D&D General How do you sandbox ?

[iserith]

Yeah, it’s different strokes and all that. I do an initial burst of prep of the starting locations, but couldn’t imagine fully prepping everything else ahead of time!

I guess it comes down to scale and how much you outsource to random generation. My current hex crawl campaign has only 20 keyed location in 150 hexes. I believe this will be about twenty 4-hour sessions for the whole campaign. A lot of the content will come from roll tables which are pretty easy to prep in my experience. WotC even does some of that work for us in D&D 5e (XGtE).

 

[transmission89]

I guess it comes down to scale and how much you outsource to random generation. My current hex crawl campaign has only 20 keyed location in 150 hexes. I believe this will be about twenty 4-hour sessions for the whole campaign. A lot of the content will come from roll tables which are pretty easy to prep in my experience. WotC even does some of that work for us in D&D 5e (XGtE).

Oh yeah, I use plenty of tables. But I also like to use those as starting points. Let’s see, for this hex, I’ve rolled a ruined village, an abandoned mine and a pack of ghouls. What happened here? Let’s build something cool. So it always takes a bit longer.

 

[payn]

The Traveller map is pretty great for sci-fi sandboxes. I dont know how many star systems are on the map, but its a butt load. 9/10 systems have basic wiki entries that just indicate what type of planets, people, and resources can be found in each system. It gives the GM and players a nice foundation to build on.

 

[kenada]

Since campaign length been mentioned a few times, I wanted to add that there my sandbox game doesn’t have a planned length. The core conceit (members of an expedition) allows us to have a cast of characters, so the active adventuring party can change based on who is available to play that session as well as if anyone is bored of their character and wants to try something new. This also helped facilitate a system switch from PF2 to OSE because the old characters were able to step back into leadership roles.

 

[iserith]

Since campaign length been mentioned a few times, I wanted to add that there my sandbox game doesn’t have a planned length. The core conceit (members of an expedition) allows us to have a cast of characters, so the active adventuring party can change based on who is available to play that session as well as if anyone is bored of their character and wants to try something new. This also helped facilitate a system switch from PF2 to OSE because the old characters were able to step back into leadership roles.

I also have a pool of players who each have multiple characters. So there's a lot of switching that goes on.

 

[Xetheral]

I guess it comes down to scale and how much you outsource to random generation. My current hex crawl campaign has only 20 keyed location in 150 hexes. I believe this will be about twenty 4-hour sessions for the whole campaign. A lot of the content will come from roll tables which are pretty easy to prep in my experience. WotC even does some of that work for us in D&D 5e (XGtE).

Scale definitely makes a difference, although I think it's complicated by the resolution of the preparation. From my standpoint, just-in-time creation lets a DM run a much (geographically) larger sandbox without the prohibitive levels of advanced prep required to create sufficient detail to run the game in such a large sandbox. Instead, the early prep can be all high-level, with the time spent making details on the fly limited to only those areas that ultimately show up in the campaign.

For illustration/comparison, the last sandbox campaign I prepared covered a substantial portion of a continent. There were dozens of major items of interest (e.g. cities, settlements, named geographical features) on the full map that the characters had (although that map was IC known to be outdated) and dozens more in my notes. Given that the areas of interest were days or weeks of travel from each other, each would have been surrounded by its own set of hundreds of hexes (at one-hex-per-day scale).

But my total up-front prep was probably less involved than what you're describing, as each point of interest only had a couple sentences in my notes, or a bulleted list. If/when the PCs actually decided to head for any of these points, that's when the just-in-time creation would take place, expanding my high-level notes into low-level detail. Based on previous campaigns, only a small fraction of the high-level areas of interest would actually be visited, so this method avoids wasting low-level prep time on areas that are never seen up-close.

It sounds like in your game, a much higher percentage of the original map will likely be visited, so maybe the need to avoid wasted prep is less pressing? What fraction of the map do your PCs typically visit?

Also, I never use random tables for content generation, so just-in-time creation lets me limit myself to what is in a given location now instead of spending even more time in advance having to figure out a schedule of what would be in that location at any arbitrary time.

 

[Scott Christian]

I know I differ from many about the definition of sandbox, but in my mind, it is literally just a series of adventures on a single map.

An example would be something like this:

You write everything out for the civilizations that exist. Have that thoroughly planned and plotted and remembered. Then, just like an adventure board in a video game, have a bunch of "adventures" tied to the map. They might be tied to people in the city, rural people, specific locales, random encounter tables, items, etc.
Then, in true sandbox fashion, the PC's decide where to go. Who to talk to. And what they encounter is a direct result of their actions and/or decisions. There is no large overarching plot, and more important, specific motive driving the story. There might be several locales tied together to complete one of the adventures. But, it is not a pathway adventure where everything is written out with several different ending and the absolute endgame of the players.

 

[aco175]

I like the idea of starting with NPC goals and motivations. Not sure how many, may depend on how long the campaign is lasting. Prep phase may only need 4-5 NPC bad guys and a few NPC good guys. Maybe loosely outline their goals such as goblin chief wants X, and giant living in the mountain wants Y, but goblin chief may contact the giant after a while to make Z happen. The PCs can have the option of stopping the goblin threat at lower levels, but they will develop to a larger threat if the giant becomes involved after level 5ish. The whole concept of the world going on without the PCs.

Good people have their own designs as well. Saruman started off with good intentions and aided the PCs for a while before being being corrupted to evil by the Eye.

 

[iserith]

Scale definitely makes a difference, although I think it's complicated by the resolution of the preparation. From my standpoint, just-in-time creation lets a DM run a much (geographically) larger sandbox without the prohibitive levels of advanced prep required to create sufficient detail to run the game in such a large sandbox. Instead, the early prep can be all high-level, with the time spent making details on the fly limited to only those areas that ultimately show up in the campaign.

For illustration/comparison, the last sandbox campaign I prepared covered a substantial portion of a continent. There were dozens of major items of interest (e.g. cities, settlements, named geographical features) on the full map that the characters had (although that map was IC known to be outdated) and dozens more in my notes. Given that the areas of interest were days or weeks of travel from each other, each would have been surrounded by its own set of hundreds of hexes (at one-hex-per-day scale).

But my total up-front prep was probably less involved than what you're describing, as each point of interest only had a couple sentences in my notes, or a bulleted list. If/when the PCs actually decided to head for any of these points, that's when the just-in-time creation would take place, expanding my high-level notes into low-level detail. Based on previous campaigns, only a small fraction of the high-level areas of interest would actually be visited, so this method avoids wasting low-level prep time on areas that are never seen up-close.

It sounds like in your game, a much higher percentage of the original map will likely be visited, so maybe the need to avoid wasted prep is less pressing? What fraction of the map do your PCs typically visit?

Also, I never use random tables for content generation, so just-in-time creation lets me limit myself to what is in a given location now instead of spending even more time in advance having to figure out a schedule of what would be in that location at any arbitrary time.

My hexcrawls are smaller in scale and the PCs visit every hex or nearly so before we're done, often backtracking or retracing steps more than once. I prefer smaller, fully-prepped hex crawls with 20 to 30 sessions of content rather than larger ones where I'm having to keep up with the players every week or two weeks.

 

[nevin]

Ideally you define the starting area well, if youve git tume the big stories going on across the world, to guve you a bigger structure. Then turn the players loose and throw stuff at them till they start getting defined plans, enemies etc. Then they'll be driving and youll be filling in the story.