Warning: this one kind of gets away from me somewhere in the middle. In particular, I think the bit about Chekhov's Gun is getting at quite a good point, but doesn't necessarily make it very well. Plus, I think I've said much of this before...
After the game on Sunday, we spent some time discussing the adventure, and the weaknesses that it showed in actual play. Such as the PCs lacking in motivation to actually get involved, the 'blocks' put into the adventure where one bad roll can derail the whole thing, and so forth.
So, the question: how did published adventures get to be so bad? Was it always thus?
Personally, I'm inclined to identify the publication of "Ravenloft" as the key moment. It's not that it's a bad adventure; on the contrary, it is one of the best that has ever been published. However, what Ravenloft marked was a point where adventures shifted from providing locations in which to set adventures to providing stories in which PCs could have their adventures. Now, there's obviously nothing wrong with stories, and there's a fair amount that adventure design can learn from storytelling techniques.
But an RPG adventure is not a novel, and cannot be structured like one. And I think that's a key mistake that too many adventure writers make.
Adventures to be read, not run
I get the distinct impression that a lot of adventures are intended to be read, rather than to be run. Thus, they have a distinct story-like structure, with a clear beginning, middle and end; they support only a small number of PC paths through the story (sometimes as few as one); they presuppose key PC motivations, and sometimes actions; and so on.
Now, in fairness, it's important to note that any adventure's first test will always be in the reading - unless I can wade through the text, I'm never actually going to run it, so if it plays well but reads like a VCR manual, it's useless. And it doesn't really help that many (if not most) reviews are done on the basis of a read-through rather than a play-through - so an adventure that has some fun looking encounters and some good ideas, and has clear text, will score extremely highly even if actual play throws up major issues.
(It also doesn't help that a lot of reviews are also skewed by the names involved - an adventure is likely to review better if it has the name "Monte Cook" on the cover than one without, for example; while the 4e D&D adventures actually review pretty well despite being almost universally awful. But that's another rant.)
But one of the glorious things about RPGs, that sets them apart from reading novels, watching films, or even playing other types of games, is that you can't assume PC actions. Most likely, they'll go off in some direction you didn't expect, or spend a lot of time meandering about. Which plays merry hell with your lovely three-act structure, and can really mess with your pacing. Just as not every novel adapts well to film, so too do many adventures fail to play out the way they're written.
The Logical Leap
Probably the place where the difference between adventure-as-story and adventure-as-game becomes most apparent lies in the use of mysteries. The thing is, Sir Arthur Conan Doyle can have Sherlock Holmes solve a complex and convoluted mystery based on the nuance of a half-heard conversation and a bit of fluff, because SACD has full control over making sure his character has access to all those clues, and also has perfect insight into what they mean.
The writer of an adventure module simply cannot assume that. Not only can't he assume that the players will even overhear that conversation at all, but he cannot be sure that the GM will present it exactly as written (so that the nuance is even present). Nor can he be sure that the players will pick up on the meaning of it all - in fact, most likely, they won't.
What this means (as discussed at great length previously, and elsewhere) is that the adventure designer really must insert loads of redundant clues into his mystery. So many, in fact, that on a read-through it will seem like a really lame mystery - if you've got three clues all pointing to "the Baron's a bad guy!" then that's hardly a revelation! And yet, that's pretty much what is required.
(In fact, I'd go further than that. In addition to putting lots of clues in place, the adventure designer should also tag every clue with the mystery to which it is attached, as a reminder to the GM, but also so that they can be sure they can recheck their working later. Which will make it read like an incredibly lame mystery.)
The Red Herring
Don't. Just don't.
Chekhov's Gun
I think this is a big one, and a big mistake for an adventure writer.
Anton Chekhov said, "Remove everything that has no relevance to the story. If you say in the first chapter that there is a rifle hanging on the wall, in the second or third chapter it absolutely must go off. If it's not going to be fired, it shouldn't be hanging there."
For a script-writer, this makes sense. Cutting superfluous detail just makes for a better story - it's tighter, it is more likely to hold the audience's attention, and it's probably much more effective at making its point.
But for the adventure writer, it is absolutely poisonous, at least if taken literally. See, it gives rise to the temptation to remove everything that doesn't fit your pre-determined narrative - that beginning-middle-end structure you have envisaged. So, you cut the number of clues for your mysteries down to the absolute minimum, because more is just redundant. You cut out all those extraneous encounters that sit on your non-prime paths through the adventure. And, once done you have the 'perfect' adventure... provided your players are four clones of yourself, and provided they follow the exact path from encounter A, through B, C, D, and E, and on to the final confrontation with your BBEG at encounter F. Huzzah!
But then we sit Andrew, Mark, Jill, and me down at the table, fail to equip us with the helm of reading Monte's mind, and expect it all to work out. But of course it doesn't, because after encounter A we wander around for an age, miss the clues in C and D, don't ever get to B... and end up just lost.
And Another Thing...
Protagonists in novels and films are forever getting themselves captured, or infected, or otherwise into scrapes. Naturally, adventure writers want to include much the same in their adventures. That's railroading, of course, but I'm not convinced it's the worst thing in the world. But...
The writer of an adventure needs to be very careful about including such events, for two reasons.
The first of these is that events like captures or infections are very hard to engineer if played 'fairly', and run a very significant risk of alienating players if they are made inescapable. The issue is one of control: the GM has absolute control over every single thing in the game, with only one exception. Conversely, the player has control over only one single thing in the entire game - his character. Enforcing a capture, or infection, or whatever, removes that control, which some players will find utterly and completely unacceptable - possibly to the point of walking out of the game over it.
Now, this is a place where a game like Serenity or Numenera has a distinct advantage - here, the game builds in various storytelling mechanisms, where the GM can basically say, "this is a plot point - just go with it." And, in return for doing so, for accepting the intrusion, the player is rewarded with XP or plot points, or whatever, which they can then trade in later for narrative control of their own. It's not a perfect solution, but it seems to work pretty well.
The second reason it's a problem, though, is that if you have a truly mandatory event in your adventure, whether a capture, infection, or actually anything else, you've created a bottleneck in your design - every path in the adventure must either lead up to that point or must lead away from that point. That's easy said, but not easily done - directing PCs is much like herding cats, and it's really hard to make sure every path leads through your bottleneck without it being obvious what you're doing.
So... How to Fix It?
Unfortunately, I don't think there's an answer for that. Published adventures are both low-margin and, actually, low-use things - even those of us who buy lots of adventures will read many more than we ever play. And there's no magic formula that will make for a good adventure, both to read and to play.
Ultimately, I fear it's going to come down to GMs doing a fairly large amount of leg-work - taking an adventure that may be good to read, and refitting it to be good to run. Which, ironically, begs the question of why you'd want to use a pre-gen adventure in the first place, since they're supposed to be labour-saving tools for the GM...
Oh well.
