OK, now to answer
@Hussar 's question as to how I'd prepare my next sandbox campaign. In very short, probably much like I prepared for the campaign I'm running now......
Long essay warning.
Step 1: initial inspiration. This can come from anywhere but in my case it's usually a map, or some scribble that looks like it maybe could be a map, and I start thinking "What could be there, and there, and there? What would this look like? What type of campaign might this result in?"
Some coke once got splashed across a piece of paper, I found it after it had dried. Just for kicks I traced around the edges of the dried coke and came up with what I thought was one hell of a good map (coke was land, blank page was sea) that could have become an excellent maritime-based campaign. I spent the rest of the afternoon imagining what that silly map could become, but didn't go any further with it as I already had a campaing on full boil and didn't need another.
My current campaign was seeded in almost as random a scribble: a party in my previous campaign had somehow got themselves stuck on a different world and I needed to bang out the very basics of this world in a hurry. I quickly sketched out a rough map, threw a few cities on it, decided it'd be a faux-Greek culture, and within 15 minutes I'd (unknowingly at the time) sown the seeds of a 17+-year campaign. It would be two years until I'd start fleshing it out any further than needed for that party's fairly brief visit.
Step 2: the hard and sometimes tedious stuff. For me, if I'm doing a new setting from scratch, there's a series of steps in no particular order other than forced (e.g. can't determine climate until I know the basic geography):
--- flesh out whatever map or other initial inspiration got this started to the point where the local-to-starting-point region is done to a level of detail that the PCs would reasonably know about. This takes ages if a) like me, you don't have mapping software and b) you want to do it halfway right
--- make sure this map has a wide variety of "interest levels" in its naming including numerous obvious sites for potential adventuring; also make sure the map includes all kinds of different sub-settings such that in case adventuring takes them to a jungle or mountains or a desert or wherever, the map has that built in
--- also make sure the map has some less-known areas on it for later use as ideas arise
--- do a few sub-maps of what are likely to be important areas to start
--- copy all these maps then augment them with DM-only info for possible later discovery by the players/PCs
--- build the pantheons
--- design the world's astronomy, and from that, its calendar
--- determine what types of cultures (be they faux-historical or completely made-up) exist in the starting area, including those of fantastic creatures and monsters, and have a vague idea of what else is out there; after this, figure out languages and their relative commonality
--- work out an actual world and local history enough to explain how-why things are as they are; then write out a player-visible version to the degree the PCs would reasonably know, if some of that history is to be unknown to begin with
--- determine the basics of climate, weather, etc. for the starting area
--- anything else that leaps to mind
Step 3: rules review. For my own game, as we use a 95+% homebrew rules system anyway, I'd look at what worked and didn't work rules-wise in the previous campaign(s) and tweak-adjust-kitbash to try to fix what didn't work. If I was using a more standard system this is where I'd go through and kitbash it, including decisions on what would and wouldn't be allowed for classes, species, feats, abilities, splatbooks, etc. etc. and including any significant decisions re optional rules and-or homebrewed changes.
Step 4: storyboarding. "But wait," you say, "a sandbox can't have a storyboard!" Well, please hear me out.
Even if the campaign is to otherwise be a full sandbox, I-as-DM will still predetermine (or, if you like, railroad) two things: 1) where they start, and 2) what the first adventure will be. I do this in order to get things started and the ball rolling, pure and simple. What they do with that first adventure wil be entirely up to them, but I'll somehow make sure that adventure is what they do first (and I've never had any problem getting buy-in for this).
After that, here's where the storyboarding comes in, in three ways:
1 - regardless of what they decide to do or not do once play begins, I need to know what's going to happen when in the setting if no interruptions occur to the flow of events. This includes the actions of factions or nations, the actions of specific NPCs, sometimes the actions of geology, and so forth.
2 - have some developing stories going on in the background such that if-when the players find themselves at a loose end I've got hooks to dangle
3 - for each of the potential adventuring sites on the map, have an idea of what canned module or homebrew idea that site represents e.g. this island is Isle of Dread, this swamp holds Tomb of the Lizard King, I'll write something myself if they go to these ruins, etc. etc.
Step 5: recruit players. Now that I've got a setting, rules system, and campaign premise ("anything-goes sandbox after the first adventure") I can pitch, and know what character options are available to play, it's time to find some players. Once I've got four or five players and a set night of the week to play on, we'll get together for roll-up night and drop the puck.
Step 6: ongoing development. Just because play has started doesn't mean steps 2 and 3 are finished. Setting expansion is often necessary when the PCs decide "where the map is blank, we'll go", and rulings made on the fly become incorporated into the rules system going forward.
Hussar, I hope that's what you were asking for.
