How I did conflict resolution in Dungeon World, thanks to D&D 4E

I think I learned the concept of a longer-scale conflict resolution (that is, one that is made up of small tasks) from the 4E Skill Challenge rules. In a skill challenge we are tracking numbers of successes to see if a particular conflict is resolved in our favor. It’s similar to modern usage of “clocks” in Apocalypse World or Blades in the Dark, although the devil is in the details.

When I arrived to Dungeon World around 2012-2013, I think I intuitively understood that that type of resolution is necessary to get anything out of the very “task-like” moves that Dungeon World offers.

Apocalypse World’s basic moves frame resolution around conflicts that are inherent in the types of situations AW tends to generate. This means that oftentimes the intent in triggering a move is immediately apparent from the situation the move was triggered in, and you’ll naturally end up resolving the overall conflict or intent rather than just the character’s action or task. The more removed a game is from AW, the more in my opinion it tends to fail to generate the necessary context for this kind of “implicit conflict resolution”, and the moves become more like prompt-generators for genre emulation. We were partially talking about it here.

DW is one of the first games that showed this problem. Essentially, DW’s basic moves frame resolution around a fictional action taken by the character, and produce more content that is supposed to be genre-appropriate. But there is no guarantee that they actually resolve anything that we’re here to resolve. Maybe I hit the guy successfully with Hack & Slash, but he manages to snatch the idol by GM fiat. Essentially, we just spent a whole lot of effort resolving something we didn’t care about.

The two ways of solving this problem are:

  • Framing the task-like move as a small-scale conflict, clarifying conflicting intents before rolling, injecting conflict resolution into the move, even if it’s not explicitly said.
  • Framing the task-like move as an element of a larger scale conflict—e.g. a Skill Challenge. Each success brings closer to achieving the overall conflict goal.

I had some experience running D&D4E skill challenges (with extensive house ruling), and I inherently found myself applying the same principles to extended confrontations in Dungeon World. This resulted in this article that I wrote in 2014 for an Italian DW zine, which is not how I would phrase these concepts now, but I can definitely see the intent behind them and I remember how I used the method in practice.

Particularly, notice how each adversary is framed as having multiple “layers” that have to be overcome. These are essentially equivalent to Skill Challenge ticks. And if I think about it in practice, I definitely used the layers to define stuff like “can you steal the idol”.

Machine translation by

The Onion Method

More and more often I find myself introducing Dungeon World to novice players. One of the most difficult knots to untangle is explaining how to make an obstacle (a monster, for example) more or less difficult to overcome. In fact, the basic mechanics of Dungeon World involve rolling two six-sided dice to overcome a fixed threshold (7 for partial success, 10 for complete success) regardless of the difficulty of the obstacle. This is counterintuitive to some, particularly those accustomed to games such as Dungeons & Dragons, where the threshold to overcome is proportional to the difficulty of the task.

Actually, there is a method to differentiate the challenges. In explanation after explanation, I came up with an effective way to explain the process: I jokingly called it “the onion method.” It is nothing additional to what is already explained in the Dungeon World manual, but it is a way of presenting the game that I have found useful for clarifying ideas for both newbies and myself.

“The Onion Method” expands the concepts presented in the article “A Dragon with 16 PF,” translated by me for the first issue of “Underground World,” which is now required reading for anyone approaching Dungeon World for the first time.

With this article, in addition to providing a good tool for veterans, I hope to convince the doubters that, yes, there are harder and less difficult feats in Dungeon World and that, most importantly, this method works excellently.

Dungeon World is a fiction-first game. Each detail you establish at the table is important, and can be both a foothold to launch into otherwise impossible deeds and a danger that comes between the protagonists and their goals. In this article we will discuss the second type of detail.

Meet Sir Reginald the Valiant. Hello Sir Reginald! The valiant knight is about to face a dangerous troll. Not the nice, love-savvy ones from Frozen. Sir Reginald is not afraid of trolls and is looking forward to bringing home the head of one of these behemoths as a trophy.

GM: You see a troll in the distance. The troll charges you with his club. What do you do?

Sir Reginald: The troll doesn’t scare me! I charge him with my sword and try to drive it into his torso before he lowers his club! Do I roll to throw myself into the fray?

GM: Of course! Why not? There’s a mutual exchange of blows, so that’s fine.

Sir Reginald: I made 11: complete success. Yay!!! I do a thousand thousand damage.

GM: The troll dies in excruciating agony as you eviscerate his heart with your noble weapon.

Sir Reginald: Hooray!

Boring, isn’t it? The threshold of 10 is easy enough to cross if the character has +2 or +3 modifiers, and while “a thousand thousand” is not a regular damage number, some warriors come pretty darn close.

It may sound strange, but what I have just presented is one of the most common mistakes of novice Dungeon World GMs: presenting opponents who do not threaten characters in any way.

Now let’s see how to play hardball.

GM: You see a troll. The troll is a huge monster, as tall as three normal people. With every step he takes, the ground shakes and you struggle to keep your balance on the vibrating ground. His colt-sized hands wave a man-sized club toward you in an attempt to crush you even before you can get close to him. What do you do?

Sir Reginald: The troll does not scare me! I charge him with my sword and try to drive it into his torso before he lowers the club! Do I roll to throw myself into the fray?

GM: Well, no. You cannot reach the troll with your tiny weapon, however noble and glorious. To get close you’ll have to at the very least expose yourself to the troll’s blows: his club has immense range. What do you do?

Sir Reginald: My glorious shield will protect me from the troll’s blows! I defy danger with COS: I made 11.

GM: You receive a solid beating on the shield, but both you and it are intact. You come running to the feet of the troll, which begins to pound mightily on the ground, knocking you off balance. What do you do?

Sir Reginald: I plant my faithful blade in the ground to support myself! I defy danger with COS and make 7: partial success. Damn, that’s a bummer, damn them.

GM: You manage to stay on your feet, but remain unarmed by your noble blade that remains stuck in the ground. Taking advantage of the opportunity, the troll moves one of its huge hands in an attempt to grab you. What do you do?

Here we’re starting to get somewhere. What is the main difference? Well, on a mechanical level in the second case the player had to make an extra roll.

So is that the secret? Splitting the challenge into multiple rolls? Good stuff: I’ve been doing that for a thousand eons with twenty-four different games. Besides, wasn’t Dungeon World supposed to be the game that limited the GM’s absolute power? So, in the end, isn’t the GM still the one who decides when the confrontation ends, deciding when to have the danger challenged or not?

Wait, dear hypothetical interlocutor. See, I said that’s the mechanical difference, and that’s true. But observe how the GM presented the troll in the second case. He went out of his way to highlight its dangerousness. He listed a number of dangerous details that only later turned out to be an obstacle for Sir Reginald.

Let’s list them for a moment.

  • The troll is huge: as tall as three normal people;
  • When the troll steps on his feet the earth trembles;
  • It has hands as big as colts (thus capable of grabbing a whole man);
  • He is waving a very long club, keeping the valiant knight at a distance.

Now let’s see what happened next. All these details reappeared during the fight: first the club and the enormity of the troll, then the ground shaking with his every step, and finally the foal-sized hands.

The GM did not decide that Sir Reginald should challenge danger every time, because it is not in his power to decide how an obstacle should be faced. Instead, the GM set as an obstacle the details he had established at the beginning, leaving it up to the player to choose how to deal with them.

Consider the troll as an onion and the details as its layers. To get to the heart of the onion, one must get through its layers. In short, each detail needs to be neutralized (cutting off the troll’s hands), overcome (dodging the club) or circumvented (attacking the troll while it sleeps).

My advice is to write them down, if there is time and the obstacle is important enough, on a sheet of paper.

As a detail, not only the things the GM says when introducing the monster count. All details established in advance count. Whatever has already been said in the past or is on the GM’s fronts counts. For example, one might already know that the troll has skin as hard as stone or that it regenerates quickly if it is not burned by fire.

The onion method works well, very well. It allows you to define the difficulty in terms of micro-obstacles to overcome, and as you tackle them, it generates exciting scenes, just like those in a good book or action movie.

You did not answer my second question. What prevents a GM from removing or adding details as and when he pleases, in a way that pushes the story where he wants it to go?

Well, first of all, the GM cannot add and remove details whenever he wants. He can only do so in correspondence with a move. Let’s remember when the GM can make a move:

  • When players are watching him to find out what is going on, such as when nothing interesting is happening;
  • When players provide him with a golden opportunity, i.e., ignore a clearly presented threat;
  • When a player fails a roll.

In some cases the GM intervenes on a partial success, following the instructions indicated on the move activated by the player, but these cases do not concern us.

Now, the introduction of the troll is clearly an example of the first type. The player has moved to a new area and it is up to the GM to introduce the troll to him. In this case, the GM can describe as many dangerous details as he wants: as long as it is consistent with what has been said and agreed so far about the nature of trolls.

If, on the other hand, new details are to be added when things are now in motion, it is more complex. The most important moves the GM can make to add dangerous details are show signs of an approaching threat and reveal an inconvenient truth. Also offer an opportunity, with or without a cost and explain the requirements or consequences and then ask can they be applied, depending on the context.

The GM can only make these moves in one of the three cases listed above. Thus, the GM’s ability to make things worse will always depend only on the characters’ choices or the outcome of the dice. Therefore, the GM cannot increase or decrease the difficulty of the confrontation at will. Why is the GM constrained and limited in this way? The reason is simple: so that he can be deprived of responsibility for the outcome of the challenge. In this way, it is no longer the GM’s job to decide, by more or less nudging and judging the actions of characters, how a challenge should end. The GM can worry about fielding interesting obstacles to overcome and can have fun with the players finding out what happens next, relying completely on the rules of the game to determine the outcome.

Wait, but does that mean that if a GM has prepared hidden hazards in advance, he cannot pop them out when he wants to, but only when he has a chance to make a move?

The answer to that question is a very clear yes, at least literally interpreting the manual. I play this way and it allows me to fully enjoy the development of the story, knowing that I have no responsibility for how it will turn out.

I have noticed, however, that some GMs think that having prepared in advance for a threat (such as a trap or ambush) means that it must manifest itself at all costs, regardless of the conditions necessary for the GM to make a move. My personal opinion is that all such dangers should still be introduced at least by a soft move. The manual is imperative about this: any threat must be made clear to the players before it is carried out.

Of course, this issue can be interpreted in ways other than my own, but the important thing, in my opinion, is only one: that the GM is honest and committed to finding out what is going on together with the players.

Thank you very much, I finally understand how this wonderful game works and I am ready to keep to the straight and narrow! You are beautiful and very smart!

You are welcome, dear hypothetical interlocutor, although I must begin to accept that you are a mere projection of my ego. I should stop doing drugs.

5 Appreciations

Ah, this is excellent! I like your analysis here of what happens with PbtA design; I’ve noticed a similar thing, although I really like how you describe it here.

That’s a good and interesting insight as well, about applying those two techniques. Excellent!

3 Appreciations

This is indeed a nice way of making Dungeon world be less about game master just reading the emotional reactions of the player group and adjudicating as a response to those. Framing it as having to get past these four obstacles, and doling out consequences as per the rules, seems nice.

Did you meet any situations where players completely sidestepped some of those difficulties there? Someone who was flying would not be bothered by the shaking ground, or an archer might not care about the big hands.

1 Appreciation

I don’t want to answer for Claudio, but in my experience that’s a bit feature of this approach (I wrote a similar thing a long time ago at the Gauntlet): you can flexibly take into account maneuvers and fictional positioning to sidestep or overcome challenges and dangers, which adds some emergent content and player-challenge or character-character challenge to gameplay, which can be satisfying (particularly in a game where otherwise this is difficult to do).

1 Appreciation

Huh! I found the Gauntlet thread. It’s about harm and hit points, specifically, but the issue is somewhat similar to the “onion” here, just sort of approached from the opposite end. Not as general or flexible as the onion technique, but I think you’ll see the similarities and/or overlap.

My onion-like idea is near the end - there are lots of different approaches and idea discussed there:

1 Appreciation

I am writing the draft of a game (since too long now) where damage is marked on a Harm track. At its core, each arm is a layer of the encounter. The cool thing is it can represent reach, dodging a hazard, or even a coolness field (that moment where the end level boss realizes the heroes are indeed a threat to consider.
Hiding the harm track itself, but clearly announcing soft and hard moves is fun in and of itself. Of course, valid creative solutions are encouraged.