Before I get into this week’s blog, I forgot to mention last week that I recently the newest installment of my series of role-playing articles on HubPages, “Seven Rules Every RPG Player Should Know." Also, just yesterday Hex released our first novel, Carter Newton’s Suicide’s Run: A Tale of the Hobomancers. Carter does a great job of bringing the world of Hobomancer to life, so it’s a must-have if you’re a fan of the game. We’ll be posting a 3-part interview with Carter here on the Death Cookie starting Wednesday. Next week, we’ll have an interview with Jeffrey Johnson, who illustrated the novel.
Earlier this week, a poster on the not-quite-abandoned QAGS subreddit asked for tips about using flash-bang style magic in a QAGS game. As he correctly pointed out, the system from Magic Rules! (and the variations on the theme that have appeared in Zoe the Zombie Hunter and a few other supplements) are really written for ritual magic and aren’t a great fit for the “wizard chucking fireballs” variety of spellcasting. That’s by design. As I am contractually obligated to point out, the “flash-bang” style of magic found in most games has no precedent in real-world magical tradition (at least as a force that can be harnessed by regular humans), and even the magic in early fantasy fiction was usually much more subtle than the kind of magic you see in a typical RPG. The Magic Rules! system was designed to encourage magic that mostly manifests in subtle ways that aren’t necessarily verifiable as the result of magic. Overt magical manifestations are difficult to accomplish, unpredictable, weird, and a little terrifying.
Sometime in the 20th Century, fictional magic started getting more overt and predictable, and wizards began to act more like scientific experts or master craftsmen than crazed weirdos who harnessed (and made deals with) powers beyond human comprehension. I don’t know the exact timeline, but I suspect that Jack Vance and comic books were key influencers early on, and the popularity of D&D did the rest. By the time we reached the era of fantasy video games and D&D-derivative fiction, flash-bang was the default style of magic. Today, even stories that aren’t directly descended from the 20th Century fantasy mainstream of Tolkien, pulp, comics, D&D derivatives, and video games tend to use flash-bang magic. Take the Harry Potter series, for example: Rowling’s main influences seem to be fairy tales, mythology, and earlier young adult fiction, but take away the wands and pseudo-Latin and Harry and the gang would fit in as well at Xavier School as Hogwart’s.
So, since magic-as-superpower is a thing now, how do we handle it in QAGS? That’s simple: Just like any other Job. If restrictions on magic aren’t dictated by the genre, setting, tone, or style of the story you’re telling (which we’ll discuss more later), there’s no reason to treat wizardry differently than any other profession, at least mechanically. Warriors slice people up and bash their heads in, thieves lurk about and steal things, and wizards cast spells. If a wizard wants to cast a spell, he makes a Job roll, just like a con artist would make a Job roll to sell somebody a bridge. Usually, there are mechanics for doing something similar to the spell’s effect without magic that can be used if you need rules: mind control is really insistent persuasion, invisibility is ninja-level stealth, and a magic missile is just a bullet with better special effects.
For the most part, the idea that magic needs its own specialized rules to keep wizards from becoming “too powerful” is a gamer superstition wrapped up in the myth of game balance. Once you accept that game balance isn’t real, the necessity of specialized magic rules mostly falls away. We don’t worry that giving a gunslinger too much ammo will make a character too powerful or make a hacker list which programming languages he knows, so why worry about how many Magic Missiles a wizard can cast and make him keep a spell list?
There is still one pitfall of treating magic just like any other ability: its versatility. For most Jobs, there’s a certain baseline understanding of what a character can do based on fictional and real-life examples, usually across a wide variety of genres and settings. A warrior’s job is to kill things regardless of the genre, so the only restrictions necessary are genre and style conventions that we trust the player to follow and, if necessary, the GM to enforce. The Bride can decapitate a whole circle of ninjas with one sword swing, but John McClane needs a grenade or machine gun to kill a bunch of people with a single attack. Since magic differs considerably from story to story, there’s a lot less consensus about what it means to be a wizard, so the “wizard” Job can come dangerously close to “Jack of All Trades.” We know a cop fights crime and thief separates people from their valuables, but there’s no common theme for a vanilla wizard’s abilities. He might talk to a ghost one minute, shapeshift into a giant eagle the next, and set a bunch of people on fire a little later.
While characters don’t have to be “balanced” in terms of power, a story about group of characters does require each character to play a role that differentiates him from the other characters and justifies his inclusion as a protagonist. Audiences tend to notice redundant characters (*cough* Merry & Pippin *cough*) and it can take them out of the story. In RPGs, making sure each character has a clear role also helps keep players from stepping on one another’s toes. A character without a niche makes the other characters seem redundant or comes across as either a universal understudy or a portable deus ex machina. While abilities aren’t the only things that define a character’s role, they’re often the starting point, so sometimes it is necessary to limit an extremely versatile character type to make that character fit the story and the group.
If the players all have similar sensibilities about what constitutes good fantasy, or if you’re playing a game where magic is closely modeled on an existing ficton, you can rely on the same kind of mutual agreement that keeps John McClane from going all Wushu. For example, if you’re playing a Buffy game, it will usually be obvious to everyone involved whether a particular spell is par for the course or veiny yellow crayon territory, as well as whether a spell can be cast with common household items or requires searching for ancient urns on ebay. If the players can’t trust one another to act in good faith (and talk through any disagreements like adults), your group has problems that all the rules in all the games won’t solve.
So what if you’re playing a game where the parameters of magic are too nebulous and the players’ preferences are too varied to rely entirely on mutual understanding and good faith? Fortunately, fantasy authors realized early on that having a character with the power to alter reality made it difficult to build conflicts that couldn’t be solved with some Latin and bat guano. To keep their stories from falling apart, these authors built limitations on magic into their fictons. Some of these limitations were based on fundamental laws of the universe, some were based on the way humans interacted with or learned magic, and some were based on societal, cultural, religious, or professional rules, taboos, and traditions. While these limitations varied considerably from story to story, there were some common themes, and most of them can be easily adapted to a game setting. Since the discussion of how to do that gets pretty involved, I’m going to save that for next week.
It’s not unusual for RPG players to find themselves in situations where, if they’re being honest, the character they’ve established would make a choice that the player knows is a bad one. In these cases, a good role-player acts according to his character’s established background, goals, and personality even when it isn’t the most advantageous course of action. In such cases, “I was just playing my character” is a perfectly valid reason for doing something that complicates things, and generally the other players aren’t going to be too upset about the problematic turn of events. In fact, some will applaud the bad decision because it fits the character and makes for a more interesting story. The GM may even give you Yum Yums.
Those aren’t the kinds of “just playing my character” moments I’m talking about here. I’m talking about the player who constantly disrupts the game and annoys the other players and, when confronted about it, says, “Hey, I was just role-playing my character!” In these cases, “just playing my character” isn’t a reason for making a particular choice, it’s an excuse to avoid taking responsibility for bad behavior. While there are as many types of bad behavior as there are annoying players, there are five general categories of fuckery where “I was just role-playing” is a common excuse.
Sabotaging the Game Premise
This is when a player interferes with the group’s ability to play the game they’ve all agreed upon, usually by creating a character that is woefully unsuited for the kinds of stories the group has decided to tell. It’s the guy who decides to play the Druid in the city-based campaign, or, more often, the evil character in a game whose premise assumes the characters are good guys. Or Samuel L. Jackson the Barbarian in a game that’s supposed to be straightforward heroic fantasy. Or a character who talks like Scooby Doo for absolutely no reason in an atmospheric horror game. Characters don’t always have to be a perfect fit for the party and game concept. In fact, well-crafted outliers can often make the story more interesting, and even stupid one-joke characters don’t always ruin a game. However, if your character concept forces major changes to the nature and focus of the story or ruins the tone of the game by merely existing, you’re sabotaging the premise.
Sabotaging the Party
This brand of dickishness is often the logical extension of creating a character that sabotages the game premise. It happens when “being true to the character” requires the player to constantly do things that interfere with the goals of the rest of party. As I mentioned in the introduction, good role-players often have to make decisions that result in negative repercussions, sometimes for the whole party. The main difference between them and the party saboteur is often a matter of who suffers the repercussions of the character’s bad decisions. In the former case, most of the fallout usually hits the character who makes the decision, or at least affects him and other party members equally. The negative backlash of a saboteur’s actions, on the other hand, nearly always fall primarily on the rest of the party. In many cases, the saboteur himself benefits in some way from the decision. Having a mole or double agent in the party can be fun in some kinds of games, but only if that traitor’s player understands that things are destined to end badly for his character. When the party saboteur gets caught, “I was just role-playing my character” isn’t going to save him, usually because retribution is the only way for the other players to stay in character.
Taking An Idea Too Far
If the DragonLance novels didn’t make you hate Kender, the first game you played with a Kender PC almost certainly did. The player annoyed everyone else and constantly sidetracked the game so he could steal shit from the other party members. One-note characters like Kender (at least as they’re played in a typical D&D game) are a common example of taking an idea to far. Another surprisingly common one-note character is the mad bomber--the guy with a demolitions skill who blows things up constantly and gratuitously, usually to the detriment of the party. In these cases, the problem is mainly one of bad character design. In other cases, taking an idea too far is a matter of not understanding context (most often as it applies to humor): the player gets rewarded (through game mechanics, story outcomes, or just positive feedback from the other players) for a particular action, so he repeats the action over and over again in a bid for additional positive reinforcement, blissfully unaware that not all actions are appropriate for all situations and that a lot of things are only cool or funny the first time.
Monopolizing GM Time
Players should always try to introduce subplots and supporting characters, so there are always going to be scenes that focus on individual characters, often with no involvement from the other PCs. This character takes it too far by constantly creating situations, either through character design or character actions, that force single-character scenes where the other players are demoted to audience members. To add insult to injury, the player regularly drags out such scenes far beyond the point at which they’re still interesting to anyone else. Some players don’t even give the other players potentially interesting character development to watch; they just split the party at every opportunity because “that’s what my character would do.” The worst example of the latter is the “loner” character. The loner is also to some extent also guilty of sabotaging the game premise and taking an idea too far. RPGs are by definition group activities, so if you create a character that’s going to constantly split with the group, you’re intentionally creating a problem. If you want to play a character who “isn’t a group player,” follow Batman’s example of getting the point across by constantly reminding the group you’re working with (or in Batmans’ case, the half-dozen or so different groups you regularly work with) about it. Don’t expect the GM spend half the game running a solo adventure for you.
Just Being A Jackass
Some of the most beloved characters in fiction (especially action heroes) are complete jerks, so it’s only natural for RPG players to create characters who are jackasses from time to time. In most cases, there’s nothing wrong with that as long as the character, like most fictional assholes, has some redeeming qualities to make the audience (in this case, the other players) appreciate him despite his douchebaggery. Without the redeeming qualities, nobody’s going to want to be around him, so the other players are legitimately just role-playing when they move to another town without leaving a forwarding address or volunteer the jerk character to find out whether the dragon’s breath is hot. Particularly dickish jackasses, either for shock value or because they’re terrible people, use the jerk characters as an excuse to say and do things that they know the other players will find offensive, disturbing, or hurtful. Those people should usually be asked not to come back.
There are a lot of situations in which role-playing requires a player to make choices that are detrimental to the character and the party or distasteful to the players, but nine times out of ten those situations are going to make everyone involved enjoy the game more. If your “in character” actions are constantly interfering with the other players’ enjoyment of the game, however, “I was just playing my character” is a bullshit excuse. It’s not the character; It’s you. The character, after all, is a fictional construct with no agency of its own. You’re the one who created a character who annoys the other players or doesn’t fit with the group’s goals for the game. Don’t blame your own bad decisions on imaginary people unless you’re trying to beat a criminal charge with an insanity plea.
Sometimes, instead of a blog, you get a widget. This is one of those times. I made up a random generator script to generate one-shot (or if you're really brave, campaign) ideas based on the tables in The Book of Dumb Tables 1 & 2. You can play with it on this ugly-ass page here. I've also linked this and the other random generators from the menu at the top so you don't have to bookmark the link to find them later. One of these days I'll get around to making them all prettier and easier to navigate, but not today.
A few of my favorite results so far:
- The PCs are legendary heroes who work as cat burglars who are opposed by pulp villains
- A rock opera version of A Midsummer Night's Dream as directed by Tommy Wiseau
- A cyberpunk version of Snow White as directed by Woody Allen
- The PCs are monsters who work as con artists who are opposed by a mad billionaire
- A teen movie version of MacBeth as directed by George Lucas
- An art film version of Don Quixote as directed by Sergio Leone
- The PCs are ponies who work as cops who are opposed by ninjas
Edit: Speaking of the Book of Dumb Tables, it's part of the Game Masters Memorial Fund Charity Drive Bundle, which just went up at Drivethru. The bundle is to benefit fellow gamer, GM, and game designer Christopher J.N. Banks, who recently lost his 5 1/2 year old daughter, Persephone. The bundle is full of good stuff including Shadowrun 5E, some Savage World Books, and lots more, so it's a great bargain as well as a good cause. All proceeds go to the Banks family to help pay for medical expenses and to put together a proper tribute for their daughter. If you've got $20 to spare, please consider picking up the bundle.
I recently finished writing the section of M-Force that deals with conducting investigations. Since investigation scenes are common in all types of RPGs, I decided to share it here. Since the details are more about process than mechanics, it should be fairly easy to adapt this to other game systems if for some reason you don't use QAGS. This hasn't been through the Hex editing machine yet, so if it's a little bumpy or there are errors, that's entirely my fault.
Investigation is one of those weird borderline areas of RPGs where the line between players and characters becomes very thin. For most players, solving puzzles is part of the game, so having the characters find the solution with a roll of the dice isn’t very satisfying. On the other hand, the characters have a lot more experience conducting investigations than the players do and they’re actually seeing and hearing what they’re investigating, not getting all the information through an intermediary. The GM’s thankless job is to find a balance that lets the players enjoy some sleuthing while still allowing the game mechanics to fill the gaps between player and character knowledge.
Role-Play Before Roll Play
When characters begin an investigation, role-playing should come first. The GM should describe what the players see, role-play the GMCs they interact with, and give the players a chance to pick up on clues and ask questions on their own before she has them roll any dice. If the players are satisfied after the initial scene investigation or interview and all the vital clues have been found (if not necessarily recognized as clues), move on to the next scene. If there are still missing pieces that the PCs will need to solve the mystery, or if the players want to conduct a more in-depth investigation, let them roll some dice.
The Clue Hierarchy
To make the mechanical end of looking for clues easier, we recommend that the GM create a clue hierarchy for every investigation scene. To create a clue hierarchy, the GM makes a list of clues that the characters can discover from investigating an area or interacting with a particular witness. She then groups the clues into categories based on how difficult they are to discover or recognize. The groups we recommend are:
- Obvious: These are things the characters are going to notice immediately, like dead bodies in the middle of the floor and swarms of flies covering the walls. Obvious clues should be included in the GM’s description of the scene or in the GMC’s dialog.
- Gimmes: Gimme clues aren’t immediately obvious, but are usually in places where someone will think to look. Answering machine messages, bodies stuffed in closets, and anything in a victim’s underwear drawer are gimmes. In terms of witness testimony, gimmes are pieces of information that the witness doesn’t include in their account but will reveal if asked the right (and fairly obvious) questions.
- Easy To Miss: Easy to miss clues are either somewhere the investigators might not think to look or out in the open but hard to spot. Things between the couch cushions, small scratches on the furniture, or tiny spots of blood are examples of easy to miss clues. Minor inconsistencies in a witness’s story or details that are downplayed or mentioned in an off-hand manner would count as easy to miss. This is the middle tier of the hierarchy, so if you’re not sure where something should go, stick it in this group.
- Inconspicuous: Inconspicuous clues are very difficult to notice, and in some cases merely noticing them doesn’t necessarily get the investigator anywhere. Examples include a spot on a shelf where something is obviously missing (it’s a vacation photo, which the demon stole to use in a spell against the victim) or the fact that the fireplace was recently bricked up (because that’s where the body’s hidden). In witness testimony, inconspicuous clues are usually hints to things the witness is hiding, either intentionally or subconsciously.
- Conditional: Conditional clues are ones that are nearly impossible to stumble upon accidentally. They can only be found if the investigators take specific actions, either independently or in response to other cues. The body hidden in the fireplace is a good example; Unless the investigators tear out the bricks, they’ll never find it. Other examples include email on a victim’s computer, security camera footage, and things hidden in places where investigators would not normally be expected to look (a crawlspace, for example). Conditional witness testimony is anything that the witness will only reveal under very specific conditions.
In some cases, a single clue might occupy multiple spots on the list. For example, the fact that the wall is covered with blood is obvious, but the fact that the splatter pattern reveals that the victim was killed by something 10 feet tall is at least easy to miss and possibly conditional on investigation by someone with forensic training.
In addition to categorizing all the clues the characters may find, the GM should make a note of which clues are vital for solving the mystery. A vital clue is one that the PCs are very unlikely to solve the mystery without. If the story will (or is very likely to) come to a screeching halt if the characters fail to discover a clue, it’s vital.
Last but not least, keep in mind that for purposes of RPG investigations, “clue” means anything that might be important to solving the mystery. Not all clues are guaranteed to pan out, so make sure to include a few red herrings in your clue hierarchy.
The first die rolls the characters make are for noticing clues. For crime scene investigations, players may use M-Forcer or Brain, whichever is better. For witness interviews, they may use M-Forcer or Nerve, whichever is better. This roll allows the characters to recognize things that might be important (in the case of a crime scene investigation), or recognize inconsistencies, missing details, or items that need clarification in a witness’s story. Players get one roll at the end of the initial scene investigation or witness interview. If they conduct a more thorough search, expand the search area (by searching the whole house instead of just the room where the body was found or canvasing the neighborhood for additional witnesses, for example), or ask more questions, they can make additional rolls. If the amount of time the characters spend on the investigation is important to the story, assume 30 minutes for the initial “once-over” investigation (including talking to witnesses at the scene) and an hour for each additional roll.
Players find clues based on how well they roll, as follows:
- Obvious clues should be given to the players regardless of what they roll.
- Players with low Success Degrees (5 or less) find a gimme clue.
- Players with medium Success Degrees (6-12) find an easy to miss clue and a gimme.
- Players who get high Success Degrees (13+) find an inconspicuous clue, an easy to miss clue, and a gimme.
- Conditional clues can typically only be found if the characters take an appropriate action, but in some cases a Lucky Break may lead to an unlikely series of events that reveals a conditional clue more or less by accident.
- Vital clues should be found first, and the GM should try to give them to players who rolled appropriately (this is usually easy to do if there are several characters investigating the scene) and mix them in with some non-vital clues so it’s not immediately obvious which clues are most important. If PCs fail to find a vital clue, the GM should try to find some way to introduce it, either during the investigation or later on (for example, by having a police investigator GMC notice it later). Since only doing this kind of thing with vital clues will meant the players know anything introduced in this way is important, it’s a good idea to occasionally have non-vital clues and red herrings turn up without a roll as well.
- If there are no clues left in a category that a PC’s roll has “earned,” the GM should give him clues from lower categories first and higher categories only if they’re the only clues left uncovered. If all the non-conditional clues have been found, inform the player that there doesn’t seem to be anything left to discover.
When giving out clues, the GM should take the players’ descriptions of their actions into account. For example, if a player states that his character is going to search the kitchen, that character should only find clues that are in the kitchen or that the character is likely to notice on the way to and from the kitchen. If there aren’t any clues in the there, that character won’t find anything regardless of what he rolls.
Once the characters have found some clues, the next step is to try to figure out what they mean. In most cases, that’s entirely up to the players and doesn’t require any rolling, but there are two exceptions:
- If interpreting the clue requires the character to perform some sort of analysis, research, or other action, the player should make the appropriate roll. Examples include testing a slime sample to identify what kind of monster it came from, analyzing blood spatter, and searching through The Herrick Pocket Guide To Cults and Secret Societies in hopes of identifying that symbol that was drawn on the wall in blood.
- If properly interpreting the clue requires knowing something that the character would probably know but the player doesn’t (or that the player seems to be forgetting), the player may make the appropriate roll (usually Brain) to remember the necessary information and recognize that it’s probably pertinent to the investigation.
I rarely do very much planning for the games I run. One of the benefits of using a rules-light system is that it’s easy to make things up on the spot, so I don’t see the point of spending a lot of time drawing maps and looking up stat blocks that might or might not get used. That doesn’t mean I improvise everything, just that I try to keep my plot ideas as open as possible so the players have plenty of room to make their own decisions. What kind of planning I do depends largely on what kind of game I’m going to run. Since a depressingly high percentage of games I run are convention demos, most of them fall into five broad categories for planning purposes.
These are games where the basic premise is just a genre, like “Sword & Sorcery” or “Super-Heroes.” I usually run genre demos either because I want to show players how QAGS can be used for the genre or because when I’m coming up with games for the convention I think “I’d like to run a [genre] game.” Since I tend to run genres I’m familiar with, I usually just think of a conflict, often a specific bad guy with some kind of scheme (“Anathema The Dark is trying to steal the Cauldron of Despair to raise a zombie army”) and rely on genre conventions to get the players caught up in it. I might also think about a few scenes, locations, or supporting characters that are likely to show up or that I’d like to use, but I rarely try to plot out exactly where and how they’ll appear in the game.
Setting demos are games I’m running to try to get someone interested enough in a setting to buy the book or PDF. For obvious reasons, these make up the majority of my convention games. Nine times out of ten, the game description in the program is generic, so I’m not locked into a particular plotline. In that case, I usually have an idea of which existing scenario I can use (“seems like a Sharkcano kind of day”) or a vague plot idea similar to the ones I come up with for genre demos (“I’ll have them go up against Al Capone’s voodoo master”), but those are just back-ups. I prefer to rely on setting knowledge to build a plot around the characters that the players create at the table whenever possible. For example, a couple of years ago at GenCon I was running a generic M-Force game and went in with the fallback of "I'll throw some goblins their way and let them track down a massive nest." When the players decided they wanted to be members of the Honolulu field office, the goblins got tossed out for a plot about a secessionist cult who used a magical tiki mask to command native monsters to carry out a political murder, then summoned a volcano god. The goblins would have worked fine, but the plot probably wouldn't have been resolved in part by PCs doing "white hula magic" to counter "The Hula of the Damned." In addition to having a plot I can use, I also go into setting demos with a mental list of setting-specific tropes that I want to try to include, either because they show off the setting particularly well or because I just happen to like them.
These generally fall into another category as well, but if you write up a con blurb that promises a mystery, making it up as you go along won’t cut it. At the very least, you have to go into the game knowing what the mystery and the solution. For example, if you’re running a murder mystery, you need to know who’s dead, who killed them, and why. You might be able to create the clues based on player actions with just that, but it’s best to have at least a few big clues in advance, along with likely ways the characters can find them. The key is to make sure you don’t create a mystery that requires the players to follow a specific flowchart to reach a solution. Wherever possible, make sure there are several different sources characters can use to arrive at key information. Getting the PCs involved isn’t usually an issue, since most mysteries are either “mission” type games where solving the mystery is the characters’ job (a private detective or CSI game) or “bottle” games where the characters have little choice but to solve the mystery, either because there’s really nothing else to do or to save their own skin (the standard “how to host a murder mystery” set-up).
Character Driven Games
When it comes to conventions, character-driven games require pre-generated characters. Otherwise you’re basing your whole game premise on the hope that the players will create interesting characters with their own plot hooks and that you’ll be able to improvise a game based on those plot hooks. This does happen from time to time, but not often enough for it to be a safe bet. Most of the character-driven games I run are based on pop-culture, either as direct adaptations of some book, movie, or TV show, or as a mash-up of two franchises. The most important aspect of preparing for this kind of game is to make sure that the characters are designed with built-in plots and reasons to pursue them. You’ll probably also want to think about some likely scenes that will give each of the characters a chance to be in the spotlight, along with possible stories that can lead to those scenes. The amount of plot planning required depends a lot on the premise. My usual plot outline for “Muppets of Sherwood” is along the lines of "Sheriff is up to something dastardly; NPC merry muppets will get captured and possibly nearly hung; make sure Gonzo gets a chance to do crazy stunts and Kermit Hood has plenty of opportunities for swashbuckling and sharpshooting." On the other hand, running "Lock, Stock, and Two Ruby Slippers” for the first time required very detailed (for me) planning because the concept needed to at least somewhat adhere to the Wizard of Oz story structure. So in addition to translating Dorothy, Tin Man, and the others to the modern-day London crime underground setting and giving them reasons to want to see the Wizard, I had to work out, for example, what "Follow the Yellow Brick Road" would mean in a Guy Ritchie movie.
Note: The characters for both "Muppets of Sherwood" and "Lock, Stock, and Two Ruby Slippers" can be downloaded from the Hex Pre-gen archive.
Sometimes I run games like "Players Choice" (the players choose what genre they want, roll up characters, and the GM runs it) and "Life is Random" (the premise of the game is generated randomly using The Book of Dumb Tables or The Book of Dumb Tables 2). The whole purpose of these games is to show off the versatility of the QAGS system. Since by design it's impossible to plan them, those succeed or fail entirely based on the ability of the GM (and players) to improvise. The only time the GM has to plan is while the players are making characters, and usually at that point you’ve just got the genre or setting to go on. I spend that time trying to come up with a good conflict or villain, but also trying to think of what parts of that genre the system can handle especially well so that I can try to include ways to highlight how well the system works for that particular situation.
And When I’m Not Running A One-Shot...
When I actually get to run a campaign outside of a convention, I combine all these plotting schemes as needed, but for an overall storyline I usually rely on knowing what’s going on in the world independent of the PCs. If you’ve got a world where lots of different characters and factions have their own agendas, and those agendas intersect with one another or the goals of the characters, the adventures basically write themselves. I usually also come up with ideas for a few generic stand-alone adventures/encounters (for example, a haunted house to investigate or highwaymen attacking people along the roadway) that I can drop in when things get bogged down or I feel like players need a break from the ongoing story.
Unless you’re running a dungeon crawl or railroading the players, or are lucky enough to players who can generate character-driven plots that you just have to flesh out as the game goes along, you’re going to have to do some planning. Most people probably plan more than I do, but if you want to keep the game open-ended, it’s still important to think about what you’re trying to do in the game and plan the things you need to get the point across.