Have tips, key considerations, or methods to help you devise wilderness, dungeons, or adventures? Please share and discuss! 
As for me:
I. Know your players' interests and their PCs' desires. If not, give them some by starting with assumed PC desires like their lives or homes.
(Ex: Players like pet monsters? Put the PCs in charge of finding and taming beasts. Don't know much about them? The game's world will explode if they don’t find the crystals). Just a simple excuse to play a game.
II. Give them situations where they must weigh their desires against each other. It thus follows that risk and reward are always coupled.
(Ex: I like the dragon's treasure, but I also like staying alive, but if I don't steal the treasure, the bandits will kill my family, but if the dragon gets upset, we could all be immolated anyway.)
Choice and conflict (applied to video games, but assuredly applies here, too)
Give some direction (leads, motivations), especially when your dungeon has NO walls.
Give choice, especially when your dungeon HAS walls.
III. What, in-game, provides this conflictive choice to the PCs? Well, what situations and subject matter excite you and the players?
You can build an area from different perspectives:
Area by scenario (best when you have ideas but no map):
- Devise about 7 area situations that require player consideration. Stop when you hit a wall.
- Arrange them in a location web that makes travelers think of the risk and reward of traveling different ways through it. Stop when you hit a wall.
- See if there are areas that need filling in, thus repeating steps 1-3 until it’s filled.
Area by interaction (best when you have a map but no ideas):
- What lives here: locate notable resources, creatures they've attracted, and resultant settlements/lairs. (I like to do this like an ad-libbed How to Host a Dungeon, but applied to wilderness too)
- Situations among these creatures
- After making a few areas, what are the situations among these areas?
How to set up campaign paradigms; how to make the way the game works encourage activity you want to see:
West Marches
I've been GMing for over 30 years now and that has made me pretty set in my ways with what makes a good adventure, but they are just refined ideas based on my personal preference/experience and are not for everyone. I also improvise quite a bit, which makes a big difference in how I design adventures - so take all of this with a grain of salt:
I usually design adventures as a series of modular encounters, my compromise between railroads and sandboxes. Railroads typically have better stories but bug experienced players with the lack of choices, and sandboxes give plenty of choices but often lack direction in the story development. With modular encounters there is often an 'expected path' that the players will take: A - B - C - D, but by making the encounters modular the players can make their own choices about how they want to proceed and it is easy for the GM to adjust on the fly. Usually they are real choices but sometimes it's the illusion of choice, which is often good enough if used sparingly. Another advantage of modular encounters is that you don't lose your prep, if it is skipped early it can often be changed and re-purposed later. However, I prep as little as possible - one of the biggest mistakes new GMs make when writing their own adventures is overpreparing. A few notes and some stat blocks are typically all I need. I also avoid creating stories involving the players - instead I create a situation that the characters are going to interact with and influence, where no one knows for certain how things will turn out. The best way to do this is to work out the story of what would happen if the PCs weren't around and proceed in that direction until the characters alter things with their actions.
As far as encouters go, here is the general pattern I try to follow when designing individual encounters of all types:
1.) I make sure the encounter clearly presents a challenge of some sort that needs to be overcome,
2.) I make certain that the characters care about the problem by tying it in to their player and/or character motivations,
3.) I give the players interesting and meaningful choices to make in order to solve the problem,
4.) I show the players the consequences - both positive and negative - of their choices, and
5.) I keep things simple - the players will complicate it on their own!
I'm also a big proponent of collaborative world building (of a small starting area) and collaborative character creation to kick off new campaigns. GMs usually over-design their worlds even more than they over-design adventures. A collaborative start takes much of the initial burden of world creation off of the GM, it makes the players care much more about the campaign world, it gives them stronger bonds to it and to each other, and it just makes for a much stronger foundation to the campaign overall. You could be the most creative GM in the world, but a player is still going to care more about a simple idea that they came up with themselves than anything you have to try to explain to them and get them excited about.
*crickets*
That's too bad this topic didn't explode - adventure creation is one of my all time favorite things to read about and there are so many ways that different GMs approach it.
Not going to go through a whole list of DMing tips and strategies, but one of the most useful activities I've found is to play test the content I intend to subject my players to. Looking at a tactical map and crunching some numbers isn't going to reveal the many surprises an inventive party can bring to the table.
I also try to give the players something to play with. Say the player is attacked by a wizard or magical trap. They could be plain zapped, but what if they were instead turned into a house cat? Tho' it still inconveniences them in some ways, they may also find it offers opportunities: they can now fit into spaces they couldn't check before, or communicate with some of the lesser dungeon denizens who speak cat but not human. Unless it is temporary, better find a way to turn back before you need to be human again. Of course, if you want to be comical, maybe the shock trap electrifies metal armor, allowing the character to charge some forgotten machinery or golem they found earlier inside the dungeon.
Say you made a calendar in your setting, and you want the players to care. Perhaps it was prophecied dragons would emerge in great numbers on a certain month in the future, and the PCs should mark their calndars as the campaign wears on. Perhaps every first of the month, the goddess of seasons blesses those who come to certain holy spots.
I'd rather show than tell: Barrow of the Evensong.
That's a simple little location that showcases certain aspects of design considerations that I use.
I have quite a few, but here are a couple I'll share here:
1. Build an adventure that I think is fun and would like to play. If I don't think it would be fun, I can't expect anyone else to think it's fun.
2. Build in a little mystery so there are plot points to discover. I don't mean just having a dungeon layout that players discover as they adventure, but plot elements that are revealed a little at a time by things they discover as they progress.
3. While designing, create details about the world including sights, sounds, temperatures, textures and even smells to help bring the world to life as you describe it.
However more there are players, less choices must be available, at least those affecting the party, because it raises more conflicts and then most of the time goes to argue, halting the adventure and many players getting bored. Sandbox is single player game, so if there are more players, there cannot be a sandbox anymore. With many players, railroading is better, because the story goes quicker and more smoothly.
I have a huge sandbox for my group, and they love it! Railroading isn't an option for me, but as I've buildt my own world, I can move things around, as long as the players haven't been to the area before.
The way I do it, is I build the framework first, as I've done with my world (see link in signature), then I let the players explore an area. So far, they are content with being within range of the refugee camp where they started, as they have no horses or any means of transportation, and they are low level, so they know (or rather think they know) that roaming monsters will have them for supper. I'm expanding the world like a bubble, so when the players decide that their PC's will go further, or in another direction, I wing it the rest of the session, and then go home to write out the area they're headed into.
This is how I've always done it, and after a while, the world sort of runs itself. It does however mean I have to be proactive, and always one step ahead of the players.
PetardoMauriti wrote: