I’ve run Hoard of the Dragon Queen and Rise of Tiamat three times each, Curse of Strahd twice, Out of the Abyss and Princes of the Apocalypse once each, and I am currently running Storm King’s Thunder and Tomb of Annihilation. By now, I’ve had a lot of experience running the campaign adventures of Wizards of the Coast.

Many people evaluate adventures based on how good the encounters are. I don’t. My main interest is in the story structure. I want to know if the encounters fit together and provide an environment for the players to explore. This is not to say that encounters aren’t important – they certainly are – but I feel the art of making a good encounter is far better understood than the art of creating a good story structure.

It is tremendously important not to expect a published campaign adventure to function as a homebrew campaign. They don’t. In a homebrew campaign, the DM creates material in response to the actions of the players and their desires. The published adventure needs to include that material beforehand; the DM can add additional material of their own, but I presume that one of the reasons a DM might use a published adventure is so they don’t have to keep creating material. I know it’s a reason I run so many.

In general, there are two basic types of campaign: The linear adventure (such as in Hoard of the Dragon Queen) and the sandbox adventure (such as Curse of Strahd). Both forms present challenges of design.

A linear structure is not a sin. It’s a technique of adventure design. Each encounter area follows the next is a line. This can be frustrating for some players, but many players prefer an adventure with a linear structure and well-defined goals. Being able to go wherever you want isn’t always a good thing. One advantage that the linear adventure has is that the characters get to the encounter areas in the right order. You don’t suddenly find yourself in the lair of Orcus at first level.

The drawback of the linear structure is that it can feel too forced (the dreaded railroad). One technique for combatting that is to include a lot of choice in the encounter areas. Many of the individual episodes in Hoard of the Dragon Queen give the players a lot of choice how to approach them. Do they go for a frontal assault, sneak in, or try to persuade the guards that they’re fellow cultists? The links to the next part of the adventure are set, but there’s much decision-making to make in-between.

The sandbox structure gives the player freedom of choice but courts the risk of the adventurers going into areas that are too difficult for them, or that don’t continue the story. This is very apparent in Princes of the Apocalypse, and it’s made worse by the goals of the story pushing the PCs to confront monsters too tough for them.

There is a basic question you need to ask about any encounter area: What reasons do the PCs have for going there? Curse of Strahd looks like a wide-open sandbox, but story-based reasons tend to push the PCs towards areas that work for them. These new goals are missing in Princes of the Apocalypse.

If the rationale works each time, and there is at least one path from the beginning to the end of the adventure, then you have a successful structure. When the reasons to go to a new area fail, then you have a problem.

My feeling is that in a multilevel adventure, sandboxes work only for a limited range of levels. It’s why Princes of the Apocalypse has trouble and Tomb of Annihilation is more successful. In Tomb, you’re only in sandbox mode until level 5 or so. With limited levels, the range of challenges is manageable. A level 1-15 sandbox? That’s difficult!

So, what makes a good sandbox adventure? I think it needs lots of plot-hooks for side-quests, multiple hooks for the main quest, and different paths from one story encounter area to the next. When a hook for area D appears in areas A, B and C, then you’re getting somewhere.

A major problem for published sandbox adventures is that good sandbox settings aren’t static. The DM constantly alters them and updates them in reaction to the actions of the PCs. What actions does the villain take in reaction to the PCs? Discussion of that is something missing from many sandbox adventures. Consider the factions in Tomb of Annihilation. There are many of them, and the DM can use them to take the story in interesting directions. How much does the adventure aid the DM in planning their actions? I don’t think there’s much there.

For all the problems I have with Princes of the Apocalypse, this is one area it does cover: the reactions of the villains to the adventurers’ actions. This is also something you’ll find in The Rise of Tiamat, where a chapter details the reactions of the cult to the characters. (They want them dead!)

So, how do I approach a published adventure? The first thing I do is work out what encounters are essential to the plot. Then I start ignoring everything else!

That’s an exaggeration. However, I think it’s important to not include a LOT of other encounters between important story encounters. If you have three sessions between advancement in the plot, it’s likely the players will forget what the plot was! In Tomb of Annihilation, I replaced the entire “wandering around the jungle” introductory section with an adventure by Shawn Merwin (Return of the Lizard King) that very quickly got the players to a level where they could explore Omu: the real heart of the adventure.

Try to discover why the players might go to each area, and how the areas link together. When those links are absent, and the players are wandering randomly, it’s likely you need to step in and create clues and goals for the players. Whether there’s just one or several, most players appreciate having a sense of where to go next!

For a critique of the structure of the published adventures, have a look at this article by DM David. I don’t agree with all of his points, but it’s worth reading.