So my first game of Dungeon World has hit the ground and we're all having fun, but something seems off in handling Bonds at the end of session.

One of my players has the bond, "Able always has my back when things go wrong." while another has the bond, "I am working on converting Baker to my faith." And these are great, the players are having great fun in exploring them and acting on them, etc.

But then we get to the end of session move:

When you reach the end of a session, choose one of your bonds that you feel is resolved (completely explored, no longer relevant, or otherwise). Ask the player of the character you have the bond with if they agree. If they do, mark XP and write a new bond with whomever you wish.

This seems like it favors the bond with Baker, because that appears to be a goal with a more-or-less well-defined metric for completion. I can easily understand how this bond can resolve in a way the players are happy with.

In contrast, the bond with Able is more of an existing, recognized status, almost more like an alignment to be acted on than a bond to resolve (and in fact Able's alignment is to protect those weaker). I can't really see how this one can properly resolve if it's getting acted on constantly, which seems to put them at a disadvantage, denying them both the xp and the chance to establish new bonds with the party.

How do I handle these bonds that don't seem to lend themselves to any kind of resolution?


4 Answers 4


Personally I've never found the starter bonds for classes to be all that compelling for exploring a party's inter-relationships. And for new players they can lead to some poor assumptions about what constitutes "good" ones. To alleviate that:

Add something actionable

Like you've noted, bonds like the Thief's "_____ has my back when things go wrong" tend to just describe states that two characters are in rather than offer them something actionable to draw upon in play. My go-to solution to prevent this is to simply have them add what their character intends to do about it to the end. Bonus points for making it something that can be done in the short term (i.e. next session). For example:

  • Able has my back when things go wrong. I'm going to make it up to him.
  • Able's misguided nature dangers their very soul! I must show them the way.
  • I worry about the ability of Able to survive in the dungeon. I will keep them safe.

You'll notice that many starter bonds have this element in them anyway, but in my opinion, all of them should. Doing this will also set up good habits when developing bonds of their own—so long as there is an element to them that can be acted upon, character relationships can grow and develop naturally as each one is resolved.


the bond with Able is more of an existing, recognized status

In which case, it’s resolved.

You should think of the bond not as:

Able always has my back when things go wrong .

But, rather:

Able always has my back when things go wrong ?

If, in the session, there was an opportunity for this to be demonstrated and it either was or wasn’t i.e. Able showed he did (or didn’t) have your back, then the bond was resolved. If he has shown he always had your back then there is nothing more to be explored about it. If he showed he didn’t well, that’s even more interesting but there’s still nothing more to be explored about this - although it does raise some new bond opportunities.

Now, conceptually, it's no different from your other example, "I am working on converting Baker to my faith", it's just expressed passively rather than actively. This bond is resolved when Baker is converted or when it becomes clear that Baker will never be converted - at that point it's "totally explored".

  • 2
    \$\begingroup\$ Might be worth pointing out that "completely explored" is one of the reasons for resolving a bond. If you've seen all the ways in which Able has your back, it's time to move the spotlight. \$\endgroup\$
    – Erik
    Commented Oct 13, 2020 at 4:32
  • 5
    \$\begingroup\$ @Erik if say “seen enough” rather than “seen all”. “All” includes a lot of things - everything for a start. \$\endgroup\$
    – Dale M
    Commented Oct 13, 2020 at 9:10
  • 2
    \$\begingroup\$ The impression I'm getting from this answer is that Bonds should act something like stakes questions in a sort of 'party cohesion' front. Am I reading you right? \$\endgroup\$ Commented Oct 13, 2020 at 19:05
  • \$\begingroup\$ @StopBeingEvil not exactly, a bond is a link between characters that still has something unknown about them - whether that is a "stake" or not depends on the particular bond. "Joe is my brother" could be a bond if there's something to be explored in that relationship or it could just be a fact in the world. \$\endgroup\$
    – Dale M
    Commented Oct 15, 2020 at 0:54

The example bonds are just that, examples

Alignments, Bonds, Moves, Monsters, Races, even whole classes: Everything in Dungeon World can be made up and decided by you as a group.

Make things fit your preferred playstyle

In my groups, I run bonds as goals that you have for specific characters. So they should always be phrased like the Baker bond. Or they start out as 'I don't know this Baker guy, but he seems the friendliest so far - I want to know if he will have my back against X' Ultimately of course, you might have an amazing, trusting friendship with a character and that's great! You can still have bonds there though if you think of them as things you want to do between the two characters, things you want to happen, scenes that you want to play out which ultimately all affects the dynamik and the story. f.e.

"Baker is my friend, I trust him! But he really doesn't get along with Able, maybe I can help them find common ground?"


"Baker is my friend, I trust him! The last few battles seemed to be really hard on him though and he think's he is no good. I will try and help him feel good about himself again." (this could be a big party, three successful, pivotal rolls by Baker's player or other things, as decided by the group when the bond is written / presented)


"Baker is my friend, I trust him! He recently lost all his magic powers, oh no! We need to fix that and chase down the evil Malakai who has them!"

Or all kinds of other things :-)

Back to basics is fine too, though

Or, if you just want to run with the basic stuff for now, refer to Adam Koebel (Co-Creator) - Dale M already kind of referred to this, but here is a link and a quote. https://twitter.com/skinnyghost/status/1211457430980218880

So if you've got a Bond like "Sage and I have a con going." but you're not all that into the Bond, you just say "Hey, Sage, I'm not feeling this Bond." at the end of session and Sage says "cool, ditch it" and you get XP. If Sage says "Actually, I was thinking we could explore that some more." you keep the Bond and you and Sage work on making it more fun together.

As you can see, games in DW are very mod-friendly and do tend to play drastically different depending on the GM (not that that's not the case with more crunchy systems but they do tend to feel more strict, generally).

Personally, I obviously prefer the bonds as group-person goals approach though as I feel it makes them more meaningful and shows a development of the relationships in the group... especially if you keep old bonds with that character in your notes, it can be a nice feeling to look back at how a relationship grew into what it currently is.


"We need to go deeper."


You're right in that a lot of the reference bonds seem to suggest their own endpoint. Wizzrobe did a prophecy about Fightgar, and the prophecy happens, or it doesn't, or there's a twist no one could have foreseen. Shanksworth stole something from Clericsdottir, and she finds out, he comes clean, or she gives up looking.

But some bonds seem like they're more about "facts in the world". Stringfellow sung the tales of Sir Justice long before they met. Or, yeah, Fletcher has Shanksworth's back when things go wrong.

Just because a bond seems more of a fact than a plan, though, that doesn't absolve you of your responsibility as GM to ask questions and use the answers. The first class of bonds aren't suggesting their own endpoint as much as they are suggesting their own questions.

There's a prophecy about Fightgar! Cool, so what's the prophecy about? What's Fightgar do? Shanksworth stole something from Clericsdottir! Cool, so what did he take? Who does she suspect? Does she suspect?

(Fleshing out those questions in the bond is a solid part of my first-session worldbuilding.)

The bond resolves when those answers are tested, and for the bonds where the questions are obvious, the circumstances where the bond is tested and resolves are also going to be pretty obvious. This is it! The prophetic moment! Can Fightgar be everything Wizzrobe saw? But for bonds that seem like more of a background thing, the circumstances are going to depend on the answers you got.

So you think Fletcher has your back when things go wrong? Why, what happened? Who'd he back you against? How badly did things go wrong? And the circumstances that test that bond are going to pose a question of the general form "Things have gone worse. Does Fletcher still have your back?"

And when it happens, don't stop asking questions and using the answers. On a rewrite that bond could wind up anywhere from "Fletcher sees something better in me than I do in myself" to "Fletcher is a fool, but a useful fool", and those are two very different stories but I'd love to follow both of them.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .