D&D 5E Consequences of Failure

[Ovinomancer]

Okay, so, to me, the crux of a check is that the fiction changes after it's made. On a success, the fiction changes towards the goal. On a failure, the fiction changes away. The away part is the consequence for failure, and take many forms, but must always, always, always be a change in the fiction.

So, to that end, if you're having a bit of trouble coming up with whether a change should happen, get formal with your stake setting. Before you ask for a roll, clarify what you think the stated goal and approach is, say what you think a success looks like and then what a failure looks like, then set a DC. If you get to the middle and can't come up with what a failure looks like, don't ask for the roll, just award the success. If you get to the DC and think, "oh, I'll set this at 30!" Don't ask for the roll, narrate the failure. The trick is to get in the habit of thinking "what changes with this roll" every time. If you need to do this out loud and formally, go for it -- it won't make your game worse. I prefer explicit stake setting, and it helps me remember because I, too, still occasionally wrestle with old habits.

As for knowledge checks -- these are hard because they don't function like other ability checks in the 5e rules. They're kinda vague, and don't seem to require a check, and go into telling the players what their characters think. Failure cases seems to be either "you don't know" or making up wrong information. I dislike the wrong info path, as it may erode trust in the DM's provision of information, which is critical for a good game. To get around this, I tend to make knowledge style checks active. If you want info, you can have it if you're appropriately trained. "I recall what I know about demons to identify this one and determine it's weaknesses" gets a pretty direct answer if you leverage "using my long training in planes under my Arcana Masters" or "because I grew up hiding from them in the Abyss as part of my planar urchin background." Whatevs. I do not predicate my challenges on not knowing things about the foe, so it's not a big deal to provide this kind of information freely.

What I try to do with these skills, though, is provide framing that encourages active use of the skills. I may say, "this demon is one you've not discussed in classes or study before, but given the depths of the Infinite Abyss and it's chaotic nature, that's to be expected. You do recall the taxonomy of demons from the lectures, though, so you can ask one question for free from your study or it's a DC 15 INT check. Succeed and you get 3 questions. Fail and you still get the question, but you'll grant advantage to the demon as you get momentarily distracted by your study." Or, maybe it's disadvantage on a save. Depends. Either way, I make the check an active interaction with the environment and this helps ground the fictional positioning for success and failure.

This is a poor example, though, as I usually don't do trick monster encounters like this. I do build in plenty of features where use of the "knowledge" skills turns into direct activity. I try to give what are otherwise single niche challenges a broader appeal by grounding them into the fiction a bit more. Say, maybe, you'd have a trap on a chest that they party needs to get the stuff inside while there's a fight going on (otherwise, traps aren't as much fun). Instead of just a challenge for the rogue, maybe provide that the trap is based on an ancient society and is, in fact, a puzzle based on that society's legends. Then the cleric who's trained in history can jump onto solving the trap by leveraging that while the rogue gets to go fight for a bit. I try to do this often -- make situations have more than one obvious hook with one often tying into a 'knowledge' type skill so that these get screentime in ways that aren't "is it okay if my character knows something?' and more in the 'My character knows something, and will use it to do this thing." In other words, design so that you lean on the characters having knowledge and use the knowledge skills that way rather than just use them to ask questions. This both increases the depth of the fiction by becoming more rich but also lets players that take knowledge skills feel like that skill is something that says how cool their character is rather than a button to press to get information from the DM.

I do the same with Insight -- I use it as the "elicit information" skill. Persuade is for convincing others, intimidate for scaring them into compliance, and deception is for... deceiving. Insight is for find out things. If you build your social challenges such that the NPCs have handles like I suggest for the traps above, then Insight is super engaging as a skill that gives you a leg up or a different in to this NPC rather than just punching the CHA skills. Don't make "Is this NPC lying to me" a part of the challenge -- make why the NPC is lying, or what specifically they're lying about the challenge. If you do this, then Insight isn't the button mash "are they lying skill" because your'e telegraphing the deception because your challenge isn't about detecting a lie, it's what you do anyway. Then hang some stuff on the NPC -- a bond or flaw (you don't have to do all of them, one or two usually does for most NPCs and makes them richer in play) that the players can Insight to pick up on and then leverage. Use the social rules in the DMG -- they actually work pretty well and create a deeper interaction. Don't just freeplay social interactions, formalize them a bit. Get explicit with the players. I'm working with my group and fighting my bad acting habits to get social interactions further away from us acting and someone dominating the social roleplay through superior acting to making the social part of the game actually part of the game. The funny voices are still there, they're just no longer standing in for the resolution mechanics.

All in all, to steal from PbtA, be a fan of the PCs. This doesn't mean give them stuff, but instead means be interested in who they are and what they do. Give them interesting things to interact with so you can enjoy the characters. But, don't shy away from laying on the hurt, because they're even more interesting in adversity, and bringing adversity is your job as DM.

 

[Guest 6801328]

Nice. At least there is some acknowledgement in this thread about the shortcomings of goal and approach.

So can we make goal and approach (as meant here) work for knowledge checks.

I think it's possible to make anything have a meaningful consequence of failure so long as you aren't too picky about how that failure looks.

Consider an arcana check to learn how to shut off the magical portal. Perhaps a failed arcana check gives you the information but also has an enemy come through the portal or gives info that the portal is more difficult to shut off than it would have been if you succeeded.

In other words, if you don't maintain that the consequence of failure must directly follow from whatever your character is trying to do then you can have exactly what you are looking for. Personally I dislike that technique, though I still like having it in my toolbox just in case.

The tricky thing here is (assuming you have open rolls) that the player know he/she failed the roll, so giving them false information isn't going to lead to a bad result, and therefore isn't really a penalty for failure. (Caveat: if your table enforces the sort of roleplaying where it is expected that the player will act on the knowledge anyway, because to do otherwise would be "metagaming", then this might work for you.)

I suppose...hearking back to @Charlaquin's comment about stealth...you make them roll at the critical moment. That is, when the character actually attempts to shut down the portal, and commits to it, that's when the roll is made.

And maybe that's the general case answer to these scenarios? (Not that you handle ALL knowledge tasks this way, but when the stars align and inspiration strikes.)

Let's try it out with monster abilities/features. As long as the DM is willing to be flexible, I could see something like this:
Player: "Thinking back to my course in Elemental Combat Fundamentals in Wizard School, I try to recall if these things have any special vulnerabilities."
DM: "Well, what do you remember?"
Player: "That they are especially vulnerable to sonic damage!"
DM: "Cool, try it out."
(Player casts Shatter)
DM: "Let's see an Intelligence check."
Player (using @iserith's method): "I'm going to apply Arcana, since this was something I learned in Wizard school....oh, sheet, I rolled a 2. So 7 total."
DM: "Well, lucky for you it wasn't a total fail* and you managed to not heal the thing, but you can tell that not only did it not do as much damage as you were expecting, but it barely seemed to affect it."

Two things about this scenario:
1) Whatever the stat block said, the DM was willing to switch it up on the fly. (Although from now on these creatures probably have resistance to sonic damage.)
2) The penalty for failure came in the form of an underutilized spell slot, and turn. (EDIT: In my book this qualifies for Ovinomancer's criterion, as the odds have just shifted in the monster's favor.)

And I'll point out that the above scenario, with higher stakes and more dramatic effect, could very well turn into the sort of story that gets told around that table for years to come. "Remember the time...?" Which is in some ways my ultimate test.

*By the way, I don't intend to keep promoting gradations of success. I tossed that one idea out there a while ago, and that seemed to have gotten interpreted as "Elfcrusher is arguing for degrees of success."

So the above conversation could have had this inserted:
"Sure. Tell me what you remember, and when you act on it you can roll Int with a DC of X. If you fail it will produce the opposite intended effect."

 

[Guest 6801328]

@Ovinomancer and I are both about to have our examples dissected by lichen specialists.

 

[Guest 6801328]

Great post, but really the highlight is that first paragraph:

Okay, so, to me, the crux of a check is that the fiction changes after it's made. On a success, the fiction changes towards the goal. On a failure, the fiction changes away. The away part is the consequence for failure, and take many forms, but must always, always, always be a change in the fiction.

Yes. Yes. Yes.

 

[FrogReaver]

@Elfcrusher

It's like you totally ignored my actual point I'm sure it was by accident though.

tldr: the meaningful consequence for failure doesn't have to have anything to do with the action the character is performing.

So that failed History Check can give you legitimate information, but you might just be beamed in the buttocks by a falling meteorite (probably not the best example, but it does illustrate my point).

Simply stop trying to tie the check with the consequence and instead force a consequence to happen simply because of the failed check. Now your playstyle works.

 

[Ovinomancer]

The tricky thing here is (assuming you have open rolls) that the player know he/she failed the roll, so giving them false information isn't going to lead to a bad result, and therefore isn't really a penalty for failure. (Caveat: if your table enforces the sort of roleplaying where it is expected that the player will act on the knowledge anyway, because to do otherwise would be "metagaming", then this might work for you.)

I suppose...hearking back to @Charlaquin's comment about stealth...you make them roll at the critical moment. That is, when the character actually attempts to shut down the portal, and commits to it, that's when the roll is made.

And maybe that's the general case answer to these scenarios? (Not that you handle ALL knowledge tasks this way, but when the stars align and inspiration strikes.)

Let's try it out with monster abilities/features. As long as the DM is willing to be flexible, I could see something like this:
Player: "Thinking back to my course in Elemental Combat Fundamentals in Wizard School, I try to recall if these things have any special vulnerabilities."
DM: "Well, what do you remember?"
Player: "That they are especially vulnerable to sonic damage!"
DM: "Cool, try it out."
(Player casts Shatter)
DM: "Let's see an Intelligence check."
Player (using @iserith's method): "I'm going to apply Arcana, since this was something I learned in Wizard school....oh, sheet, I rolled a 2. So 7 total."
DM: "Well, lucky for you it wasn't a total fail* and you managed to not heal the thing, but you can tell that not only did it not do as much damage as you were expecting, but it barely seemed to affect it."

Two things about this scenario:
1) Whatever the stat block said, the DM was willing to switch it up on the fly. (Although from now on these creatures probably have resistance to sonic damage.)
2) The penalty for failure came in the form of a wasted spell slot.

And I'll point out that the above scenario, with higher stakes and more dramatic effect, could very well turn into the sort of story that gets told around that table for years to come. "Remember the time...?" Which is in some ways my ultimate test.

*By the way, I don't intend to keep promoting gradations of success. I tossed that one idea out there a while ago, and that seemed to have gotten interpreted as "Elfcrusher is arguing for degrees of success."

So the above conversation could have had this inserted:
"Sure. Tell me what you remember, and when you act on it you can roll Int with a DC of X. If you fail it will produce the opposite intended effect."

If we were in a slightly different game, I'd be cheering this approach. It's putting some authorial control into the hands of the players, and that can be a powerful thing. However, this works best in systems that use a mechanical resolution mechanic designed with this kind of play in mind, and 5e isn't that kind of game. Its resolution system is too game-able to unduly weight such outcomes, and it lacks a useful reinforcement mechanism for formal stake setting. Can you do it anyway? Yup, and maybe even well, but you'll be constantly tending to the system to make it work, and likely end up with a set of table rules to really make it work.

5e is just too prep heavy. The encounter balance maths are too tight and it's too easy to tip the game over into a tailspin by playing with that balance on the fly in the middle of encounters. Even if you just eyeball stuff, the encounter maths still lurk behind it all and you might accidentally tip a challenge over into the unwinnable. It's a lot to balance and consider while in the heat of play. Other systems that are built with this kind of play in mind use different balancing systems so that the changes are easily manageable in play.


That said, I applaud the thinking here -- it's subversive and trying to do new things with the game.

 

[Ovinomancer]

@Elfcrusher

It's like you totally ignored my actual point I'm sure it was by accident though.

tldr: the meaningful consequence for failure doesn't have to have anything to do with the action the character is performing.

So that failed History Check can give you legitimate information, but you might just be beamed in the buttocks by a falling meteorite (probably not the best example, but it does illustrate my point).

Simply stop trying to tie the check with the consequence and instead force a consequence to happen simply because of the failed check. Now your playstyle works.

This suggestion actually fights against one of the strengths of goal and approach, though, which is that the players feel more connected and in control of their characters because you start and end with the fiction. But having the players state actions instead of ask for resolutions, there's a tight coupling of the action with the fiction. Similarly, if you flow through the process and consider if the approach and goal have a consequence for failure and is uncertain, only then ask for a check, otherwise narrate success or failure as appropriate, that fiction flows through the entire process and the end is an obvious result of the action.

Your suggestion would eliminate one of the legs for determining if a check is needed (ie, is there a consequence for failure) but providing a fixed yes, decided arbitrarily by the DM independent of the declared action or current state of the fiction. As such, it would act to cause more rolls than otherwise needed and disconnect the player's understanding of the fiction from the results of their declared actions.

I hesitate to say this is a terrible idea, but only briefly.

 

[FrogReaver]

This suggestion actually fights against one of the strengths of goal and approach, though, which is that the players feel more connected and in control of their characters because you start and end with the fiction. But having the players state actions instead of ask for resolutions, there's a tight coupling of the action with the fiction. Similarly, if you flow through the process and consider if the approach and goal have a consequence for failure and is uncertain, only then ask for a check, otherwise narrate success or failure as appropriate, that fiction flows through the entire process and the end is an obvious result of the action.

Your suggestion would eliminate one of the legs for determining if a check is needed (ie, is there a consequence for failure) but providing a fixed yes, decided arbitrarily by the DM independent of the declared action or current state of the fiction. As such, it would act to cause more rolls than otherwise needed and disconnect the player's understanding of the fiction from the results of their declared actions.

I hesitate to say this is a terrible idea, but only briefly.

And yet, this is the only way to assign a meaningful consequence of failure to a History check that otherwise has no obvious consequence of failure.

Don't get me wrong, I'm not fond of that solution either. I hated it when iserith first proposed an example with that methodology in it probably a year or more ago. However, it does solve the whole meaningful consequence of failure problem for knowledge recall right?

 

[Monayuris]

Maybe. The point of this thread is to explore exactly that question.

I'll admit (and have been admitting) that I wrestle with the knowledge check. I don't think it's a terrible crime to just ask for a check, and then say "Yeah, you know..." or not. But I also don't think that's very interesting or particularly fun. As I expressed in another thread, you're unlikely to ever recall with your friends, "Remember that time I made that History check?"

So if it's possible to make "History checks" more exciting, in a way that can be applied in various situations, I'd like to figure it out.

Good question. I'm not super sure how to make history or knowledge checks exciting. My approach is that if a player asks "What do I know about X?" I first recognize it as a valid question. The player in the real world may not know the same things as their character who lives in the campaign world.

My current approach to this is to figure out the 'rarity' of the X. If it is something common enough, I will just tell them what they want to know. If I don't know what they really want to know, I'll ask the player... "What is the purpose for needing this information?". If its esoteric, I'd ask for a roll. I guess this is where it is difficult to make it exciting... right now, my idea is a success means the player recalls the information and can act on it, if a failure then perhaps provide a situation where the character does not know, but I'll give info about how they can get the info they need (which may lead to more adventure).

I know this kind of steps on the Sage background ability, but I think that ability should be the default for failed checks.

Insight is somewhat easier. Want to figure out if somebody is lying? As I've stated elsewhere (in many threads) I simply don't have "lie detection" in strangers in my game. Not without magic. But please let's not digress into whether that's a good idea or bad idea, and just start with that baseline. If you want to "search out a lie" in my game I want you to actually search out a lie, and try to push/trick/provoke/lure the subject into making a mistake or otherwise showing their hand. And doing so carries a risk of them figuring out what you're up to. That might involve History, Sleight-of-Hand, Stealth, Animal Handling ("It's clear the horse has been ridden recently, even though he says he's been here the whole time.") or prac

I don't really run high intrigue style games, but I kind of put insight into the same bucket as combat. I can't expect myself as a DM to act in such a way as to provide proper 'tells' that the NPC is lying (without being obvious about it) and I can't expect my players to be so skilled that they can read my proper role-playing to accurately ascertain whether the NPC I'm playing is lying or not. So, its something that is best abstracted.

The real question is why are you lying to the PCs and what does it gain you to deceive the players through lies? As a DM, I think you are the players full senses in the game and I think they should be able to trust you. If you have an NPC that is lying to the players, its probably just better to tell the players that the NPC is likely lying. I think it is more interesting to just say they are lying and let the players figure out what to do about it, rather than flat out give them false info and let them act on that and then pull the rug out from under them, later.

If it has to be an ability check, then you have a very interesting situation... do you do the check in secret (so they don't know whether they succeed or fail) and then tell the players lies or truth based on the result? Do you allow the check in the open with a known DC and ask that your players role play the outcome?

Personally, I'd lean to the second choice. I'd rather be transparent to the players about the situation and let them choose to role-play their characters' ignorance.

I guess I'm pretty hard-core about player information. I don't like to lie to the players, even if it creates a 'realistic effect'. I guess this is why I'm not good at or not so interested in heavy intrigue games. Just not my thing.

 

[Ovinomancer]

And yet, this is the only way to assign a meaningful consequence of failure to a History check that otherwise has no obvious consequence of failure.

Don't get me wrong, I'm not fond of that solution either. I hated it when iserith first proposed an example with that methodology in it probably a year or more ago. However, it does solve the whole meaningful consequence of failure problem right?

It isn't. What's required is to change your conception of what a History check can do. Skip using it as a "Can I Please Know Things, DM?" skill and use it to actually do things in the environment. This requires the DM to do a little bit of extra thinking in prep to write down a note to say "remember to mention the historical significance of this trap/puzzle/secret door."

I've had history be used to find secret doors, disarm traps (a curse written on a sarcophagus), and influence nobles at court. I've also had it used by a player to analyze an ongoing battle and make a recommendation to the army's general on tactics.

"Do I Remember Stuff, Please?" is boring. Just give information. Stop building encounters and scenes where hiding information is part of the scene. You can have very exciting scenes by not gating information behind checks -- even mysteries and intrigue (in fact, I think these get easier when you stop confusing mystery and intrigue with gated information).