How does one write a game, anyway?

March 5, 2008 12:11 AM

I have played a lot of D&D over the years, but I've never been the DM. My turn in the current group is coming up, so I'd like to write a campaign for them. The current one is a bit hack-and-slashy, I have in mind something more like a multi-arc murder mystery.

My google skills fail me: I'm looking for resources on how to structure a campaign and story arc. Something a little more detailed than 'Intro->encounter [...] encounter-> BIG BAD DUDE -> happily ever after'. Pointers especially for non-dungeon-crawl style campaigns especially welcome.

I'm familiar with the rules, I know the generic stuff like have maps, decide on encounters ahead of time, etc. I also have a general story sequence in mind ... I just need help getting it from an idea to something playable.

posted by ysabet to sports, hobbies, & recreation (12 comments total)

First off, I assume you realize that D&D has some issues as the system for a mystery campaign, since it's replete with solve-the-mystery spells -- divination, contact other plane, raise dead, detect evil, zone of truth, etc. There are a couple ways to deal with this, ranging from the low-level solution of just having PCs that don't have the spells (but what about the concerned NPCs?) to the higher-level solution of assuming the PCs will have them and writing a mystery that requires they be used.

Next off, I assume you're running this not just because you want something more cerebral but because the players want it too -- there is no surer way to ruin everyone's fun than to try to run a mystery for people who just want to kill stuff.

Ok, so those out of the way, there are basically two ways to design a mystery in RPGs. One is the moving-clues approach: the players do investigation type stuff, and when they do well at their interrogation or search roll or whatever, you give them a clue. That is to say, where they search determines where the clues are. You can further expand this approach to a totally solipsistic one, where the solution to the mystery is also determined by the players. Just listen to their discussion and if they think the butler did it, throw in more clues to prove he did. Or put in a clue to prove he absolutely didn't, and then have it be the *next* guy they fasten on. This approach has the advantage of being really easy as a GM, and there is never any worry that the players won't find the clues. It has the disadvantage that if the players don't know you're doing this and then come to realize it, it can be a major disappointment for them and they feel like you've invalidated their work. But for relatively casual players and/or a GM who's good at improv and storytelling it's a great way to go.

The other approach is the fixed-mystery approach. This is the standard approach in a lot of Call of Cthulhu games, where if there is a clue in book X in the library on the third floor, by golly, the only way to find it is to read that book. This has the advantage that the players really earn everything they get, but it often means pacing suffers as they wander around trying to figure out where the next clue is or What It All Means once they do. There are techniques for making this work better, but they require some practice (a good place to start for them are in the posts by clehrich on this forge thread (starting with post #3 or so)
posted by inkyz at 12:41 AM on March 5, 2008

Uh, and having re-read the question, maybe you were more interested in the design of the campaign, not how to do mysteries as such. Basically, the main things to keep in mind are
- Have a handful of interesting NPCs (4-8 is a good range) with conflicting goals. They should have things they want to accomplish, and (this is important) things they want the PCs to do. They should also have limitations -- reasons why they can't accomplish their goals themselves.
- Foreshadowing is awesome. If they're going to fight a lich in the last act, have them fight some skeleton guys wearing a weird amulet in the first one. Later on, have some friendly merchant guy be happy to buy all their spare magic items. Still later, have him turn out to be wearing the weird amulet himself.
- Things that are really obvious to you will not be obvious to the players. This is related to the previous one. If you want players to notice something, put it in at least three times and expect them to maybe miss it anyway.
- Real humans are more important than fictional ones. The point of the game is for the players (including you!) to enjoy themselves -- don't put anything into the game, whether it's an NPC, a setting, or a plotline, that you're not ok with the PCs stomping all over if they think of a way to do it.
- If you have a Big Storyline in mind, throw in occasional adventures that are unrelated to it. B-plots about the characters' personal lives are good, as are parties and festivals and other chances for the characters to just hang out and chat. Sometimes these even turn out to be relevant down the line.
posted by inkyz at 12:55 AM on March 5, 2008 [2 favorites]

You can further expand this approach to a totally solipsistic one, where the solution to the mystery is also determined by the players. Just listen to their discussion and if they think the butler did it, throw in more clues to prove he did. Or put in a clue to prove he absolutely didn't, and then have it be the *next* guy they fasten on.

I've heard of this approach and - although it requires some agile thinking - it can work well. Throw out some random clues at the start, without any idea of who the murderer actually is. Whoever the players accuse first *won't* be the murderer - some additional evidence will come to light that exonerates them. After that the party will be much more careful.

My other bit of advice would be to beware of "rolling against skills" to solve the mystery. That is, progress should depend more on the party making correct deductions, imaginative leaps and putting things together than the use of a given skill or power. Otherwise the game degenerates into a series of repeated rolls to be given the solution: "I roll to track the footprints. I failed? I try again. This time I take my time. Fail? Okay, Sarah has a better at that so she can try ..."
posted by outlier at 1:00 AM on March 5, 2008

Whatever you do, it would be really interesting if it is somehow related to Gary Gygax (co-creator of D&D) who recently passed away. Maybe something in lines of Simorgh.
posted by caelumluna at 6:18 AM on March 5, 2008

For your first few turns as a DM I would suggest that you run a premade campaign, think of it as training wheels.

IANADM, but I have played in some really crappy homebrew story lines and would suggest you let someone else do the heavy lifting while you are still figuring this out.
posted by BobbyDigital at 7:15 AM on March 5, 2008

The most important thing to remember as a DM/GM is: No matter how much time you spend planning and plotting, the players will do something you didn't think of.
So, be prepared for spontaneous tangents.
posted by jozxyqk at 7:30 AM on March 5, 2008

To expand on jozxyqk's point, sometimes players just can't be induced to bite on the plot hook. Either they just miss it (see inkyz above), or they latch onto some totally unimportant bit of color you threw out. It means nothing to you, you can hardly recall what you said, and they are on it like ugly on an ape. We call that "Look, Shiney!" in our campaign.

Depending on how important the plot hook is, I will either tweak it up and troll it back in front of them again in a different situation. If, after 3 attempts they still won't bite, I generally drop it and let them run with what they are doing.

Try to make sure you're keeping track (in your head, at least) of what's going on in the world around the group. They had a fight in some dive bar and some poor bystander gets killed? Too bad they don't realize he was the favorite nephew of the local summoner. Then, later, you can hit them with an enemy they didn't know about from out of the blue.

Another fun one is to follow a plot hook they ignored to it's conclusion. The players never got around to investigating those stories about intelligent trolls, because everyone knows that trolls are feral monsters? If they had spent the time, they would eventually discovered that all trolls were intelligent once. They practiced a voluntary population control regimen to keep the number of trolls constant at a rate the planet could support.

Trolls have a taste for human flesh. Humans know trolls eat people, and that trolls are tough, so they burn troll corpses. Unfortunately, the regenerative abilities of trolls far beyond human undertanding - they can regenerate from as little as a single undamaged cell. It takes years sometimes, but they come back. They come back feral - nothing more than an almost indestructible platform for a ravenous hunger for human flesh. And, a burning drive to reproduce.

So, over thousands of years, almost all the trolls have become feral. And, the population is exploding. The few remaining intelligent ones have spent hundreds of years trying to ride herd on this problem, for fear that soon enough, the feral trolls will denude the planet of everything edible. They are loosing this battle, so they have researched a ritual to grab intelligent spirits out of the ether and stuff them into the feral troll bodies. Some brains is better than no brains, right?

At this point it can take a couple of different paths:

• The ritual works, and all the trolls have intelligence again. Great! But now we have a troll nation that wants to rebuild its long lost civilization.
• The ritual works, and all the trolls have intelligence again. Great! But now all the recently feral trolls all have their own agendas. Thus troll adventurers spread out across the globe.
• The ritual fails! The intelligent ones grabbed malevolent spirits by accident, so now all the feral trolls are low level undead. Great! They won't reproduce anymore. Or will they..."
• The ritual fails! The intelligent ones grabbed malevolent spirits by accident, so now all the feral trolls are high level undead. Now the intelligent ones must convince the other nations of the world that it's in everyone's best interest to destroy just the feral ones.

posted by Irontom at 9:53 AM on March 5, 2008

I'm going to echo BobbyDigital and say that you should really consider running a module first.

DMing is a whole skillset that overlaps, but doesn't coincide completely with campaign design. If you run a module or two first, or even a whole premade campaign, you'll get the DMing skills down while getting a real education in how a good adventure is made from the inside out.

It gives you a completely different perspective than playing does.

I've been part of any number of groups where new DMs wanted to do everything, ambitiously, all at once. It always fails, and then they don't want to DM anymore. Do yourself and your group a favor, and rely on some of the tools out there that can make you a good DM. This is what modules are made for.

Look over reviews, find a module you like, and study it. Put your effort into running it well. Gain a whole new appreciation for what your usual DM goes through every week. Lean the art form of DMing, and you'll be a lot further on the road to learning how to create adventures.
posted by MrVisible at 10:03 AM on March 5, 2008

I'm going to argue against running a module, because that's not going to teach ysabet what she needs to know about building a game, but I will say that maybe trying something smaller in scope might be a good trial run.

Running a short storyline as a first-time GM is probably your best bet. If you're going to make it a mystery, as others noted, you'll have to be very aware of what spells your players have access to that would let them short-circuit the normal course of events. This is key, because if they whip out a truth spell and you balk, because it "ruins" your plot, you'll lose their trust. Now if you had already planned in advance what protections people had (and within reason), it'll show forethought.

So, I'd say, focus on a short adventure, built around the PCs backgrounds' that may last a session or two, just as a warm-up. Plan a few scenes, and try to imagine a few different outcomes for each important scene - since as others have pointed out, the players will come up with something you haven't thought of. Just take a moment, get a drink of water, and roll with it.
posted by canine epigram at 12:29 PM on March 5, 2008

Hmm.

Just for some persepective - I really don't want to run a module. The group as a whole wants to experiment with a less combative form of D&D, and I've yet to see a module that really fits that description.

Also, this group has had precisely one DM that had run a campaign previously - most of the campaigns we've run have been homebrew, first-time-DM - sort of trial by fire. It's the traditional way to do things around here. So the players won't be worried by me getting it wrong or taking a while, or swearing about them going completely the wrong way. I know I'll drop my dice, have to pull an encounter or six out of nowhere, and I have a live-in not-first-time-DM to help with the preparation side of things if I need it. That's covered.

More about the story - I have a story arc sort-of planned. Something gets nicked, players have to get it back and figure out why that particular thing got nicked. Someone survives an assasination attempt. Figure out why they are a target, catch the assassin - while someone seemingly totally unrelated gets hit (but it ties in with the theft). Motives, intrigue, drama. Timing will be very important. Obviously I need to firm up the storyline - but I'm most interested in the writing, not the mechanics.

So, please - anything about writing a campaign? Structure? Pace? Whilst I'm looking at doing a specific sort of campaign, campaign design in general beyond 'have encounters prepared' seems hard to come by.
posted by ysabet at 2:41 PM on March 5, 2008

A lot of this stuff is something you need to work out per-campaign and per-group -- how much interaction and how much investigation and how much combat can you fit into a session without feeling rushed, that kind of thing. You might want to poke around on Treasure Tables some. In particular, this looks like a relevant post (especially the comments).
posted by inkyz at 7:48 PM on March 5, 2008

If anything I say is old-hat to you, I apologize, I'm just going to toss stuff out there:

Are you using a setting they're all familiar with, or is this your world? If it's old hat, no problem. If it's new... drop them into it gently. Don't drown them with information about the world. Maybe start them as outsiders, so they have game reason to be ignorant, while you slyly slip in game-color in your descriptions of how people act and how things look.

I think pace is really important, particularly for the first session of a game - it sets the tone for everything afterward. You want to set things up so the session starts off with a bang. Don't do the old "the old guy calls you together and asks you to retrieve blah trinket lost by his...zzzzz..."

Instead, figure out how you can make as much of the intro directly relevant to the PCs.

Maybe the person to whom the object belongs is a relative or friend of the PCs. Make it personal - give them direct reasons to care (instead of just oh, it's a job.)

Maybe they see the theft happen, or arrive just after it happened, or hear it happening and arrive too late - make it matter.

If someone seemingly unrelated gets hit, think a lot about what clues you're going to put out to give them a chance to figure out the big picture. Why will they care if this unrelated person gets attacked (unless they're the town guard or police)?

It sounds like you've got ideas for your scenes brewing. The tricky thing is remembering what inkyz said above - stuff that seems totally obvious to you, will not be at all to your players, unless they know you really well. You cannot, I repeate, cannot just toss a single clue out there and expect them to figure it all out. Things will grind to a halt, and they'll get bored. Be prepared to think up several different clues to whatever puzzles you have, always keeping in mind the question why should they care? Make sure you've mapped out what the capabilities of any npcs or creatures involved in any of the scenes, and have a few alternate ideas for how to connect those scenes.

So for any scene, ask yourself - what's going on, how are the PCs involved, why should they care, and what's at stake.
posted by canine epigram at 8:00 PM on March 5, 2008