D&D 5E Consequences of Failure

[FrogReaver]

Last paragraph, entire paragraph: You'd be very wrong, and your formulation is wrong. A climax is the pinnacle of the action. Having tense actions before, that are interesting and engaging at the table, are not climaxes. If every roll is meaningful and some are even fraught, this doesn't speak to climax at all.

I'm not going to get drawn into a definition debate. Suffice it to say that others hear read and understand what I'm saying.

Example: my last session -- the party rushed a group of gnolls lead by a flind, but, due to initiative and actions declared, ended up in a position where the dwarven barbarian was in the room killing some gnolls but the rest of the part was still in the hallway. Two of the gnolls rushed the door, closed it, and dropped the bar on it, separating the party. The barbarian quickly went down to the flind and the gnolls. In the meantime, the party ranger tried to bash down the door, figuring this would be the quickest path to access. The other party members either tried to assist or tried to think of a different way through the door. The door resisted the bashing until the dwarf went down on the other side, at which point the cleric suggested the vial of acid she'd picked up on a previous adventure. As the immediate time constraint was gone, I ruled that applying the acid and letting it do the work so that the door could be broken open would take a minute of time.

Here we have an example of a check that had a big reward for success and a bad consequence for failure (in this case, just time, as it was another round the dwarf had to fight all the gnolls alone). This was a very fraught moment, as I'm the kind of DM that will kill characters in this kind of situation. It was not a climax, though.

Cool story. Did you tell it just to tell me that it wasn't a climax?

The dwarf was knocked unconscious instead of killed because the flind, wily as it was, had other problems in the complex and so took a hostage to bargain with the party to go soften up the other problems (assuming the party would die, because he'd already beaten them). The party managed to engage in some negotiation by noting the arrogance of the flind (insight check) and then leveraging that into advantage on the ask for the dwarf back instead of being kept as hostage ("You've already defeated the dwarf, mighty warrior that you are, surely you don't need to keep him and he will help us help you."). The agreement struck, the dwarf was returned to consciousness and the party was locked into a different hallway (triple barred door this time) with a Stone Golem. Which was, amusingly, exactly where they wanted to be.

Again, checks with consequence and reward, but not climaxes. The ask for the dwarf was fraught, though, as it meant leaving a party member behind with murderous gnolls and lacking his strong axe arm in any fights.

Thanks for the play recap. Sounds fun. I don't get why you told me that story just to add on the end "no climaxes here."

You do realize we are using the word differently? So I say there were climaxes by my use and there weren't by yours. So what?

 

[Guest 6801328]

If I do have a investigation roll planned, then I'll probably let the character know, "Ok, roll Investigation, but if you fail, then you've spring the trap." If the player meta-games and says then they won't bother since me saying that admits the lock is trapped, then I'll remind them that player knowledge and character knowledge are different things.

In that case, why tell them the consequence for failure? I like to make the consequence known so that the player can make an educated bet, or choose not to. But if providing the consequence negates their ability to make a decision, what roll does the information serve?

 

[FrogReaver]

Oh oh oh.

I think I have miscommunicated.

(This explains so much.)

I did not mean to suggest that every roll needs to herald an event that gets told and retold around the gaming table for years to come.

Maybe you didn't miscommunicate, maybe I did. I wasn't suggesting that. I was just using that to illustrate that to have fun and memorable game or moment in the game that you need the more generic occurrences to really put it in perspective.

I just meant that the structure of the task resolution should support this being possible, even if the majority of the time the stakes are too low for it to become a legend. But I think there has to be some kind of gamble, some kind of negative consequences you are risking, for this to be possible.

Agreed. I'm just suggesting that maybe there's not really as much benefit as you think there will be in giving every check that opportunity.

By way of contrast, "I'll check for traps." "Gimme an Investigation roll." "X." just isn't going to ever produce a great story. Maybe what ensues because of the roll will be a great story, but that could as easily have been determined by the DM without dice, so it's not the same thing.

Agreed. I'm just suggesting that not every check has to be capable of being something that produces a good story by itself. Sometimes you need the bland to appreciate flavor.

And, no, there's nothing inherently wrong with checking for traps. (God knows I've done it enough times.) . But there's nothing great about it, either. So maybe it's superfluous? That's all I'm saying/asking.

The game can be played without checking for traps. Heck, it can be played without traps at all. So yes, it's superfluous in that sense. But I'm not so sure that it's an activity that is valueless in relation to fun. Even though it may not be particularly fun by itself, it may enhance the overall fun of the game.

Does that make more sense?

I think so, does what I'm getting at make more sense?

 

[FrogReaver]

The latter scenario is entirely possible. Not every telegraph is correctly interpreted by the players.

If the audience doesn't pick it up then it definitely wasn't a successful telegraph. If the notion is that you attempt to telegraph such things and if the players pick up then they do and if they don't then they don't then that's fine, but that's not how telegraphing was spoken of earlier. It was spoken about such that there couldn't be an unknown trap because the characters would always be aware of it due to telegraph.

 

[LassiterKeshell]

In that case, why tell them the consequence for failure? I like to make the consequence known so that the player can make an educated bet, or choose not to. But if providing the consequence negates their ability to make a decision, what roll does the information serve?

Again, I'd base it on importance. If, in this example, the trap was important, then I would have more than likely made it a puzzle (and left clues about it along the way).
If the trap isn't important, then letting the player know what happens if the Investigation rolls fails imparts the following things: they know the roll is more than just staring at a lock with their naked eye, and whatever is on the other side of the lock isn't important to the story (but if it's trapped, then there might be something valuable anyway).
As far as what role does the information serve, it's a boobytrapped lock... how many ways can that play out?

 

[Bawylie]

If the audience doesn't pick it up then it definitely wasn't a successful telegraph. If the notion is that you attempt to telegraph such things and if the players pick up then they do and if they don't then they don't then that's fine, but that's not how telegraphing was spoken of earlier. It was spoken about such that there couldn't be an unknown trap because the characters would always be aware of it due to telegraph.

Hang on. “It was spoken about” might not have included a good faith explanation. Might’ve been a colorful criticism instead. (Honesty I don’t remember).

But a telegraph isn’t a guy waving a sign that says “traps here!” on the street corner. It’s a hint or a clue that may or may not be picked up by the players but IS included in some way in the DM’s setup of the scenario.

If the players don’t pick up on it, it MAY be because it was insufficiently described (every DM should check to make sure the hint is not so subtle that it’s unfairly impossible to pick up on), or it MAY be that the players just don’t quite get what it is or otherwise get distracted by something shiny.

 

[Charlaquin]

Sorry, this reference is lost on me - what's Metal Gear, other than electric guitars and a wall of amps?

Metal Gear Solid? It’s an extremely popular video game series. Involves a lot of stealth, and when the main character, Snake, is at risk of being spotted, a red exclamation point appears above the head of the NPC that is about to detect you.

Sometimes doing this is cool. But if there's something behind the door that's also stealthy, do you telegraph that too?

Does this come up often for you, that a PC is sneaking around, and a hidden NPC is searching for them? I can’t say it does for me. I would imagine the NPC would be too focused on remaining hidden to also be searching. If players have to pick one task or the other when exploring, shouldn’t NPCs have the same limitation? I don’t know, doesn’t seem like a major concern with the way I set up challenges.

I don't want to know what they do about something in reaction, I want to know what they proactively do in general before any of those somethings come up.

Sure, that’s interesting too. Preparation can be a good way to get yourself Advantage on a future check, or even eliminate the need for a future check.

And now. This...
...directly conflicts with...

...this.

I disagree strongly.

Without a strong element of realism backing things up it becomes impossible to make decisions based on what people would do in that scenario. Put another way, without realism as a goal your other goals become unattainable.

And here, by 'realism' I don't necessarily mean things being true to the real world (though that's always a good start for those who don't design their own settings), I mean a setting that has its own built-in realism based on strongly enforced internal consistency and reliable cause-effect loops.

I’d call that verisimilitude rather than realism, but otherwise I agree. I just don’t believe that players having access to knowledge their characters wouldn’t actually harms verisimilitude. Verisimilitude is about insuring that the world behaves and reacts to the players’ input on believable and consistent ways, not about the players’ experience and knowledge perfectly matching the characters’.

And some of that internal consistency is represented by the PCs not always knowing everything...and sometimes barely knowing anything...about their situation.

Sure, but again, the PCs not knowing doesn’t mean the players can’t know. And while verisimilitude is important for empowering players to be confident in their decisions, so is access to pertinent information.

You set out to sneak down a hallway past 5 doors, all closed. You don't know how many, if any, of those doors might have observers or threats (or treasure!) behind them - hell, for all you know one or more of the doors is an illusion!

If there's a dog behind one of the doors it might smell you and rattle its chain a bit, or even bark.

Ok, so taking it as a given that the dog has a chance to hear a person who is actively trying to move silently through this door, it seems reasonable to say that the character could pretty easily hear a dog that is not attempting to conceal its presence through it. So I’m going to narrate that and ask the player what they do about it. Assuming what they do is attempt to sneak past by walking carefully and quietly, I’d call for a Dexterity check. The player could add their proficiency bonus for Stealth if they have it. Until they stop hiding or are detected, the result will set the DC for Wisdom (Perception) checks to find them.

If there's a guard behind another door she might hear you and raise the alarm; or not. The prisoner (who, if released, could become a useful ally) behind a third door might hear you and try to get your attention. A fourth door is an illusion with a trap behind it - a pure thief-catcher. And the fifth door is, well, just a door; with nothing behind it but a closet.

How many stealth checks?

We’re already making a lot of assumptions just to get as far as one check made in response to hearing the dog. What else might happen from there is contingent on so many variables, including what further actions the player takes and what the outcomes of those actions are, I can’t tell you with any degree of accuracy how many Stealth checks.

Conversely, all five doors might hide no threat or problem at all.

How many stealth checks?

None, because there is nothing to detect the character; no risk, no challenge, no point rolling dice.

As the PCs don't know what's behind the doors, the answer to "How many stealth checks?" should be the same in either case.

I disagree. If you want to do it that way, that’s fine, but the game won’t be ruined if you don’t.

My point is that it in fact doesn't serve that goal well at all.

Your goal is to "create opportunities for the players to imagine themselves as other people, in a fictional scenario". That's cool! But it only works if they players are exposed to the fictional scenario as it would really be, i.e. with info that the PCs don't and can't know kept hidden from the players and not unduly telegraphed or hinted at. And yes, this naturally means that sometimes you're gonna hit 'em with a 'gotcha'; but that too is only realistic.

I don’t think you’ve really demonstrated that to be the case.

 

[Charlaquin]

I had a thought, and I gave up reading each post around page 9 so bear with me. On any kind of check without a relatively obvious consequence (history, knowledge, whatever) on a success we get the info we want, on a failure we get helpful but indirect info. Not rolling we get nothing.

As an example. Players find a book written in an obscure language, maybe in code maybe not. They ask questions of the DM, something about makes the wizard think it might be magic in some way (trained in Aracan, no check needed). Wizard wants to decipher here and now. DM decides it is possible, but difficult so a check is needed. Success results in a decipher tome of whatever. Failure results in the wizard not deciphering, but recalling a master cryptologist lives in Waterdeep... six week journey away. Not rolling means no info of any kind.

Neat! I like that idea. What would you say if the wizard asks to try deciphering it again?

 

[Charlaquin]

If the audience doesn't pick it up then it definitely wasn't a successful telegraph. If the notion is that you attempt to telegraph such things and if the players pick up then they do and if they don't then they don't then that's fine, but that's not how telegraphing was spoken of earlier. It was spoken about such that there couldn't be an unknown trap because the characters would always be aware of it due to telegraph.

Who ever spoke about telegraphs that way?

 

[Lanefan]

Metal Gear Solid? It’s an extremely popular video game series. Involves a lot of stealth, and when the main character, Snake, is at risk of being spotted, a red exclamation point appears above the head of the NPC that is about to detect you.

Ah, OK.

My video game career began and ended in the days of plug-in-the-quarter arcade games.

Does this come up often for you, that a PC is sneaking around, and a hidden NPC is searching for them? I can’t say it does for me. I would imagine the NPC would be too focused on remaining hidden to also be searching. If players have to pick one task or the other when exploring, shouldn’t NPCs have the same limitation? I don’t know, doesn’t seem like a major concern with the way I set up challenges.

Cat-and-mouse is an old trope. But a more common situation than an NPC searching for them is that someone (somehow) knows the PCs are coming and a stealthy NPC has been sent out to scout them - or attack and expose them.

I’d call that verisimilitude rather than realism, but otherwise I agree. I just don’t believe that players having access to knowledge their characters wouldn’t actually harms verisimilitude. Verisimilitude is about insuring that the world behaves and reacts to the players’ input on believable and consistent ways, not about the players’ experience and knowledge perfectly matching the characters’.

Where my point is that in order to be able to give that input in a consistent and believable way the player's knowledge should (as far as possible) match that of the character.

Otherwise the input is either filtered through some sort of self-policing or is liable to be incongruent with what the character would do otherwise.

Sure, but again, the PCs not knowing doesn’t mean the players can’t know. And while verisimilitude is important for empowering players to be confident in their decisions, so is access to pertinent information.

I guess we disagree on the definition of 'pertinent'. If I-as-DM fail to tell the player about something relevant that the character would have known or observed, that's on me. But that by no means suggests I should be telling them about things the PCs don't (yet) know.

Put another way: I hold and retain the right to now and then surprise the players and PCs together.

Ok, so taking it as a given that the dog has a chance to hear a person who is actively trying to move silently through this door, it seems reasonable to say that the character could pretty easily hear a dog that is not attempting to conceal its presence through it.

Not hear, smell. But in either case I'd probably do a series of things:

First, I'd roll to see how much noise the dog happens to make; as this is a somewhat random element and informs what comes next. Then, based on that roll I'd give the PCs a Listen (or Hear Noise) roll if the dog was quiet, or skip straight to narration if the dog, say, started barking (i.e. a really high how-much-noise roll).

I'd then narrate from the results of the Listen roll.

We’re already making a lot of assumptions just to get as far as one check made in response to hearing the dog. What else might happen from there is contingent on so many variables, including what further actions the player takes and what the outcomes of those actions are, I can’t tell you with any degree of accuracy how many Stealth checks.

I was referring to how many Stealth checks to get down the entire hallway - one per door? One overall? Or?

None, because there is nothing to detect the character; no risk, no challenge, no point rolling dice.

And right here is where I would want to get off the bus. The hallway should be treated exactly the same in the mechanics, regardless of what's actually there, until and unless something happens to change that. So, if it's one check per door it's one check per door, end of story; until and unless either a check result causes the PCs to be noticed (if there's anything to do so) or something else intervenes (e.g. an unexpected guard walks around a corner).

And why is this? Because neither the PCs nor the players know what if any threats lurk behind those doors, and nor should they until and unless a) a threat makes itself obvious (a dog barks; or a door opens) or b) the PCs notice something that indicates a threat might be present (they hear a dog; or see light coming out from under the door).

It sometimes generates a big sigh of relief from the players when they absolutely butcher a sneak check and my response goes something like "Well, fortunately for you you picked a good time to mess that up; as nothing seems to have come of it."