D&D 5E Hot to handle level progression in Princes of the Apocalypse (or sandbox adventures in general)

[Li Shenron]

The problem (for me) with a sandbox that doesn't adapt to the party level is that it becomes as much a railroad as the adventures that they proponents claim to hate, it's just that the party have to self-railroad or die horribly.

The whole choice thing becomes a thinly veiled fiction, the characters may become aware of things that they can't handle but they still have to wait until they can handle it before they can engage with it and hope that they don't attract the attention of a big bad too early.

I think this can happen only if the sandbox is too randomized in challenge levels (i.e. widely different encounter difficulties in the same general area), and the rules system used is such that different challenge levels quickly diverge.

For example, I found World of Warcraft being well-randomized but horribly diverging: you have areas where most encounters are of similar challenge level, but then an encounter of just a few challenge levels difference is either trivial or impossible. You just can't go to another area until you're ready.

The randomization is totally up to the DM (or author). So if you design a sandbox where in the same forest you have lots of trivial encounter but also some epic monsters, this can cause problems. OTOH, the system is very important in how much freedom it allows, and 5e has been said to work better than previous 2 editions in terms of "flattening" the difficulty of encounters a few levels apart. This means that the players have more freedom to choose what to do / where to go in the sandbox, as the risks of death (or boredom, in the opposite case) is lessened.

Yeah, sandboxes generally allow parties to face disproportionate threats. If the party investigates something out of their depth, I would recommend a combination approach to keeping them alive. Play the encounters as written but:

1) Offer appropriate clues as to the danger. "These foes seem particularly hardbitten, fierce, magical, etc." "You feel the intense waves of heat coming from that tunnel."
2) Be a little lenient with interpretation if the PCs are in over their head. Ok, so the PCs are caught in the cage trap. Maybe they have a little time to get out before the Bringers of Woe show up.
3) The cultists could easily have reasons for wanting to keep PCs alive, at least in the short term. They might hope to sacrifice or interrogate the PCs at a later occasion. Interrogation makes an excellent opportunity to justify reprisals against the PCs' party or to capitalize on PC weaknesses later on. IF the PCs are kept alive, there might well be opportunity for escape or rescue. Keep in mind that the area is filled with monsters and rival cultists that are not necessarily friendly to potential captors. It would be relatively simple to stage an ambush against a PC's captors that might allow them to escape.
4) Be SPARING with all of the above. If the DM backs down from every potential threat or character death then the sense of danger quickly fades away. PCs death is not infrequently more painful for the DM than the players.

Add:

5) Remind the players that encounters don't have to be killed-or-be-killed. It is not only perfectly valid but also narratively very realistic, that when facing something too hard the PCs decide it's better to withdraw. IMO the main problem of 'hard' encounters (or why not even 'impossible' sometimes!) is the stubborn refusal of many players to consider alternatives to a killed-or-be-killed strategy.

And then as I mentioned above:

6) Harmonize the sandbox CR randomness at least a little bit. IMO it feels narratively sound that the world has areas of greater danger, where creatures of higher powers tend to cluster just because lesser creatures don't survive. You don't go hiking in the Himalayas until you think you're ready! When a higher-CR creature is placed in a lower-CR area, make it somehow 'enclosed' (for example, in a dungeon/prison/cave, behind a spell etc.), which would also explain why animals and people haven't already fled.

 

[Greg Benage]

Regarding level advancement in PotA and the most open forms of sandbox campaigns, I prefer session-based leveling. Most any other mechanism -- monster xp, quest xp, milestones, gold, etc. -- will incent the players to focus on certain activities (e.g. killing monsters, completing quests, hitting the next plot point, looting treasure). Session-based advancement means they're rewarded with the same level progression no matter what they do, and that's kind of the point of a sandbox campaign. It also makes it easy for the DM to plan: "I want this 1-20 campaign to run about a year of weekly sessions, so we'll average 2.5 sessions per level."

(Yes, this will get weird if the PCs sit around in the bar drinking ale and playing poker for several sessions. I've never had that problem, because I've found that most players play D&D to go on adventures.)

Finally, for me as a player, the possibility of level-disproportionate challenges is a big part of the appeal of a sandbox campaign, because this trait prioritizes some of my favorite things: research/lore, planning, recon/scouting. Perhaps even to the point of subordinating the combat pillar to the exploration pillar. Love it.