The Noble Art of Adventure Design

[Warning: spoilers for the Pathfinder modules Wake of the Watcher and House of the Beast below.]

So, recently, I've been reading through a stack of old Pathfinder adventure paths. The system is far too complex for my tastes, and the modules are always enormously padded, so I do a lot of skimming; but inside some of these baggy sixty-page monsters is a tight, punchy twelve-page adventure screaming to get out. Others, however, really do just feel like 50+ pages of nothing; and this blog post is my attempt to puzzle out both why this is, and why so few people seem to mind.

As a project in RPG adventure writing, the sheer scope of the Pathfinder Adventure Paths has probably never been equalled. One module per month for over a hundred months, all set in the same setting; seventeen-and-a-half campaigns and counting, each one going from level 1 all the way up to level 15 or beyond. It's all in there somewhere: wildernesses and dungeons, sandboxes and railroads, mysteries and murder-fests, ninjas and pirates. From my perspective, they can veer from 'almost usable as written' into 'completely worthless' and back again from one module to the next, but the editors and authors don't seem to recognise the shifts. My surprise is not that they don't happen to share my tastes in adventure design - why should they? - but that they seem to regard modules which, to me, seem so qualitatively different from one another as effectively interchangeable.

Two modules I read recently cast the matter into sharp relief: Wake of the Watcher, and House of the Beast. On the surface, Wake of the Watcher looks like much the more promising of the two: it's a Lovecraftian bonanza set in an Innsmouth-style village, featuring everything from crazy cultists to the Hounds of Tindalos, while House of the Beast is basically just a ruined temple full of gnolls. Wake of the Watcher is ostensibly an investigation, whereas House of the Beast presents itself as a straightforward kill-the-baddies dungeon bash. And yet Wake struck me as much the weaker of the two; so much so that while I'd happily run a scenario based on House (although not using the Pathfinder ruleset), I'm honestly not sure that there was anything in Wake which I could salvage for use in one of my own games. After some reflection, I think it mostly comes down to one factor: their relative levels of dynamism vs. stasis.

This sorceress never does seem to have any luck. Image from Wake of the Watcher.

Wake presents what, on paper, is a highly unstable situation: necromancers, Cthulhu cultists, Deep One knockoffs, Mi-Go brain-stealers, a haunted mansion, and a backwater village that's being deliberately kept in the dark about exactly where all the girls they keep giving up for 'fostering' are ending up. (Like The Hook Mountain Massacre, the module tries to avoid following through on its own premise here: by the time the PCs find the missing girls, they're all dead, so no-one actually needs to engage with the consequences of a lifetime of sexual exploitation.) But instead of throwing it all down in one huge storm of violence and horror and chaos and leaving the PCs to reap the whirlwind as best they can, the module assumes a completely linear approach: first the PCs go to the cult's temple and kill everything, then they go to the mansion and kill everything, then they go into the Deep One lair and kill everything, and finally they go to the Mi-Go base and kill everything, ultimately having a big fight with a big monster in order to get their hands on whatever damn thing it was they were looking for in the first place. The same sense of stasis pervades the individual encounters: most of the monsters just wait in their rooms until the PCs arrive, at which point they attack on sight and fight to the death. The whole thing reminds me of a World of Warcraft dungeon: a series of set-piece encounters positioned along a linear corridor. The locations might be impressive, the visuals might be cool, the fight might be challenging, the backstory might be good... but there's no free will, no room to manoeuvre, no chance to get off the rails. The only meaningful decisions to be made are tactical ones.

House, in contrast, presents what looks like a very stable and boring situation: there is a ruined temple, it's inhabited by gnolls, it's run by the Gnoll King, you need to kill him, the end. But scratch the surface, and there is so much stuff going on! There's been a recent slave revolt: one band of escaped slaves are currently hiding in an outbuilding, while another is barricaded into the tunnels under the temple. The shaman of a nearby troglodyte tribe, allied with the gnolls, has had his identity stolen by an evil shape-changing genie; the trogs have just found the body of their real shaman, and they are furious. The genie has no real loyalty to the gnoll king, and is using his disguise as cover for his attempts to find a secret treasure chamber hidden somewhere beneath the temple. The gnolls are living in fear of some strange whispering creature which is killing them one by one: this is actually a crazy goblin who thinks he's on a mission from God, and who assumes that the PCs have been sent to help him fulfill his sacred duty to kill the gnoll king. Every part of the situation is balanced in a kind of precarious equilibrium, just waiting for some outside influence (i.e. the PCs) to come along and push it all over into chaos.

Here she is again, in House of the Beast, clearly seconds away from yet another wardrobe malfunction...

The same sense of teetering on a tipping-point extends down into the individual encounters. The gnolls have captured an enormous scorpion to use as a source of poison, but they have no way of controlling it; if released it attacks the first thing it sees, and as a result it's obviously a resource waiting to be exploited by clever PCs. The gnolls are using giant, bad-tempered mutants as shock troops, but can only keep them in line by giving them regular sedative injections: I can think of a whole bunch of ways to exploit this just off the top of my head. (Swap the sedatives for poison! Swap the sedatives for stimulants! Pour their entire sedative supply out of the window! Drop the sedatives into the temple's water supply!) Of course, if your PCs are feeling uninspired, then any or all of these can become straightforward combat encounters: kill the goblin, kill the mutants, kill the scorpion, kill the trogs, and so on ad nauseam. But come on: giant angry mutants kept in line with sedative injections? If your PCs can't find a way to exploit that then I'd start to wonder whether they're even trying!

Now, these two ways of designing adventures seem to me to grow out of very different assumptions about what a session's play is supposed to look like. Wake clearly assumes that the point of the game is set-piece combat encounters, in which a group of powerful PCs with a range of special abilities attempt to overcome various groups of powerful enemies with special abilities of their own. Within this playstyle, the kind of dynamism which appears in House is actively undesirable, because it messes up the balance of the encounters: where's the fun in testing your tactical abilities in battle with a bunch of mutants if you've already rigged the outcome by tricking their handlers into accidentally poisoning them? If this is your paradigm, then the task of the adventure designer is simply to present you with a suitable sequence of appropriately challenging fight scenes, along with some pretence of a narrative to string them all together. House, on the other hand, seems to expect PCs to treat the game-world as a world, and to engage with it in a much more fluid and dynamic fashion. Like most OSRians, I prefer the latter; but what I'm slightly bewildered by is the fact that the editors and authors of Pathfinder don't really seem to recognise the difference at all. They sort-of understand the difference between railroads and sandboxes, and write about it in their module introductions, even if they're not very good at actually writing the latter. But they really seem to think that 'here is a place full of monsters, now go and kill them all' is functionally equivalent to 'here is a place full of seven kinds of craziness existing in precarious balance, now go and start poking things and see what happens', and that strikes me as a little odd.

Whenever I come across something in a Pathfinder module that I might like to use, I make a note. 'Ruined castle full of degenerate bird-men, ruled by a demon who fancies himself as a playwright and spends all day forcing them to stage his nonsensical dramas', for example, or 'opium-addled spider wizard seeks ancient knowledge from demon's library in the dungeon below, but can no longer distinguish between memories and hallucinations'. Weird little scenes which can be dropped into adventures with a minimum of fuss in order to create a good, memorable encounter, which could unfold in any number of directions based on the actions of the PCs, but are likely to be colourful and fun no matter what ends up happening. The better modules, like House, give me maybe eight or nine of these. The worse ones, like Wake, give me none. I can write my own dramatic combat scenes. Dramatic combat scenes are easy. 'The PCs are stranded on a burning boat when, suddenly, ninjas attack!' I could have thought of that myself, thanks. I'm much less sure I'd ever have come up with the demon playwright and his cast of degenerate bird-man 'actors' - and that, ultimately, is what I'm looking for in a published adventure. I just wish that Paizo could be as clear about which of their modules are dynamic situations as opposed to strings of set-piece fight scenes as they are about which of them are railroads or sandboxes.

And while I'm making wishes, I also wish they'd stop putting in trash fights and filler dungeons. Those are super lame.

And get that poor sorceress some underwear!

'Fifteen levels in and we still haven't found any? Not even a bra? Come on...'