D&D General [rant]The conservatism of D&D fans is exhausting.

[CreamCloud0]

"Fail backward" would be, "not only did you not succeed, but now things are worse than if you hadn't tried at all."

You want a disincentive to trying things? You got one right there.

Oh but think of the story they're creating!

 

[Campbell]

If your issue is that fail forward means we're no longer in task resolution / sim mode, cool. You do you. An example of what fail forwards looks like is not needed to make that point.

But if your issue/point is that fail forward means the results will lead to inconsistent or implausible fiction than I'm going to say that depends on how the GM frames both success and failure. I don't think providing an example of a GM making a poor framing choice is all that condemning because there are all sorts of coherent moves/framing choices that will be available in any circumstance. Choosing to use examples where we are juxtaposing poor fail forward GMing to experienced trad GMing is not fair and incredibly misleading (and I am much more concerned with the misleading part).

The other point I'd make is we don't really know what success would look like or the framing decisions a GM is going to make as the explore the rest of the house, if they do so. Personally, if I'm introducing the cook as a possible complication that means I've set up the idea there might be people here and that making noise might draw their attention. That telegraphing applies to the whole house and will likely be relevant in any future framing I do.

Once we leave sim land, we leave sim land. Our ability to determine what moves/frames will and will not be made disappears. It's a whole lot of undefined.

I'm also not really sure why our examples are always the sorts of conflicts that are much more likely to be seen in exploration-oriented play. I cannot remember the last time a locked door was relevant in any of the games I run or play.

 

[FrogReaver]

You're hung up on the cook.

In lock picking, fail forward means, you get in, but there's a complication of some sort. Such as:

• you lose or break your tools
• you startle someone who is in the room and they make a noise, alerting others
• you cause an inordinate amount of noise doing so, alerting others
• you leave evidence behind that can be tracked to you
• you hurt yourself in the process
• someone sees you and doesn't immediately scream, but they can ID you later on

Pick one. There doesn't have to be a quantum anything; three of those results require nothing but the PC.

That you are picking one instead of generating the result fully via a mechanic (like a table) is the quantum-like state (honestly describing this as quantum like probably harms the explanatory power more than helps but it’s a fairly common description so what can one do?)

In an ideal, maybe ‘platonic’ traditional game the result is going to be either 1) prefilled random table driven or 2) extrapolated directly from established and preauthored facts.

For 2) there is an implicit plausibility test such that extremely implausible scenarios are filtered out and in the case more than 1 remains they are weighted appropriately and rolled for. But this part is mostly beside the overall point.

The recurring issue is in how you are picking the particular result. (Which isn’t an issue in itself, but it does have pros and cons for different player experiences.)

 

[zakael19]

Because this doesn't move the game along or provide an interesting consequence. It's merely punitive. Or worse, if the thief had checked for traps and rolled well--yes, there are traps the thief might not be able to see, but you're making one up here on the spot for no reason other than to "fail backwards."

(The last one is, IMO, pretty dumb. You're assuming that part of the thief's door-checking routine doesn't involve checking to see if the door is opened. If you did that to me, I'd start narrating every time I inhale, just in case you decided I was holding my breath. I'd make a little recording and have it play on a loop.)


Lockpicks are very thin and fragile; having them break often is fairly logical, unless they're made of, like, mithril.


Fail forward takes this into effect. Here's something I found on an old reddit post, with the idea that the PC is chasing a mysterious assassin:


Note that none of these involve the PC automatically catching the assassin. What they all do is make the adventure continue to move.

When it comes to picking the lock, if there's no consequences for failure other than "nothing happens"--there's no deadline they have to beat, no monsters or guards that might find them if they take too long, nobody relying on them, nothing to learn from the lock itself, a failure simply means "nothing happens"--then why have them roll at all? Why not just say that the thief picks the lock? Having them roll, whether it's a failure or a success, is just a waste of time.

What really “fails forward” IMO is the scene and the fiction. It’s not “fail succeed,” it’s a conscious avoidance of “nothing happens.”

 

[LordBP]

Couple more links on locks

The history of padlocks

Lock Picking Masterpost

 

[Micah Sweet]

Or to ask an even broader question, could you enjoy a trad-style game if you knew the DM was just making everything as they want along?

I ask because no-prep trad/neotrad type games are probably my most common GM style.

I would enjoy it less personally, but my players would probably be fine with it.

 

[Faolyn]

::sigh:: Missing the point. I'm not looking for a reason for the cook to be there or not. I'm saying that when I DM I limit the result of a failure to what makes sense in the context of the action taken. Picking a lock makes as much noise as opening the door with a key, someone would need to be practically adjacent to hear it. How do you think it works?

If the character breaks a window because they can't open the lock, the noise of breaking glass may wake someone up. The in-world fictional result follows the action taken.

That's why there's usually multiple things that can happen in a fail forward, not just "noise."

 

[Micah Sweet]

But you are only considering circumstances in which it DOESN’T make sense. That is the point.

If the situation was that a cook was rolled up on a random table, you would make a proactive effort to come up with a reason that it makes sense in context. Why aren’t you doing that in this case?


This is a perfect example of what I mean. I live in a house built in the 1920s. Whenever someone opens the front door, I can clearly hear it even if I am in the kitchen 50’ away.

Rather than taking as the assumption that many doors are pretty noticeable when unlocked or opened, you proceed from the assumption that only someone adjacent to the door can hear it. Starting with this assumption, the conclusion is invariably that fail-forward is implausible.



Skills in D&D are fairly broad, so rolling a Lockpicking check is more than simply physically adjusting the tumblers. Here are 3 reasons it might make sense:

• The cook is an elf that is trancing and doesn’t need to sleep. The house is pretty quiet at night, so the unlocking and the opening of the door is noticeable and they come to investigate.
• The rogue unlocks the door but fails to grease the hinges. The door lets out a loud squeak as it is opened.
• The rogue unlocks the door, but fails to notice the small alarm Sigil. The silent alarm notifies the cook in the other room.

Isn't that just the opposite? Looking for reasons for the desired effect to make sense rather looking for reasons for it not to?

 

[Micah Sweet]

But they weren’t. They were only there because the GM made a random encounter roll and rolled 82: cook. And the GM only made the random encounter roll because the PC was trying to break into the house!

But it wasn't in response to the actual attempt to break in. It was in response to needing to know who was in the house and where.

Again, you probably don't see the distinction. Others do.

 

[Micah Sweet]

If that is the case, doesn’t the fail-forward GM have exactly the same amount of latitude to decide that the consequence of the failure is something that makes sense?

If they're happy with the other implications of the technique, then sure.