• The VOIDRUNNER'S CODEX is coming! Explore new worlds, fight oppressive empires, fend off fearsome aliens, and wield deadly psionics with this comprehensive boxed set expansion for 5E and A5E!

Guide to Adventure Writing


First Post
Here are some finished passages from a larger writing project that I'm abandoning. I posted this on my blog last week, but I thought it would be interesting to post here as well. I do believe this is my first post to the discussion area here, but I've been reading the "oldschool" posts for a few weeks now and I suppose this is a good time to put my neck out there and see how this is received here.

Success and Failure

The most important thing to remember when constructing an adventure is to not assume that the PCs will succeed at any point during the adventure.

As a referee, your job is to be completely impartial during game play. You have absolute power at the game table and can bequeath success or mandate failure at any time. Doing either of those things ruins the game, as both give no incentive to play well.

Do not fudge the dice. Ever. Luck is a part of the game, and the dice are there for a reason. Resist the temptation of sparing characters that fail or even die due to “bad luck” or a “stupid die roll.”

Would it be acceptable to tell a player that just rolled a stunning success that you’ve decided, just because it’s more fun, that the die roll doesn’t count and he instead failed? I don’t think so. So why would ignoring the dice in the players’ favor be acceptable?

Good game play will tip the scales of fortune and those that rely on pure luck deserve what they get – either way. At the same time, if an incredibly lucky roll derails the entire adventure and gives the players a quick victory, it should stand. It needs to work both ways. When the dice go badly for the players, they should be thinking of how to not let a roll of the die be the sole determiner of their fates. And when the dice go a little too well for the players, the referee should note what he needs to do to prevent a single die roll from determining the course of an entire adventure.

Traditional games are all about the players (and referee) learning to play better over time. The characters’ experience gains are secondary. Demand and reward player excellence and the game will be more challenging in the long run.

So what are the consequences of deciding to play this way?

The party is just lost and sitting around because they didn’t find the secret door that leads to the next section of the dungeon? Tough. It goes unexplored.

The party missed a vital clue and has no idea where to turn next in a murder investigation? Tough. The killer gets away.

There are too many options to choose from, and the players are disorganized and can’t agree on an option and look to the referee for guidance? Tough.

This only works if the referee is willing to realize that sometimes, all his work on an adventure is going to be wasted. The players are sometimes going to be unwilling or unable to see it all. The referee must contain his ego and resist the urge to introduce some way of being able to show all his work off. And the referee must not take the unused, unexplored parts of his adventure and plug them in elsewhere, as this negates the choices the players have made that led to them, intentionally or not, failing to explore the areas in this particular location.

Playing this way also means that the game can “stop” at any time because a battle wipes out the PCs, or some other disastrous result that means the mission will come to an abrupt end. Oh well. Of course success is always more fun than failure. But if failure is not an option, then the success is but an illusion, it’s fake, it’s a lie. And by taking the attitude that the end result determines the fun of the game, then suddenly the process of playing the game is not fun in and of itself.

I don’t need to say anything about how stupid that is, do I?

Deadly Situations

Every adventure must have situations that directly and truly threaten the lives of the characters participating. If there is no true threat, it is not an adventure, it’s a tour.

I'll go so far as to say there should be situations designed specifically to kill characters. A monster that's way too tough. A trap that's going to claim a victim. Save or die. These sorts of things. Every. Single. Time. The key is to put these "expected death" situations in places where it isn't necessary to encounter them. The players must choose to engage in these areas and situations.

Teach them that the game world isn't scaled so they can kill everything.


Every adventure must have meaningful choices that the players must make, and these choices must significantly alter the flow of the adventure for them to have any meaning.

The absolute key to good gaming is the ability of players to choose their character’s actions. Any adventure which dictates what a character thinks or feels or does (barring magical enchantments, of course) is a terrible, terrible adventure.

The choices made must be real choices. “Floating locations” of the “Well, whichever inn they stop at will be where the adventure happens” sort is not a real choice, it’s a mere illusion. This is worse than railroading because it is dishonest in its methods.

Choices should not only be offered, but forced: Things are happening, and the players have to do something, and none of the options seem to be all good. Of course, if they choose to not do anything, they’ve still made their choice and the consequences should be different (and more severe!) than if they’d done something.


There are two standards that adventure rewards must meet: They must be enough, and they must be not enough.

Enough that everyone involved doesn't think that they've completely wasted their time... and not enough to leave anybody really satisfied with what they have. They need more! Where next to plunder?

Note that concealing the rewards well may wind with the players not finding it. Tough. As a referee, just make sure it's there. Don't help the players to actually find it.


A player-driven adventure challenges the now-common philosophies of good adventure pacing. Common wisdom today states that if the action has slowed and the players either don’t know what to do or don’t want to do anything, the referee should make something happen to give the players something to react to. I declare that this ruins the pro-active element in traditional gaming, causes the referee to be biased towards character action, and creates a disincentive for players to control their own destiny.

But what do you do if all the obstacles described in the Success and Failure chapter actually stop the party?

You do nothing.

If a player complains that he’s bored and that nothing is happening, look at him and say, “I agree. So are you going to do something or not?”

It is not the referee’s job during a session to provide excitement for his playing group. His job is to administer the setting and resolve character actions. If the characters are taking no action and are not interacting with the setting, then the referee has literally nothing to do. The players are wasting his time.

Other common standards of pacing become obsolete when dealing with a player-driven adventure. Traditional games commonly feature a “retreat, rest, and recharge” element of play, and in fact almost demand such a thing. This creates a bit of difficulty in trying to structure an exciting adventure if the party is going back to rest after every fight of even slight challenge.

Don’t let the players turn the game into a series of “Scout out the next room, ambush the beasties, collect the loot, and then retreat back to camp and get all the spells back.” Or don’t let them complain of monotony and boredom if that is what they choose to do. There are a variety of ways to prevent this, although some may seem heavy-handed. Cave-in traps or other methods of blocking exits can be useful, once, before it becomes a crutch instead of an idea. Pit or slide traps that dump a party to a lower level and teleporters that move a party somewhere unfamiliar are old tricks that might be acceptable a time or two. Missions with time limits are another possibility, but the meticulous planning needed to make an adventure just challenging enough will tend to cause the referee to become too invested in the adventure outcome.

The first reliable way to control this is through the proper use of wandering monsters. Never skip a wandering monsters check, and never hand-waive the results. Do this for the area that the PCs decide to rest as well. If their recuperation is not just a matter of saying, “We go back to camp,” maybe the players won’t be so quick to do so.

Keep a strict record of time! This wisdom was presented in bold in a major publication and has been laughed at ever since. But it’s excellent advice. Endless searching for secret doors and traps takes but a second to roll for the players, but a good deal of time for the characters. How long does that torch burn? And that lantern? So many referees simply make sure that there’s a torch or lantern present (and if the referee is on the ball, he might make sure that somebody with a free hand is actually the one carrying it) and then ignore it. Players will pick up the pace if the torches and lanterns keep going out… and keeping close track of encumbrance means they can’t just buy a hundred flasks of oil, either. These oft-ignored rules aren’t there to be a pain the ass, they are there to push play along in a system that otherwise rewards characters moving at a snail’s pace.

But when the players go looking for adventure… you’d better have some for them to find.

Dungeon Design

You’ve been given a big pile of philosophy concerning adventure design, but now it’s time to put it together into a coherent adventuring environment. As a nod to this hobby’s traditions, this environment will be called a “dungeon” here, but this remains true if the environment is a dungeon or not.

The first thing to remember when creating an actual adventuring area is to forget the idea of “encounters.” The “encounter” has become known as the standard unit of “excitement” in an adventure. It’s an awful terminology, and it influences adventure design in an adverse way as referees stop to think of adventures as a flowing, natural sequence of events and more like a flowchart where players travel along boring lines in order to get to the “encounters.”

Never place a secret door that you intend to be found.

Never place a trap that you do not intend to be set off to its full effect.

An important factor in designing a dungeon is allowing for the fact that under the guidelines presented in this book, characters will die. Perhaps often. Replacement characters are often rolled up very quickly, but there needs to be an in-game explanation for how to introduce these replacements.

Create challenges for every primary class in the game, especially those that are not present in the player character group.

Spellcasters, particularly clerics, always have a number of spells available to them which they simply never prepare. This is due to referee laziness; off course they are never chosen if they are never used! Create situations where such spells easily solve the issue at hand. They’ll gripe and moan at first when they realize they have to come back to the situation the next day with the proper spells (and complain yet again that they are doing so at the expense of “useful” spells… you know, the type used in combat), but a referee being diligent in this course will remove the idea of “useless” spells from his campaign altogether.

Interaction versus Combat versus Traps versus Tricks

It is usually better to present an encounter with a greater number of enemies than it is to give the players one opponent at a time. It's attrition versus “The Big Fight.” Make smaller, less threatening opponents the order of the day, so that the decision to continue on or stop and rest is actually meaningful. If every encounter is a big one, then continuing on is stupid. This is advice that I really have trouble with in my home games. I can usually eyeball a Single Big Monster for suitability against my group. But there get to be a lot of dynamics when it comes to group encounters. It's a bit difficult with groups of creatures, because for all the "kill 'em all and sort the character sheets later," tone all this advice takes, the ultimate goal is to challenge, not annihilate, the players and their characters.

Make use of terrain and “set pieces” when coming up with encounter areas.

Kill Them and Take Their Stuff – Complicate it! Vary what the treasure is, hide its value, make it inconvenient to transport.

Random Encounters

Random encounters are a wonderful tool. They keep players from ever feeling secure about their position in an adventure location, they can turn tense situations into complete chaos, and they are just good all around fun. Never fail to create a random encounter table for your adventuring locales. While most of the random encounters should not be major battles, there should be at least one possible encounter that will be a roughly equal, major fight, and one entry which will probably be too much for the party to stand toe to toe with.

Note that many old modules poo-poo wandering monsters by advancing the idea if a random encounter depletes the party too much or detains them from their final goal, the encounter should be ignored. This sort of thinking is drenched in the notion that the game is somehow a failure if the characters do not reach the pre-scripted conclusion in just-so condition so that they can deliver a satisfying climax to the adventure the way they are supposed to. Isn’t that the sort of thinking this entire essay is trying to avoid and prevent?

Take care that the random charts make sense within the adventuring environment. These creatures roaming around will also be coming into contact with the placed creatures. Why aren’t they killing each other? If they’re random monsters, it’s a good bet their lair isn’t keyed on the map. Where do they live? How do they get from there to the dungeon? If the party is closing doors behind them as they go, many creatures won’t be able to “randomly” appear.

One solution is to make random monsters connected to a keyed area. This can happen in several ways. The first is to just assume that every (or most, or whatever’s appropriate) keyed area’s inhabitants have an extra member or two running around the environment.

Also, not every random encounter needs to be a battle. Adding in neutral or friendly encounters into the table can provide an unexpected twist. The encounter need not even be with anything living. A cave-in, flash flood, or other random event can easily fill a random encounter table slot.


Think before placing traps. Really, there is no quota for placing traps and they should never be thrown in there “just because.”

Three things must be thought through before placing any trap. First, what triggers it? Second, how do people who are supposed to be in the area avoid the trap? And third, why hasn’t the trap been triggered by all the wandering monsters (and regular nearby inhabitants)?

In instances such as a tomb or mad wizard’s lair or some such, these are easily answered. Nobody is supposed to be there, period, and it makes sense to booby-trap the living hell out of the place. Locations with living inhabitants, not so much. But each trap should have a clear purpose.

Be descriptive about placed traps. It should be possible to detect and disarm almost any trap without making a die roll. In fact, if the proper way of dealing with a trap is nothing more than a couple of thief skill rolls, then the trap is boring and no good. You can do better.

“Gotcha” traps keep players on their toes, but are also detrimental to game play. Merely entering an area shouldn’t be enough to trigger a trap. There should be some specific action that triggers it. Poison needle traps are a perfect example here. If a character does not attempt to open a chest or pick its lock, they have no problem. It’s only by taking a specific action that they put themselves in danger.

Not that this is a screed against pit traps and the like. They have their place – especially if nobody is bothering to use a ten foot pole anymore. The problem with such traps is that they are often in areas where many creatures travel. Not even the most diligently trained or fiendishly clever beast will walk amongst traps unless there is an ongoing siege or hostile information. Any “triggered just by standing or walking right there” trap that does more than sound an alarm is simply not going to be found in areas where people, or creatures, ever go.

Obvious, no-roll-needed-to-find-them traps are simply awesome. They dare the players.

The last consideration to make is whether this trap is effective. Too many referees place traps as “obstacles” in their adventures to be “overcome.” Traps should be placed with the full intention of being triggered. Whoever set the trap was certainly aiming to kill (or imprison, or immobilize, or whatever) whoever set it off, and certainly trying to keep people out of a specific area, so it must be able to do what it sets out to do or the whole thing’s worthless. If you’re going to place a death trap, set it up in a way that will kill, and count on a character dying from that trap during the adventure. When (if!) the traps are discovered and bypassed, it becomes a real accomplishment (even if it was dead easy and the players don’t understand what might have happened), and not just something that happened because it’s “supposed to” in these types of games.

log in or register to remove this ad


I run Compose Dream Games RPG Marketplace
Thanks for sharing your thoughts. It's given me a fair bit to think about.
I especially like the section on Pacing and player driven games. Another way to combat the attack and retreat symbols is to have organized resistance, and possibly proactively. (If the PC's leave the Dungeon, a group of 'monsters' tracks them down.)

All it really needs is a brilliant conclusion.

I might mention that there should be multiple reasons to travel to any dungeon, and Dungeons with multiple enterences and exits should be the norm. This is especially true if most wandering monsters are to make sense.


First Post
Thanks for the great post, Jim! I'm running a 1e game, and it's pretty much all adventures of my own making - some good stuff here to keep in mind when I'm writing up the next adventure.


I agree with most what you say, but it entirely depends on the players.
Some players are very happy with being railroaded through a story as long as they can kill monsters and write plusses on their charsheets.


First Post
because for all the "kill 'em all and sort the character sheets later," tone all this advice takes, the ultimate goal is to challenge, not annihilate, the players and their characters

Lots of good advice here for keeping the "challenge the player" element in a game, as well as adding meaning to "wining" or at least surviving.

The above quote points out, however, that this style of DMing requires the DM and players to both agree that they are ok with both the player and DM learning curve. For instance, some DMs I know don't fudge dice just because they want the PCs to win at any cost so that the story can go on. They fudge dice to remedy a poor situation they created unintentionally. For example, they are not familar enough with the rules, monster power levels, etc. and create a situation that is 99% doomed to give a TPK when what they thought appropriate to the location, etc. was a 50% chance of success.

Still, I get your main point which is an important one. How do you maintain a sense of accomplishment in the game if the players know that no matter what, they are going to beat these giants, then go to the drow city, then kill loth (even if it takes them 20 characters to do it...)?


More than a few good ideas and reminders, Jim. I especially like the theme of not building in player-tailored challenges and circumstances that virtually guarantee success. Such a design approach ultimately leads to shallow, meaningless victories.

I like the tough love, AD&D Dr. Phil tone, heh.


Victoria Rules
Only thing I'd add...but I'm not sure where...is that one aspect of design that can make adventures more interesting is to have some variety in opponents. Most modules do this as a matter of course, but if you're writing a homebrew adventure and the idea of it is to take out a city of Goblins, you can bet the players are going to get mighty tired of fighting (or avoiding) Goblins after the first 3 or 4 or 10 times they do it. Mix it up somehow...have the Goblins own some interesting pets or slaves that get sent out against the party, whatever.

Otherwise, excellent work. :)



JimLoTFP; Thanks for this. A total breath of fresh air after all the mollycoddling advice we have had since 3E was published.

Makes me want to break out the Megadungeon again after all these years.

Remove ads