Sunday, May 29, 2011

Running a Better Sandbox Game

As a bit of a PSA, and because most folks probably scan my blog through a reader and don't bother to read comments, I'm going to repost something Matt Finch said the other day in response to the "megadungeon monotony" post.

There are tons of house rules, class variants, new monsters, etc, littering OSR blogs (two thumbs pointing to this guy - just as guilty).  Tons of resources on building a sandbox, too. But keeping a sandbox game going, keeping it interesting?  Yep, not so much - that's why I thought a note with solid practical advice on *running* the game is worth a special shout.

It is absolutely key that a megadungeon has distinct areas -- and I mean distinct enough that the players can make the distinction based on play, not that the DM happens to know that there's a greater chance of giant rats there. Not only that, but the distinctions need to be meaningful. This allows players the ability to make a decision about where to go in the dungeon. Without that, it is a fungible, undifferentiated, who-cares landscape.

It's incumbent on the DM to either create a dungeon where future missions can be perceived and planned, or to provide some missions the players might choose. Blind exploration gets old if not leavened with some purposeful expeditions that have planning behind them.

Upshot of both those points is that players have to be given meaningful choices and the ability to pre-plan at least some of their activity. 

The major potential failing of a megadungeon is not to realize that pre-planning and preparation are fun and integral to the player experience, and a megadungeon risks removing this aspect of the game if it doesn't provide information. Information flow and discoveries are critical components of a megadungeon campaign, or else there is no meaning behind player choices. Might as well flip a coin about our equipment, spell, and direction choices? Not fun.


  1. That's a very valid point and one well made. But I think one point in particlar needs elaboration: the size of the particular distinct areas. Too small and the missions become too easy (unless the location is difficult to reach), to large and they almost become a seperate dungeon in themselves.

    It helps if the way the denizens in neighbouring areas interact with one another makes sense as well.

  2. The whole concept of sandbox gaming is kind of new to me, but seeing what different people define as part of the concept is fascinating to me, and really opens my eyes on different methods of plotting beyond traditional linear or branching narratives. "Mission" based gaming like Top Secret or Shadowrun reminds me a lot of sandbox, and I'd like to refine the concept in my own mind better, to improve my storytelling skills.

  3. @Doc - think of sandbox gaming like a computer game where you can go anywhere - a wilderness sandbox might have a home base, a wilderness map, and the DM has sketched out various lairs, dungeons, etc in the wilds. Through rumor and information gathering, the group learns enough to plan excursions into the wilds.

    A megadungeon is like the dungeon-equivalent of a wilderness sandbox - a dungeon so large a group can plan similar excursions.

    Lots of (other) bloggers have tips on making a sandbox or a megadungeon - I wanted to repost Matt's comment because it was good advice for *running* a megadungeon.

    And since you're new here - my Black City project (link at top) are my efforts to put together a campaign setting that will include a wilderness sandbox and megadungeon - bit by bit, at least.

  4. Yes, that makes sense. I didn't do that with The Black Pyramid of Cha'alt. It's kind of an anti-megadungeon... perhaps. However, I'm about to create a 2nd megadungeon, also on Cha'alt (or underneath it, rather). That will have distinct areas. Useful blog post. Thanks, hoss!