I've been playing Dungeon World for a while and I encourage players to collaborate to make the game really "our game" by using the principles to the letter. But when I look back at the world we created, it looks nothing like a coherent settings. I think this is the consequence of improvisation mixed with cooperative storytelling with a lack of central theme. The world we created is weird and crazy and there's no way that I can get my head around it. That's basically where we are right now. I don't understand the setting I'm supposed to portray.

I'm about to start another game of Dungeon World and this time I want to keep things simple. By simple I mean something that you would recognize as something Tolkien, G.R.R. Martin or Salvatore would have written. The main reason is that I'd like to avoid dealing with moon dwarves fighting orcs wearing jetpacks again. Some of my players told me that was the downside of the last campaign. It went too far on the crazy side. I have a mix of crazy and more vanilla players (what I consider vanilla is tolkienesque) in my group and I want to make sure things stay "under control".


So we've been playing Dungeon World using improv and cooperative storytelling and now my world is populated by jetpack-wearing orcs fighting moon dwarves and wizards are known to be lactose intolerant. How can I prevent this kind of sidetracking in my next game of Dungeon World?

  • 10
    \$\begingroup\$ "Simple" and "Tolkien, G.R.R. Martin" do not belong in the same sentence. ;) \$\endgroup\$ Commented Oct 10, 2013 at 18:51

8 Answers 8


To some extent, this is the pitfall of collaborative storytelling mechanics: when you have a lot of cooks, the pie can come out weird. There are a number of ways you can get around this, depending on the style/attitude of your group:

  • Let It Ride. While this sounds exactly what you do not want to do here, I present this merely as one option among many. Letting things play out how people want them to play out can create some disjointed stuff but it can also be a learning experience: if players are free to make plot-related mistakes and live through the consequences, they will know in the future not to make this mistake in the future. Another advantage of this is that in narrative-based games, everyone, not just the GM, is contributing to the story. That often means that things will not go in the direction you wanted them to go. You may not like orcs in spacesuits but another player might love them.

  • Collaborate on some ground rules. Erik Schmidt basically summed this up already but in an effort to be exhaustive, I will include it too. If you want a low-magic setting, you need to state outright that the setting is going to be low-magic. That alone will probably eliminate 90% of the issues that might arise, as people will not be confused or misled as to the eventual nature of the milieu. The first session of FATE and many of its offshoots (like DFRPG), for example, require the players to sit down and carve out their milieu.

  • Put it to a vote. It may be a little bit too late for your group to do this but when someone comes up with something that sounds a bit on the silly side, by all means feel free to say "I don't know about that. Guys, what do you think?" If you've already collaborated on ground rules, you can, for instance, make a pretty decent case that orcs in spacesuits do not belong in your low-magic fantasy world. The downside of this, of course, is that the players might disagree with you but such is the life of narrative-based RPGs.

  • GM Fiat. If all else fails, you can always just say "no, that's dumb" and move forward. I think that in most cases you're better off using one of the other options but if you're playing with a group of people who are used to adversarial based RPing this may be a required thing... at first, until you "retrain" them into narrative okayness.

  • Retcon. If you've gotten yourself into a really big mess, you can always say (hopefully by vote/agreement but perhaps by GM fiat if all else fails) that Giant Episode X, which made the whole game go from interesting to too silly, just plain never happened / was a dream/hallucination / etc. Find a good reason for it if you can, but if you can't, it's better to get objectionable material out than to keep it in because you couldn't figure out a good reason for excising it. It's not like you're writing a novel here (and take it from a semi-failed novelist here: that kind of thing gets chopped out of first drafts all the time).

  • "Yes, but only if..." Admittedly, I have not played Dungeon World (although I've played a lot of other narrative-based and straight up story games, both short and longer term), so I'm not sure if this kind of thing is in there or not, but a lot of story/narrative games do have it. In the game Polaris there is this whole mechanic where Player A proposes something and Player B can either say (usually at some cost) "No, I won't allow it" or "Yes, but only if (insert clause here)". Then the original requestor can continue on with "...but only ifs" of his own, accept the terms, or (rarely IME) say "You ask too much" and withdraws their original request. How this might apply to your scenario:

    Player: I want orcs wearing space suits!!!

    GM: Okay, that sounds... potentially interesting, but only if instead of space suits it's magical orc armor, and instead of space you're on the Ethereal Plane.

    Player: Okay, but only if they have lightsabers.

    GM: Okay, but only if those "lightsabers" are swords of fire. High technology being indistinguishable from magic and all that...

    Player: YOU ASK TOO MUCH.

    The advantages of this method are that a. nobody is actually telling someone that they flat-out can't do something they want to do, just trying to make it fit within their own narrative, and b. it gets people thinking in terms of the larger narrative itself rather than "OMG this would be sooooo coooool."

  • Make sure everyone feels comfortable with everyone else. Above everything else, my experience with games like Fiasco and Microscope - full-on story games without even a GM element - is that with new players things get very, very silly, and you only get those good, hardcore, serious sessions once everyone is pretty well acclimated to either each other or the concept of role-playing (as opposed to the standard kind of "roll playing" that most people do in DnD). Humor is a much easier thing to creatively express than other concepts. For whatever reason, a failed joke is just a failed joke but a failed attempt at drama can feel creepy and weird.

    I wish I had a magical answer to solve this. I'm afraid that the only cure to this that I know of is time and experience. If you "force" your players to be serious before they're ready to be serious, you'll lose them: either they'll stop showing up or they won't contribute stuff at all*. If you have a couple people coming up with weirdness whereas everyone else is doing fine, my suggestion is to use the "but only if" technique to corral their ideas into something resembling normalcy and hope that they eventually "get it".

*Which, no, is not a good thing, even if said person mostly just contributes stupid things. Think of this whole exercise as though you're trying to create a game through the lens of Stephen Covey's Seven Habits of Highly Effective People. Thinking win-win is a plus but equally valuable is having a "mission statement" everyone feels like they are a part of. In this case, the "mission statement" is the narrative itself.

  • \$\begingroup\$ Possible alternative to retconning is to reveal real consequences to what the players thought was whimsy. "All the jetpacks were great, but they created a worldwide shortage of residuum, so we're effectively in a low-magic world now." It's still a bit weak, but not as weak as "and then you all woke up and vowed never to eat mushrooms again." \$\endgroup\$
    – detly
    Commented Oct 11, 2013 at 5:50

"Ask questions and use the answers" is a principle. What's not as obvious is that this principle doesn't tell you to just take whatever they answer as written in stone. You can "use" the answer to ask a follow-up question. The only thing you can't do by that principle is ignore the answers.

"Huh, how is that supposed to work?" and "I can't picture that. Can you explain more?" are valid follow-up questions, and necessary ones. Keep asking questions until what they're saying not only is understandable, but makes sense as part of the world (or at least could make sense). You don't have to sort out every last detail – you can leave holes unanswered for later, as long as they're not genre-breaking holes. But for outlandish answers, keep asking questions until the answers make sense or they offer to "redo" their answer with something more fitting.

Players, especially players put on the spot suddenly and who are not use to the giddy power to create the world, often give silly or plain stupid answers just because serious answers take more thought and time. And what these players absolutely want to do is get rid of the question and the need to think hard as quickly as possible. Don't stop at the first answer, dig deeper, and let them either replace their answer, or explain to everyone's satisfaction why their weird answer does actually fit and make sense.

On a social psychology note, by continuing to ask more questions because they gave a silly answer, you show the player that being silly will do the opposite of making the question go away quickly. Instead, being silly will make you keep asking them to think hard. After a few repeats of this, they'll stop trying to take the (non-working) shortcut of silliness and actually take the time to think of serious answers when you ask questions.

  • 2
    \$\begingroup\$ This is kinda like the GM move "Tell them the requirements or consequences and ask." The requirements being that the world doesn't become ridiculous, the consequences being that the GM doesn't understand you. \$\endgroup\$
    – As If
    Commented Mar 11, 2014 at 0:33

You may want to have a discussion with the players about the basic parameters of the world. Collectively you're determining what sort of campaign you want it to be, without drilling down too deep.

Beyond discussing which books and movies will inform the game, you can pin down a few elements that will keep everyone on course. Is technology going to be limited or unlimited? Is religion going to be a strong component of the politics of the world? The answers to those sorts of high-level questions can determine a set of parameters that will allow for all sorts of improvisation without being unduly confining.

  • \$\begingroup\$ Excellent points. The Same Page tool can help with this- as it definitely appears that your group was not on the Same Page in your current campaign. \$\endgroup\$
    – Chuck Dee
    Commented Jan 21, 2015 at 16:53

Since the players seem to be unsatisfied by all these things, ask them to be cooperative. When you will ask questions around ask things like "how does this interacts with that other element we mentioned?" and don't be afraid of letting them change what they just answered you. "Are you sure?" is a perfectly legit question to ask.

These two things togheter could help you avoiding weird incongruencies such as those you faced in your previous DW campaign.


As a GM you do get to set the setting in advance. I started a DW game yesterday, and the first thing I said was something along the lines of: This is going to be a cliche fantasy setting, but magic is relatively rare, and there aren't dragons around every corner. Think Tolkien, or maybe G.R.R. Martin with Elves.

That way, it's implied that there won't be any jetpack-wearing space orcs.


Well, the simple thing would be to just totally eschew cooperative storytelling, but I take it that you want to avoid stuff like that at all costs.

Don't fear, it's possible to prevent this.

I like to use a three-part criteria list to prevent stuff from getting out of hands; game, story, and veto.

If you've ever looked at Game Chef, you'll note that a lot of the "games" are really just loosely ruled narrative development exercises. When I allow players to have a say in the setting, which I'm not a huge fan of, I make sure that there's a specific game purpose for something. There's not really a universal rule about adding silly elements to a game, but say what players can and cannot introduce explicitly. For instance, say: "It's okay to add a historical event, a major defining weakness, or a major power, but not necessarily write on and on about quirks.". Turn it into an intentional exercise of adding in details until you're done, like Dwarf Fortress's megabeast generator with a human guide, and don't be afraid to cut out things that don't work or suggest alternatives. You're the GM. Likewise, you don't have to do this ad infinitum; have limits when you've hit the furthest extent of what you're worried about. Don't have goblins, orcs, griffins, dwarves, illithid, and more all in the same session.

Also, look at the story. Warhammer 40k has two distinct flavors; the earlier power-metal football hooligan one and the overtly fascist-dystopian developments from later on in the setting's development. Is your game taking itself seriously? If so, then don't allow lactose intolerance (at least not as something that comes up in play), and if not then feel free to go zany within limits. But know when you'll make a compromise; is it for the rule of cool, for a specific style, or to fit withing the conventions of a genre. You mentioned the lack of a central theme. Make a central theme. Rely upon it. It'll save you much angst down the road.

Finally, whenever anything is introduced, write it down and look at it for a few seconds. Ask yourself if you will remember it. If your players want to tell quirky, wild, anarchic stories, that's fine, but you need to restrict the setting to the things you can prepare and handle. If you have to, just say outright "I'll probably never work this into the game", or "This won't work, I won't remember it." and let your players know that they've left the acceptable boundaries of setting expansion.


Sort of a combination of SevenSidedDie's and NotVonKaiser's "yes, if..." answer.

You have to "use the players' answers"; this is not identical to "what the players say is true in the fiction", at least with respect to anything outside of the characters themselves.

The players can define what they've (thought they) heard, seen, believe... but that does not need to be the truth of the fiction. Player says "jetpacks and lightsabers in space" — that is how to describe what the character believes is going on, but the fiction might be more like magical arms and armour on a different plane — the GM gets to make that call. Following through with additional questions, especially leading ones, might help the player get at the true fiction, or realize that what their character thinks that they know might not be the full story.

This differs from NotVonKaiser's in that there is no explicit bargaining, and from SevenSidedDie's in that it emphasizes that the GM gets to decide the characteristics of the environment (which is not indicated in that answer).


this is also known as the meat-grinder syndrome.

as is, you are asking "when i put my hand in the meat grinder and i turn it on, it hurts. how can i stop hurting?"

answer: you stop inserting your hand into the friggin' meatgrinder, dude! you stop, as a group, inserting weird details into the setting, but you all have to be interested in doing so.

it's simple like that. you have to care about the fiction. you have the power to make the game about what you want, so make the game about what you want instead of what you deem funny at the moment. you also need your fellow players to address the game in the same way.

there is no loophole to circumvent this, even if you can mitigate this by adding rules or withholding questions.

however, you are probably already doing this. this has probably happened as a result of being power-drunk with this playstyle the first times, it should fade with time and habit.

  • 3
    \$\begingroup\$ Hello, and welcome to RPG.SE =] I downvoted this because I think the author of the question understand what is happening, but doesn't realise how to stop it from happening. While your answer isn't wrong, it doesn't really address the question (beyond saying 'just stop'). If you can edit your comment to add strategies or tools the author can use, I'll remove the vote =] \$\endgroup\$
    – kadu
    Commented Nov 10, 2014 at 5:57
  • \$\begingroup\$ yes please help me: do you think that pointing out the he should simply say: "hey guys let's stop doing this" is enough? because the point i wanted to convey is that complex in-game strategies are not gonna help him as much as simply aknowledging: "hey, we have this problem, let's use more restrain" i also thought the bit about this happening so strongly, because this was their first game, would have been helpful. \$\endgroup\$ Commented Nov 10, 2014 at 14:20
  • \$\begingroup\$ Sorry, I just saw your response. Pointing that out is a valid response, but probably your answer would be perceived as better if it had more substance. If you offered some strategies on how to approach the subject, for example. \$\endgroup\$
    – kadu
    Commented Nov 18, 2014 at 5:34

You must log in to answer this question.