'll ever run', if you have no experience with running a successful Sandbox with this group, and have no idea what will and will not work with this group".

Instead, run a series of tests, a series of one-shots, moving to adventures of increasing length, as you calibrate the skills and expectations of everyone in the group, yourself included. That - proper testing, not going in blind, making informed decisions - is key to a lot of successful things, games included.

I'm just grouchy that my player doesn't seem to be well calibrated for this "IRL" thing we've got going on (they thought Charisma was a dump stat, for example!), and am hoping that they'll either run me in something I'm more suited to, or revise my character sheet for the reboot.

[False God]

This is completely superfluous. The term "sandbox game" has nothing to do with a cold uncaring world independent of players. It never had anything to do with simulating a whole world. Instead, it straightforwardly appeals to the idea of self-motivated play in a relatively free, interactive environment, like the kind children engage in when they're playing in a real sandbox. Everyone who has ever played in a real sandbox knows they're finite and that the players have to do interesting things for play to be interesting.

This concept you outline is called a simulation radius. The 3x3 box can be perfectly literal if using a square grid map. Arguably it also serves as formalization of how human memory and information processing works anyway. There are, however, several easy methods to hack this. A global calendar, for example, allows for plotting events like "town [1][1] attacks town [7][7] after X game turns have passed" regardless of where players happen to be. Another way is to keep a separate chart for relationships between characters or political bodies (such as towns) that's updated separately, for example, using a Chess game to model how a war between two countries is going on in the background (moves in the Chess game are made at rate of 1/X turns).

Huh, didn't know there was a term for that. Cool to know. IME though, this is not how sandbox games run by others have operated. They have typically been a poorly or completely un-directed tossing to the wolves of the party, while the world continues to turn around them and often in spite of them.

I agree with everything you said about what a sandbox should be, just IME what I've played from others that isn't what they've been. Maybe I've only played bad ones and it's skewed my perspectives.

[SimonMoon6]

On a related note, a key to running a successful Sandbox is having content that is worth exploring.

I would certainly agree with that that helps.

I'll mention a few of my sandbox games (none of which used D&D):

(1) Back when TSR's Marvel Superheroes RPG (with FASERIP) was reasonably new, I used to try to make the setting (the Marvel comics universe) into a sandbox, which is hard because superhero stories are almost always "A bad guy is doing bad things; the heroes have to go stop him." And the fact that the heroes are just trying to maintain the status quo, rather than overthrowing a big bad evil guy (even someone like Doctor Doom), also makes it hard.

The things I did to help were to give a "newspaper" that I would write up, which would not only show how the world saw the events that the heroes had been involved with, but also gave lots of teasers about possible plot threads for the PCs to investigate.

And as much as that helped, it was still sort of "choose which railroad you want to go on, but it's still a railroad once you choose it".

But the thing that made it more of a sandbox was that the world was the friggin' Marvel universe. Most of the player were intimately familiar with the setting, while everyone else was at least somewhat familiar with the setting. So, when the players were finding things too difficult, they could go to another part of the world and meet up with other known characters. One PC even begged for help from Doctor Doom in return for an "unspecified future favor" which I think came to bite him in the ass (but I don't recall because that was over 30 years ago).

But standard comic-book superhero worlds have problems with being a sandbox unless the world really gets shaken up (like it was when I let one of the players play a master villain).

(2) I had a game set in a multiversal setting, where each multiverse was an established well-known universe (Call of Cthulhu universe, DC universe, Star Trek TNG universe, Doctor Who universe, a generic D&D universe, several universes from Michael Moorcock's novels, etc.)

There was so much to explore that once the players got powerful enough to travel at will, they could go anywhere to do anything. This was one of the most fun and craziest campaigns I've ever run (or seen). After the main villain was defeated, they even got to make their own universes to play in. Like one guy raised an army of zombies to attack a city in the Dreamlands that another PC was ruling after marrying a princess. When that failed, he studied genetics and tried to make an army of superpowered servants by splicing genes from various superheroes. Eventually, he built bombs that were too powerful and ended up exploding the entire planet a massive battle took place on. Meanwhile, another PC was just trying to study brains and psionics and stuff in one of the more science-fictiony universes, along with his mind flayer girlfriend.

(3) Another sandbox game took place on a mosaic world, where there were lots of 1000 mile by 1000 mile square areas, each with its own setting (like generic fantasy, cheesy horror movie, fairy tale, GrecoRoman fantasy, superhero, sci-fi robots, opera, beach party, martial arts tournament world, etc.). While there were McGuffins to seek out, there were some in every patch of the mosaic, so they could go where they wanted to find these McGuffins. And sometimes they split up to check out different parts of the world.

[Vahnavoi]

@Quertus: it isn't necessarily intuitive, but a game master making the player characters and their backstory isn't opposed to sandbox play. The reason is simply because control over game set-up isn't the same as being controlling during play. Mr. Baker's freedom to flip off Sir Imperius doesn't go anywhere even if Mr. Baker's premade character sheet reads he's the Hero of Space, Champion of Princess Hilda, wielder of the Rectangle of Bravery.

[KorvinStarmast]

... wielder of the Rectangle of Bravery.

Is that, by any chance, a toaster?

[Quertus]

@SimonMoon6 - those games sounds like a blast. Still in contact with your old players, to see if they remember their Doom?

@Quertus: it isn't necessarily intuitive, but a game master making the player characters and their backstory isn't opposed to sandbox play. The reason is simply because control over game set-up isn't the same as being controlling during play. Mr. Baker's freedom to flip off Sir Imperius doesn't go anywhere even if Mr. Baker's premade character sheet reads he's the Hero of Space, Champion of Princess Hilda, wielder of the Rectangle of Bravery.

Depending on the background and its perceived purpose, I don’t disagree. It’s simply the stated good of “give the GM more control”, and the subsequent implications, that I was objecting to as being opposed to successfully running a sandbox.

[King of Nowhere]

On a related note, a key to running a successful Sandbox is having content that is worth exploring.

this is certainly one of the most important parts of running a sandbox. perhaps the one most important part: if the content is worth exploring, generally the players will want to interact with it meaningfully, and all manners of virtuous cycles will start.

unfortunately, it's also very uninformative for a wannabe-dm, because we are very bad at judging our own content.
I mean, how many dm think "i made a lot of super duper interesting stuff! I'm so good at this "?
How many, instead, think "awww, my world is stupid and pointless, my players will puke all over the table and then quit and find a better dm"?

you can never know if your world is interesting until you unleash the players on it

[SimonMoon6]

@SimonMoon6 - those games sounds like a blast. Still in contact with your old players, to see if they remember their Doom?

Not really, sorry.

One more thing I remember was how, when I let a player play a master villain, he gathered together an army of villains from the Marvel Universe. In my particular campaign setting, five DC villains had accidentally crossed over into this universe and got stuck here, so they were included in the villain's army. Those five were the Fatal Five.

One of the heroic PCs had a plan to use illusions or shapechange or something to pretend to be one of the villains in the army that had been assembled. The plan was to use this particular villain's influence to try to talk the other villains into... doing something, something like fighting against the leader or going their own way again or some other clever thing like that. But this plan would only work if the villain was a tough badass villain that the other villains would be scared to fight against. So the player (who wasn't really a comic book reader) asked me which of the villains was the most physically intimidating. Well, naturally, I said Validus of the Fatal Five.

So, the assembled villain army was soon addressed by Validus, who gave a very intelligent, sophisticated, erudite speech on the matter.

The other villains instantly knew that this was an impostor because Validus generally doesn't speak in anything other than growls and screams (at least not since he was first mind-controlled by Tharok in his first appearance and even before then, he spoke like a child or the Hulk).

[Vahnavoi]

you can never know if your world is interesting until you unleash the players on it

Sure, there's empirical element of trial and error to making interesting content. But the basics are hardly rocket science. You find some things interesting because of your personality and background, so it's a reasonable bet that people with similar personality and background will find similar things interesting, more so if they have known shared interests.

Considerably harder part of holding a sandbox game is actually living with all those decisions your players find interesting, but you don't.

[Yora]

Probably the most important thing about making any sandbox campaign work are solid incentives. If the campaign is about progressing a story, incentives don't really matter that much. The plot dedicates what the players need to do next, or at least need to accomplish next. A sandbox doesn't have that.

The players continuing to play because they are having fun is not an incentive. Having fun playing campaign is the goal that we are trying to achive. Fun isn't the cause the of fun.

Incentives come in two forms, the pulls and the pushes. A pull is everything that provides the players with a reward for taking an action, reaching a goal, or accomplishing a task. Money for the PCs can be a pull, if they can use that money to buy things they want. But in many RPG systems, this is a fairly weak incentive. Usually much more important is character advancement in the form of experience points or something similar. Players always enjoy that. The critical part for a sandbox campaign is that there always need to be multiple possible sources of XP, but also that the players know what such sources look like and how they can find them when there's currently none within reach. Players waiting for quests to come to them makes for poor sandbox play. Players knowing how to go looking for quests is much better. This is why the treasure hunting exploration campaign style works so very well for sandbox games. The players know that there are many dungeons in the wilderness, and that those dungeons have treasure, and when they find the treasure they get XP. That is something the players can easly act on on their own initiative. "Let's go find a dungeon!" is an easy thing to do for players, even if they don't know anything about the game world they are playing in yet. They can always walk into the next village and ask around if anyone knows of a ruin, cave, or monster lair in the area.
Another really great thing about rewarding the players for getting treasure is that it's an objective that has no clear "completed/failed" state. You only fail at treasure hunting if you don't find a single piece of treasure. But if you are running a rules system that in addition to treasure rewards also offers minor combat victory rewards (like roughly 3:1 for example), then even a treasure hunt that produces no treasure still provided a few XP to the characters from the enemies that got into fights with while they were searching. When exploring a dungeon and collecting treasure, it's up to the players to decide if they have found enough, or if they want to continue looking for more hidden stuff. This is different from a task where they have to slay a beast or rescue a prisoner, which is something that they either manage to do or don't. And unless they have failed with certainty, the players generally feel compelled to keep working on it until they either reaches a completed or failed state. Which usually is a state that has been defined by the GM when the task was given. But in a sandbox campaign, the players being able to shift their priorities and head to greener pastures is a big element of their agency.

But another really important factor, that is mostly specific just to sandbox campaigns, are the pushes. A push is everything that makes continuing to wait for or looking for something fun unsustainable. When players have infinite choices for tasks to pursue and can always continue to keep looking for something that seems more interesting and fun if none of the current options is really pulling them in, there's a real risk of things starting to drag and meander, with players reconsidering their options in circles and continuing to go looking for more rumors. A push is anything that makes it impossible to keep doing this for long, or at least makes indecisiveness come with a price. The typical forms of this are some kind of permanent upkeep that constantly goes down slowly. Running costs for the PCs staying in town considering their actions. As GM, while the players are debating their options how to continue after a completed adventure, occasionally announce that the PCs have been staying in the current town for another week, casually looking for work but mostly burning through their money. Maybe not every 10 minutes that the players are talking, but peehaps every 30 minutes or every hour, depending on how these things play out in each campaign. Eventually the players will get sick of loosing money and decide to go and do something that is available, even if it's something that isn't sounding particularly exciting right now. Which is exactly what we want. Continuing to debate options will not make the choice any easier. Being out on adventure and having things happening shakes up the current situation, and after the current quick adventure of neccessity is done, the players might have a much better idea of what they want to do on their new situation. If not, the cycle continues and their newly acquired supply soons runs out again. Doing those quick jobs that nobody is terribly excited about doesn't have to be mindless grinding. It absolutely should not be mindless grinding. It should still be a fun adventure. But it doesn't need to be the big epic dream adventure that the players wish they could go on but can't define. After all, they never now what they will actually encounter when they go to explore a place. Even if it doesn't sound terribly interesting to check out from the description they have, actually being there won't be nearly as drab. The important part is that the players need to keep playing adventures instead of going in circles looking for adventures.

The key idea behind all of this is that with both pulls and pushes, the ultimate decision to act always comes from the players. As GM, you never tell the players "You have A, B, C available, please finally pick one now!"

[Quertus]

this is certainly one of the most important parts of running a sandbox. perhaps the one most important part: if the content is worth exploring, generally the players will want to interact with it meaningfully, and all manners of virtuous cycles will start.

unfortunately, it's also very uninformative for a wannabe-dm, because we are very bad at judging our own content.
I mean, how many dm think "i made a lot of super duper interesting stuff! I'm so good at this "?
How many, instead, think "awww, my world is stupid and pointless, my players will puke all over the table and then quit and find a better dm"?

you can never know if your world is interesting until you unleash the players on it

You’ve misunderstood. While “interesting” is a good goal, and I agree that it’s… somewhat subjective and audience-dependent, I was talking about the content being built in solid foundations. Crunch, not fluff; substance, not style.

Gar don’t use guns. Why? Because it’s worth 2 character build points to take that flaw.

Batman don’t use guns. Why? Because his parents were shot & killed in front of him at a young age.

Both have the same fluff (“don’t use guns”). And while, yes, it’s nice if you’ve picked fluff such that some of it is interesting to the players, only one of those examples has fluff that is rewarding to investigate further, even though it’s the same fluff.

My advice for building a sandbox with content worth exploring was to start at that founding event, and work out to the visible effects. That way, if anyone decides to dig, there’s actual substance there already.

My examples were how, on Placia, “deity” is just a race, and only uses followers to crowdsource their magic, whereas dragons are actually dependent upon belief. No players have ever interacted with those mechanics, but, had they dug into the player-facing bits (the unusual nature of Placian religions, the way dragons acted when finally awakened (which wasn’t actually initially player-facing; that became player-facing after one group chose to dig into several other players-facing bits, like how traditional elven martial arts (based off giving penalties) made absolutely zero sense for the constantly outnumbered fey race)), they would have found actual consistent substance awaited them.

A lot of GMs, their worlds have all the substance of, “doesn’t use guns is worth 2 build points”, and just aren’t worth exploring.

(EDIT: insert obligatory “Armus moves to protect <PC with better defenses>” reference)

But another really important factor, that is mostly specific just to sandbox campaigns, are the pushes. A push is everything that makes continuing to wait for or looking for something fun unsustainable. When players have infinite choices for tasks to pursue and can always continue to keep looking for something that seems more interesting and fun if none of the current options is really pulling them in, there's a real risk of things starting to drag and meander, with players reconsidering their options in circles and continuing to go looking for more rumors. A push is anything that makes it impossible to keep doing this for long, or at least makes indecisiveness come with a price. The typical forms of this are some kind of permanent upkeep that constantly goes down slowly. Running costs for the PCs staying in town considering their actions. As GM, while the players are debating their options how to continue after a completed adventure, occasionally announce that the PCs have been staying in the current town for another week, casually looking for work but mostly burning through their money. Maybe not every 10 minutes that the players are talking, but peehaps every 30 minutes or every hour, depending on how these things play out in each campaign. Eventually the players will get sick of loosing money and decide to go and do something that is available, even if it's something that isn't sounding particularly exciting right now. Which is exactly what we want. Continuing to debate options will not make the choice any easier. Being out on adventure and having things happening shakes up the current situation, and after the current quick adventure of neccessity is done, the players might have a much better idea of what they want to do on their new situation. If not, the cycle continues and their newly acquired supply soons runs out again. Doing those quick jobs that nobody is terribly excited about doesn't have to be mindless grinding. It absolutely should not be mindless grinding. It should still be a fun adventure. But it doesn't need to be the big epic dream adventure that the players wish they could go on but can't define. After all, they never now what they will actually encounter when they go to explore a place. Even if it doesn't sound terribly interesting to check out from the description they have, actually being there won't be nearly as drab. The important part is that the players need to keep playing adventures instead of going in circles looking for adventures.

The key idea behind all of this is that with both pulls and pushes, the ultimate decision to act always comes from the players. As GM, you never tell the players "You have A, B, C available, please finally pick one now!"

I’ve been in a lot of groups where your “pushes” would have resulted in the party murdering the entire town, animating the corpses, and calculating how many decades they can keep planning based on the loot obtained (especially if the food can be preserved / “purify food and watered”).

[Yora]

Would that have been bad?

[King of Nowhere]

You’ve misunderstood. While “interesting” is a good goal, and I agree that it’s… somewhat subjective and audience-dependent, I was talking about the content being built in solid foundations. Crunch, not fluff; substance, not style.

yes, consistency is important for worldbuilding. perhaps the most important factor - at least for some people. I am certainly a very inquisitive person, I think of consequences, and if something does not hold to scrutiny, you lose me. So I think a lot on how elements of the world interact with each other.

On the down side, that alone does not guarantee the world is interesting.
Personally I am good at working at a macro level - I have built wikipedia-style articles on several in-world nations, and I have a good idea of their reciprocal relationship, governance, and social issues. But the small scale eludes me.
Here is the big elven nation of Tal Calel. They have a major magical disaster zone in their middle, spawned a century before during a large battle between lots of spellcasters, and they enacted various measures to contain them. they have some social friction between the elves and the humans, the elves are afraid that with human high reproductive rate soon they'll no longer be an elven nation, human resent that the elves have longer career and get all the important jobs. i can give relations with other nations, i have a pretty good idea of how much high level power they can call upon in need.
So you are in the capital, there is the wall keeping the creatures from the magic area out, i have good details for how it is defended. is the wall never bypassed by flying or tunneling creatures? actually that's part of the plot, because there is a villain in the magic area that's controlling those creatures and wants the defenders to grown complacent into their wall. all great and detailed.

Ok, you enter the pub. err.... yes, there is a pub. You talk to someone. ok, i'm sure there's someone in there with whom you could talk. what does the city look like? well, besides a couple of monumental places, it looks like a city. remarkably city-like.
I need a plot hook for the next session... well, I have a couple major villains plotting in the shadow, but as far as concocting some decent adventure for a mid-low level party, I'm shooting blanks.

[Quertus]

Would that have been bad?

Depends on the group, and how they feel about that, I suppose.

But I've known numerous groups who were really into planning, and several who would happily level even friendly cities just to not have to worry about minutia during their planning.

Of course, so long as I'm the one animating all the dead, I suppose I won't complain too much about the loss of one town.

yes, consistency is important for worldbuilding. perhaps the most important factor - at least for some people. I am certainly a very inquisitive person, I think of consequences, and if something does not hold to scrutiny, you lose me. So I think a lot on how elements of the world interact with each other.

On the down side, that alone does not guarantee the world is interesting.
Personally I am good at working at a macro level - I have built wikipedia-style articles on several in-world nations, and I have a good idea of their reciprocal relationship, governance, and social issues. But the small scale eludes me.
Here is the big elven nation of Tal Calel. They have a major magical disaster zone in their middle, spawned a century before during a large battle between lots of spellcasters, and they enacted various measures to contain them. they have some social friction between the elves and the humans, the elves are afraid that with human high reproductive rate soon they'll no longer be an elven nation, human resent that the elves have longer career and get all the important jobs. i can give relations with other nations, i have a pretty good idea of how much high level power they can call upon in need.
So you are in the capital, there is the wall keeping the creatures from the magic area out, i have good details for how it is defended. is the wall never bypassed by flying or tunneling creatures? actually that's part of the plot, because there is a villain in the magic area that's controlling those creatures and wants the defenders to grown complacent into their wall. all great and detailed.

Ok, you enter the pub. err.... yes, there is a pub. You talk to someone. ok, i'm sure there's someone in there with whom you could talk. what does the city look like? well, besides a couple of monumental places, it looks like a city. remarkably city-like.
I need a plot hook for the next session... well, I have a couple major villains plotting in the shadow, but as far as concocting some decent adventure for a mid-low level party, I'm shooting blanks.

Well, yes. There's layers. My experience is that many GMs don't start at a good foundation, and so, once you try to dig through the layers, what gets returned is valueless. Which only matters when there are those who actually do the digging.

But once you have a setting built from firm foundations, to run a successful Sandbox, you need to position the PCs at a point in time and space such that there are plenty of interesting toys for them to play with.

Let's look at your Tal Calel elves.

If they had been *my* Tal Calel elves, I would know...

• Who they were beforehand, and why
• How the major magical disaster zone in their middle, spawned a century before, changed them
• Why there were lots of spellcasters in their middle; what the outcome was
• Why there is friction with the humans; what that looks like
• How this event has changed them

OK, so, it looks like the elves and humans live together, but apart. When did this start, and why?

I don't generally like trying to back-fill to pattern match, but this isn't terribly hard to work with. So, let's say that the Elves in question were your somewhat generic forest tree-fort-dwelling hippie warrior mages, following traditional patterns laid down in a time beyond memory, presumably by one or more of their ancient heroes / deities. A portal from Beyond opened in the middle of Tal Calel, and a huge fight erupted between the native Elves and the incursion of Illithid Slavers, wielding their strange "magic" (psionics). The Illithids were defeated (boo!), but some of the human former slaves remained. Because the humans understood a little of their masters' craft, they helped devise and still help maintain the anti-psychic obelisks around the affected area.

The Elves, being super long-lived, have the huge advantage of centuries of experience (and XP). They're individually more powerful than the humans, and the humans die off so fast, that doesn't look to be changing, like, ever. So of course they hold and will always hold all the positions of power. So they're not worried about losing leadership roles. Unfortunately, that's exactly what has the humans upset. And the upset humans breed like cockroaches. While they might individually not be a threat to elven superiority, as they continue to grow, they may someday become a majority, and flood the elves through sheer numbers, like the Goblins of old.

Now, if this is a Sandbox? Then you don't "need a plot hook for the next session". If the players want to play Elves trying to create non-psionic versions of the obelisks, so that they can then move to exterminate the cockroach humans? Go them! If they want to play disenfranchised humans, who invented Ice Cream, but then had Elves use their superior skills to produce superior Ice Cream, and pushed them out of the market, and these humans now seek to make their mark on the world with Psychic Ice Cream that the elves *can't* reproduce? Go them! If they want to play Descendants of slaves and like-minded young elves, who want to take the fight back through the rift to the Illithids? Go them! If they want to play the human beastmasters who "encourage" monsters to rampage through the human slums (on ground level), and try to blame it on the Elves, to fan the flames of friction between the races, in order to seize power themselves? Go them!

In a proper sandbox, you don't need to create an adventure or a plot hook, just a setting rife with possibilities, for the players to tell their own story in.

As for the appearance of things... I kinda choose / steal aesthetics whenever I create a race / culture. And ask how events would change that.

So the elves I picked are kinda generic, favoring woodland colors and clothes ranging from "serviceable" to "flowing" (I think modern LotR has this covered). Their "buildings" are part of the trees, usually added on with lots of rope walkways, but in rare cases actually grown from the trees.

Because fire is such an issue, they build close to rivers and streams, are skilled with fighting fires and making fire breaks, and metal is rare. What metal is forged is done with magic (and "mined" by elementals, who just earth glide down, and bring up the desired metals).

After encountering psionics... gems and crystals went out of fashion with the Elves, being considered "low" (ie, human) things (as humans are literally low, living on the ground and all). This is also pragmatic, as the human Psions need the gems to build the psionic devices (like their portion of the obelisks that keep the rift closed).

The humans live in stone shacks (created via Wall of Stone & Stone Shape) beneath the elven trees, or in designated clear-cut areas (where their forges won't endanger the forest). The general aesthetic is rounded, natural looking (Yoda's hut, I think?), not boxy like one might expect. Their furnishings are likewise stone, favoring benches over chairs, and pillows are a big thing ("BYOP"). Their clothing is generally primitive (bear skins and the like), as the elves don't have enough of the source of their clothing (spider silk) to clothe the humans, too (and haven't taught the humans the trick of it, to further classism), and the humans don't have the source of their Illithid-provided clothing (sheep, for wool, which the elves have never seen, and laugh at human about their ancestors' stories of "clouds with legs"), either.

Members of both races may wear little obelisks, although the exact meaning of doing so isn't quite as simple for either race as the other believes.

So, if you wanted me to describe an area, or a group of people, or wanted to interact with a specific person, it's easy for me to work with when starting from that base of "who are these people?".

But it's starting with that foundation, with "Bruce Wayne's parents were shot and killed in front of him when he was young, and therefore...", and building from there to a setting rife with toys, that I was advocating.

[Yora]

Continuing to give the players plot hooks for prepared adventures is not going to get them to become proactive themselves. It can work quite well to give the players an introduction to a new world they don't know, by introducing them to several people and places and the local structures of power.
But to become a successful sandbox, the players have to take over defining their own goals eventually.

[King of Nowhere]

But to become a successful sandbox, the players have to take over defining their own goals eventually.

yes, but I still need to prepare stuff for them to do, if I don't want to rely solely on my (limited) improvisational skills.

So, the party decides to investigate the magical hazard. They already know pretty much everything that's common knowledge about it, due to two of them picking backstories directly related to it. People have been trying to experiment with it for a century, so there's a well-established body of academic literature there. so they ask in the university, and they don't really find anything more than they already knew. they go to several experts, but none of them actually knows much more. If they can think of any experiment to try, you can be sure someone else already tried it. Maybe they notice that only certain kind of monsters attack the wall, or they notice a handful of other hints I left - but everybody assumes it's normal. it's how things have been for the past 70 years, since the wall was built. Some people may agree it's weird, but they certainly have no other input to offer.
Meanwhile, I know that the main villain is biding her time deep within. Nobody knows of her exhistance. Nobody even suspects she may exhist in the first place. She keeps to the inner area - a place so rife with magic radiation that no divination, teleportation or other long range recon magic works, so deadly that even 20th level people have disappeared in it without traces - true resurrection didn't work, miracle didn't work, the gods themselves cannot see in there.

So, I have a few possibilities:
1) i let the players bash their head against this stone wall of lack of information until they get tired
2) after they randomly try stuff for a while, i let something work
3) i figure out some way to trickle informations in a way that would be sensible

In a completely realistic scenario, 1) should happen. There are other instances of similar magical accidents, and they all baffle researchers. They have been a major issue worldwide, and a lot of resources were devoted to finding ways to deal with that problem. There's no way a bunch of greenvines can unravel it.

At your tables, maybe 2) would happen. the players try stuff, until you decide something works, even when it really shouldn't. Or perhaps you created a mistery that was easier to unravel - and just pretend that the villain has been able to hide there for 70 years despite those glaring weaknesses in security that allow a low level party to discover everything. Or perhaps you are better than me at going from large to small scale, and you have figured out a few ways for the players to be successful already.

Me, I have to go with 3). I have to figure out some way for the players to find stuff that nobody else could find before, in a way that won't look ridiculous in hindisght. Either they are at the right time in the right place for something that is just happening this one time. Perhaps I can pull off something from their backstories. Regardless, I need to find some ways to break down this huge, impossible task into smaller tasks that the players can solve.
Regardless of the way, I need to be prepared with something. Otherwise, there is a strong risk of devolving into a dm-may-i scenario where the players sstumble around until i decide to let them succeed in some way. And this has happened, and sometimes it has worked greatly - sometimes the players making the right questions allowed me to figure out some stuff I was still missing. Other times, it was indeed a dm-may-i game, and the best I could hope for was that it wouldn't be too obvious.

In the specific case, I decided - plot hook - that there was a plague of undead animals coming out of the wild magic area. the players investigated and found a lot of discrepancies, those weren't common undead and didn't behave as common undead. So they investigated, and discovered that those undead were not caused by normal necromancy, but by bacteria. further investigation discovered that those bacteria were reacting specifically to uses of turn undead.
long story short, those bacteria were meant to gather and store divine power. they were invented by two brothers who wanted to ascend to godhood; they disseminated those bacteria around to cause zombies, the zombies would get turned, the bacteria would absorb some divine energy from that and would carry it to the brothers, who would store it until they had enough for ascension. But those brothers were not evil people, and they did put a lot of safeguards to make sure the plague wouldn't get out of hands; it only affected large animals (a zombie bear looks bad, but it's an easy fight for a mid-low party or for a platoon of soldiers; can you imagine zombie mosquitoes spreading with their bites?), and they did not become aggressive, and they were still afraid of man. And nobody had discovered it before because everyone knows that the wild magic zone is full of weird stuff; zombies behaving strange? must be tuesday already.
in turn, the brothers were hiding inside the wild magic - one is a druid and invented the bacteria, the other is a wizard and he invented the best abjuration for it, stuff that nobody else knew; and the wild magic was the only place that could hide the signature of so much divine energy.
the party eventually found them by tracking the bacteria, and basically got bribed into letting them hide somewhere else. but the new hideout wasn't as safe, and the main villain could now detect their presence. so she kidnapped the brothers, but she was used to hiding in the middle of wild magic where nobody could come close to her without growing tentacles or worse, she wasn't any good at spy games, and so she left a ton of clues. And by now nobody had as much direct experience in the wild magic area as the party, and so it would make sense that nobody else tried to take the investigation from them...
And it worked great, and there were no glaring plot holes. And all the while I was scrambling to figure out the details before the next session. How exactly were the bacteria going to the brothers? If I make a misstep, I risk contradicting myself. I had to think things through to make sure all the details checked.


tldr
After this long winded story, I'm realizing that probably what I'm calling plot hook (hey, there's a bunch of zombies coming out. want to fight them? But hey, they did behave real weird, want to investigate?), you guys may call clues. As in, sprinkle clues that the party can follow. I am bad at that, I don't have them figured out in advance when I set up the world, and it's better if I don't have to make them up on the spot, else I'm likely to create plot holes.
If the party had no bitten into the zombies, I'd have had to devise something else.

[SimonMoon6]

yes, but I still need to prepare stuff for them to do, if I don't want to rely solely on my (limited) improvisational skills.

So, the party decides to investigate the magical hazard. They already know pretty much everything that's common knowledge about it, due to two of them picking backstories directly related to it. People have been trying to experiment with it for a century, so there's a well-established body of academic literature there. so they ask in the university, and they don't really find anything more than they already knew. they go to several experts, but none of them actually knows much more. If they can think of any experiment to try, you can be sure someone else already tried it. Maybe they notice that only certain kind of monsters attack the wall, or they notice a handful of other hints I left - but everybody assumes it's normal. it's how things have been for the past 70 years, since the wall was built. Some people may agree it's weird, but they certainly have no other input to offer.
Meanwhile, I know that the main villain is biding her time deep within. Nobody knows of her exhistance. Nobody even suspects she may exhist in the first place. She keeps to the inner area - a place so rife with magic radiation that no divination, teleportation or other long range recon magic works, so deadly that even 20th level people have disappeared in it without traces - true resurrection didn't work, miracle didn't work, the gods themselves cannot see in there.

So, I have a few possibilities:
1) i let the players bash their head against this stone wall of lack of information until they get tired
2) after they randomly try stuff for a while, i let something work
3) i figure out some way to trickle informations in a way that would be sensible

Or, (4) give them information in a non-sensible way.

In a world where the PCs aren't special and can't investigate things in any way that hasn't already been tried, then you can't rely on them getting anywhere in that sort of plot. So, if non-special characters are going to get involved, it's probably going to take an act of god or ridiculous coincidence.

So, the players walk along and the ground gives way and they find themselves at the entrance to the big baddie's basement. Nobody else walked exactly on that spot before.

Or a meteor crashes down and makes a big crater. What a coincidence.

Ok, maybe not a meteor, but a purple worm traveling through the area leaves big open holes in the ground.

Or one of servants of the big bad suddenly quits and shows up near the PCs who can now investigate that character.

Incredible coincidence is one of the best tools of a DM, as long as it's not abused too much.

On the other hand, maybe the PCs can't find anything. Not everything that can be investigated needs to bear fruit. Just let them know that they're not getting anywhere so it doesn't use up valuable play time.

tldr
After this long winded story, I'm realizing that probably what I'm calling plot hook (hey, there's a bunch of zombies coming out. want to fight them? But hey, they did behave real weird, want to investigate?), you guys may call clues. As in, sprinkle clues that the party can follow. I am bad at that, I don't have them figured out in advance when I set up the world, .

And for me, that would be the problem. When sprinkling these clues or plot hooks, you have to have in mind what the PCs can do to investigate and where things are going. You don't have to know all the details yet, but you have to be able to flesh things out once the PCs get there or else the clue is worthless.

For example, I might give PCs a clue of "People have gone missing when traveling through the forest. You know, that forest over there where a bunch of druids live." And the PCs could go there or not. They might even find bodies killed by animals... so obviously the druids are behind it! The PCs could talk to the druids or fight the druids. If they talk to the druids, they'll say that they're not responsible but a bunch of gnomes have been acting weirdly lately (and using their ability to talk to burrowing animals to sic those animals on passersby). The PCs can then investigate "Gnome Man's Land" and find that there are blue-cap gnomes and red-cap gnomes. The blue-caps are friendly, but the red-caps have been going kind of crazy lately.

After investigating the red-caps, the PCs will learn that the gnomes have been infected with chaos, especially the gnomes in a certain area. That leads to a gnome-filled dungeon, where the gnomes are super-annoying with sneakiness, illusions, and traps. Also riddles and puzzles galore because gnomes are effing annoying. The gnomes are also infected with chaos mutations.

Other monsters to encounter are chaos beasts and gibbering mouthers. Later bosses will include slaads, who have a plan (to be figured out later if necessary, something like gathering an army of chaos gnomes or turning gnomes into chaos beasts or something) and there may be a boss beyond the slaads (maybe a vampire or something).

Not all the details have been worked out, but I can flesh that part out easily once the PCs start to interact with the story. I still only give the PCs the plot hook of "maybe the druids are murdering people" but it can lead somewhere if the PCs care to investigate.

[Thrudd]

Has anyone mentioned "Worlds Without Number" yet? One of the best toolboxes for building a sandbox campaign I've seen, with tons of tables for determining the features of your world, locations and adventures off-the-cuff, or to help you prepare in advance. It comes with its own setting, but the world-building and adventure building features are applicable to any swords & sorcery D&D-like game. There's a free pdf version that includes a lot of useful stuff (that's the only version I've personally looked at). I'd recommend everyone check it out.

[Easy e]

Well, I maybe crazy but the best "Sandbox" I ever ran had zero planning. The world evolved around them as the players built their backstories and characters and then started to play.

As the GM, I only introduced potential hooks and then we followed the paths the players decided to follow to see where they led. As they did stuff, new hooks appeared, and as they followed those they learned more and more about the world. I had no idea what was going to happen and where it led, until they started doing it.

This was co-operative world building as much as anything else. Creating a lot of "foundation" for your players to maybe dig through is a lot of work. Instead, only build on what players want to interact with, and get their help in building it as much as you can.

Of course, I am a lazy GM and pretty good at improvising on the spot.

[Quertus]

Well, I maybe crazy but the best "Sandbox" I ever ran had zero planning. The world evolved around them as the players built their backstories and characters and then started to play.

As the GM, I only introduced potential hooks and then we followed the paths the players decided to follow to see where they led. As they did stuff, new hooks appeared, and as they followed those they learned more and more about the world. I had no idea what was going to happen and where it led, until they started doing it.

This was co-operative world building as much as anything else. Creating a lot of "foundation" for your players to maybe dig through is a lot of work. Instead, only build on what players want to interact with, and get their help in building it as much as you can.

Of course, I am a lazy GM and pretty good at improvising on the spot.

Never forget: being crazy doesn't mean you're wrong.

Being good at improvising, and having a group onboard with cooperative world-buildng? Yeah, that's definitely a good recipe for success. Just as one may want tomato sauce in a pizza or in lasagna, but probably wouldn't want it in cheesecake or General Tso's, what ingredients you use / what ingredients you will find useful will vary based on what your intended outcome is.

So, yes, I agree that that is a good recipe for a different type of sandbox.

However, I do have to disagree on the idea that creating a lot of foundation is a lot of work.

First, because one need not have a lot of foundation to create a lot of player-facing content. That is, "gods are a race, not a status" directed the creation of pantheons and religions - something I was kinda going to do anyway. It's just that *starting* from that foundation and working out to player-facing details meant that those player-facing details had the solid foundation to produce consistent answers, should the players have ever started digging. So it's not *necessarily* necessary to create a lot of foundation, just enough to get you to the necessary player-facing details (like "what religions are available in the world").

Second, because... it's not a lot of work? Like, "Bruce Wayne's parents were shot and killed in front of him at an early age, and therefore..." ... a) he decided to train b) to become a vigilante superhero c) and doesn't use guns. "He had a really scary experience with bats while exploring a cave, and therefore..." ... a) he dresses up like a bat b) to scare criminals c) because he thinks bats are scary d) and that scaring people is effective, and e) he made his secret lair in said cave, for reasons.

Is that really harder than... uh, "Rat Man"... dresses as a rat, shoots criminals, lives... somewhere... gets his guns... and ammo... uh, somehow... uh... and, when his guns fail, he... uh... I donno? What skills does he have? Uh...

Or, for a "real" character, and one that I can maybe answer a lot of the "what" but not the "why", Superman... dresses in red and blue... and probably deals with criminals (and not just major disasters)... somehow (probably not by punching them if they're normal humans)... and lives... in an apartment... and has a fake identity as a reporter... and generally acts like a big boyscout?

Is it really harder to invent a thing (a character, a piece of world-building, whatever) by starting at the root causes, than by just randomly picking things?

Even if you somehow find such random generation easier, my senile self sure doesn't find *remembering* those random details anywhere near as easy as ones that I can trace back to / from root causes. (Which is why my players often commented on how good at "improvising" I was, when, in actuality, I think I'm pretty terrible at improvising, and I was just remembering all the core components & working out to / thereby remembering the calculated bits).

[King of Nowhere]

laying out the fundations is not a chore to me, because it's the kind of stuff that I think in my free time. I have ideas rolling around in my head.
it's figuring out the minute details before a session that's the chore.

[kyoryu]

Well, I maybe crazy but the best "Sandbox" I ever ran had zero planning. The world evolved around them as the players built their backstories and characters and then started to play.

As the GM, I only introduced potential hooks and then we followed the paths the players decided to follow to see where they led. As they did stuff, new hooks appeared, and as they followed those they learned more and more about the world. I had no idea what was going to happen and where it led, until they started doing it.

This was co-operative world building as much as anything else. Creating a lot of "foundation" for your players to maybe dig through is a lot of work. Instead, only build on what players want to interact with, and get their help in building it as much as you can.

Of course, I am a lazy GM and pretty good at improvising on the spot.

One of the advantages of not writing a whole bunch of stuff at the beginning is that you can figure out what the players find interesting, and build off of that.

If you're running a game for a consistent group, it's pretty safe to say that they will be more engaged with the things they find interesting than the things you find interesting.