The fifth edition of Dungeons & Dragons describes the core elements of the game as "exploration, interaction, and combat". In 4th edition, we spent a lot of time building clear "encounters" tuned in design and power to challenge our party of characters. Often the adventure revolved around these clearly defined battle arenas.
In 5e, combat can be less clearly defined. It can flow into the rest of the story of the game. It can spontanously erupt in a lot of different scenes, whether its random encounters or a negotiation gone bad. This gives us a lot of flexibility in how we build our games. We don't have to build clear definitions around combat, exploration, or interaction scenes. We can just build situations.
In fact, there are a lot of advantages to building scenes this way.
Tom Lommell, the Dungeon Bastard, discussed this topic in one of his excellent dis-organized play videos, referring to it as "Prepping Outcomes, Not Encounters". We can let go of our definitions for the types of scenes and define sitiuations and outcomes intead of clear combat encounters.
Here's an example.
In the adventure Legacy of the Crystal Shard, our heroes have an opportunity to deal with a pirate ship from Luskin called the Howling Fiend. In our old way of designing adventures, we might set up a series of well-balanced encounters where the characters fight groups of pirates, pirate veterans, and the pirate captain along with her personal guard. We might have a scene where the characters meet with some drunken pirates who aren't happy with this whole situation in Icewind Dale and might be willing to negotiate. The characters might also explore some nearby caves that the villainous wizard Vaelish Gant had used to hold his hostage—the speaker of Bryn Shander.
There's another way we can handle this design though. Instead of breaking it down by scene and encounter, we just build the situation. The Howling Fiend is docked in a huge glacial overhang along Lac Dinneshere. There's a crack in the glacier overhang that leads up to the surface and a guard posted there who can collapse the ice above and seal off this entryway if need be. There's some caves carved out of the ice that Vaelish Gant has used for his own personal chambers when he's not busy pulling some politics at Bryn Shander. He also has the speaker of Bryn Shander tied up there. There are a bunch of ramshackle huts built on an iceflow down by the water's edge and a winding path to an overlook near the crack that leads up to the surface.
Our pirates include two dozen bandits, a bandit captain, a gladiator, and a mage. The captain and her barbarian gladiator bodyguard often oversee repairs to the ship and shipments being moved from the iceflow to the ship. At any given time there are a dozen pirates on the iceflow, another half dozen up along the overlook and the guard post, and a half dozen sleeping one off either on the ship or in the huts.
We can add in some nice details like a teleportation circle used by Gant in his cave, a frozen body in the ice with its hand out that may not be as dead as it looks, a store of booty aboard the ship, some explosives the pirates planned to use to cause trouble to Ten Towns, and some other fun things.
There are no defined "encounters" for this situation. There is only the situation. It's up to the players to choose how their characters deal with this situation. They might use stealth to sneak in and assassinate as many pirates as they can. They might negotiate their way in disguised as pirates or barbarian mercenaries or something. They might just decide to go in with swords drawn and hope for the best. They might sneak in through the crack in the glacial wall or come in on a boat painted black. Who knows!
Rather than defining the scene, we DMs can instead develop situations with lots of options, lots of hooks, and lots of interesting things the characters can interact with. The video game Dishonored did a great job building scenes like this. In Dishonored you had lots of options in any given scene. You could fight your way in, you could mind-control one of the guards, you could stealth around, or you could turn into a rat and go through the sewers. The scenes weren't built expecting any given course of action. Some courses might be a real pain and others might be the easier path but all of them were options.
In general, scenes should have a few different ways the characters can interact with it. You nerds out there might consider this the scene's API—the interface between the world and the characters. We have a few big interfaces between characters and the world: skills, combat abilities, backgrounds, and spells. We don't have to build environments with specific hooks for all of these but it helps us keep a few in mind so we know the scene won't be a complete dud. If we build a scene that looks like a set for American Gladiator but none of the characters are particularly athletic or acrobatic, that might not be a lot of fun. Likewise, if we take our group of five barbarians with an average charisma of 8 into the wedding reception of the duke's daughter, that might not be a lot of fun for folks.
It's better when we design scenes with lots of options and lots of complications. Here's another example:
The wedding of the duke's daugther is rumored to be the site of an assassination attempt by the Order of the Black Kris. The wedding is taking place in the temple of Oslandia built near the little-known cliff-side ruins of the dark temple of Avrix. All the well-to-do members of the city will be there, including likely the unknown leader of the Black Kris, whose agents will probably be in hiding somewhere within the ruins. The duke himself has hired the Red Fist mercenary company as his bodyguards, the very group that Volund, the fighter of one of our players, left a decade earlier under bad circumstances.
Now we have lots of hooks. There are options for a bunch of skills including history, athletics, acrobatics, performance, persuasion, and intimidation. There's room for combat, both with the Red Fist mercenaries and the Black Kris assassins. There's even a hook to one of the PC's backgrounds.
This is a pretty robust scene, and might even take up an entire adventure. Not all of our scenes need to be this big in scope and detail but any of our larger scenes will benefit from including all of these potential hooks.
Avoiding the Single Roll Failure
When we build a situation like this, we have to be careful not to let a single roll completely ruin the situation or, by default, drop it right into combat. We can see how a bad roll in either the ship infiltration situation or the wedding situation could send the whole scene tumbling towards mass bloodshed. We don't want that to happen, at least not right away. We need to keep the idea of "failing forward" nearby. Just as no single attack roll usually determines the success or failure of an entire fight, no single skill roll should determine the entire outcome for a whole big scene. Instead, it could be a whole pile of skill checks that move the needle one way or another or steer the whole path that the scene takes. When we run scenes like this, we need to keep in mind that we can't leap to pure sucesses or failures and instead let each roll nudge the scene down one path or another. It's harder to do but builds much more rewarding scenes.
Breaking Away From Assumptions
All of this is part of our exercise to break away from our own assumptions about how a game is going to go. D&D is a lot more fun when we can watch scenes unfold in new and interesting ways well beyond what we originally intended. In order for that to happen, however, we need to build environments with all of the right elements to give characters, and their players, the chance to take things in lots of different directions. Give it a try.