The debate between proponents of “sandbox” and “railroad” design have long been a mainstay among gamers. Of course, the argument is not nearly as cut and dried as we tend to make it. In a section on adventure design in FASA’s Doctor Who Role Playing Game, Wm. John Wheeler describes precisely how both linear and free form scenarios can be put together for a more entertaining and robust design:
The best published scenarios combine the two types, using some linear encounters and some free-form encounters. Linear encounters are used to introduce the scenario, drawing the players and the characters into the action, giving them a reason to enter the scenario environment and meet the scenario NPCs. After the ‘hook’, as the introductory encounter is sometimes called, the linear encounters lead the player characters into a situation that gives them free choice about where they will proceed. The actions in each of the free-form encounters affect the players in the short term. In the long term, another set of linear encounters lead the players into yet another area of free choice, perhaps the climax of the scenario. Linear encounters are often used to wrap up the scenario, bringing it to a satisfactory conclusion.
Using encounters of both types is like building a structure with tinker toys, with the sticks being linear encounters and the knobs being the free-form encounters. The linear encounters give some structure to the free-form encounters. The combination allows the scenario to have a well-defined story line– not as well-defined as purely linear scenarios, but much more defined than those that are purely free-form. The combination also allows players freedom to choose their action– not as much as in purely free-form scenarios, but far more than in those that are purely linear.
He follows this up with a breakdown of three campaign types:
- Ones “where the players characters are all the same even though the scenarios do not have much to do with one another.”
- Ones where “the scenarios all have to do with the same topic, perhaps approach it from different angles, possibly with different characters.”
- Ones where “the same characters follow the same plot from adventure to adventure, solving puzzles along the way and discovering more and more about the plot as the adventure scenarios progress.”
Wm. John Wheeler’s remarks on adventure design are about as cogent as any that I’ve seen. What he is describing here is what I call situation oriented play. It’s a natural fit for creating adventure in the context of an otherwise infinite universe. A certain segment of players don’t want to be able to go anywhere and do anything– they want to, as the old Infocom ads put it, “get inside a story.” Indeed, you can see the tinker-toy structure quite clearly in many of their games. Planetfall, for instance, opens up with a linear sequence that leads in to a more free-form situation.
Wm. John Wheeler’s outline here is practically a template for how many of the adventures that were written for second edition GURPS. Given that his game was published a year beforehand it’s interesting to see how his views compare with those of Steve Jackson. It’s odd, but he spends more time in talking about the wrong way to do it than he does in spelling out actual techniques and design principles:
- “In children’s fantasy games, every encounter may be rolled randomly!” He’s calling out the random dungeon generation sections of Advanced and Basic D&D here without any any recognition of how these sorts of techniques are essential to creating more sophisticated, free-form play in a notionally pure sandbox.
- He makes snark at the “hack and slash” dungeon in which the rooms are stocked willy nilly with no thought given to why the monsters are there or what motivates them.
- He sneers at the “plot of the story” of the typical dungeon adventure even though a session where only one player character makes it back can be quite engaging, generating narratives that are told and retold at conventions.
- He hammers the point that earning character points for good role playing is far more sensible than XP for the “amount of wealth you drag home.”
Steve Jackson must have been traumatized by some pretty awful D&D sessions! A lot of digital ink has been spent since then explaining why his views here are wrongheaded. He just didn’t seem to “get” old school play… even though he was involved in the industry at the dawn of the hobby! He can perhaps be forgiven for this due to the fact that the case for why things were being done the way they were was never made plain in any of the rule books for the games that he was reacting against. In any case he certainly spoke for a lot of people that had a similar reaction and that wanted something more.
What was his ideal for good adventure design then? Given the thrust of GURPS design, it’s clear that he was really taken with making it easier to move characters between different game worlds and campaigns. He practically describes the structure of Infocom’s Deadline when he discusses more advanced adventure plots that have things going on apart from the player characters while they engage in the situation. His best advice centers on nailing down the climax of the adventure, but leaving the players to get to it whichever way they choose:
The players earlier actions affect the details of the finale, but its basic nature remains the same. If the players make “wrong” decisions along the way, it will take them longer to finish, and they should have a harder time dealing with the situation — but they should make it to the finale eventually. The exception might be a case where they have blundered so badly that the finale would certainly kill them all — in which case, the merciful GM will drop a hint that they are over their heads, and let them give up and run for home.
A more sophisticated adventure will have several possible finales, depending on decisions made by the players during the adventure. This sort of “branching path” adventure is harder to design, but sometimes easier for the GM to run — less improvisation is needed. Such an adventure can be played several times, making it especially suitable for “programmed” adventures.
Taken together, Wm. John Wheeler and Steve Jackson provide a comprehensive breakdown on how to craft an adventure plot that typifies the sort ideals that mid-eighties gamers were looking for. This type of game is very different from the older “sandbox” style of the seventies. While not necessarily better or more mature, it does acknowledge player autonomy while creating something much more recognizable as being a story. It accomplishes this without having the more blatantly obvious “rails” of something like Tracy Hickman’s infamous Dragonlance series.