How did/do you run your most successful sandbox campaigns?
Do you use a setting that you know light the back of your hand? Improvise large chunks of the world? Jump in with minimal prep time?

I'm not going to say I'm an expert in the field of hex crawls but I have done long campaigns with hex maps and here's some things I did that worked for me.

1. Have landmarks that can be seen from a distance.  I put mine about 5 hexes apart and I try to make it so that you can see at least two other landmarks from any one.
2.Populate the map and get the NPCs doing things.  Trade, conflict, rivalries, and intrigue should all be going on before your players see the map for the first time.
3. Put in your dungeon entrances. 
4. Fill every hex with some kind of encounter. I do about 20% combat type and the rest with other things that tie in with major factions and world events.  I also put in clues to various things I've hidden in the map for the players to find.  If you're rolling on a random table, now is a good time to do it.
5. Repopulate hexes as the campaign goes along.  If someone escaped arrest in a previous session, pick a hex where he's hiding out.  Animals and monsters will go to areas that have been previously cleared.   
6.  Keeps notes.  I keep one notebook for the campaign and a separate one for the map.  The hex map book has two pages for every hex coordinated by hex number.  I keep a lot of spare room to write so that I can update the hex.  The campaign notebook should keep a running chronological event table that includes the hexes visited. 
7.  You can have the players involved with complex intrigue and plots while living in a hex map.  Adventures can be layered into events on the hex map and hexes can contain elements of your adventure in them.
8.  Hexes are living places.  A creature can wander around multiple hexes.  Maybe the dragon isn't home when you discover it's nest.

Again, I'm not telling anyone that I'm doing it right, these are just the things I consciously do to keep my game running smoothly.  I hope you find this helpful.

I am not an expert, but am currently running one.

Will try to summarize some tips, but let me start with this:

"Improvise large chunks of the world? Jump in with minimal prep time?"

I do not think it is a good idea. I almost ruined my campaign by improvising a situation and making some impromptu rulings.

I'd do the opposite: take pre-written modules, change them as desired BEFORE you start, and then STICK TO THEM. If the PCs live or die it is their own choices, not yours.

Other than that: have plenty of NPCs, encourage hirelings/retainers, start with a "home base" and a few obvious targets/rumors (I give them at least three most of the times).

Let the PCs control hirelings and allies.

Keep notes of past events and (possible) consequences of the PCs and NPCs actions.

Use encounter tables and reaction tables. Reaction is easy to forget, but assuming every orc will attack will quickly cause a TPK for no good reason.

I have been using this for my hex map:

https://hextml.playest.net/

And scattering several classical dungeons. Some examples:
https://methodsetmadness.blogspot.com/2022/10/sandbox-quest.html
https://methodsetmadness.blogspot.com/2022/11/sandbox-quest-part-ii.html
https://methodsetmadness.blogspot.com/2023/03/doom-of-savage-kings-dcc-665-actual.html
https://methodsetmadness.blogspot.com/2023/02/the-god-that-crawls-actual-play-review.html

Some world building tips:
https://methodsetmadness.blogspot.com/2021/08/easy-worldbuilding-with-pareto-price.html

I'd rather have a finished sandbox, written by someone else, than to create my own, but I did'nt find any good ones with OSr rules - 5e's CoS and ToA are half-decent.

Quote from: Eric Diaz on October 18, 2023, 09:35:50 PM
"Improvise large chunks of the world? Jump in with minimal prep time?"

I do not think it is a good idea. I almost ruined my campaign by improvising a situation and making some impromptu rulings.

I absolutely agree with this.  I don't think you need to build the whole clock before running a game but you should have a solid world filled out with all your major NPCs and a few events and adventure ready to go.  If a town is on the other side of the map, you don't need to name the baker's wife but you should have the town ready to run without coming up with it on the fly before the players get there.

Have a bunch of bite sized content. A mini-dungeon. An involved random encounter. A mysterious map or magical compass. Toss this stuff in the way of the PCs when things are slow.

That's how I was running my Starfinder Mercs campaign before Covid killed it. Everyone seemed to be enjoying it. The idea is, if the players find a chunk of content interesting, you can elaborate on it for future sessions. If they aren't interested, you only spent a few minutes on any specific bit.

I've never run a true hex crawl sandbox.  I've run various things that are in the 75% to 80% sandbox territory, and some of them have been close to straight hex crawls.

There are two things that I do that I think are useful and appropriate for something approaching a sandbox, and should be useful in one:

1. Make a distinction between things that are placed and things that will be placed but not yet.  For example, I've got a bear cave, with 3-5 bears in it, and some odd chunk of treasure left from a mauled humanoid.  I haven't placed it when the session starts.  When it does get placed, randomly, on purpose, whatever--it stays.  Moreover, getting "placed" is decided at the first moment that it is relevant.  So that could be someone stumbling on the lair, but it could also be a rumor, or a map, or just some garrulous old codger talking about seeing some bears "back up in them hills".   If the players ignore the old codger and go on their merry way, that lair is still right there where it was when they talked to the guy.  This is a compromise between completely random and everything mapped out ahead of time, but being true to the idea that the world doesn't move around to have interesting stuff where the players are now.

2. And related to that, if you play it right, some "boring" is appropriate and useful. You don't have to spend tons of play time dealing with the boring stuff, but an uninhabited hex or nothing but a village of peasants that do nothing interesting mixed in with the dangers makes the world seem more real.  The players are supposed to be pushing or backing off as needed and to fit what they want.  Sometimes a place to retreat and rest if helpful, especially in a system that allows characters to get meaningfully depleted or resources and health in ways that take time to regain.  As soon as the players want more excitement, they are free to go find it.

Thanks for the tips folks!

It might be my lack of D&D knowledge showing here*, but why are so many of you talking about hexes?? I didn't think that sandbox=hex crawl?

(*I've played mostly Rolemaster, CoC, and Runequest/BRP-derived games)

Quote from: Trond on October 18, 2023, 10:24:28 PM
Thanks for the tips folks!

It might be my lack of D&D knowledge showing here*, but why are so many of you talking about hexes?? I didn't think that sandbox=hex crawl?

(*I've played mostly Rolemaster, CoC, and Runequest/BRP-derived games)

A sandbox doesn't have to be a hex crawl but a hex crawl is the standard sandbox structure for D&D.  Whether you use a hex map or not, you need to have a map of your sandbox and then put the toys in.  I can be as abstract as you want as long as it's a consistent layout that players can navigate.  You can easily use a city map and do all the same things that are mentioned here.  Hell, you can make it a social network map with each junction a high strung social thriller event.  It's just that advice you get is most likely going to be about hex crawling because that's how most of us started.

However, get and read some of the old hex crawl modules and read them.  They are excellent text books to learn how to run a sandbox. 

One other book for sand box play that's not a hex crawl that I've personally run successfully is Chrome Berets for Cyberpunk 2020.  Set aside the setting and look at it as a structure for sandbox play in contrast to traditional hex crawl modules for D&D.

Quote from: BadApple on October 18, 2023, 10:58:25 PM

Quote from: Trond on October 18, 2023, 10:24:28 PM
Thanks for the tips folks!

It might be my lack of D&D knowledge showing here*, but why are so many of you talking about hexes?? I didn't think that sandbox=hex crawl?

(*I've played mostly Rolemaster, CoC, and Runequest/BRP-derived games)

A sandbox doesn't have to be a hex crawl but a hex crawl is the standard sandbox structure for D&D.

Yeah, I had to mentally stop myself from thinking hexcrawl because a hexcrawl is probably the easiest format for a sandbox, and it springs to mind right away when talking about sandbox play.

Which brings up an interesting topic you touch on. How to sandbox without a hexcrawl.

The Holy Grail, IMO, of a sandbox is self-directed play. When player characters decide to do stuff because it interestes and involves them, instead of some NPC telling them to care about saving the world so go here and do this.

I like to write a short "newspaper" with articles that give some fun places to visit or mysteries to investigate. A sandbox campaign doesn't have to start from ground zero, but if I have a number of possible plot threads I like to put some ideas and clues down on paper for the players to ponder.

A friend of mine has a "job board" in town where there are some possibilities listed.

In both cases the players can choose what to do, but there are some ideas for them that represent things that might have been prepped.

Quote from: Ratman_tf on October 19, 2023, 03:40:13 AM
Which brings up an interesting topic you touch on. How to sandbox without a hexcrawl.

I don't really understand the question, but then I don't understand why "hexcrawl" and "sandbox" get conflated in the first place. I can't speak for how "most of us" started, but I can say with certainty that no one I've played with ever used or heard the word "hexcrawl" until the 21st century, some 30 years after we started playing.

To me, "hexcrawl" is the wilderness equivalent of a "dungeon crawl". It's a crawl -- it's right there in the name. It's an adventure, not just another name for a map.

As to how to do a sandbox without a "hexcrawl"? Just make a map (or don't even make a map, if you know the area in your head well enough). It doesn't need to have hexes on it, and it doesn't need to be a "crawl", and it doesn't need to be wilderness adventure of any kind. The sandbox contains adventure locales and non-adventure locales too.

Hex crawling is generally done in the true wilderness, when the players want to explore what lies beyond civilization. A sandbox is simply an area for play that is living and moving that the players can interact with in numerous ways. You can have hexcrawl sandboxes but they do not have to be. A sandbox can be set anywhere that you wish. A village, a town, or even a city can be the central focus of a sandbox. Once you have your sandbox area picked out, fill it with things of interest that the players can involve themselves in. Always remember that the place keeps moving and living no matter what the players decide. I like to sketch out a rough timeline of what will take place at certain times barring player interference. I like to sprinkle small adventure sites around the area rather than one big dungeon that is the only thing in the area.

Quote from: Trond on October 18, 2023, 07:49:51 PM
How did/do you run your most successful sandbox campaigns?
Do you use a setting that you know light the back of your hand? Improvise large chunks of the world? Jump in with minimal prep time?

The best advice I can give to anyone running a sandbox campaign (other than trying out 1:1 time with patron play because that's really fun and surprisingly easy to manage) is to actually not start it as a sandbox. Start in a small area with a tight narrative. A single street of the city, a small village, that kind of thing. Give them a clear cut object that establishes the world. Apprehend a dangerous wizard, find the missing jewels, etc. Funnel them down the first couple of leads in this small and controlled environment where their choices are limited and they still believe you're dragging them through a heckin epic story.

The entire time you're still asking them what they want to do. And making sure whatever they pick gets them through what you've made. Then start gradually opening things up as the plot thickens. The wizard was working for a shady guild of demon worshipers, the jewels belong to a notable merchant. Where do you go next? Intriguing choices are made and resolved. Then you open things up a lot more suddenly they have three different objectives, one in a nearby cave, another in a lord's manor, and one more in a different neighborhood.

By the time they've completed a couple objectives, they start to really make decisions for themselves. By about the tenth session you've resolved the plot thread you started in the first session in such a way that you open things up even more with a map and several plot threads to pursue. It's the party's choice on what to do. They're now playing a sandbox campaign with the confidence to make decisions.

Quote from: Trond on October 18, 2023, 10:24:28 PM
Thanks for the tips folks!

It might be my lack of D&D knowledge showing here*, but why are so many of you talking about hexes?? I didn't think that sandbox=hex crawl?

(*I've played mostly Rolemaster, CoC, and Runequest/BRP-derived games)

The original RPG sandboxes were hexcrawls IIRC (from  Wilderness Adventure)... but TBH I would consider using a pointcrawl instead.

A hexcrawl means exploring a territory that is divided by hexes, with no clear paths, and a pointcrawl means exploring a territory through preexisting paths and points of interest.

I'm convinced that pointcrawls are more useful and hexcrawls are only good for an specific (but very popular) type of campaign: one in which the PCs go exploring the unknown wilderness beyond civilization.

Pointcrawls, on the other hand, are useful if you're dealing with roads, cities, caravans, or even when going though the wilderness with a guide; if the guide knows a place, it knows a good path to this place. A dungeon, with rooms and corridors, resembles a pointcrawl.

One thing about pointcrawls is that you should focus on paths, instead of only points of interest. How does the Old Road look like, and what kind of creatures use it, how long does it take to travel it? Etc.

https://methodsetmadness.blogspot.com/2021/11/hexcrawl-x-pointcrawl-when-to-use-them.html

Here is another technique I use: write open questions for me and my players about the consequences of their actions or loose thread in general.

E.g.:

- How will people react to the death of the mayor?
- Will the enemies you left in town X come search for you?
- Etc.

EDIT: one more - have a big, evil faction as an end boss. Let the PCs see some sings early on.