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

[Maxperson]

GM: So, knock has a verbal component, right? [rolls some dice in secret, checks their notes, and manifests their signature grin]

It also has a loud gong that goes off when cast.

 

[Pedantic]

This would require enumerating actions? In that case that is completely counter to the TTRPG medium. We are then quickly into boardgame / structured story game territory.

Oh I couldn't disagree more. The important part is the open ended nature of goal setting, not undefined interaction. A robust set of rules, if anything, makes it clear what players can achieve with each action and in the rare cases an activity outside then comes up, gives the GM more to model a novel interaction on.

The skill typically rolled to pick a lock in D&D5 ed is thieves tools. This skill is potentially used for a wide range of other actions and activities - some of which might be very noisy by nature. Using the hammer to break the window for instance. What would I write down as parameter to my thieves tools skill? Silently +2DC do not make any sense for breaking a window, but it might make sense for picking the lock.

Sure. An intrinsically loud thing should have how it's hard specified under the perception skills, and whatever you use for silent action should clarify what can reasonably be muffled.

Also of course it is possible to design a system that require negotiate exactly how a success outcome will look like ahead of time, and try to come up with a huge list of possible custom modifiers. However I think this smells very much of clunky design that have been attempted quite a few times in early RPG history, but not really catching on as most people find it far too tedious.

The whole point of a big set of modifiers is to avoid negotiation. I've talked before about how DCs should be "derived" instead of set. You get the DC out of the situation; the lock is off X quality, the thief has professional tools, they're spending the normal time and making no special provisions for noise or stealth; that should procedurally output a DC the layer can then interact with.

Ideally, that should all be written down, so that once a player has gathered information about the situation, barring anything hidden they didn't get, they can do the derivation themselves, and go in to the action declaration knowing the odds. Success is clear; the lock is open, failure it isn't. In a stress free scenario, we can start applying Take 10/20 rules and possibly just noting the extra time taken. Frankly, I'd prefer players be spending resources in system scenarios to avoid rolling when possible altogether, probably some rogue ability to gain significant bonuses to the roll to overwhelm the RNG to keep pace with similar effects like knock and silence.

The "game" bit is in how players apply that broad palette of actions to get their desired results. The unbounded play, where that goal will change over time and be ones will be set when it succeeds or fails is where the role-playing comes in.

 

[Maxperson]

Why can't it more consistently be "fail backward", though?

When breaking into a kitchen and failing the roll, why can't the narration be that a previously undetected alarm goes off, or that an unexpected electrical trap in the lock shocks the thief for [a bunch of damage, however the system does such things], etc.*, while - honouring the root 'fail' roll - the door remains locked? One of my favourites for such things is that the thief, thinking she was unlocking the door, has in fact just locked it i.e. for some reason it wasn't locked to begin with.

That's what fail forward actually is. You've failed, but the story moves forward in some way that is interesting to the group. The situation doesn't remain the same, with the group just standing outside of a locked door like there were before the attempt.

* - or even the old standby "the thief breaks her tools in the lock", but that one gets overused and thus a bit stale.

After you've broken all of your tools, that stale bread is looking mighty useful!

 

[FrogReaver]

It's only "quantum" if you move the goalposts and include the DM. The context of this discussion is the PCs entering the kitchen and maybe encountering a chef. Not the DM. So while it may be "quantum" in that until I roll to see if there's an encounter, there might or might not be, this isn't about me.

When the PCs show up to enter the kitchen through the locked door, in your method it's quantum, with the existence of the chef being determined by the success or failure of the lockpicking attempt. In our method, though, the chef's existence or lack thereof has already been determined prior to the PCs making the attempt, so there's nothing quantum about it for them.

I was struggling to follow your reasoning on why quantum and why not earlier but this makes more sense. It's that you are viewing the scene from the players perspective.

Essentially players in D&D don't encounter any player facing quantum resolution mechanics in their play and so from their perspective the cook was always there.

To say it a bit more poetically, Schrodinger looked in his box before they arrived on the scene and so for them the superposition of the cat was never in question. The DM being Schrodinger and the cook being the cat of course.

 

[AlViking]

Or they live there. A feature of many manors that employed a maid or cook would be a maid’s room just off the kitchen. They might be in a position to hear someone breaking in.

Which is fine if the lockpick fails and they decide to bust the door down. But picking a lock? That's not loud. It also doesn't matter whether or not the check is successful - it might actually be less noisy if they fail because if it's a deadbolt type lock the deadbolt isn't sliding.

It's not being discovered the is the issue. It's being discovered in a way that is unconnected to the declared action taken. It's "an interesting narrative" overriding the logic of the setting.

 

[AlViking]

That's what fail forward actually is. You've failed, but the story moves forward in some way that is interesting to the group. The situation doesn't remain the same, with the group just standing outside of a locked door like there were before the attempt.

The issue that I have is that the story moves forward in some way that is interesting to the GM, the group has no say in how things move forward. It feels too close to a railroaded moment, the GM has decided the group will get through the door so therefore it happens but there has to be a cost. As I said above, I'd much rather provide alternatives for the players to choose from.

 

[FrozenNorth]

What happens when the character progresses in lockpicking? The world lacks for cooks and the orphans go hungry.

Same thing that happens if a trad GM rolls on a table and no encounter (or a different encounter) happens. The cook was elsewhere.

 

[FrogReaver]

Yes


And no. This is explicitly more quantum than the other. In the wandering monster case, the events are independent--your skill at lock picking does not affect your chances of encountering a monster. In the fail forward case, they are dependent--a skilled thief encounters fewer wanderers.

The thief's observations play a role where they didn't before, hence quantum.

I think the dependence/indepedence of the roll and the quantumness of it are 2 separate things.

 

[Micah Sweet]

Easy. I know the party is going to be traveling through a forest for three days. So I roll random encounters for all three days prior to the session ever starting. I know what the random encounter will be, when during the day it happens, what direction it is coming from, and so on.

A day is a day. If they are in the forest, the random encounter is happening at 6pm that day, whether they are futzing around looking for herbs or traveling straight through. Random encounters are not preset encounters with a specific location, but rather a specific time. They wander into the group, which is why they are also called wandering monsters.

With regard to the cook, I've established through rolls that she is going to be in the kitchen from X time to Y time, then she is going to bed. If the party shows up during that time period, they will encounter her regardless of their pick locks roll. If the show up outside of that time period, they aren't going to encounter her, even if they fail the roll. She is a preset encounter, not a random one.

Still not quantum, because a preset encounter remains in that location whether they ever go there or not. It's existence is not based on whether the party wanders there.

Nothing about what I am saying is quantum. Nothing is based on whether the party does something or not in order to be in a particular spot or time.

All of that checks out for me. I followed that exact procedure a few days ago while prepping for my Thursday A5e game.

 

[Micah Sweet]

I think it might be helpful to seperate between the concept of random encounters and random stocking. It appear to me like what you describe amounts to random stocking using random encounter tables.

In common parlance "random encounter" refers to exactly the procedure of play you say you are not using - making a roll on the spot during play based on some trigger (time passing being the most common, but on loud noise is another classic). So when someone are talking about properties about random encounters, they are not talking about anything that appear relevant for your game; but it is relevant fot a lot of traditional play.

Fair enough. I use "random stocking" techniques in my games.