Dungeon World, Difficulty and Finding Waldo

waldoOne of the most interesting things about the -World engine is that it seems to be an argument that difficulty isn’t relevant. This is a pretty bold position – a lot of games have wrestled with how to handle the idea of difficulty, and that has lead to ideas like automatic success, taking 10 and failing forward, but the -World solution is, in effect, to forgo difficulty entirely. When your their comes upon a lock, it doesn’t matter if it’s a mid-grade padlock or a quadruple locked mithril puzzle lock built by ancient gnomes. The Tricks of the Trade roll is going to have the same odds no matter what.

On the face of it, this seems disruptive as hell. It upends a huge number of assumptions about how a game should work, and basically says that the distribution of success and failure (and compromised success) should be spread out along a certain curve, irrespective of what’s going on. This has some specific benefits – because there’s not increase in difficulty, there is no need for an increase in capability to represent progress. This is why Dungeon World characters generally get broader (in the range of what they can do) rather than better (absolute bonuses don’t increase a lot).

It is, however, all a little bit more complicated than that. a different way to look at it is that -World system exist within a specific sweet spot, mathematically, and the point of a specific design is to align that sweet spot with an ideal vision of play. That may seem like a weird concept, but consider for a moment the idea of high level D&D play, real Paladin in Hell stuff. Dungeon world, as presented, doesn’t really do that, even though it emulates D&D. Instead, it emulates a particular band of the D&D experience that maps pretty well through levels 1–9ish, depending on edition.

But Dungeon World could do Paladin in hell. You absolutely could have epic, world-shaping playbooks without changing the rules and numbers.[1] The fiction of the movies would be more sweeping and dramatic but the underlying system would not need to change. But you cannot practically mix those two levels of play.

That seems like an obvious thing to say, but if you play Dungeon World, you might think about it and think “why not? He may be an archmage and I may be a peon, but we both face a hard choice on a 7–9, and since we’re not actually rolling against each other, doesn’t that just mean that we actually can mix power levels successfully in accordance with cinematic logic?”

And it’s a good question. But no, or at least mostly no, for two reasons that both come back to difficulty.

See, -World may look like it forgoes difficulty, but it can’t quite pull it off, so instead it moves it off someplace discrete.

The first place is damage. In any roll where damage is a result, it is the stand in for the margin of success. Damage output is a critical part of a playbook’s design and of the game as a whole. This is the first real barrier to mixing “tiers” of play. it is totally cool if an ogre has 19 hit points in Dungeon World and a Balor has 19 hp in Paladin in HellWorld because they are comparable challenges. But the fiction becomes hard to maintain when the shatterer of worlds is only as tough as an ogre (or vice versa).[2]

The second, and more subtle barrier, is the question of when you roll the dice. To go back to the question of the two locks at the outset, the reality is that in play, the GM would probably say “The ancient door is too complicated for you to open with your lockpicks, you’ll need to [DO SOMETHING AWESOME] first.”[3]. And that’s not a bad thing, but it’s a bit of a blunt instrument. Rolls occur in a fairly narrow band in -World, and while that’s true in every game, it’s a bit more of a hard boundary. In games with difficulty, the GM can lean on the dice in grey areas, and allow rolls with a slim chance of success rather than rule one way or another.

And that reveals a fascinating bit of sleight of hand. Often, when a GM calls for a roll with a very high difficulty, she is counting on failure. This is a bad habit, but a common one, because inevitably there will be an unplanned success, and if the GM is not ready for that, then things can go badly. The -World system does not allow the luxury of that mistake[4]. If the dice are on the table, then so is success[5].

This is the real barrier to mixing tiers – the different fiction means radically different rules for when you get to roll and when you don’t. The archmage will largely be able to solve all the peon’s problems without ever rolling the dice, which feels kind of crappy for the Peon. This is made worse when the Peon doesn’t even get to roll against the Archmage’s problem.[6]

This is, I should add, not a criticism of the -World system, and it should only be taken as one if you think any one system must do and be all things. -World made some specific tradeoffs that allow it to excel within its intended space rather than to be all encompassing.

I think of it a little bit like one of those Where’s Waldo pictures. Most games are like the whole picture – interesting, dynamic, but maybe a bit overwhelming. -World takes a piece of paper, cuts a hole in it, then covers up most of the image so you can really pay attention to the part it reveals. Don’t like that chunk of the picture? Hack it to move the hole somewhere else. You’re giving up the broader picture, but at the same time you’re increasing your engagement with it. And depending upon how you look at things, you might be more likely to find Waldo that way.

And to bring it all back to difficulty, it teaches an interesting lesson that’s been on my mind when I look at things like 4e, Final Fantasy or Exalted through a -World lens.  When you come upon a design problem where the traditional solution is to change the difficulty, in the -World mode, the right solution might be to change the game.

That’s kind of cool.


  1. I mean, you might change the numbers cosmetically to make it look more awesome, but if you multiply everything by 10, then you’re pretty much changing nothing.  ↩
  2. There are ways to hack this, but this require embracing that the tiers are different. And it’s a very, very old problem in RPGs, best summed up in 2 words: Mega Damage.  ↩
  3. Which is, by the way, why the second most important move in the entire game is, to my mind, the Wizard’s ritual move. If I only had one move, it would be defy danger. If I had two, it would be ritual. It works just as well for planning heists and winning wars as it does mastering arcane powers.  ↩
  4. This is one of those great examples of how the -World system solves problems – if you’re a GM who is not ready for success and failure on every roll, or who does not really think about these things before rolling the dice, then it will cure you of that habit. Of course, if you don’t have that problem, then you might find it restrictive.  ↩
  5. And failure. Everything I’m saying about high difficulty applies equally to low difficulty and the decision of what you don’t need to roll for. I’m just not qualifying it as I speak because it feels awkward to qualify it every time.  ↩
  6. And, of course, if you “solve” the damage problem, then you probably just perpetuate this problem into the arenas where they do get to roll. The Peon gets to roll against the Pit Fiend, but does such trivial damage that he might as well have done nothing.  ↩

11 thoughts on “Dungeon World, Difficulty and Finding Waldo

  1. Rob Donoghue Post author

    And just to add, this is part of why I’m intrigued by -World for supers. If you can solve the damage problem, then the other tiering issues are less relevant in supers, where Batman and Superman rub shoulders.

    Reply
    1. Michael Phillips

      I know the power discrepancy isn’t quite as large, but I think vanilla AW does a good job of handling that discrepancy. It is a game where a Skinner and a Space Marine Mammal or Gunlugger can meaningfully interact, even though the damage that the combat forms can take and deal would just obliterate a skinner. I think AW makes a strong argument for the Glitterboy and archeologist party. Sure, the Skinner needs to go hide in a case where the SMM is fighting, but when the Skinner is doing social things? The SMM is totally outclassed.

      Reply
      1. Dave

        I have to agree. I have played the Rifts setting via AW reskin and found the power levels were easier to deal with, but that might be because the systems are so different that you can have social interactions that are mechanically meaningful in AW.

        Reply
      2. Rob Donoghue Post author

        I think AW makes a different argument, with a different set of problems. Thing is, you can ALWAYS have the Angel Summoner and BMX bandit in the same group if everyone is willing to bend things to make it work (and if everyone enjoys that bending) but barring that, the best thing you can do with the damage system is just say it’s a manifestation of the tiering problem rather than a separate problem.

        Reply
        1. Michael Phillips

          AW is very much competence porn. It encourages characters to push the areas they excel at. It bends things by default by valuing situations where massive firepower, or hordes of angels aren’t the best solution as much as it values the ones where they are. I think Dave’s comment illustrates the problem. In AS+BB, really all of the conflicts are the same sort, and you could do that in AW as well, but the game actively encourages an array of problems that need different kinds of solutions.
          That said, going to the dungeon world model, that is one of the problems I’ve had conceptually with that game as well, In focusing on “The D&D I played when I was 12” it has made a lot of the other parts of *World games that I love less accessible. You can do the cool social stuff, but even if you aren’t fighting the system, it is certainly not helping you along like a lot of other *World games.

          Reply
          1. Rob Donoghue Post author

            Multi-axis competence is a traditional way to handle a lot of different type of competencies, and I’m on board with that in general, but you get a certain amount of weirdness when the arenas are structured differently. Classically, D&D illustrates this with fighting being SO MUCH more important than anything else that the idea of balancing on other areas of competence is just a non-starter. But that’s not the only imbalance.

            Another problem comes when the *scope* of a given arena gets too huge. If you are a social character and I’m a fighting character in leverage, that’s cool, because you talk better and I fight better, no biggee. If we take the same roles in RIFTS, then you still talk better, but I fight SO SO SO MUCH BETTER than you do that we might as well be in different arenas entirely. The practical impact is that if we have a social scene, we can both participate, but you will excel, which is fine. But if we have a fight, then odds are good that you are effectively not even in the scene.

            Obviously, there are ways to address that, but the need to do so introduces a tension that makes things more work.

            Similarly, arenas often bleed. The classic example is spellcasters having spells that effectively allow them to trump other arenas. If i win every social engagement, I may be able to never get in a fight. if I win every fight, I may never need to worry about social stuff. The boundaries are rarely as clear as we like.

            And, again, it can be addressed. But it’s work.

            Not to sleight Apocalypse World, but it has to deal with these issues as much as any other game. 🙂

    2. Craig Payne

      Maybe the way to do supers is to eschew damage and look towards some sort of “states” that you put the villains in, as it’s often a case of incapacitating the villains from escaping in some way: Supes may wrap an iron bar around while Bats hangs them by their feet from a gargoyle, but essentially they both have a “tied up” status at that point.

      Reply
  2. Paul (@princejvstin)

    And that reveals a fascinating bit of sleight of hand. Often, when a GM calls for a roll with a very high difficulty, she is counting on failure.

    –Which goes over to Torchbearer, and players taking dice against themselves with the expectation that they are expecting failure.

    Reply

Leave a Reply to Michael Phillips Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.