Geek Part III: The Homebrew

November 8, 2011 8:36 PM

Help on building an investigation-based, long-form D&D Homebrew campaign.

So I've been playing 3.5 for a few years now, and would like to try DMing an idea I had. I've DMed before, for white-wolf, but never D&D, but I know the material pretty well, and will have a co-DM who knows the books backwards.

My inspiration was to do a noirish, Veroinca-Mars-style campaign, with a large mystery arc, but with each session more or less acting as a stand-alone mystery unto itself. So, for instance, the party will begin, at least, as a low-level sort of detective-squad for the city of Ravens Bluff, which recently suffered much war and strife but has returned to peace and enormous prosperity and greater infrastructure for all upon the return of their victorious Prince(ss). As the game goes on, it will become clear that this Prince(ss) made some deals they didn't quite understand in order to bring about the peace, which have led to a massive and secret slave-trade of magic-users into the galleys underneath the palace, with a powerful devil behind it all and the possibility of massive upheaval and wars should the truth ever come to light.

In the first session, however, the party would be investigating he attack on the city's champion archer on the eve of a major tournament, leading through several possible leads until they find the right guy. Etc. There will be things which come up which will seem fairly meaningless at the time, but which will prove later on to be very meaningful indeed. (for instance, that the guy they got was a patsy whom they unwillingly delivered into slavery, and that the attack was actually a message sent from a pissed-off city-state they started cutting out of the trade.)

My questions here are pretty vague. Basically:
1. What tips do you have for building homebrew adventures?
2. How do you keep the players on track in an investigatory campaign without railroading them?
3. How can you make sure they remember minor clues?
4. What are some organic ways to bring combat into such a setting?

Anything else you can offer.

Thanks!

posted by Navelgazer to sports, hobbies, & recreation (24 answers total)

My single biggest advice for any DM is this:

Do not attempt to plan a plot.

The party will not cooperate.

They will ignore the evidence.

They will kill the witnesses.

(Sometimes by accident)

They will walk past the crime scenes and interrogate a random hat merchant you mentioned for local color.

They will guess wrong about who did it, and kill the wrong guy.

They will guess right, and not care, because the villain can totally get them all badass gear, and besides, the wizard's backstory has them being part of the same guild, don't you remember that coming up during character generation?

Come up with your foundation, your history - "here's what happened, here's the major players, with their motivations" - and have some basic events lined up in case things get dull or stuck - "And then, six goblins wearing bright yellow armor burst through the door!" - but if you try to plot things out beyond a thin sketch, you'll either bore your players because it's not their story, or you'll get frustrated because the stupid party ignored your painstakingly designed story in favor of stealing that awesome ship in the harbor because, come on, you said the figurehead on the prow was a mermaid, and I bet that mermaid-obsessed guy we met in the last town would give us a lot of money for it!
posted by Tomorrowful at 8:45 PM on November 8, 2011 [6 favorites]

Oh, just to clarify, I know you're aware of the risk of railroading - what i mean is, investigatory campaigns as such don't really work unless you basically just let the characters find evidence you make up on the spot; if they have to know exactly who to talk to it'll fall apart, so let their hunches be retroactively correct. If they think that shifty tavern-owner knows something, he should. Build the facts of the narrative around their decisions and their actions, rather than having a predetermined reality that they have to figure out or not.
posted by Tomorrowful at 8:48 PM on November 8, 2011 [2 favorites]

Heh. Thanks, Tomorrowful. Thankfully, I know the crowd I'm working with, and they are much more cooperative than that. At least, they will be at first. They are experienced gamers who will understand my situation as I bring them into it, and will be playing along at first. (They are also all lawyers, which presents its own set of problems, but that's another discussion.)

I plan to use the advice I found elsewhere here of having "scenes" rather than a set reality. i.e. if my players don't go where I need them to go, they can still hit the same encounter, or find the same information, etc. And if they go completely off the rails, I plan to have the larger series of events in the world go on without them. We are starting at level 1, and as such the world effects them much more than they effect it at this point.
posted by Navelgazer at 8:51 PM on November 8, 2011

Check out this essay proposing the so-called "three clue rule" for designing mystery scenarios. Like many things, I'd take it more as a guideline than a rule.
posted by Edogy at 9:10 PM on November 8, 2011 [2 favorites]

Use objects for clues more than people. They are easier to lead parties around by. I never did mysteries, but I had one campaign where the players found a ruined elven keep with an ivory wall with a bas-relief of a tree on it. Several of the leaves on the bas-relief were places where a leaf like object (a gem, they learned later) to open the door, they had to collect all the leaf gems, which I scattered about the local kingdoms and which they had to track down and usually go into a dungeon or town adventure to get.
posted by Ironmouth at 9:18 PM on November 8, 2011

Um, play Call of Cthulu instead?
posted by bardic at 9:39 PM on November 8, 2011

bardic, I see what you're saying, but this isn't actually that type of thing. The session-to-session adventures will have levity and monsters and whatnot. The overarching plot will be darker. But thanks.
posted by Navelgazer at 9:41 PM on November 8, 2011

Specific mechanical advice:

In general, investigative/city campaigns have fewer combat encounters than a dungeon crawl (the typical use case for the d&d rules) and more skill use. Therefore:

- Don't award xp in the usual way (ie, based mostly on killing things). Instead, you can arbitrarily declare the characters level up every once in a while (after N sessions, or after a major plot point, or whatever), or use a quest-based xp system (find out who attacked the archer, 500 xp; capture 5 criminals, 100 xp -- you can also provide individual quests, like the party wizard gets 50xp each time they learn a new spell; this eventually turns into something like Clinton R Nixon's adaptation of the Shadow of Yesterday's Keys system)

- When you do have combats, make them count; there's probably going to be at most one combat per game day, so spellcasters are going to be unloading all their spells, characters will be fully healed at the start of the fight, etc. Use encounters that are at least CR+2, and CR+4 for tough fights. (Unless your players aren't good tacticians, then drop it by a CR)

- Don't let the classes with 2 skill points/level be made useless by that and not get to do stuff. Strongly consider one or more of: everyone gets +2 skill points/level, everyone gets a minimum of 4 or 6 skill points/level (seriously, it's not that unbalancing -- there's still a skill cap), some skills get merged together (classic examples being Hide and Move Silently, Acrobatics and Athletics, etc)

- Don't let skill-using classes get overshadowed by spellcasters: fly trumps Climb, invisibility trumps Hide, knock trumps Pick Lock, etc. This is mostly only an issue if you're going to hit level 10 or so, since before that spellcasters can't really afford to have the huge range of spells memorized, but it can still be a problem. The best way to solve this is really with a gentleman's agreement among the players, and/or just make sure everyone is on the same page with respect to the situation.

- Remember you're likely to see a lot more human/humanoid opponents in this kind of game: that means things that rely on the opponent being a particular size or type are more effective (Trip builds don't work great in normal campaigns because eventually the opponents are too large to trip; Charm Person works against a higher fraction of enemies; "favored enemy: humans" may be a big win). Not necessarily a problem, but something to keep in mind when picking enemies -- you either want to run with it and let your players know, or be prepared to be creative in thinking of ways to put in more monsters (arena fights, sewers, etc).

- I wrote these rules for an eberron campaign up a while back which had a similar scenario; it might be worth seeing if there's anything you want to grab.
posted by inkyz at 9:42 PM on November 8, 2011

A friend of mine set up a campaign like this once upon a time. There were five or six smaller things going on that may or may not have linked to the big thing. We were running in a system were you got extra points for having things like enemies (who might track you down) friends (who you had to take care of) and other personal issues and obligations and he tied many of these to something going on in the campaign.

I would caution about making too many things retroactively correct or they feel like they're on a different kind of train ride.

Not sure what the D&D rules look like these days, but now is the time to which rules you can subvert as an excuse for handing out clues. A fumbles table, for example - every time a character rolls a 1 in combat make them roll a save. Then roll some dice of your own (hidden) and declare that nothing happens, or the character makes his next attack at a -1 or misses his next attack all together, or the strap on his shield breaks. Now you can send them on any number of errands when they get back to town. While in the harness maker's shop to get the strap fixed, the harness maker expresses joy that he can take a break from on this pile of little leather cases that the earl of Flinghovia's captain of the guard ordered and said the earl needed them right away. And later on one of them shows up in the context of the mystery. And at the Earl's castle you learn the captain of the guard is missing. Hmmmm.

I agree with what Ironmouth says about objects, but only if you consider some of your NPCs to be more object like. The harness maker is a fixed thing - he and his family and two apprentices are a self contained unit that doesn't need to be tracked down or anything. It's a relatively permanent and self contained unit that doesn't go anywhere.

Also, if you've described what should be a clue and the players aren't making the connection with something that happened a month ago, don't be afraid to have them make an intelligence roll and tell them that their character remembers blah blah blah.
posted by Kid Charlemagne at 9:43 PM on November 8, 2011

Doh - that's another thing my friend did, doled out experience based on mysteries solved as much as (or even more) than combats won.
posted by Kid Charlemagne at 9:46 PM on November 8, 2011

Be prepared to recap and reintroduce subtle clues or clues that were meaningless w.o. context from five or six sessions ago. Remember six sessions is like a season in real life with bi-weekly play. Too much of that though, and it feels like the mystery is being solved for you. It can be anti-climatic.

Also, be ready to ramp up the end game of your seasonal plot when the players get into it. They'll get tunnel vision and head straight for it.

You might want to look at the Gumshoe system for some tips on clue planting. Add a minor 'interest' subsystem so you can have everyone specialize in various types of clues. That way the dumb fighter can still find some clues in his specialty.

Don't worry too much about railroading. In a Monster/Murder of the Week campaign, it's pretty much expected, and if you let them do neat unexpected things to solve them, or to go completely off the rails in the seasonal arc things, it'll be fun.

Don't worry about combat. It will happen as much as your players want it to :-P

I might not even give out xp, just have the characters level up every few sessions, after they solve a case particularly well.
posted by Garm at 9:50 PM on November 8, 2011

Specific scenario/play advice:

- If you want players to remember clues, you have to decide what kind of people they are and what your scenario is. Me, I play mostly online, so we have a wiki, which is great. Some other GMs bribe players with xp to write up diaries or summaries. Other GMs just start off each session with a "The story so far" recap (which is a good way to subtly re-emphasize important stuff). Other GMs just don't worry about it and let people sink or swim. Any of these will work, but you have to pick a method that matches the level of effort that your players are willing to put in, and that keeps you happy too.

- Similarly, you have to decide what your policy is on railroading. Personally, I never, ever move clues around -- if the bartender has the scroll with the cult password on it, and they search the bakery instead, they're not going to find it. But at the same time, I'm not afraid to make things happen to make things happen. Like, if they search the bakery, it doesn't seem crazy that the bartender might notice and get freaked out and flee in a suspicious way. I feel like this pays off a lot better in the long term: the characters' actions influence the world in a meaningful way, and they don't have to make good choices all the time because the world keeps on moving regardless.

- Be ready to ad-lib. The amount of world prep I would do before the first session is exactly what you have written in this post and no more, but I'm also ready with some rules of thumb if the players want to explore some part of the setting that I haven't specced out. Like, in my current (space opera) game I decided that the world is a mix between star wars and dune and lensmen and HHGTTG. If the players say "hey, how do elevators/fighter ships/teleporters/cuisine work?", I decide which influence that falls under and that lets me know how to answer: "Elevators are built by Sirius Mega-corp, and are always failing" "Fighter ships have a protective forcefield, go faster than light, and fight with lasers and missiles" "Teleportation is on the fringes of science and risky when it works at all" "Cuisine is weird and frequently hilarious and very different per-planet, but anyone can eat anything". Similarly, if you decide the city's architecture is like a d&d-medieval version of 20's new york, you don't have to design any buildings in advance because you can work them all out on the fly.

- Establish some specific NPCs and setting elements on the fly as you need them, and if the players think they're interesting, then bring them back more. Give them just one or two character traits to start out with (the bartender is an elf who acts like a dwarf, the Church of Zogo is run by jerks) and round them out more as it comes up. Then next time you need somebody or something, look at your list of established things and pick one from there if appropriate (or even if not appropriate) -- the next murder victim is ... the bartender. The float in the parade is sponsored by ... the Church of Zogo.

- Finally, mix things up genre- and setting-wise. If this is a primarily city-based campaign, have a few sessions where they're sent out of the city on a mission and you get to see how your urban characters cope in the forest. If it's mostly dark, put in a low-stakes comedy bit once in a while. If they're low-class thugs, make them infiltrate the nobility for an adventure or two.
posted by inkyz at 10:12 PM on November 8, 2011

General home-brew campaign advice - not specific to a mystery campaign, but it does work well...

Characters know people in their world. They have motivations, and they have a voice. It is important to know what that is. Likewise the NPCs that they know, do not necessarily know everyone in the party.

For my best campaign (AD&D 2nd Ed - like 18 years ago now), I asked the players to write/justify their stats and equipment through some gentle prose about themselves. I asked them to also give me 3 *very short* descriptions of people they knew in the town and what their relationship with that person was. I'd then review it 1-1 with the player to get a few things straight and then incorporate as much of the character into the story (over the long arc) as I could. I'd also make some notes, as to what that person's motivations really were... (This meant I made a childhood best friend, rekindle a friendship only to betray it later.) I made the bartender far more aware of the thief's true motivations. It meant that the players had a vested interest in the environment and they didn't run in and start cutting off heads or drawing swords when unnecessary. Before the session, I summarized a few people that everyone should know (with the player that wrote them knowing *more* than I stated). Each week, we'd add more NPCs, or a player would flesh out a bit more of an NPC I had introduced that didn't have (much) of a back-story. Then, I encouraged people to keep their mouths shut about motivations and things that they knew that were special, and gave everyone a notepad to pass me notes should they need to. We built the world slowly, and if there wasn't anyone that it was appropriate that they would add, I would get them to flesh out themselves a bit more - most importantly why they were still on this godforsaken quest...

With a noir campaign, you need NPCs that have some multiple dimensions. You need some that people can "trust" that they don't trust. You need to have them betrayed, or split by motivation. This was a good mechanism for me to accomplish some certain goals. Also, people will pay more attention to the characters they help create, meaning you can drop far more subtle clues to your party.
posted by Nanukthedog at 4:38 AM on November 9, 2011

This is creepy... I'm working on an almost identical campaign concept.

I'm still in the early stages, but I can throw out one tidbit in response to your item 4:
Have an undercity. You mention "recent, enormous prosperity" — that would seem to suggest plenty of opportunity for crime. Whether it's a single treacherous Kowloon-like segment of the city (separate from the rest but accessible from many less underworldy neighborhoods), or a literal undercity in the sewers, you want a place where the savagery of monster-oriented D&D combat is likely to occur, which the surface streets of a civilized city are not.

In my campaign, the city in question is the most ancient and cultured in the world, heavily influenced by elves, probably the nicest place to live in my game world. Nevertheless, I decided early on that I wanted to have regular "Law & Order"-style footchases or "U. S. Marshals"-style standoffs, and perps fleeing into the Undercity is a great way to facilitate that.

Otherwise, I agree with the remarks above regarding generous XP rewards for victories in problem-solving and roleplaying. On this topic, the Book of Exalted Deeds has a useful section on "Waging Peace."
posted by AugieAugustus at 6:10 AM on November 9, 2011

Oh! and since you mention you've never DMed D&D before, here is the #1 thing I wish I had known about from the beginning. And the #2 thing is this. I was screwing up XP for months, maybe years.
posted by AugieAugustus at 6:32 AM on November 9, 2011 [1 favorite]

1. create a situation, not a story. Dont think about the outcome of anything ahead of time. Play to find out what happens.

2. To avoid railroading, play a couple of sessions in which you bombard the table wih plot hooks. Think of these sessions as 'pilot episodes' for your game. Develop those plot hooks that engage the table.

3. If these ideas seem interesting, check out Dungeon World, a D&Dish distilation of the great new-school RPG Apocalypse World. DW has terrific, succinct adventure and campaign building advice, useful for any RPG.
posted by Sauce Trough at 8:19 AM on November 9, 2011

Here's something to think about concerning investigation-based campaigns:

In my experience, they don't work all that well in non-historical settings. Call of Cthulhu has the advantage of being in a historical setting that really isn't all that different from our own, and that fact carries tremendous subtle advantages for investigation. Investigation is about the finding of clues and piecing together of facts. Both are problematic in an invented world.

But where do you find the clues? In Call of Cthulhu, interested investigators eventually figure out the basic places to look for information: the library, the newspaper, city hall, the police station, the neighbor, the man on the street, the dealer in antiquities, the pawn broker, and so forth. Some of these places don't exist in a fantasy world, or if they do players can't be sure of it. Because so much more of the world is the creation of people who are not the players, they can't be as sure of its workings, and so will hit places where they don't have ideas about where to go next more often. This means giving players the system's version of CoC's Idea Rolls more often, which means the referee is more often manipulating the game to keep the players moving, which feels subtly more like the referee is playing than the players. I have this problem in our Tuesday game, and it's set in the 1920s -- I can only imagine how much worse it would be when libraries aren't public, there's no such things as newspapers, and literacy rates are low enough that recordkeeping is little-used if at all. Of course, none of those things need necessarily be true -- but then, the setting isn't quite as fantastic any more. It can be made to work, but it's a lot more work on the part of the referee in fleshing it out, and a lot more work from the players on imagining it and keeping the new setting consistent in their heads.

And, how do players piece together facts? The simple fact of magic makes detective-style investigation tons more difficult. In the real world, Sherlock Holmes-style deduction is partly a myth created to facilitate storytelling, and don't get me started on CSI. In a role-playing game, the myth works well in giving players leads. For all the horrible monsters and strange magics of Call of Cthulhu, these things are still very rare and often provide visible signs of their passing: the trail of slime, the haunting stench, the insane witnesses. The setting does an excellent job of not having "disappear in a puff of smoke"-style magic, wholly without evidence. If a murder takes place in a locked room with no apparent means of exit, if the answer turns out to be, effectively, "a wizard did it," it doesn't feel very satisfying. If it happens enough eventually wizard becomes the default, go-to answer. If there actually are wizards in the world, then, the problem becomes a lot more challenging. Any Sherlock Holmes-style investigator will necessarily have to make a comprehensive study of magic, its capabilities, its limits, the incidental signs of its working, to be sure he can invent an accurate causal chain of occurrence. Even if players don't go that far, it still presents major difficulties when even a low-level magic user could easily circumvent the physical in so many ways, when custom-made magic items are so wide-ranging and weird in their operation, and when a character's free will can be subverted using mind-affecting spells.

Finally, and to put the exclamation point on my argument, the following section comes directly from the Call of Cthulhu rulebook, 6th edition, page 138, under "Reasonable Deduction," which sticks out in my mind as among the most insightful RPG refereeing advice I have ever read:

"When the keeper sets a scene, his or her most important ally is invisible, one which no scenario-writer ever puts on paper. 'Reasonable deduction' consists of all which is in the room, cavern, aircraft, or other physical setting which is not described as being there, but which can be logically inferred as being there.

"For instance, the investigators are in the library of a mansion. Specifically mentioned are the many massive bookcases lining the walls, two leather armchairs, a desk and a chair, and a billiards table. What else might be there?

"Books, certainly, and lots of them. Cues, chalk, and billiard balls. Writing material. Paintings on the walls. Lamps and light switches. Windows, maybe lots of them. Rugs on the floor. A fireplace, and fireplace tools. Lots of odd things in the desk drawers, including scissors, a letter opener, glue, stamps, twine, tape (if it's the right era), and an address book. Perhaps correspondence files. Matches for the fireplace. Cigars in the humidor. Brandy on the side table.

"Some of these things are useful weapons--thrown billiard balls are dangerous, for instance, and a swung cue is an excellent club, as is the poker beside the fireplace. The windows offer entry or escape. One might bind a captive with twine, or with sash cord from the drapes. (There are drapes, of course, on the windows. This is a mansion.) One might spirit out a captive by rolling him up inside a rug. One might torch the room, or build a torch from a rug fragment wrapped round a pool cue, held with twine and soaked with brandy. The key to the library door may not be in the door but, if not, is very likely in a desk drawer. An electrical cord can be stripped to contacts and used (cautiously) as a weapon, as long as the electricity lasts.

"Ingenious players or keepers can come up with much more. The point is that every setting offers things potentially of use in specific situations, things which may not be specifically mentioned in the text."

All this, a thousand times over. But this works because players might be reasonably expected to know it from personal experience. The further a setting gets from the player's lives, the harder it becomes to operate meaningfully in that setting. Most of the times you can gloss over the details in the name of play, but investigation demands comprehensive of the details, there is less you can handwave away, and that means you and your players must do a lot more work. Beware of those situations where you end up having to prompt the players -- I have had to learn the hard way, if this goes too far, it feels shallow. If you as referee have to give the players ideas at every turn as to what to do next, then you should probably writing them a story instead of running a game.
posted by JHarris at 8:25 AM on November 9, 2011 [1 favorite]

"comprehensive" should be "comprehension" of course. There are probably other typos in the above, but I presume you can still understand it all.
posted by JHarris at 8:32 AM on November 9, 2011

On rereading, my answer sounded like an advertisement. Sorry about that!

I am not affiliated with the Dungeon World crew, though they are friends-of-friends. I'm pretty much just a fan.
posted by Sauce Trough at 8:55 AM on November 9, 2011

With the caveats I gave before still in place, I'll quickly answer the questions you offered:

1. What tips do you have for building homebrew adventures?

Think of what happened first, the thing being investigated, then think of the ways this occurrence affected the world. Write those down, then think of ways players can discover them, starting from zero information. One clue should lead to others, in a kind of web-like arrangement. Don't worry if some clues seem to lead back to other clues; this just gives players more ways to explore the web.

In the classic Call of Cthulhu style, players can discover what happened relatively easily, but at some cost (sanity, Cthulhu Mythos points gained), but if they then charge in with guns ablazin' they'll face a stiff fight and probably some of them will die. The real key to the investigation is one or two little facts that can be used to make the monsters much easier to handled (like a useful spell, a secret weakness, information on a hidden ally, information on the true nature of a place, etc). *That* stuff can be missed in investigation, but if it's found it makes it much more likely for the players to survive. That's a good way to make investigation relevant to the play.

2. How do you keep the players on track in an investigatory campaign without railroading them?

By giving them more ways to get to each "node" of the web, and making the essential facts relatively easy to find, but the stuff that makes it more likely they'll survive optional.

3. How can you make sure they remember minor clues?

You can prompt them to use a notebook. I sometimes bring a whiteboard and write down important facts on it, although players really should be doing that themselves. Ultimately you can use an Idea Roll -- in Call of Cthulhu, a player multiplies his INT by 5 and tries to roll under it on D100. For D&D purposes, you can easily invent a similar roll by just trying to roll under a player's INT on D20. Idea Rolls are looked down upon by some keepers, but once in a while they're invaluable for nudging players back on track, and because you CAN nudge them there sometimes, you're free to let them go further off sometimes, which can produce interesting play and sometimes avenues of investigation you hadn't thought of, which are very rewarding times if you are able to improv them out.

4. What are some organic ways to bring combat into such a setting?

Investigation is able to truly display combat in Call of Cthulhu -- there are occasional scenarios with no combat at all. When there is combat, it should either be very dangerous, or it should have the possibility of reducing some important player resource in a way that isn't trivial to refresh, like hit points or spells per day. A popular format has just a single combat -- at the end, the confrontation, where the player's skills and the knowledge they've acquired combine to possibly save the day. The same format could work in D&D. Anyway, the 4E formula (DISCLAIMER: I loathe the game) of "tiered" monsters from minions up to bosses is unnecessary -- you could just have two or even one real fight.

The "Keeper's Lore" chapter in the Call of Cthulhu rulebook (most editions) is full if invaluable advice, for pacing, atmosphere, making investigation fair, and keeping players interested. There is a D20 version of Call of Cthulhu; it might be useful to hunt it down.
posted by JHarris at 1:29 PM on November 9, 2011

Should be "Investigation is able to truly DISPLACE combat". Gah. I write these comments fairly quickly, sorry about that.
posted by JHarris at 1:30 PM on November 9, 2011

Here are some more thoughts....

Nanukthedog's comment about making the players describe three NPCs they know is kind of what I was getting at with my bit about dependent characters, enemies and what not. It's a solid idea.

Another thought along these lines, create a large list of skills that would be useful a D&D type world if you were NOT an adventurer. Things like farming, blacksmithing, herbalism, masonry, etc. and dole a few points in these things out among the player characters based on their back story. This gives you little lead ins to things that might be clues (or not) but also to personally engage the players. Players tend to get real good at back story generation if they think they're going to get skills and abilities out of the deal.

For example, when the players are visiting Lord Aesric, you can tell the character who knows about blacksmithing, "You notice the ironwork on the gate and the door is kind of sloppy, like it was done in a hurry or by a beginner." and then tell the character who knows a bit about carpentry, "You notice the same thing about much of the woodwork in the house. Either Lord Aesric is a miser and was trying to save a few silver, or he's not as rich as he let's on."

Maybe this kind of thing is a clue, or maybe not (and you need to have enough maybe nots that it's not like every time you do this you are waving a big red flag that says CLUE! on it) but it makes the players feel like their character's skills and abilities matter and let's you tell them things about a world they might not be intimately familiar with, without making it feel like you're reading your manuscript to them.

Another thing - any time a game session ends in the conclusion of the plot of the day, have the players give you a short list of what they're going to be doing before next game session. You can respond to these via e-mail so you can do the shield repair I mentioned without soaking up game time, but it also let's you think about ways the characters might brush up against some plot hook or exposition without having to come up with it on the spur of the moment. Again, put some random throw away stuff in here, too. The city guards stopping the ranger, but then letting him go because he's definitely not their man, er, elf, er on your way friend, sorry for the bother... is just the city guards trying to catch a cut purse. Or maybe the cut purse with a cloak that looks just like the ranger's is a plot to come.

Thinking about what JHarris said about magic, don't be afraid to prune the spell list, or add spells of your own to the list with an eye to de-deus ex machinaing the magic system. A spell that let's you know if any spells have been cast in a place or on a person (with the look back window, information on the type of spell, and exactly when all increasing as the caster gain level) or wards against certain types of magic. Be careful to to create things that don't unbalance the game - an anti-magic ward, for example, should have all kinds of conditions attached to it so you can't just cast it on a rock, and then the fighter carry the rock in his pouch and go off to attack high level wizards with impunity.
posted by Kid Charlemagne at 2:31 PM on November 9, 2011

One more thing you can do about magic is enforcing spell components, even adding them to many spells that don't ordinarily have them, and leaving traces of them when they are used.
posted by JHarris at 2:49 PM on November 9, 2011 [1 favorite]

I ever so fondly remember a (starting at) first level campaign sort of like this. Our characters got approached by a wealthy ex-merchant/burgher type that came to ask us to investigate rapid, radical changes in the local leadership. We each gained 1 - 2 levels while investigating (xp per plot milestone per the ref). I'd say it was about 33% railroaded as the surprise was rather subtle (the local lord had been replaced with a doppleganger - yes, the creature - that was reasonably well trained) and the ref wanted us pointed in the right direction (it was yet another "we should finish this tonight" but spiraled madly out of control adventures this group was awesome at - remind me to tell you about the missing girls/winery adventure that became a truly awesome campaign, thanks to yours truly being so impressed with the aforementioned group running with it he didn't have the heart to end it). It ended up with my (dumb as a dumb stump) level 3 berserker that had been SERIOUSLY munchkinized via massive layered buffs throwing down with the level 5 kensai (but not by that name) master of the guard so the party could get to the "lord". With the 'serker rules we were using + the buffs, it was still the closest, most nail-biting encounter I think I have ever had. The whole group was blissful that not only had we tracked down the doppleganger, but had found the still living lord (kept for information), and still enough credibility with the locals that they would believe us that the lord was the doppleganger and that we weren't using magic trickery to fake it. Oh, also appreciated was the time we had to prove all this so the local guard didn't just cut us down for murder. (My berserker would have ended up stumbling in his own lights had he had to fight at that point.)
posted by Samizdata at 6:19 PM on November 9, 2011 [1 favorite]