Monday, January 11, 2010

Some reflections on sanbox campaigns

If you hang out by the blogs, especially those OSR oriented, you have probably heard the word "sandbox" mentioned a gazillion times. I don't know if it's as frequently used on the rpg forums. Maybe it is.

The idea behind a sandbox is nice. You have a world to explore as a player, and the GM have the freedom to develop just as much as is needed, since the players are going to be exploring and thus be the engine that takes the campaign somewhere. Often, freedom from "story" seems to be an objective when people set up to talk about their sandbox campaigns. I think a few things is worth mentioning about this.

Now, I am just as bored by GM railroading as the next guy. But, that is just as extreme as a world map for sandbox and having everything that happen by driven by the players. Simply put, elevating the sandbox style of play and disparaging "story based games" is taking one extreme and making it an ideal while calling the other bad and "extreme". You see what I mean?

I have had some experiences that highlight some things worth thinking about, regardless of game style. I once had a game prepared where the characters were walking down the street and a NPC jumped into the river right in front of their eyes. I expected my friends to do the obvious thing and try to rescue that fellow and try to find out why he was trying to kill himself. Let's look at this from two perspectives.

From the sandbox perspective, I was a bad GM. I had a story and I wanted my players to walk the path. In a way I agree with that description. It might have been better if I had asked the players what they wanted to do.

Let's put it another way. We had gathered to play a game, and I had prepared some stuff to entertain my friends. They re-payed that by acting like jerks, just being contrary and refusing to follow along. Weren't they just ignoring my kind of fun and trying to strike out on their own instead? No, I maintain they broke the social contract.  The rest of the session was a meandering mess where they walked around town ignoring any kind roguish adventure. No fun was had. While I certainly failed to make the game fun for them, they failed to make it fun for me as well.

If your players wont grab a plot hook you'd better let them create their own adventures. In the same vein, if your players seem to wander about and not doing much like adventure you'd better show them a hook or two. If they wont do any of those, they might need a reminder that you agreed to play a game and they don't. Too bad it took me quite a few years and my own attempt at a sandbox campaign to realize that.

It's very popular in the blogosphere to talk about the liberating effects of a sandbox. This experience of mine, and my latest attempt at a sandbox campaign in Traveller have shown me that neither the open sandbox nor the GM-story ends of the spectrum works for me. Nuances don't come across as well in this medium, but for me it seem to be the way to make a campaign work. A guided "story-sandbox" kind of works. I wanted to toss in those two cents in the big sandbox conversation.


  1. Hi,
    I can't see why you've been a bad GM if you provided hooks (such as the river-dive) and players ignored them as in a sandbox the responsibility is shared. If you kept on running the game for them and they just ignored anything you put on their plate, well, shame on them.

    Last night I posted about sandboxing being a high-risk/high-reward method for nunning games as the "inner" social contracts in it are non-existing compared with more railroaded or "bespoke challenges" games.

    When i see my players doing that (just wandering around without doing anything and dodging hooks/places/npcs like the black plague) i just take anything they might say (if you know them, try with an old favourite topic of them, it works almost every time) and spin off from that. Players make a comment about the local mob? well, let them run in the mob. Using hooks about topics they are thiking about places less cognitive load on them,so the "barrier of bothering about it" should hopefully be lower.

  2. Frankly, both a 100% railroad and a 100% sandbox arent that different. Both are just lists of encounters, one with a pre determined order and a backstory and one with no particular order or connecting back story. But they are just encounters and the players have no meaningfull way of determining what they want to do. Go to the ruines for some encounters or enter the dungeon for some encounters.

    Freedom of choice for players only has value if the options are more then just A, B and C and the choice they make has a consequence. As 100% railroading doesnt allow for choices and 100% sandbox makes everything rather pointless a mix seems the way to go.

    In my current (modern delta green) campaign the players have a bunch of leads and the bad guys have a time table before they make their next move. Each investigation takes time and travel takes time too. They dont really know how much time they have and they dont know what leads are the best. They do know (or think they know) certain leads will lead to different types of new information. They can focus on what they want to investigate, do they want to know who the bad guys are or where they are operating from or what their next target could be or do they want to know more on how they managed to pull the first major crime off?

    This setup creates discussion amongst the characters on what leads to investigate and if a trip to Singapore is worth the bother or that tapping the phone for another day might be a better option. Most important, I have some control over what they do (the focus remains tied to the operations onf the bad guys) and they dont feel restricted, as they enjoy the ‘meaningfull freedom’. Their actions have consequences and they can bodge the investigation and that will not be the end of the campaign, as the bad guys are not commencing a ‘destroy all’ campaign. As both failure and succes are fine options for me to continue the campaign I dont have a need to ‘push’ them on the railroad to succes.

  3. Good feedback both.

    I sometimes think the shared responsibility of sandboxes are glossed over. I think the best games work like Pingwin describe it. Very few probably play sandboxes as free like you sometimes might think from their blog posts.

  4. I agree with tsojcanth. Just because you provided a situation that you wanted them to investigate doesn't mean you're a bad DM. It just means that the players ignored a possible hook and now you have an outcome to adjudicate.

    In a sandbox, things happen and continue to happen. I may have a player driven "plot", but that merely means that the actual things the players do and investigate and affect is up to them.

    However, the world continues moving on. Things continue to happen in indifference to the players. If they choose to interact with the world/sandbox in a way that affects what is there, then I adjust to that. It's up to the players to engage and develop that engagement.

    I can run random encounters all day. It's when the players reach out to really do stuff that the world comes alive for all of us.

  5. One problem of a sandbox that is truly open, is that the amount of work for the DM increases as the square of the distance from the starting area.

    In a story arc game, you have a linear progression from one place to the next. The DM doesn't prepare areas off the path. But it's clear where the party needs to go and that's where they will find the most fun.

    My sandbox hex map is 500 miles across (a little over four times the area of Washington State) and is made up of 5-mile hexes. Even with a lot of empty space that's still a whole lot of little dungeons and such scattered around. I made it easy on myself by including a wide variety of terrains - I don't know what I would have done if it were Dark Sun. It's not kitchen-sink, but I'm stealing ideas from everywhere :P

    These adventure areas tie in to each other. Things are going on. People have their agendas. But there isn't a Big Story in the background. And I think that's the difference between the Story Arc and the Sandbox: if the players decide to say "screw this noise" and leave your Story Arc rails, what is available for them to do?

    If your Story Arc game has enough separate threads tangling around interacting with each other, it's no longer on rails. Congratulations, you've created a Sandbox game.

    A secondary feature of an open game is that the players probably won't experience everything you've prepared. In fact, it may be impossible for them to do so. In a straight Story Arc game the DM has an incentive to show the players every little things, like a grand tour, and often if they don't see everything they will fail the adventure. They miss a clue, or don't search for secret doors, or accidentally kill the vizier before learning that he replaced the princess with a hen in a wig.

    So that's a good thing to think about when designing a sandbox or a complex story arc.

  6. That wasn't bad DMing, by your description, that was non-immersive role-playing. Absolutely, the players (as most decent folk should) would have rescued the NPC. However, after doing that, they could have chosen to ignore the hook, if they weren't interested in that story. But they then should have given you some signal as to what they wanted to do.

    Niether the sandbox nor the railroad style work without the active participation of the participants.

    It doesn't appear to be a failure of the sandbox system, but of the agents within.

  7. This comment has been removed by the author.

  8. I listened to the Open Design podcast, where Robin Laws was talking about Gumshoe. That system is designed to make investigative rpgs work, by giving away all the core clues for the mystery for free, and not make them depend on a die roll.

    That attitude makes me wonder about Story Arc games. Is it necessary to have that approach? It is a Grand Tour game, but without any rails. Is it still a railroaded game?

    I feel there are a lot of small nooks and crannies in the design of scenario structure that can be discussed here.

    Robin's answers to the interview in that podcast made me think, and gave me the idea for this post. All the feedback posted is most interesting.

    Thanks all!

  9. In that concept (not making key clues roll dependent) I have some agreement. I've not listened to the podcast in question, so I'm basing the following on the word "free" which to me indicates without any player requirements.

    I think there is a trade-off to be had, otherwise I would feel like I might as well open the treasure chests and remove any obstacles, if all I'm doing is reciting clues.

    What I've been doing is resource or situation trade-offs. Yes, you can have the clue, but you have to earn it through role play, or through some exerted effort and/or some resource expenditure - whether it's time spent (and resulting wandering monster throws) or gold/food/spell components spent.

  10. Sounds like my attitude.

    In Gumshoe it is indeed free. Be at the right place and the GM will feed you everything needed for the Story.

    Do listen to that podcast. Even when I strongly disagree with Robin he is very good providing food for thoughts.

  11. Gumshoe? Free clues? Well, I suppose so, you still have to have the appropriate skill and ask the right questions but there's no hit or miss die roll required...

  12. Well, if you listen to that podcast, you'll see that what Robin say is basically free clues as the design idea.


Copyright 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016 Andreas Davour. All Rights Reserved. Powered by Blogger.