Monday, September 26, 2011

Plot Hooks and the Sandbox

One thing I've been mulling is the limiting factor "plot hooks" place on player agency in the sandbox.  There's a common sense element at work here; if a person is given limitless choices, it can be hard to make *any* choices.  But give someone three interesting options, now there's something to weigh and analyze and make a decision.

I kicked off a discussion about illusionism (The Shell Game in the Sandbox) a few weeks ago to work through how far a DM could go in having premade content to drop into the sandbox, before crossing over into illusionism and railroading.  There are similar pitfalls with "plot hooks".  There seems to be a slippery slope between a plot hook and being led along the DM's "story".

To establish a common lexicon, let's say that plot hooks are a subset of information about the game world.  Players can gather information Actively or Passively.  Active information gathering involves creating a plan to find information - it's a pull model; passive information gathering is pushed by the DM.  A "plot hook" is passive information pushed on to the players that provides a definitive chance to have an adventure - the classic plot hook is a job offer or similar patron encounter.   A plot hook is not a predefined story; it is a way of delivering specific, actionable information about the entry point to an adventure.  However, I'd be glad for folks to chime in on refining these views on presenting sandbox information - these definitions were off the cuff to help frame the discussion on plot hooks.

One of the biggest criticisms regarding the free form sandbox is identified in that very first paragraph; the players start in a tavern, with no idea what to do next, GO.  And they just sit there, mouth agape.  They could get jobs as dishwashers, or they could plan to rob the bank. They could walk in a straight line north-north-east out of town, for as long as they wanted.  They could start a bar fight.  If they have too many options, there's a good chance they'll seize up with the analysis paralysis.  The complaint "nothing happens in a sandbox campaign" really just means that the players are unprepared for the responsibility of active information gathering.

But my interest here lies in passive information:  the plot hook.  Plot hooks are practical, because they promise direct routes to adventures.  That's why players show up, to do something interesting, and not for living out their lives as fantasy dishwashers or scullions.  But how far can you go down the road of using plot hooks before it crosses into illusionism or railroad - ie, you have total freedom of choice, as long as you make my choice?  It clearly seems to be a type of railroad if there's only one plot hook offered; how about if there are only two plot hooks? Three?

I don't see much discussion regarding the use of plot hooks and patrons and job offers, and their intersection with a player-driven game.  If there are multiple (distinct) opportunities to pursue at any give time, I can see an argument that a game using plot hooks is still player driven; it's up to the players to pick which opportunity to pursue, and an open-ended plot hook should let them create their own plan on how to go about exploiting the opportunity.  But don't lose sight that the players have allowed their nearly limitless choices (in the active scheme) to collapse to a much smaller number of options when using plot hooks.

Megadungeons present an interesting problem; the DM has made the biggest choice for the players in the campaign - "This is a megadungeon campaign, if you show up to play, you're expected to explore the megadungeon".  It's the smaller choices within that larger decision where the players can execute the player-driven game  - preparing for each dungeon incursion, gathering information, deciding which entrance do they use, where do they explore.  This seems to cross over into a "social contract" area - they accept the game's gigantic plot hook and premise up front, out-of-character - it's a game about a megadungeon.  We certainly would consider these players bad sports if they showed up to the DM's new megadungeon campaign's first session, and promptly decided to leave town and head for the mountains, ignoring the detailed dungeon for the loosely defined wilds.

Thus, there's a question of fairness and equity in the sandbox, and it's related to the sandbox triangle (player freedom versus the DM's level of effort versus the amount of detail).  Imagine these two situations:  the DM prepares the next level of the megadungeon ahead of the game, the players show up, start planning the night's adventure, and decide to head north, into the wilds, to find the island of Avalon.  In the second situation, the players show up, continue their exploration of the dungeon, but at the end of the session, they mention to the DM that next week, they plan to head north and find the island of Avalon.

As DM's, we keep tools handy for generating content on the fly, so perhaps the first DM could improvise the overland journey to Avalon.  It's likely Avalon is just a name on the map, loosely detailed if at all; how well can the DM wing it?  And the DM's work on the megadungeon was a bit wasted.  Is there a duty on behalf of the players not to waste the DM's time and effort?  In the second situation, the journey to Avalon and the situation on the island will be more detailed before next game - more detail should also make it more interesting.

I've been thinking about these problems around plot hooks as I work through how a "wide area sandbox" would be structured.  In this case, the wide area sandbox would be the size of the Western Hemisphere - imagine Europe, the New World, the Spanish Main, the Mediterranean.  Quite a bit larger than a micro sandbox (the Keep on the Borderlands) or a regular sandbox laid out as a hex crawl.

In a micro sandbox or even a regular sandbox, I don't see why the DM needs any prepared plot hooks at all; those sandboxes can be so well documented up front that player choice is only limited by their active information gathering skills.  Plot hooks are only necessary to bail out players that fail Sandbox 101 - when the DM needs to light up the big neon sign that says, "Adventure is Here -->", via a patron, quest, or similar info dump.

In order to run a game that covers a much larger geographic scale, the plot hook seems to be a pragmatic necessity.  For instance, it's possible the players might randomly choose to outfit a ship and an overland expedition to the distant jungle colony on their own, but if that's where the DM has developed a number of future adventure sites, he might want to tip off the players a bit more directly - through a patron, rumors, a job offer, any number of approaches to make it clear, "there's gold in them hills".  My experience has been that most players don't mind following plot hooks to interesting adventure sites as long as they still have complete freedom in the execution - including the freedom to leave.

I don't know that I'm totally ready to reduce this to any game mastering principles, yet.  For now, it's enough to establish that there are different approaches to information gathering, which I'm calling active and passive; the size and detail of the sandbox defines the utility of the different information gathering approaches; the use of passive information gathering (plot hooks) constrains choice; however, plot hooks might be acceptable due to equitable factors (not wasting the DM's time) and fun - the players get to places that are more detailed and interesting to explore.

If I develop these ideas further it would be around that last sentence - acceptable use of plot hooks, and the trade offs between detailed content and freedom.


  1. It may be wandering from the post topic directly, but reading this made me wonder if there's a way to develop something like a flowchart system to keep a sandbox world alive in a broad sense, seperate from the player's activities. So for example, let's say you have a whole country, free to explore, with many towns, wild areas, trade routes, etc. At the start of the campaign time, you do checks on tables to see if various events are happening in each area. Then flowcharts are used to find the progression and outcomes of events as time progresses. Other events start up, and some finish, all at different times. It would ensure there's often something happening wherever the players go, something that's not predetermined and forced on them, but could be fleshed out as they approach the area - maybe the events suit a full scenario the DM has available to throw in, or he could wing it, or plan for the next session - various options are available here.

  2. Hi Jason, yeah, what you're suggesting is important. Earlier in the month I posted the calendar techniques I use in my current game - Happy New Year, Greyhawk - and I agree the DM should have a schedule of events that happens in the sandbox regardless of player action. In a wide area sandbox, it would make sense to have events generated on the calendar for the major areas where the campaign could traverse; at the same time the plague is striking Paris, London could suffer a major fire, for instance.

    However, campaign events are also tangential to the main topic because events become information, and information requires the DM to decide whether Active or Passive information gathering is in effect; do the players decide it might be interesting to go to Paris, and so visit the docks and ask questions from knowledgeable travelers, learning about the plague that way; or does an NPC or patron provide a job offer or rumor or some other passive method of dropping the information?

    If the DM has built an adventure that takes place in Paris, you can be sure he'll make sure hints get dropped about what's going on there (and how it opens up opportunities). When does this cross over into "leading the players"?

  3. As long as the players can bow out at some reasonable point at the beginning of the DM's idea, there's no railroading. Remember that ordinary life is full of 'plot hooks'...only we call them 'opportunities.' My personal feeling is if the player is simply responding yea or nay to opportunities as they come along, where's the harm?

    There's something to be gained by having the world move along without the players, but unless that directly affects the players, what does it matter? Big picture, there are ongoing wars the players hear about, and encounter if they move to those parts of the world, but why should I resolve the war if there's no impact the players feel? I'd rather save up that war for the day its needed...and the same can be said for every earthquake, volcanic eruption and so on.

    I'm not saying the volcanoes wait to go off when the players step off the boat, but if I'm going to mention that a far off island just blew up krakatoa-style, shouldn't that have some impact on the campaign? Otherwise, it's just bafflegab. Most games have enough of that.

  4. I agree that there's benefit in 'plot hooks' done right, but we should be able to establish principles for their use. As you say, a minimal bar is allowing them to be declined.

    The player-driven game is still my goal, where players ferret out their own adventure leads, although it seems elusive at times. Playing by Committee identified some of the issues nicely - even when players actually have a clear goal, elucidating clear actions can be a chore.

  5. Personally, I would have no trouble at all ferreting out my own goal...but from my personal life I know that opportunities come along, having nothing to do with my goals, and most times I grab them. So no world could exist rationally without both personal searching AND plot hooks.

  6. "where players ferret out their own adventure leads" Ferret out from what? There is no pre-existing objective reality here. Everything in the world is either created by the DM or volunteered by the players and then passed by DM approval. So really there is no difference between drawing a landmark on the map and adding a plot hook to the world. As Alexis says, as long as there is no Expected Outcome the players still have total agency to act and react in any way they wish. I try to think of these as "situations" rather than plots and quests, as it places less expectation on an ideal outcome.

    Creativity needs walls to bounce off - it is incredibly hard to create anything in a vacuum. Placing stuff in the world - whether locations, entities, events or situations - provides players with launchpads to start creating their own adventure.

  7. As long as there are options, "plot hooks as information" should be harmless and that's as good a standard as I can come up with. And I'm not saying declining plot hooks should have no consequences - choices have consequences.

    We've probably all been party to games where the DM has only a single adventure prepared, and to decline the one offered plot hook would end the session; though at that point is it fair to even call it a sandbox.

  8. I'm running a sandbox where I presented players at the start with, essentially, an info-dump about their base. In character creation they gave me backstory, which I translated into affiliations. They hear about possible work through these, which means that they pre-chose even their plot hooks; they aren't DM fiat of the kind you talk about here. I also have a set of rumor tables that I roll on during in-game downtime, again determined by player choice. I also have a megadungeon in the neighborhood... that's sketched out in broad outline (no rooms yet) that's waiting for the players to indicate an interest in going there. I don't really see any problems with player agency, not even vis-a-vis DM time, with sandboxes.