Places to Go, People to Be [Next Article] [Previous Article] [This Issue] [Home]

Questions of STRUCTURE

by Antoine Dinimant

"Has the adventure started yet?"
Baloche the Mythical


So you've lovingly shaped your NPCs and cooked up a savoury plot? Well done! But before writing your adventure's story line, have you thought about how to structure it ? Be it linear or open, here are some terms and tricks for your adventure designer's toolbox.

An adventure's structure is not its synopsis, but the way it will be played, or more precisely the ways in which the events can happen. To put it more simply, it's the map of the various paths the players can follow to go from the beginning (the introduction) to the end (the goal). The degree of freedom given to players depends wholly on the adventure's structure.

The tower example and the concept of deceptive freedom (liberté illusoire) are quoted from the leading French RPG review, Casus Belli.

First we have to agree on what we mean by 'freedom'. Let us first consider the case of a perfectly linear dungeon-type adventure. The players must enter a tower and get into the treasure room at the top. Theoretically, the players are free to do whatever they want, but at each stage, they only have one way to go forward. This is the problem of linear adventures, that confine the players in a empty freedom.

Now, what happens if ever the players imagine an unexpected means of bypassing the adventure's traps and getting directly to the top? Any GM's first reaction will be to try to prevent them from implementing their idea, using more or less convincing pretexts. A second one is enforcing the principle of deceptive freedom. Here the GM turns his map bottom up and puts the guards' room under the roof and the treasure in the cellar. The players will have the illusion of getting an advantage over the adventure but when they get to the top they'll have to just make it all the way down again. They have been authoritatively put back in the planned path.

Of course, the problem is that if the players realise what the GM has done, they could feel quite frustrated - if their inventive efforts amount to nothing, why should they bother? Eventually they will always work out what is going on. Some can be satisfied[1] with the type of adventure game that consists in finding the key to each stage; others, more concerned with the role aspect of the game can reasonably think that someone is taking the mickey... [2]

Now remains the hardest thing for the GM to admit: acknowledging that the players have played well and that they deserve to skip some stages[3]. By doing this, the GM must accept improvisation and the opening up of the adventure.

Opening up a linear adventure

Opening up and improvisation come together, but, of course , improvisation is always better when it is carefully prepared. First, let's see some means of rehashing a linear structure into something flexible enough for both the players and their characters to have space to express themselves. I'll look at various ways of doing this which can be split into three groups: partial openings, time constraints and complications.

Partial openings include the open climax. Typically this is created by a moral dilemma: will they hand over the young rebel and fulfil their mission or on the contrary, will they embrace her cause and betray their sponsor? The climax is actually the easiest stage to open up, as long as the GM doesn't have to bother with the consequences. But it is precisely because an open end has no real impact on the adventure that it's an illusion to think that it alone can be enough to remedy a closed adventure.

So, why not move the moment for a choice to the very middle of the story? Then, you get a bifurcating structure which is much more interesting.

A bifurcating adventure does need a lot of design work. To begin with, you'd better not change the general plotline but just design some open stages. For each task imposed on the PCs, plan an alternative solution, and then several others. An example: the players must seize an item (an artefact, a treasure map, the key of the princess' bedroom) watched over by ten or so mercenaries. The linear adventure only includes a fight as the solution. The players could also try burglary (you can always fall back on a fight if this fails), they can try to be captured as slaves in order to poison food, they can pretend to be the nearest dragon's messengers or even bribe the mercenary sergeant into selling his little treasure.

This system of open episodes in a linear structure becomes much more meaningful if the way the players solve it has consequences on a later stage of the adventure. Not as a judgement on the players' solution but as the natural consequences of their actions in an evolving and open world. Let's say that some days later, the characters suddenly meet the mercenaries. Had they chosen fight, the enemy party now probably includes some impaired warriors, which will ease things; but had they succumbed to bribery, they could now easily do it again...

The simplest time constraint is the timed adventure. This way of putting of the players under pressure is always a good means of having a more dynamic and involving adventure. Any effort they take to go more quickly will be its own reward. And well, if you have a linear adventure, the best thing is that it flows quickly: there's no point in letting the players hit their heads against the walls of their useless freedom. The hard part is giving an suitable deadline: if it's too short, you won't be able to loosen the rope without losing much of your credibility; but of course, if too long, your scheme will be spoilt[4].

The solution could be not announcing the deadline, but to hang a Sword of Damocles over their heads, likely to fall at any time. For example, they have to find a drug to cure a dying person who can die at any moment. In theory, it's much more stressful than a known date; but in the game, it raises a big problem of credibility: the players will tend to think they will always arrive a heartbeat before the end. And if the death comes too early, they will protest against the GM's arbitrary decision. The only means of being completely believed would be to announce the chances of a fatality in every set period, and to throw the dice before the players. Make sure the suspense fits the adventure or the players may complain about being subjected to a lottery.

Another sort of timed game is competition. The idea is simple. You keep a linear adventure as it is, but you add a second party who is attempting to reach the same goal as the players. The implementation is not so easy: you have to balance the two parties very carefully to avoid that either[5] be tempted to kill off the other too early. You'd better rely on suspicion, so that the rivals discover each other very gradually. That kind of problem also occurs in chases, but the solution depends on an imbalance: the chasers must be powerful enough in order that the prey is not tempted to take them on in an ambush.

To end this piece on opening up linear adventures, let's see two sorts of complications. First, the system of composite adventures[6] consists in having several adventures played at once, one being the main plot and the others being subplots or misleading tracks. This again requires much preparation work from the GM but also a deft touch in interweaving the threads in a believable manner. A good example would be Buffy The Vampire Slayer, a bad one Sunset Beach.

Secondly, you can entrust a player, some players, or even all players with personal secret missions. These missions should oppose, at least partially, that of the party or of another player: stealing a treasure, assassinating an NPC who could be helpful for the party, or even capturing or killing another PC. An example that is good for roleplaying is when a character assumes the part of the traitor to the party. Of course, the greater is the treachery, the likelier the character is to be a one-shot, should he succeed or fail. So you have to be very careful not to go too far: in Paranoia, the rule was that every PC had the secret mission of eliminating another PC, but it worked only because there were six clones for each player...

Building blocks for an open structure

An openly structured adventure is actually quite simple and quick to design. On the other hand, it's the most difficult to GM. The risk is that because of their extreme freedom, the players have no idea of what to do and get bored. I'll here describe four designs of open adventures, which may be combined and don't exhaust the idea.

Open tourism[7] is an extreme genre, which, in my opinion, is not to be used beyond one session. Rather than on an adventure script, the GM relies on a place (usually a town) sufficiently original and well-described ; the characters have no immediate aim but exploring the place and/or precisely finding a sponsor (and henceforth an mission objective).

The USC (a personal shortcut for "up s**t creek") is a good drill to get acquainted with open adventures, and to train your players as well. It usually includes a very linear part 1, followed by a dramatic change of situation. The principle states that the adventure designer must not have any solution drafted: it's the players' job to invent one. Of course, the USC should not end arbitrarily. It's important that the trap has certain flaws, which good players can discover and use. For example, the enemy is split into two factions, a key NPC is open to bribery, to seduction, to threats against his family...

The calendar is the very tool of open adventure. Its principle is simple and well-known. The adventure script follows a set course of events, with room for PCs and reasons why they should intervene. The NPC's actions are scheduled according a precise calendar, which unfolds as is "except if the players interfere". Actually, this is designing a linear adventure for NPCs and an open one for players, so it's simplicity itself for you and tough for your players (a cool mixture).

Open on both sides, the chess game is depends upon an enemy (be it an individual or a group) whom the players will confront gradually. The adventure designer has to sketch the game board, the enemy side and the overall reasons why the confrontation will be fought indirectly, each side trying to ensure an advantage rather than attempting to win an immediate victory. Why should the PCs not go straight and bash their enemy's head in? Because he's got too many guards, his position must first be weakened; because he's not known or it's, for example, a secret society that must be uncovered or infiltrated gradually.

On the other side, why their enemy should not have the PCs arrested or assassinated? Either they're too popular, he first has to bring them into disrepute; or they're hidden, he has to find their friends and have them talk. Maybe they're being drawn in to a Machiavellian trap where they will be more useful alive and free. So, you'll have to define for the two sides a series of intermediary objectives (an alliance with a neutral NPC or group, a magic item, a hint for an enigma, an advantageous position...), and make sure that none be an indispensable key to the adventure..

Then, you can refine the game by adding a third side, or even loads of adversaries, so giving the hardly subtle but playful figure of buzkashi, that sort of Afghan polo without teams, where hundreds of horsemen are each trying to catch the "ball"[8].

The choice

From an abstract point of view, open structures can seem much more interesting than linear ones. The problem is that they demand effort from the GM, but even more so from the players. If they cannot figure out what to do, everybody will get bored. On the other hand, a nicely adapted linear story can be enough to give the players the impression they have all the freedom they want... if they think the GM is able to allow them that much.

This is the most important point: you must strive to maintain your GM credibility. When you'll have GMed them a tourism session followed by a USC they survived only by working out a madly complex set-up, I can guarantee they'll lose the habit of looking for the adventure's "rails" and they'll jump straight into it. As soon as they are not able to act as tactical gamers, or as puppets looking for the strings, they'll have no way but to act as characters in the middle of the plot. And it will be much easier for you to manipulate them without their suspicion.

When structure is concerned, the only rule is: change! As good it may be, a structure becomes a cliché the second time it is used. And moreover, as soon as they indulge in metagame thinking ("hey, he's giving us a hint"), catch them off-guard. So that no player ever again asks you if "the adventure has started yet."


[1] Well I suppose so, as most published adventures are linear.

[2] Actually, I tend to react in this way, but I don't want to force my ideas on anyone.

[3] With more or less honesty... you won't go into making the task too easy for them.

[4] If you must do this, make them believe that they dragged it from you through hard work on their part. For example, the hints they have gathered up to now can induce the governor into granting some days grace to their sentenced old friend, but only if they are convincing.

[5] The GM controls NPCs' actions, but he should nonetheless be able to justify them. Nobody will believe that a party of big blokes accepts being outrun by weaker PCs without reacting violently, unless they have a good reason, perhaps because they prefer the players to do their dirty work for them.

[6] I won't say too much about this system which is more widely developed by Mehdi Sami in his article titled "Breaking out of adventure linearity" (Briser la linéarité des scénarios, Casus Belli HS n° 25, mai 1999).

[7] Not to be confused with the linear genre of epic tourism, which consists in having the players go from discovery scene to discovery scene.

[8] Actually, a veal carcass...

Antoine is French. Yep, the land of tortuous grammar, Ray Ventura and Cheese. He has a strong dislike for fish, his favourite word is "obsidional" and he spends his time translating Steve Darlington. We have returned the compliment by helping him with an English version of this excellent article from his website.

[Next Article] [Previous Article] [This Issue] [Home]

Copyright © 2001 Places to Go, People to Be, all rights reserved. May only be reproduced with permission. Refer to the copyright page for full details. Email us: