Ability Checks: The Problem, and How To Fix It

[strangebloke]

Ability checks are something that are troublesome, particularly with non-opposed checks outside of combat. As I see it, there are three major issues:

The Problem(s)


Check Spamming: If repeated checks and/or various skill boosting abilities are in play, conventional bonuses to skill like proficiency/expertise/ability are lost in the noise. This creates unfortunate situations where the character who's nominally the 'expert' does not get to play to type and the investment they made at character creation.
All or Nothing: The reverse problem, where you only get a single chance to make a super high impact check. Succeed, and the King attaints High Vizier Stabak and casts him out, fail and the Vizier gets you cast out as traitors. You either can trivially succeed because you've got reliable talent + help + bardic inspiration + expertise + guidance, or you've got one guy with a proficiency and ability, who rolls and prays they don't get a one.
Conveyance: When I a player attempt something at a table, I've often got no idea how hard the DM will make the skill checks. A lot of DMs I play with seem to set almost everything at DC 15, since with aforementioned preconditions the DC doesn't really matter. It's rare to come up with a situation where a player will even attempt something 'almost impossible' and anything that would be below DC 10 DMs just give to you for free. Buy you have no way of knowing how your DM feels about the difficulty of climbing for example.


My Solution(s)


Take 20: If repeated/group checks are ever possible, I just give it to them. No rolls required. I think this is pretty well know as a 'best practice' in DND but it bears repeating. If the whole party has an hour to look for the secret door, raw math dictates they'll find it eventually.
Simply Tell Players the Difficulty: Before the player attempts something simply say "For a normal person, what you're about to do would be Nearly Impossible. For you its possible, but still risky." There is AFAIK no argument against doing this, as it only serves to make the player feel more confident in attempting to do things that they otherwise wouldn't, and also softens the blow when they try and fail. After all, they knew there was a chance of failure.
Pick your Action: In a complex situation, I will usually provide a laundry list of things the players could potentially do, which encourages them to split up and attempt different things, since they can only attempt one or two things during the time allotted. In such a situation, players can double or triple up on a single check. They can stack loads of bonuses, but its inadvisable because there's opportunity cost present.
Complexity and Strikes: In this context, the 'complexity' of the task indicates how many successful checks need to be made, while 'Strikes' are how many failures can occur before the thing being attempted fails. This ensures multiple checks that will offset the variance of the d20 and allow for a more dynamic skill challange.
Traits: In certain skill checks, especially social ones, it makes sense to borrow a page from FATE. Give your NPC, trap, or challenge a list of traits. Bonds, flaws, that kind of thing. If the PC attempts to do something that runs afoul of one of these traits (ex: telling the king to banish his beloved son) the check might auto-fail, or at least have disadvantage. Conversely if the PC plays to a trait successfully, they might get advantage or even an auto-success. In some cases playing to one trait successfully will allow you to make a check even if you're playing against another trait. You can maybe get the king to banish his son if you appeal to his sense of honor, for example. Some traits will be obvious from roleplay (and you can just tell the players about them) while others might require a successful insight check.




The Party is going to a party. One party member chooses to shadow High Vizier Stabak to see what he's up to, while another tries to make friends/carouse with the nobility, while another tries to get the king to to realize the threat of the coming invasion from Clefsmarch. The final party member chooses to help with the King.

The carouser gets handled first since its the simplest. The complexity of the task is low, say two, and there's traits. So you have them roll persuasion until they get two successes or failures, and then narrate the outcome. If they failed, maybe they got drunk too fast and caused a disruption, maybe they flirted too hard with someone's spouse and caused a big stir. If they succeeded, maybe you RP them meeting some new ally or patron.

The guy shadowing the High Vizier requires more information. They need to stay unnoticed, overhear conversation, and then understand the significance of it. The Complexity here is higher, say 4, and the Strikes allowed are 2. The exact flow of checks here will vary based on approach. If they want to approach the vizier under false pretenses that will be deception, insight, and persuasion. If they want to linger outside their notice and overhear the conversation, that will be stealth, perception and possibly history to understand what's being said. Failure can change things too. If you fail the stealth check you might have disadvantage to listening in, or something. The key here though is that the core structure of four successes and two allowed failures, and you can create skill challenges based on that. Failure here might mean the Vizier leaving, or dismissing the shady man he's meeting with.

Finally we deal with the King. The King has traits, and they know some of them. Irrespective of how confident the RP is, if they can articulate arguments that play into the King's biases, they can get some easy free successes, but the king isn't going to be persuaded by one argument alone. They might want to make insight checks to guess at other priorities of his, or even conduct investigations beforehand by talking to other people so that they know which topics to avoid and which to bring up.

TL;DR
Ultimately this is just a framework for making complex and engaging challenges. I mostly fixated on the social examples here but everything I said applies equally well to exploration challenges - I may put up another example if there's interest.

[Unoriginal]

Ability checks are something that are troublesome, particularly with non-opposed checks outside of combat. As I see it, there are three major issues:

The Problem(s)


Check Spamming: If repeated checks and/or various skill boosting abilities are in play, conventional bonuses to skill like proficiency/expertise/ability are lost in the noise. This creates unfortunate situations where the character who's nominally the 'expert' does not get to play to type and the investment they made at character creation.
All or Nothing: The reverse problem, where you only get a single chance to make a super high impact check. Succeed, and the King attaints High Vizier Stabak and casts him out, fail and the Vizier gets you cast out as traitors. You either can trivially succeed because you've got reliable talent + help + bardic inspiration + expertise + guidance, or you've got one guy with a proficiency and ability, who rolls and prays they don't get a one.
Conveyance: When I a player attempt something at a table, I've often got no idea how hard the DM will make the skill checks. A lot of DMs I play with seem to set almost everything at DC 15, since with aforementioned preconditions the DC doesn't really matter. It's rare to come up with a situation where a player will even attempt something 'almost impossible' and anything that would be below DC 10 DMs just give to you for free. Buy you have no way of knowing how your DM feels about the difficulty of climbing for example.


My Solution(s)


Take 20: If repeated/group checks are ever possible, I just give it to them. No rolls required. I think this is pretty well know as a 'best practice' in DND but it bears repeating. If the whole party has an hour to look for the secret door, raw math dictates they'll find it eventually.
Simply Tell Players the Difficulty: Before the player attempts something simply say "For a normal person, what you're about to do would be Nearly Impossible. For you its possible, but still risky." There is AFAIK no argument against doing this, as it only serves to make the player feel more confident in attempting to do things that they otherwise wouldn't, and also softens the blow when they try and fail. After all, they knew there was a chance of failure.
Pick your Action: In a complex situation, I will usually provide a laundry list of things the players could potentially do, which encourages them to split up and attempt different things, since they can only attempt one or two things during the time allotted. In such a situation, players can double or triple up on a single check. They can stack loads of bonuses, but its inadvisable because there's opportunity cost present.
Complexity and Strikes: In this context, the 'complexity' of the task indicates how many successful checks need to be made, while 'Strikes' are how many failures can occur before the thing being attempted fails. This ensures multiple checks that will offset the variance of the d20 and allow for a more dynamic skill challange.
Traits: In certain skill checks, especially social ones, it makes sense to borrow a page from FATE. Give your NPC, trap, or challenge a list of traits. Bonds, flaws, that kind of thing. If the PC attempts to do something that runs afoul of one of these traits (ex: telling the king to banish his beloved son) the check might auto-fail, or at least have disadvantage. Conversely if the PC plays to a trait successfully, they might get advantage or even an auto-success. In some cases playing to one trait successfully will allow you to make a check even if you're playing against another trait. You can maybe get the king to banish his son if you appeal to his sense of honor, for example. Some traits will be obvious from roleplay (and you can just tell the players about them) while others might require a successful insight check.




The Party is going to a party. One party member chooses to shadow High Vizier Stabak to see what he's up to, while another tries to make friends/carouse with the nobility, while another tries to get the king to to realize the threat of the coming invasion from Clefsmarch. The final party member chooses to help with the King.

The carouser gets handled first since its the simplest. The complexity of the task is low, say two, and there's traits. So you have them roll persuasion until they get two successes or failures, and then narrate the outcome. If they failed, maybe they got drunk too fast and caused a disruption, maybe they flirted too hard with someone's spouse and caused a big stir. If they succeeded, maybe you RP them meeting some new ally or patron.

The guy shadowing the High Vizier requires more information. They need to stay unnoticed, overhear conversation, and then understand the significance of it. The Complexity here is higher, say 4, and the Strikes allowed are 2. The exact flow of checks here will vary based on approach. If they want to approach the vizier under false pretenses that will be deception, insight, and persuasion. If they want to linger outside their notice and overhear the conversation, that will be stealth, perception and possibly history to understand what's being said. Failure can change things too. If you fail the stealth check you might have disadvantage to listening in, or something. The key here though is that the core structure of four successes and two allowed failures, and you can create skill challenges based on that. Failure here might mean the Vizier leaving, or dismissing the shady man he's meeting with.

Finally we deal with the King. The King has traits, and they know some of them. Irrespective of how confident the RP is, if they can articulate arguments that play into the King's biases, they can get some easy free successes, but the king isn't going to be persuaded by one argument alone. They might want to make insight checks to guess at other priorities of his, or even conduct investigations beforehand by talking to other people so that they know which topics to avoid and which to bring up.

TL;DR
Ultimately this is just a framework for making complex and engaging challenges. I mostly fixated on the social examples here but everything I said applies equally well to exploration challenges - I may put up another example if there's interest.

Those are DMing issues, though, not ability check issues.

As in, the game already addresses those issue so they don't happen. If the DM decides to make them happen, it's on the DM, not on ability checks.

[x3n0n]

Those are DMing issues, though, not ability check issues.

As in, the game already addresses those issue so they don't happen. If the DM decides to make them happen, it's on the DM, not on ability checks.

Ok, then "Game encounters resolved by a single ability check: key problems and a suggested encounter design remedy", perhaps?

[Amnestic]

Check Spamming: If repeated checks and/or various skill boosting abilities are in play, conventional bonuses to skill like proficiency/expertise/ability are lost in the noise. This creates unfortunate situations where the character who's nominally the 'expert' does not get to play to type and the investment they made at character creation.

Isn't this covered by Passive Checks? PHB 175, emphasis mine:

Passive Checks
A passive check is a special kind of ability check that doesn't involve any die rolls. Such a check can represent the average result for a task done repeatedly. such as searching for secret doors over and over again, or can be used when the DM wants to secretly determine whether the characters succeed at something without rolling dice, such as noticing a hidden monster.

Either that or group checks (PHB 176). You suggest 'Take 20' below, but really the standard rules would suggest that 'Take 10' is the way it's meant to be played.

[Psyren]

Isn't this covered by Passive Checks? PHB 175, emphasis mine:


Either that or group checks (PHB 176). You suggest 'Take 20' below, but really the standard rules would suggest that 'Take 10' is the way it's meant to be played.



My Solution(s)


Take 20: If repeated/group checks are ever possible, I just give it to them. No rolls required. I think this is pretty well know as a 'best practice' in DND but it bears repeating. If the whole party has an hour to look for the secret door, raw math dictates they'll find it eventually.

There are "take 20" rules too, separate from the passive check rules, just not quite called that. These are found on DMG 237:

"Sometimes a character fails an ability check and wants to try again. In some cases, a character is free to do so; the only real cost is the time it takes. With enough attempts and enough time, a character should eventually succeed at the task. To speed things up, assume that a character spending ten times the normal amount of time needed to complete a task automatically succeeds at that task. However, no amount of repeating the check allows a character to turn an impossible task into a successful one."


Simply Tell Players the Difficulty: Before the player attempts something simply say "For a normal person, what you're about to do would be Nearly Impossible. For you its possible, but still risky." There is AFAIK no argument against doing this, as it only serves to make the player feel more confident in attempting to do things that they otherwise wouldn't, and also softens the blow when they try and fail. After all, they knew there was a chance of failure.

You shouldn't call for a roll at all unless there is a chance of failure and a chance of success. Moreover, that failure needs to have a meaningful consequence.

In other words, if you call for a roll at all, the players should know success is possible, without you having to tell them the DC you set. Moreover, even if they fail to hit the DC you set, your job as the DM is to keep the action moving, so consider utilizing tools like Success At A Cost or Degrees of Failure (DMG 242 for both) rather than simply throwing up a wall.



Pick your Action: In a complex situation, I will usually provide a laundry list of things the players could potentially do, which encourages them to split up and attempt different things, since they can only attempt one or two things during the time allotted. In such a situation, players can double or triple up on a single check. They can stack loads of bonuses, but its inadvisable because there's opportunity cost present.

I agree with having multiple approaches to a given problem, but not sure what you mean by "stack loads of bonuses" - 5e was designed specifically to get away from that.



Complexity and Strikes: In this context, the 'complexity' of the task indicates how many successful checks need to be made, while 'Strikes' are how many failures can occur before the thing being attempted fails. This ensures multiple checks that will offset the variance of the d20 and allow for a more dynamic skill challange.

I agree with this one, but again, you need to craft the challenge carefully such that the individual and collective failures/successes have meaning (and preferably degrees.)



Traits: In certain skill checks, especially social ones, it makes sense to borrow a page from FATE. Give your NPC, trap, or challenge a list of traits. Bonds, flaws, that kind of thing. If the PC attempts to do something that runs afoul of one of these traits (ex: telling the king to banish his beloved son) the check might auto-fail, or at least have disadvantage. Conversely if the PC plays to a trait successfully, they might get advantage or even an auto-success. In some cases playing to one trait successfully will allow you to make a check even if you're playing against another trait. You can maybe get the king to banish his son if you appeal to his sense of honor, for example. Some traits will be obvious from roleplay (and you can just tell the players about them) while others might require a successful insight check.


I agree with this too, particularly discovering unknown traits through detective skills (e.g. investigation, insight, and perception.) Just because the players may not know a given NPC or obstacles flaws doesn't mean the characters won't.

[Frogreaver]

IMO. If it’s something your character could do and there’s no time pressure then he just does it after the appropriate time.

That stops repeated checks till success. If there is time pressure then repeated checks might be warranted, but only until time runs out.

[Unoriginal]

IMO. If it’s something your character could do and there’s no time pressure then he just does it after the appropriate time.

That stops repeated checks till success. If there is time pressure then repeated checks might be warranted, but only until time runs out.

This is explicitly how the PHB rules handle it, indeed.

[strangebloke]

Those are DMing issues, though, not ability check issues.

As in, the game already addresses those issue so they don't happen. If the DM decides to make them happen, it's on the DM, not on ability checks.
This applies to the first part of my "solution." Taking 20 is default, people just ignore this. However I brought it up anyway because it does bear repeating.

The rest though are situations that arise naturally when following guidance for most checks. Having everyone make a check to try to find the best campsite is normal and RAW and has been a consistent source of complaints for this game's entire lifecycle. Swingy all-or-nothing social checks are present in many modules as well.

Its no good to say "these are only DMing issues" when all the guidance given by the system contributes to these DMing issues occuring.

There are "take 20" rules too, separate from the passive check rules, just not quite called that. These are found on DMG 237:
Admittedly, I should have noted that this was literally RAW. Passive checks are similarly underutilized by most DMs, but represent something very different.


You shouldn't call for a roll at all unless there is a chance of failure and a chance of success. Moreover, that failure needs to have a meaningful consequence.


In other words, if you call for a roll at all, the players should know success is possible, without you having to tell them the DC you set. Moreover, even if they fail to hit the DC you set, your job as the DM is to keep the action moving, so consider utilizing tools like Success At A Cost or Degrees of Failure (DMG 242 for both) rather than simply throwing up a wall.
Sure, but "you can succeed on a natural 20" is very different from "you can fail on a nat 1." Players should know what they're dealing with. If I say to a DM "I would like to disable the trap" and I roll a 19 and fail, I'm going to be rather surprised, perhaps unpleasantly if the trap goes off in my face.

Telling the player the (approximate) DC mitigates this potentially unpleasant surprise, and also makes sense. The thief who knows how to disable traps should be able to know from experience if something looks really really hard or trivial.



I agree with having multiple approaches to a given problem, but not sure what you mean by "stack loads of bonuses" - 5e was designed specifically to get away from that.
In theory, yes. In practice, no.

BI + help + guidance = +11 effectively and that's all online from level 1, where a super-expert (good ability score, expertise) will only have a +7 total. Generic skill buffs and bonuses are more impactful than actual specialization. A skill booster with these basic tools will be able to make their party members better at every skill than a rogue will be good at any skill in particular.


I agree with this one, but again, you need to craft the challenge carefully such that the individual and collective failures/successes have meaning (and preferably degrees.)
The Complexity factors are mostly there to frame the skill check and allow the DM to have a gameplan for running things efficiently in game. Personally I find that if I purely adlib complex checks like this, I end up giving the players way too many chances to do ridiculous, terrible things to the people they're supposedly trying to cajole, which slows things down and also lessens the feeling of actual challenge.



I agree with this too, particularly discovering unknown traits through detective skills (e.g. investigation, insight, and perception.) Just because the players may not know a given NPC or obstacles flaws doesn't mean the characters won't.
Well, in general I don't care about character/player separation. If a player guesses a trait from dialogue that's fine. They might be wrong, of course, which is where insight and investigation becomes key if you want to be consistent.

[Sorinth]

This applies to the first part of my "solution." Taking 20 is default, people just ignore this. However I brought it up anyway because it does bear repeating.

The rest though are situations that arise naturally when following guidance for most checks. Having everyone make a check to try to find the best campsite is normal and RAW and has been a consistent source of complaints for this game's entire lifecycle. Swingy all-or-nothing social checks are present in many modules as well.

Its no good to say "these are only DMing issues" when all the guidance given by the system contributes to these DMing issues occuring.

Why would everyone rolling to find the best campsite be a problem? If everybody is rolling it's a group check and almost certainly worse mathematically for the party. So it does seem like a DM issue of not following the advice already in the book (It no doubt could have been written in a better manner).

[KorvinStarmast]

For the OP: another case of overthinking something into believing it to be a problem.

[Segev]

Personally, I like to solve the specific problems listed by making that final check not even accessible until the players do things to set up for it. Multiple checks to manipulate the situation to their advantage (or find that the situation is spiraling out of their control, and what they're now trying to do is just salvage things) so that there is a progression through the encounter/events, just as there is in combat.

[Ionathus]

As others have said, this sounds like mostly a DMing issue. 5e doesn't do a great job of explaining the philosophy of ability checks, so a lot of DMs are forced to muddle through and figure out their own system for judging when to call and how to set a check.

I have adopted two rules for ability checks, and they have solved basically all of my problems. I think the 1st applies to your second and third problems, and my 2nd rule applies to check spamming.

1. Only call for a check if there's a meaningful chance of success and failure. If both aren't possible, it's not worth rolling. Remember, you're the DM, you're in charge of when the PCs make a check. It doesn't matter how much they say "I roll Perception" until you decide it does.

If the character should be automatically able to accomplish their task, they shouldn't be at the mercy of a nat 1. If it's way out of their league, bonuses shouldn't be able to help them. If they're a skilled rogue trying to pick a locked door, but there's no alarm/time constraint/risk of failure, don't waste everyone's time with how long it takes to get a success. Ability checks are narrative forks in the road, each one leading to its own journey; they're not toll gates that you're stuck at until somebody pays up. If there's no fork, if the two (or three, or more!) outcomes of this roll won't change the next 10 minutes or 100 hours of play, then just let the PC show off their cool skills and auto-succeed.

2. Checks take time. If the party is in a crowded market square talking to a merchant who they worry might be a demon in disguise, they don't all get to pause the dialogue to dogpile Insight checks just because the cleric whiffed. Some of the party members might be trying to sweet-talk the merchant, or search for other clues, or blend into the crowd, or keep an eye out for ambushers. That's a Persuasion, Investigation, Stealth, and Perception check, respectively, and if someone says they want to roll Insight after the cleric whiffs, I say "you're busy with [previously stated/rolled action]; do you want to stop what you're doing and make an Insight check instead?" Sometimes they do. Most of the time, they don't, because A) they have other stuff they want to do and B) they've learned to accept that failed roll as carrying weight because it took a "turn" to roll it.

That feeds into rule 1: the failures have to matter. Failing an insight check means that you lose out on useful info, so if somebody wants to undo that failure they're gonna have to give something of equal value. This is the law of equivalent exchange It doesn't have to map out perfectly to "one check per turn" (some actions are easier and quicker to perform than others, and some checks are ongoing and require your constant attention until you decide to stop) but I find it's good to preserve the pacing of everybody's choices even outside of initiative.

As a result, my players have stopped trying to dogpile entirely, and in fact they even seem happier with their failures. Hopefully part of that is them learning to trust I won't screw them over...but part of it is the fact that every check actually carries weight, so they're not trying to just toss them around left and right.

[Psyren]

Sure, but "you can succeed on a natural 20" is very different from "you can fail on a nat 1." Players should know what they're dealing with. If I say to a DM "I would like to disable the trap" and I roll a 19 and fail, I'm going to be rather surprised, perhaps unpleasantly if the trap goes off in my face.

Telling the player the (approximate) DC mitigates this potentially unpleasant surprise, and also makes sense. The thief who knows how to disable traps should be able to know from experience if something looks really really hard or trivial.

I'm okay with signalling difficulty but this also goes back to the other point I made, about degrees of failure or success at a cost. I agree that for a thief to roll a 19 and having the trap go off in their face would not be a good feeling, but if I were the DM, I would be far more creative than to have the trap result be a binary of "20 = disarm, 1-19 = go off in face." Even if I wanted a 20 to be the only way for a consequence-free disarm, having a 19 on the die be no better than a 1, especially for a thief character, is a failure of imagination in my view.


In theory, yes. In practice, no.

BI + help + guidance = +11 effectively and that's all online from level 1, where a super-expert (good ability score, expertise) will only have a +7 total. Generic skill buffs and bonuses are more impactful than actual specialization. A skill booster with these basic tools will be able to make their party members better at every skill than a rogue will be good at any skill in particular.

BI?

[Ionathus]

BI?

Bardic Inspiration, I think.

(which, addressing the quoted poster for a second, is a limited resource. You can't apply that to every check every time.)

EDIT: (And I have other qualms about the handwringing over "BI + help + guidance" theorycrafting -- and they're shaped suspiciously like a spherical paladin in a vacuum. Of course you'll be more successful if 2-3 separate party members spend several turns and actions buffing & helping you. That's not a flaw, that's just teamwork.)

[JNAProductions]

BI?

Bardic Inspiration.

[strangebloke]

Why would everyone rolling to find the best campsite be a problem? If everybody is rolling it's a group check and almost certainly worse mathematically for the party. So it does seem like a DM issue of not following the advice already in the book (It no doubt could have been written in a better manner).
I don't follow. Group checks are a specific thing a DM can call for, and generally they mathematically disadvantage the party (but also make things more consistent with for example stealth)

But something doesn't become a group check just because multiple people are attempting something. If everyone wants to pull on a rope together, that's a group check. If everyone wants to try pulling on a rope separately in turn, that's a series of normal checks. If a party member is allowed to look for a campsite individually, it follows every party member can look individually, at which point their average roll (pre modifier) becomes 17 and the check (and the character that specialized in that kind of check) is made pointless.

You could make "setting up the campsite" a group check, but group checks don't really benefit any one character building to cover a specific skill either.

My preferred path (which I call 'pick your action') is to divy out the various tasks associated with the end of a day of exploration so that everyone can fixate on the task they're suited for. The ranger can scout a camping spot (ranger), the wizard can take a minute looking for potion reagents (nature), the fighter can get the animals comfortable for the night (handle animal), and the rogue cook a hearty meal.

This is of course just an example. I don't know that I would normally require players to do anything to set up camp.

For the OP: another case of overthinking something into believing it to be a problem.
These problems have been brought up by many people all over the internet since the launch of the game. Some of it is user error, and you perhaps don't have an issue in your personal games. But that isn't the same thing as saying none of these issues exist.

Personally, I like to solve the specific problems listed by making that final check not even accessible until the players do things to set up for it. Multiple checks to manipulate the situation to their advantage (or find that the situation is spiraling out of their control, and what they're now trying to do is just salvage things) so that there is a progression through the encounter/events, just as there is in combat.
This is the sort of thing I'm trying to systematize. Sure, you don't need to track all the above nonsense, but having a framework for how things are going to go makes ad-libbing the rest far easier. You don't have to decide mid-conversation what the king thinks about the faith, because you already have "bond:despises high priest Loran" right there in your prep because of how you've systematized things for your own use.

Basically guidelines for doing prep to make engaging skill challenges that give everyone something to do.

As others have said, this sounds like mostly a DMing issue. 5e doesn't do a great job of explaining the philosophy of ability checks, so a lot of DMs are forced to muddle through and figure out their own system for judging when to call and how to set a check.
I would argue that "the philosophy of ability checks" is mostly something people came up with via consensus. It's not really explained anywhere in the DMG or PHB, and its not how (most) checks in modules are written.

so sure its a "DMing issue" but that's where a quick guide to running checks (as opposed to a load of random bits here and there over three books) is useful, yeah?


I'm okay with signalling difficulty but this also goes back to the other point I made, about degrees of failure or success at a cost. I agree that for a thief to roll a 19 and having the trap go off in their face would not be a good feeling, but if I were the DM, I would be far more creative than to have the trap result be a binary of "20 = disarm, 1-19 = go off in face." Even if I wanted a 20 to be the only way for a consequence-free disarm, having a 19 on the die be no better than a 1, especially for a thief character, is a failure of imagination in my view.
Regardless, conveyance is important. Your player should know its a very complicated trap.

Bardic Inspiration, I think.

(which, addressing the quoted poster for a second, is a limited resource. You can't apply that to every check every time.)

EDIT: (And I have other qualms about the handwringing over "BI + help + guidance" theorycrafting -- and they're shaped suspiciously like a spherical paladin in a vacuum. Of course you'll be more successful if 2-3 separate party members spend several turns and actions buffing & helping you. That's not a flaw, that's just teamwork.)
The point is that a skill buffer can (by proxy) be better at any skill than a skill specialist. And its not 1-2 turns and actions, its a bonus action and action. It's very low cost. IMX the only reason people don't do this is because they forget. I've seen guidance get cast upwards of fifty times in a single session.

[KorvinStarmast]

As a result, my players have stopped trying to dogpile entirely, and in fact they even seem happier with their failures. Hopefully part of that is them learning to trust I won't screw them over...but part of it is the fact that every check actually carries weight, so they're not trying to just toss them around left and right.
Nice post.

These problems have been brought up by many people all over the internet since the launch of the game. People who carried previous edition baggage with them, specifically 3.x but to some extent 4. That's an own goal.

[MrStabby]

Ability checks are something that are troublesome, particularly with non-opposed checks outside of combat. As I see it, there are three major issues:

The Problem(s)


Check Spamming: If repeated checks and/or various skill boosting abilities are in play, conventional bonuses to skill like proficiency/expertise/ability are lost in the noise. This creates unfortunate situations where the character who's nominally the 'expert' does not get to play to type and the investment they made at character creation.
All or Nothing: The reverse problem, where you only get a single chance to make a super high impact check. Succeed, and the King attaints High Vizier Stabak and casts him out, fail and the Vizier gets you cast out as traitors. You either can trivially succeed because you've got reliable talent + help + bardic inspiration + expertise + guidance, or you've got one guy with a proficiency and ability, who rolls and prays they don't get a one.
Conveyance: When I a player attempt something at a table, I've often got no idea how hard the DM will make the skill checks. A lot of DMs I play with seem to set almost everything at DC 15, since with aforementioned preconditions the DC doesn't really matter. It's rare to come up with a situation where a player will even attempt something 'almost impossible' and anything that would be below DC 10 DMs just give to you for free. Buy you have no way of knowing how your DM feels about the difficulty of climbing for example.


My Solution(s)


Take 20: If repeated/group checks are ever possible, I just give it to them. No rolls required. I think this is pretty well know as a 'best practice' in DND but it bears repeating. If the whole party has an hour to look for the secret door, raw math dictates they'll find it eventually.
Simply Tell Players the Difficulty: Before the player attempts something simply say "For a normal person, what you're about to do would be Nearly Impossible. For you its possible, but still risky." There is AFAIK no argument against doing this, as it only serves to make the player feel more confident in attempting to do things that they otherwise wouldn't, and also softens the blow when they try and fail. After all, they knew there was a chance of failure.
Pick your Action: In a complex situation, I will usually provide a laundry list of things the players could potentially do, which encourages them to split up and attempt different things, since they can only attempt one or two things during the time allotted. In such a situation, players can double or triple up on a single check. They can stack loads of bonuses, but its inadvisable because there's opportunity cost present.
Complexity and Strikes: In this context, the 'complexity' of the task indicates how many successful checks need to be made, while 'Strikes' are how many failures can occur before the thing being attempted fails. This ensures multiple checks that will offset the variance of the d20 and allow for a more dynamic skill challange.
Traits: In certain skill checks, especially social ones, it makes sense to borrow a page from FATE. Give your NPC, trap, or challenge a list of traits. Bonds, flaws, that kind of thing. If the PC attempts to do something that runs afoul of one of these traits (ex: telling the king to banish his beloved son) the check might auto-fail, or at least have disadvantage. Conversely if the PC plays to a trait successfully, they might get advantage or even an auto-success. In some cases playing to one trait successfully will allow you to make a check even if you're playing against another trait. You can maybe get the king to banish his son if you appeal to his sense of honor, for example. Some traits will be obvious from roleplay (and you can just tell the players about them) while others might require a successful insight check.




The Party is going to a party. One party member chooses to shadow High Vizier Stabak to see what he's up to, while another tries to make friends/carouse with the nobility, while another tries to get the king to to realize the threat of the coming invasion from Clefsmarch. The final party member chooses to help with the King.

The carouser gets handled first since its the simplest. The complexity of the task is low, say two, and there's traits. So you have them roll persuasion until they get two successes or failures, and then narrate the outcome. If they failed, maybe they got drunk too fast and caused a disruption, maybe they flirted too hard with someone's spouse and caused a big stir. If they succeeded, maybe you RP them meeting some new ally or patron.

The guy shadowing the High Vizier requires more information. They need to stay unnoticed, overhear conversation, and then understand the significance of it. The Complexity here is higher, say 4, and the Strikes allowed are 2. The exact flow of checks here will vary based on approach. If they want to approach the vizier under false pretenses that will be deception, insight, and persuasion. If they want to linger outside their notice and overhear the conversation, that will be stealth, perception and possibly history to understand what's being said. Failure can change things too. If you fail the stealth check you might have disadvantage to listening in, or something. The key here though is that the core structure of four successes and two allowed failures, and you can create skill challenges based on that. Failure here might mean the Vizier leaving, or dismissing the shady man he's meeting with.

Finally we deal with the King. The King has traits, and they know some of them. Irrespective of how confident the RP is, if they can articulate arguments that play into the King's biases, they can get some easy free successes, but the king isn't going to be persuaded by one argument alone. They might want to make insight checks to guess at other priorities of his, or even conduct investigations beforehand by talking to other people so that they know which topics to avoid and which to bring up.

TL;DR
Ultimately this is just a framework for making complex and engaging challenges. I mostly fixated on the social examples here but everything I said applies equally well to exploration challenges - I may put up another example if there's interest.

I think I would agree with you here that there is a problem. Yes, sometimes you can mitigate it and sometimes you can avoid it, but that can just create other problems.

Whilst better DMing can help it doesn't eliminate these issues. And, to be honest though, some of it isn't just ability checks but also the whole game. One roll to determine some pretty serious consequences is familiar to anyone critically failing a death save or a save against feeblemind.

I think there will be Big Rolls sometimes. Sometimes some things will come down to one single roll of the dice; this is something I have learned to expect. As a side note, this is why I do think that there is a bit of a martial/caster disparity - resource requirements to cast spells to overcome a challenge become increasingly less of a concern when your caster uses spells to overcome the critical challenge and skills get used for the everyday affairs.

Then things like group checks... some of the rules there are squiffy. A party of two loud people is likely to be heard, but add the extra noise from four quiet people on top of that and success is easier? Or you get a character that is built for stealth but they never get to use their abilities because everyone else is loud?

Some things are also a matter of taste. I know some people don't like the role of the d20 in skill checks and think that the ratio of success from those with higher bonuses to lower bonuses should be higher. They want the wizard to ALWAYS know more about magic than the barbarian. Then I also think that the barbarian that didn't dump intelligence in order to know more about the world should also have measurably greater success than one that did.

A lot of classes are left to effectively sit out big sections of the game if delegating tasks to the character with the highest bonus becomes advisable. A narrower gap between the wizard and the barbarian means the wizard will shine somewhat in the library scenes but the barbarian still has a chance to be the one that finds the crucial information and isn't relegated.to being a glorified book-porter.

I think we have a system that will always fail. It is impossible to really satisfy all the conditions we might want from it simultaneously, but I think there are also some things going for it that a lot of DMs underappreciate.

These are contested checks and advantage/disadvantage.

Best/worst of two rolls has a lower variance than a single die roll. If you want a task where skill/stats should dominate then the fraction of the time that the better bonus will roll higher increases if both have advantage or both have disadvantage. This is why perception is such a good skill - even with dark vision you are still usually rolling with disadvantage at night and the bigger difference between characters is bonus, not luck.

On the flip side, if you want luck to play a greater role, do a contested check. Instead of s static DC check to break a code mode it an opposed check between the writer and those looking to break it.

I think 5th edition is actually quite elegant that way in giving some extra tools.

Anyway- I broadly agree skill checks are problematic. I think it's worth discussing it; even if there is no solution it can help DMs get.a sense.of what trade off will work best for their table.

[Sorinth]

I don't follow. Group checks are a specific thing a DM can call for, and generally they mathematically disadvantage the party (but also make things more consistent with for example stealth)

But something doesn't become a group check just because multiple people are attempting something. If everyone wants to pull on a rope together, that's a group check. If everyone wants to try pulling on a rope separately in turn, that's a series of normal checks. If a party member is allowed to look for a campsite individually, it follows every party member can look individually, at which point their average roll (pre modifier) becomes 17 and the check (and the character that specialized in that kind of check) is made pointless.

You could make "setting up the campsite" a group check, but group checks don't really benefit any one character building to cover a specific skill either.

My preferred path (which I call 'pick your action') is to divy out the various tasks associated with the end of a day of exploration so that everyone can fixate on the task they're suited for. The ranger can scout a camping spot (ranger), the wizard can take a minute looking for potion reagents (nature), the fighter can get the animals comfortable for the night (handle animal), and the rogue cook a hearty meal.

This is of course just an example. I don't know that I would normally require players to do anything to set up camp.

"When a number of individuals are trying to accomplish something as a group, the DM might ask for a group ability check". They are a group trying to accomplish "finding/setting up a campsite" so why wouldn't it be a group check. They aren't trying to do something sequentially (If they were there would be a time cost with every failure). Now you can just as easily say it's part of the "Working Together" section that precedes the group check section which would mean 1 character makes 1 roll at advantage. Either way your let everyone make their own individual check is not what the books suggest. So like people have been saying it's a DM creating problems for themselves by ignoring what the books say.

[Psyren]

Regardless, conveyance is important. Your player should know its a very complicated trap.

Again, agreed.


The point is that a skill buffer can (by proxy) be better at any skill than a skill specialist. And its not 1-2 turns and actions, its a bonus action and action. It's very low cost. IMX the only reason people don't do this is because they forget. I've seen guidance get cast upwards of fifty times in a single session.

Guidance could honestly be its own thread, and that's assuming it doesn't get changed in 5.5e anyway and render discussion about it moot.

Personally I think it's quite easy to make expertise beneficial regardless. Several checks can be abstracted to simply say "do you have proficiency + expertise + 1 source of magical aid? You get to skip rolling entirely." And I would argue many of them should be. It becomes less about comparing the numerical bonuses such features provide and more using them to grant martials access to nice things that the casters might need spells to pull off.



People who carried previous edition baggage with them, specifically 3.x but to some extent 4. That's an own goal.

This. The "problem" is coming most loudly from the people who haven't paid enough attention to how 5e has changed the game. That is not a reason to change 5e further so much as it is to emphasize the strides that already exist in the material, particularly everything on DMG 237-242.

[KorvinStarmast]

"When a number of individuals are trying to accomplish something as a group, the DM might ask for a group ability check". They are a group trying to accomplish "finding/setting up a campsite" so why wouldn't it be a group check. You are correct. As has happened so many times in the past on this topic, the assertions made in OPs fail to account for what's already in the book.

[Anymage]

But something doesn't become a group check just because multiple people are attempting something. If everyone wants to pull on a rope together, that's a group check. If everyone wants to try pulling on a rope separately in turn, that's a series of normal checks. If a party member is allowed to look for a campsite individually, it follows every party member can look individually, at which point their average roll (pre modifier) becomes 17 and the check (and the character that specialized in that kind of check) is made pointless.

The rules and adventure design could spell this out more clearly, but being wrong is a lot more interesting than straight up not knowing. If everybody rolls and the wizard and paladin think that a cozy cave is a good camping spot while the ranger thinks that it's probably been claimed by some other critter, either you've rediscovered group checks or they default back to trusting the person with the highest bonus.


The point is that a skill buffer can (by proxy) be better at any skill than a skill specialist. And its not 1-2 turns and actions, its a bonus action and action. It's very low cost. IMX the only reason people don't do this is because they forget. I've seen guidance get cast upwards of fifty times in a single session.

I guess the big question is how many stackable skill buffers 5.5 will allow space for. They're an issue in 5e, but I'll be curious how much of one in practice.


Then things like group checks... some of the rules there are squiffy. A party of two loud people is likely to be heard, but add the extra noise from four quiet people on top of that and success is easier? Or you get a character that is built for stealth but they never get to use their abilities because everyone else is loud?

To be pedantic: If the paladin is shouting challenges than any attempt for stealth just went out the window. The group check is the three sneaky guys trying to babysit the plate clad oaf so he doesn't wind up making more noise than absolutely necessary. It makes sense, and more importantly is a way around the idea that certain checks should either autosucceed (if you just need one person in the party to roll well) or autofail (if you just need one person rolling poorly to botch it).


This. The "problem" is coming most loudly from the people who haven't paid enough attention to how 5e has changed the game. That is not a reason to change 5e further so much as it is to emphasize the strides that already exist in the material, particularly everything on DMG 237-242.

Eh. I do think the misconceptions are popular enough that I wouldn't mind seeing some web article pointing out how to apply these things in practice, and more focus on doing it right in actual adventures. The bits are all there, it'd be nice to see examples of them being applied in practice.

[Psyren]

Eh. I do think the misconceptions are popular enough that I wouldn't mind seeing some web article pointing out how to apply these things in practice, and more focus on doing it right in actual adventures. The bits are all there, it'd be nice to see examples of them being applied in practice.

I assume you mean "from WotC" there, since web articles on this subject from content creators are pretty prevalent.

While I wouldn't mind some examples either, every one they publish ends up potentially being the locus of a table/internet argument too. So it's completely understandable why they would want to keep those to a minimum.

[jjordan]

I enjoy setting the task difficulty and then letting players come up with meaningful ways to whittle that down. Much the same as with advantage, except I make the players say exactly how they're helping. Taking more time is the big one, but there are a lot of other ways to do this. Then I have them roll and that determines how well they did the job and how long it takes. I also like to put things into perspective. RAW you can pick a lock in six seconds in a combat situation, I believe. Ten times that is just 1 minute.

[Sorinth]

You are correct. As has happened so many times in the past on this topic, the assertions made in OPs fail to account for what's already in the book.

I think a lot of the so called problems come down to letting the players call for skill checks. Stop doing that and go back to the basics of players describe what they do and the DM determines what happens.

[Ionathus]

Nice post.

Thank you!


The point is that a skill buffer can (by proxy) be better at any skill than a skill specialist. And its not 1-2 turns and actions, its a bonus action and action. It's very low cost. IMX the only reason people don't do this is because they forget. I've seen guidance get cast upwards of fifty times in a single session.

Your equation contained three things, two of which are actions, and two of which are mutually exclusive abilities (Bards don't get Guidance). You'll need at least 2 people to pull it off no matter how you slice it, and then neither of them get to do anything else with their action.

And if you're about to say "that doesn't matter: they weren't doing anything else with their action anyway" then you either need to create something to occupy that attention, or recognize that this has become a solo scene and buffing a teammate in their solo performance is fun. Seeing the advantage you granted make the difference is one of the best parts of this damned game. So yeah, I don't think there's anything at all wrong with saying "three cheers for teamwork!" and giving the checking PC an auto-succeed and moving on to something that actually challenges the PCs as quickly as possible.


I've seen guidance get cast upwards of fifty times in a single session.

I can't tell if you said that like it's no biggie or it's a horror story, but this sounds like a nightmare slog to me. Can I ask for more context of how it happened 50 times? Just one trigger-happy PC, or a particularly check-heavy encounter design?

Guidance isn't a permanent, unseen buff. It takes an action to cast, it has Verbal and Somatic components, and it requires Concentration. I struggle to think of times that it could be spammed. In social situations, people are going to notice and usually be weirded out by the cleric casting spells on herself mid-conversation. Stealth is out, because of Verbal components. Prolonged checks like searching a room are debatable; sometimes you can justify it, sometimes you can't. Stacking it with something like Enhance Ability is out, because of concentration. Initiative is only possible if you know a fight's about to pop off in the next 54 seconds and can cast it without being noticed and triggering the fight. Guidance is good for picking an occasional lock, or lifting a boulder, but it isn't a Reaction.

I've played with Guidance-spammers and, without exception, it was because the DM either didn't really know why they were allowing an ability check, or they didn't know how to control the pace of the encounter. Usually both, because the problems go hand-in-hand. The problem I always notice is that they try to retroactively say "I'd cast Guidance on myself before I did X" every time they roll for anything, no matter how reactive that check is or whether it's feasible to predict the need for it. The game gets bogged down in making a bunch of checks instead of actually understanding and solving the puzzles in front of you, including their spinoffs like fighting (violence puzzles) or roleplaying (dialogue puzzles). And usually the checks are something totally inconsequential: haggling over price at a tavern, trying to win over a guard, trying to win the pie-baking contest. They could be simple and streamlined and fun, and allowing unrealistic time for buffing in the middle kills that fun for me.

To extend my analogy from before: playing at a table with a Guidance-spammer is like sitting in the passenger seat as the driver of your car scrounges for loose change at the toll booth, insisting that they'll get this right if they can just find 5 more cents. A DM's job with skill checks is to not waste everyone's time counting pennies: DMs should learn to recognize when the scene is dead or dying and wave it through before everyone in the car passes out from boredom.


To be pedantic: If the paladin is shouting challenges than any attempt for stealth just went out the window. The group check is the three sneaky guys trying to babysit the plate clad oaf so he doesn't wind up making more noise than absolutely necessary. It makes sense, and more importantly is a way around the idea that certain checks should either autosucceed (if you just need one person in the party to roll well) or autofail (if you just need one person rolling poorly to botch it).

To be fair, I also think calling for a group check is rarely a good answer to any D&D situation. If you have a plate clad oaf you need to sneak past an open door, the sneaky player isn't going to feel cool for rolling a 29 if the only thing it achieves is balancing out the paladin's crappy check. That's not fun. That's babysitting (like you said).

In this scenario, a group check should be the least fun and most fail-prone option on purpose because it really is a cartoonishly bad idea. A better idea would be causing a distraction, or using the stealthiest player to sneak ahead solo to neutralize the reinforcements/alarm system, or using magic to disguise yourself, or any number of other options. Or - and this is absolutely a valid choice - saying "to hell with it" and just charging into the room. 9 times out of 10, group checks don't make the game more fun - they feel more like "math" than any other check, and the averaging-out of the rolls really sucks a lot of the character out of each individual roll.

I would much rather say "ok, you'll need a stealth check to get across...or you'll need 2 allies to make an additional check in your favor for you to auto-succeed." Performance to create a distraction, Deception to impersonate the guards and draw them off for a moment, Knowledge(Architecture and Engineering) to identify an acoustic dead zone...something that adds character to the story. That way, the stealthy PC is rewarded for being so dang good at stealth that they get to do more cool stuff and help the team out actively.

[Ionathus]

I assume you mean "from WotC" there, since web articles on this subject from content creators are pretty prevalent.

While I wouldn't mind some examples either, every one they publish ends up potentially being the locus of a table/internet argument too. So it's completely understandable why they would want to keep those to a minimum.

WotC has created this really weird dynamic in the community, where they create all of the rules and sources and then rarely (if ever) are the actual authority on tips/advice for playing the game. Which isn't too unusual for board games or video games, but I feel like D&D is so conceptual and narrative-based that it blurs the lines. WotC presumes to offer strategic advice for running the game with books like the DMG (which is about 10% very solid prescriptive advice for new DMs, plus 25% game lore like magic items and planes of existence, plus wayyy too much info about weird stuff like homebrewing spells that's probably been used by 12 people across 5e's entire run).

It's strange that my "recommended reading" for new DMs is Matt Colville's Running The Game, Keith Amman's The Monsters Know What They're Doing, a select few articles from The Angry GM, and finally, like, 23 pages out of a 300+ page official book that was supposedly written to Guide the Dungeon Master.


I think a lot of the so called problems come down to letting the players call for skill checks. Stop doing that and go back to the basics of players describe what they do and the DM determines what happens.

100% agree with this. Letting players ask for skill checks stifles their creativity too, in my experience, because without that crutch they're free to describe the actual thing they want to do that often falls outside the narrow guidelines of 18 skills.

[Psyren]

I think a lot of the so called problems come down to letting the players call for skill checks. Stop doing that and go back to the basics of players describe what they do and the DM determines what happens.

+1. And don't be afraid to have "what happens" either not require a roll at all, or be a roll with far more granular and bounded outcomes than pass-fail.

A master thief like Arsene Lupin or Robin Hood should not have the same outcome of a failed thievery roll as a random commoner. Nor should they even be forced to roll in all of the same circumstances either.


WotC has created this really weird dynamic in the community, where they create all of the rules and sources and then rarely (if ever) are the actual authority on tips/advice for playing the game. Which isn't too unusual for board games or video games, but I feel like D&D is so conceptual and narrative-based that it blurs the lines. WotC presumes to offer strategic advice for running the game with books like the DMG (which is about 10% very solid prescriptive advice for new DMs, plus 25% game lore like magic items and planes of existence, plus wayyy too much info about weird stuff like homebrewing spells that's probably been used by 12 people across 5e's entire run).

It's strange that my "recommended reading" for new DMs is Matt Colville's Running The Game, Keith Amman's The Monsters Know What They're Doing, a select few articles from The Angry GM, and finally, like, 23 pages out of a 300+ page official book that was supposedly written to Guide the Dungeon Master.

I honestly view this as a conceit of modern game design (especially RPG design) overall. RPGs like Divinity, WoW, Guild Wars 2 and Skyrim might have tutorials, but increasingly, you're just not going to learn all the ins and out of their systems within the game.

You might learn enough to sit down and play, but a fan guide will always have much, much more.

[strangebloke]

"When a number of individuals are trying to accomplish something as a group, the DM might ask for a group ability check". They are a group trying to accomplish "finding/setting up a campsite" so why wouldn't it be a group check. They aren't trying to do something sequentially (If they were there would be a time cost with every failure). Now you can just as easily say it's part of the "Working Together" section that precedes the group check section which would mean 1 character makes 1 roll at advantage. Either way your let everyone make their own individual check is not what the books suggest. So like people have been saying it's a DM creating problems for themselves by ignoring what the books say.
Incorrect. In the very section you quote, its said that the DM might call for a group ability check. There's no implication that a DM calling for many individual checks is 'doing it wrong.' It depends on the situation. Now admittedly, in the case of picking a campsite or navigating, group checks do arguably make the most sense. As I said above, I don't think my houserules are needed in such a case, but they could be fun to use anyway.

But there are a lot of other cases where either the characters are just individually all trying to do the same thing (not as a group) or where the DM feels that a group check is inappropriate for some other reason.

The class example here is stealth. The DM could call for a group check, but it wouldn't make much sense, since the whole point of a stealth check is that its single point of failure. If the paladin has disadvantage because of armor and and a massively negative dexterity modifier, it doesn't really make sense that he manages to sneak past the guard just because 3 of his party members are just stealthy enough to avoid notice.

Indeed, group checks aren't really a good solution at all, because they generally create a lot of perverse incentives if you enforce them strictly. Stealth is easier with a large group, but looking for a secret door as a group is harder.

The dogpiling issue comes up when the DM asks for a single check and everyone else in the party whose character isn't doing anything better asks if they can do the same action, triggering the same check. Of course the DM can just say "no, you don't get to make the check" or "If you all roll this is now retroactively a group check," but IMO these are ham-handed solutions with little justification in the narrative beyond "the rules say the DM can do whatever." If character 1 can attempt to sus out the motives of Vizier Staback, so can character 2.


Personally I think it's quite easy to make expertise beneficial regardless. Several checks can be abstracted to simply say "do you have proficiency + expertise + 1 source of magical aid? You get to skip rolling entirely." And I would argue many of them should be. It becomes less about comparing the numerical bonuses such features provide and more using them to grant martials access to nice things that the casters might need spells to pull off.
This is patching the system by not using the system. If you aren't using the full variance of the d20, and you in fact believe that the edge cases (eg expert fails on a nat 1, noob succeeds on a nat 20) are problematic then perhaps its time to admit that the system has a problem.

I think a lot of the so called problems come down to letting the players call for skill checks. Stop doing that and go back to the basics of players describe what they do and the DM determines what happens.
While I agree with this to an extent, there's more going on here. Players don't know what any given DM will allow them to roll for. Well behaved players will be very cautious about trying to do things in general because they don't want to be that player who attempts to do something ridiculous and gets told "thats impossible" by the DM. If the DM does allow one player to roll for something, the general assumption will be that its fair game for most of the players to roll, barring some character specific limitation (IE, the halfling rolling for halfling history).

Again, if character 1 can attempt to sus out the motives of Vizier Staback, so can character 2. Obviously the DM can arbitrarily say "no, you can't" but this ultimately just leads to ability checks becoming more irrelevant than they already are.


And if you're about to say "that doesn't matter: they weren't doing anything else with their action anyway" then you either need to create something to occupy that attention, or recognize that this has become a solo scene and buffing a teammate in their solo performance is fun. Seeing the advantage you granted make the difference is one of the best parts of this damned game. So yeah, I don't think there's anything at all wrong with saying "three cheers for teamwork!" and giving the checking PC an auto-succeed and moving on to something that actually challenges the PCs as quickly as possible.
The bolded is literally part of what I recommend though? Pick your action? I agree that everyone gassing up the musician before his big show is cool - but it shouldn't completely overshadow skills of everyone in the party.



I can't tell if you said that like it's no biggie or it's a horror story, but this sounds like a nightmare slog to me. Can I ask for more context of how it happened 50 times? Just one trigger-happy PC, or a particularly check-heavy encounter design?

Guidance isn't a permanent, unseen buff. It takes an action to cast, it has Verbal and Somatic components, and it requires Concentration. I struggle to think of times that it could be spammed. In social situations, people are going to notice and usually be weirded out by the cleric casting spells on herself mid-conversation. Stealth is out, because of Verbal components. Prolonged checks like searching a room are debatable; sometimes you can justify it, sometimes you can't. Stacking it with something like Enhance Ability is out, because of concentration. Initiative is only possible if you know a fight's about to pop off in the next 54 seconds and can cast it without being noticed and triggering the fight. Guidance is good for picking an occasional lock, or lifting a boulder, but it isn't a Reaction.
The verbal component is something like a quick prayer, and the somatic component is touching a person (or yourself). The idea of a cleric praying constantly, even to the annoyance of the rest of the party, is very normal and in-character. People IRL do this without receiving any tangible benefit! Heaven forbid the player combatively responds to the DM by obtrusively praying to Ao every minute in IRL time to ensure they always have guidance up.

As I said earlier, the only thing keeping these abilities from being abused is that people forget about them. :smalltongue:


To be fair, I also think calling for a group check is rarely a good answer to any D&D situation. If you have a plate clad oaf you need to sneak past an open door, the sneaky player isn't going to feel cool for rolling a 29 if the only thing it achieves is balancing out the paladin's crappy check. That's not fun. That's babysitting (like you said).
Agreed. Group checks have uses, but they create a lot weird unintuitive situations, and in most cases should require a very different DC. Mathematically a single person with a good athletics check has a better chance of pulling a rope to get the sail pulled in than the group does - it would seem as though this is not intended.

[Psyren]

This is patching the system by not using the system. If you aren't using the full variance of the d20, and you in fact believe that the edge cases (eg expert fails on a nat 1, noob succeeds on a nat 20) are problematic then perhaps its time to admit that the system has a problem.

It "has a problem" in the sense that every system made by human hands has a problem, i.e. is flawed. Is it also time to admit water is wet? :smalltongue:

[Sorinth]

Incorrect. In the very section you quote, its said that the DM might call for a group ability check. There's no implication that a DM calling for many individual checks is 'doing it wrong.' It depends on the situation. Now admittedly, in the case of picking a campsite or navigating, group checks do arguably make the most sense. As I said above, I don't think my houserules are needed in such a case, but they could be fun to use anyway.

But there are a lot of other cases where either the characters are just individually all trying to do the same thing (not as a group) or where the DM feels that a group check is inappropriate for some other reason.

The class example here is stealth. The DM could call for a group check, but it wouldn't make much sense, since the whole point of a stealth check is that its single point of failure. If the paladin has disadvantage because of armor and and a massively negative dexterity modifier, it doesn't really make sense that he manages to sneak past the guard just because 3 of his party members are just stealthy enough to avoid notice.

Indeed, group checks aren't really a good solution at all, because they generally create a lot of perverse incentives if you enforce them strictly. Stealth is easier with a large group, but looking for a secret door as a group is harder.

The dogpiling issue comes up when the DM asks for a single check and everyone else in the party whose character isn't doing anything better asks if they can do the same action, triggering the same check. Of course the DM can just say "no, you don't get to make the check" or "If you all roll this is now retroactively a group check," but IMO these are ham-handed solutions with little justification in the narrative beyond "the rules say the DM can do whatever." If character 1 can attempt to sus out the motives of Vizier Staback, so can character 2.

It's "might" because what happens and how to handle it is arbitrarily up to the DM, they might ask for a group check, they might ask for a single check at advantage (Suggested under the Working Together section), they might not even ask for a check at all and just allow you to setup camp, or they might even use the passive ability scores to determine what you sus out of the Vizier.

That's why this whole discussion is so ironic, the DM is the one who decided out of a multitude of options to let everyone roll and then complains that everyone gets to roll. Like dude, if you don't like the math behind everybody gets a roll don't give everyone a roll, it's your choice pick the method you want. Don't shoot yourself in the foot and then blame the rules, you the DM decides the method, so choose a method you dislike it's that simple.

As for how a group stealth check can logically work, the stealthy character(s) is using hand signals to silently communicating telling the paladin when to move and when to stop, where to go. That's what gives the paladin advantage (Cancelled out but still) they are partly relying on the stealthy character's skill on where/when to move and are therefore better at moving stealthily during that attempt then during an attempt where they are alone and can't rely on the stealthy player guiding their movements.


While I agree with this to an extent, there's more going on here. Players don't know what any given DM will allow them to roll for. Well behaved players will be very cautious about trying to do things in general because they don't want to be that player who attempts to do something ridiculous and gets told "thats impossible" by the DM. If the DM does allow one player to roll for something, the general assumption will be that its fair game for most of the players to roll, barring some character specific limitation (IE, the halfling rolling for halfling history).

Again, if character 1 can attempt to sus out the motives of Vizier Staback, so can character 2. Obviously the DM can arbitrarily say "no, you can't" but this ultimately just leads to ability checks becoming more irrelevant than they already are.

If the DM does say only 1 player can make a roll how does it make ability checks irrelevant? The whole argument is that dogpiling makes ability checks irrelevant because of the math, so anything that prevents dogpiling whether it's arbitrary or not is by definition going to make ability checks more relevant.

But really to fix this particular issue the answer is when the first player asks to sus out the Viziers motives don't immediately call for a check, ask every player what they are doing then resolve all the actions at the same time. If multiple people are trying to sus out the Vizier's motives then follow the rules of Working Together which specifically call for one player to make the check, or use the sub-section of group checks.


Agreed. Group checks have uses, but they create a lot weird unintuitive situations, and in most cases should require a very different DC. Mathematically a single person with a good athletics check has a better chance of pulling a rope to get the sail pulled in than the group does - it would seem as though this is not intended.

Which is why there are actually two options presented, 1 player makes the check (Possibly at advantage if it's reasonable for help to actually matter), or group checks. For sure there are cases where group checks won't make sense, in those cases you would go with the other option presented in the PHB's Working Together section, "The character whose leading the effort - or the one with the highest ability modifier - can make an ability check with advantage, reflecting the help provided by the other characters ... Moreover, a character can help only when two or more individuals working together would actually be productive."

So setting up camp/pulling on a rope, 1 check at advantage, trying to sus out the Vizier's motives probably not unless they have a creative way to actually work together.

[Sorinth]

It's also worth pointing out any illogicalness to group checks for players wanting to all do the same thing is not an actual issue if the players know in advance that it will be a group check because they will self-regulate and instead focus on trying to find a way to give one person advantage by helping them in someway.

[strangebloke]

It "has a problem" in the sense that every system made by human hands has a problem, i.e. is flawed. Is it also time to admit water is wet? :smalltongue:
Okay... but if we can agree that skill checks have more variance than is useful, perhaps we could talk about ways to fix that? Does anyone have a problem with having a structure/guideline for more complex social/exploration encounters?


Which is why there are actually two options presented, 1 player makes the check (Possibly at advantage if it's reasonable for help to actually matter), or group checks. For sure there are cases where group checks won't make sense, in those cases you would go with the other option presented in the PHB's Working Together section, "The character whose leading the effort - or the one with the highest ability modifier - can make an ability check with advantage, reflecting the help provided by the other characters ... Moreover, a character can help only when two or more individuals working together would actually be productive."

So setting up camp/pulling on a rope, 1 check at advantage, trying to sus out the Vizier's motives probably not unless they have a creative way to actually work together.
So just to be clear if the team can't cooperate (everyone checking the motives of the vizier) would you say absolutely everyone gets to make a check, or only whoever asked to make a check?

Because the former seems fair but also is pretty much 100% dogpiling which is superior in every way to group checks, the other system you generally use. I would argue that teamwork should generally be better than everyone attempting something solo yeah? As for the second reading... well, this is subjective, by I find it worse. It's just really arbitrary, you know? "Sorry Steve, Linda already tried to sus out his motives. I know this is your character's thing, but she said it first and failed and now nobody gets to roll."


It's also worth pointing out any illogicalness to group checks for players wanting to all do the same thing is not an actual issue if the players know in advance that it will be a group check because they will self-regulate and instead focus on trying to find a way to give one person advantage by helping them in someway.
"Linda, don't help us! If there's three people it becomes a group check and our chance of success plummets massively!":smalltongue:

Feels pretty arbitrary and gamist to me. I think my system is more natural. Just give people enough things to do that they don't have a reason to all hyper-fixate on one task. If they choose to hyper-fixate anyway they should be rewarded in that area, not punished by enforcing a group check.

I'll also point out that the very section that outlines group checks says that they "don't come up very often" and the example given is "navigating a swamp together" where the group check is clearly meant to stand in for dozens of lesser checks. The intent does not seem to be that literally every time multiple PCs attempt the same thing, they need to be forced into a group check.

[PhoenixPhyre]

Players don't call for checks. Individual checks can only be made if there are individual consequences. Which means that if you dogpile and fail, you, personally, suffer the negative consequences of doing so. If there aren't individualized consequences, then the correct answer is "you don't get to roll." If there are consequences for the group as a whole (but not meaningfully individually) and they can help each other, it's a group check.

And that goes for positive consequences--if only one success or failure is needed and there's time to work together without incurring issues[1], then only one roll is made for the group (probably at advantage due to Help, if that makes sense in that scenario).

As a result--
* Stealth is intrinsically individual. Because individuals are discovered or not, surprised or not. Not groups. So no group stealth checks, and no Helping with stealth.
* Determining someone's motives in the moment[2] is individual--there is no chance to converse. But each individual only knows what their own check resulted in, and playing based on what other people rolled is entirely the bad kind of metagaming.
* Most knowledge-type checks are either group checks (if the party is trying to decide things together[3]) or are single-person checks. If you're answering a question posed by a Sphinx, there's no piling in. The group must nominate someone to speak for them.
* Most social checks are single-shot checks. And there, "Helping" can backfire, badly.
* Picking a lock or disarming a trap is a quintessential single-shot (before consequences apply) check.

In cases where the only consequence is time and the amount of time involved doesn't matter[4], sure, don't bother rolling a check at all. Or even setting a DC (once you know it's doable).

Those are all from the DMG, lightly paraphrased. There really isn't an issue here other than DMs letting players call for checks and thinking that more rolling is always better. It isn't. There's a balance to be struck, and just because someone says "oh, can I do it too?" doesn't mean the answer is yes.

[1] "too many cooks spoil the broth" is a saying for a reason. More help is frequently not better than less.
[2] Doing it afterwards is a group check if it's a check at all, due to limited consequences and inability to observe more.
[3] I've seen tons of groups of students convince themselves exactly backward. Two heads are better than one...but three are often worse than 1.
[4] this is situational. But it comes down to there must be meaningful consequences for every roll, either success or failure. If there aren't, you shouldn't ever roll.

[strangebloke]

[4] this is situational. But it comes down to there must be meaningful consequences for every roll, either success or failure. If there aren't, you shouldn't ever roll.

Isn't the "consequence" often just "you wasted your time and and didn't create the benefit/gain the information you wanted?" In which case I'd argue one common way of giving a potential check consequence would be to create other things the player could have been doing instead.

If you agree with that I'd say we're on the same page, at least wrt dogpiling.

But I'm kind of curious what everyone thinks of the parts of my post that aren't about dogpiling?

[Sorinth]

So just to be clear if the team can't cooperate (everyone checking the motives of the vizier) would you say absolutely everyone gets to make a check, or only whoever asked to make a check?

Because the former seems fair but also is pretty much 100% dogpiling which is superior in every way to group checks, the other system you generally use. I would argue that teamwork should generally be better than everyone attempting something solo yeah? As for the second reading... well, this is subjective, by I find it worse. It's just really arbitrary, you know? "Sorry Steve, Linda already tried to sus out his motives. I know this is your character's thing, but she said it first and failed and now nobody gets to roll."

"Linda, don't help us! If there's three people it becomes a group check and our chance of success plummets massively!":smalltongue:

Feels pretty arbitrary and gamist to me. I think my system is more natural. Just give people enough things to do that they don't have a reason to all hyper-fixate on one task. If they choose to hyper-fixate anyway they should be rewarded in that area, not punished by enforcing a group check.

At 2 it can also be a group check but more to the point, there are two proposed ways of handling it in the Working Together section and you seem to be ignoring that and fixating on group checks. So no the two systems you generally use aren't dogpiling or group checks, the two systems you generally use are 1 person makes the check (Possibly at advanatge) or a group check. DM decides which makes the most sense given the situation so if you don't think group check is appropriate no worries stick to the other method they suggest, the player with the highest mod makes the roll most likely at advantage. It does exactly what you want, cooperating creates a better odds and there's no dogpiling.

As for your example with Linda/Steve, I addressed it in the post you quoted but you cut that portion out of the reply. To repeat don't call for a roll immediately when Linda asks to sus out the Vizier check what everyone wants to do and then ask for whatever rolls are needed. So if Linda wants to insight check the Vizier, you ask Steve and the rest of the party what they want to do, if Steve also wants to insight check because that's his thing go with the working together rules and either have one of Steve/Linda make the roll or you use the group check rules as you see fit. If Steve doesn't want to insight check at that time then he has no business complaining when Linda rolls poorly.

Because if you want to talk about gameist logic, Steve doing nothing until Linda rolls poorly and only then deciding to make a roll is as gameist as it comes.


I'll also point out that the very section that outlines group checks says that they "don't come up very often" and the example given is "navigating a swamp together" where the group check is clearly meant to stand in for dozens of lesser checks. The intent does not seem to be that literally every time multiple PCs attempt the same thing, they need to be forced into a group check.

A lot of times you won't go with a group check because you are using the working together rules that says either whoever is leading the skill check or whoever has the highest mod makes the roll.

This whole conversation feels like your intentionally proposing poor decision/rulings as DM so that you can then complain about how the system is poor. Just don't intentionally make a decision that you feel is a bad decision.

[PhoenixPhyre]

Isn't the "consequence" often just "you wasted your time and and didn't create the benefit/gain the information you wanted?" In which case I'd argue one common way of giving a potential check consequence would be to create other things the player could have been doing instead.

If you agree with that I'd say we're on the same page, at least wrt dogpiling.

But I'm kind of curious what everyone thinks of the parts of my post that aren't about dogpiling?

That's only very very rarely a meaningful consequence IMO. Especially if other people are making the same check. A meaningful consequence moves the story along in some direction. "You wasted time and learned nothing" is just status quo ante unless the loss of time is meaningful in and of itself. Which it rarely is except in peculiar circumstances.

A meaningful consequence means you can't go back and try again easily--something has changed about the situation. Often, I tend to say that once you've tried one path and failed, you can't try that path again. You already gave it your best shot using that method. You need to try something else.

[Anymage]

Okay... but if we can agree that skill checks have more variance than is useful, perhaps we could talk about ways to fix that? Does anyone have a problem with having a structure/guideline for more complex social/exploration encounters?

You're clearly taking inspiration from complex skill checks/skill challenges. And there's room for something like that in 5e. It's just that "how and when can we apply skill challenge ideas" is very different from "ability checks are a problem in need of fixing". The latter isn't really relevant.

I mean, you could also try rolling 3d6 or 2d10 if you want to cut down on the randomness intrinsic in a single d20. I think the d20 engine is so tied to D&D in everyone's consciousness that it's worth keeping for that alone.


So just to be clear if the team can't cooperate (everyone checking the motives of the vizier) would you say absolutely everyone gets to make a check, or only whoever asked to make a check?

Because the former seems fair but also is pretty much 100% dogpiling which is superior in every way to group checks, the other system you generally use. I would argue that teamwork should generally be better than everyone attempting something solo yeah? As for the second reading... well, this is subjective, by I find it worse. It's just really arbitrary, you know? "Sorry Steve, Linda already tried to sus out his motives. I know this is your character's thing, but she said it first and failed and now nobody gets to roll."

I think it's worth realizing all the assumptions baked into this situation.

First, that the whole party is present. Usually with just one person talking, because in real life it quickly gets confusing to juggle multiple people talking at once. (Whether that person is the person with the highest Cha character or just the loudest person at the table depends on group dynamics.) Everybody else is present and sitting quietly, which is its own problem.

Second, that once information is known to one person it's instantly shared across the group hivemind. Otherwise, Linda's hunch about the vizier isn't of much use, since communicating it meaningfully would also be visible to the vizier and would be an odd twist to the conversation.

Third and most importantly, that it exists in a binary where a success gets accurate information while a failure gets absolutely nothing. If Linda's character has no Insight proficiency and a Wis of 10, she may very well get it wrong. The chance to get something wrong and make it worse is precisely why people would be better off consulting an expert than crowdsourcing all their problems on the hopes that some stranger somewhere rolls that 20.


"Linda, don't help us! If there's three people it becomes a group check and our chance of success plummets massively!":smalltongue:

Feels pretty arbitrary and gamist to me. I think my system is more natural. Just give people enough things to do that they don't have a reason to all hyper-fixate on one task. If they choose to hyper-fixate anyway they should be rewarded in that area, not punished by enforcing a group check.

Situations that do scale up well with number of helpers tend to also be tasks that, for whatever reason, shouldn't need a roll. They're routine enough and/or not dramatic enough that it's worth saying that a team of qualified experts just gets the job done. For plenty of other situations, more people - especially untrained people - are more hassle to work around than they are asset.

Combine with the abovementioned point that failing a roll can mean getting it wrong, and the dogpiling problem self corrects. If Linda and Steve whose characters have no investment in Wisdom or Insight think that the vizier seems to be too concerned with the crown prince while Alex's cleric thinks that the vizier feels like a father figure due to having to act the role while the king was being distant, either you've recreated group checks (majority wins even if that's mathematically disadvantageous) or you encourage letting experts shine in their respective fields. Now you just have to go back and ask again why the whole party is sitting there with nothing to do except roll Wisdom(Insight) checks, why a check on the part of anyone who's sitting in the background will meaningfully change the situation, and how long you want to have a bunch of PCs sitting quietly in the background anyways.

[Goobahfish]

So if I interpret this all correctly, the underlying issue is a 'Law of Large Numbers (LOLN)' problem.

To clarify my perspective. Combat in D&D largely works because it follows LOLN. If killing a dragon or TPK was determined by a single D20 check (or even a group proficiency roll), I think everyone would hate D&D.

Non-combat encounters function like the above 'Cake or Death' in many circumstances.

Taking 20 is pretty bad too because it kind of means that 1 point of proficiency will always make the difference.

------------

The solution I currently implement is as follows:

Most tasks (where possible) have hit points (not really but mechanically). This can be done via the suggested success/strike method or by having an actual damage mechanic against the target. Roll to hit (Climb check), roll damage (distance climbed) etc. This is offset by some fail mechanism (I have a stamina/hp mechanic). Thus a player half-way up a wall might be low on stamina and need to think about what they want to do to overcome this (give-up or rest/wait somehow etc).

This is for actual real 'encounter' level checks (the giant cliff face) rather than incidental checks (I leap over the table to attack the orc).

For this to work...

All players need to have agency which is not 'I do the same check... again'. That is the definition of boring.

If the mechanic doesn't have this choice involved then... you designed wrong. If the choice is which ability score do I leverage for the check, you have still designed wrong. The issue is that there needs to be different risk/rewards for different player choices even if they are small.

For Kings who are temperamental, this can be a patience (run out and you are in trouble) and merit (run out and you succeed). Each time you pass/fail the description of the king's mood can tell that you might want to 'give up' (already failed twice) or risk it depending on the circumstances. Here the LOLN will help the players with decent bonuses.

For some group checks, I do the following:
Players choose who rolls. Each success means something. Each failure means something. Any nat-1's... means something else entirely. Are you sure Brutog the Moron should be searching for traps?

Now the players have a choice... the choice has meaning because the risk-reward is there.

[strangebloke]

This whole conversation feels like your intentionally proposing poor decision/rulings as DM so that you can then complain about how the system is poor. Just don't intentionally make a decision that you feel is a bad decision.

I misunderstood you, and thought you were advocating for group checks in all situations, and i can honestly only blame myself for that. I apologize.

Historically, I've mostly run with 'working together' rules. Biggest modifier gets advantage if help is present/justified. But I've found this sort of setup unsatisfying in some ways. The third person 'helping' does nothing. Advantage is an overburdened mechanic; too many things give advantage for that to be the only contribution of an entire PC. I also think its very unintuitive in places. My preference is just to avoid the situation where 3+ people are all trying to do the same thing because it inherently is going to make people feel sidelined in one way or another.

Some of what I say is a critique of the ability check system. I do think the swinginess of the d20 is a problem. I do think guidance is a problem. But mostly I was just trying to write a quick, intuitive primer on how to avoid common pitfalls while engaging everyone as much as possible. One of my notes, the taking 20 bit, is literally just RAW. So my intent is more "here's my way of running things that avoids common pitfalls" that "don't use the POO-POO system, use my GOOD system."

I'll admit I had wrong assumptions about what people's perception of common pitfalls were. Anyway. Night.

[Psyren]

Okay... but if we can agree that skill checks have more variance than is useful, perhaps we could talk about ways to fix that? Does anyone have a problem with having a structure/guideline for more complex social/exploration encounters?

My point though is what you're dismissing as a "patch," or "not using the system," IS in fact the system.

With bounded accuracy, the variance of the d20 is the whole point; it's supposed to be exceedingly difficult, if not possible, for your total static bonuses to exceed the bounds of the die - and thus those "no need to roll anymore" moments that were achieved by static math in 3.5 (e.g. the player stacking competence atop circumstance atop insight atop enhancement atop morale atop luck atop sacred/profane atop alchemical etc) are now squarely in the hands of the DM. That's intentional. Feature, not bug.

The one place they still arguably exist is the "reliable talent/silver tongue" -style effects that change the bounds of the die itself from 1-20 to 10-20 or similar - and I like those much better, because generally the classes that get that sort of thing built in are getting it for things they shouldn't be failing anyway. It's yet another signal to the DM that either the expert PC shouldn't be rolling for that stuff in the first place, or if they are, the roll is less about determining whether they succeed and more about determining other things, like "how much?"

[PhoenixPhyre]

Don't do repeated checks if one failure is enough. Because repeated checks almost guarantee failure. Just like repeated checks, if one success is enough, almost guarantee success.

I find it much better to, when building complex situations (ie not incidental things) to have a number of different blockers--issues/situations/concerns that either preclude success or make it harder. Some blockers are hard--they must be resolved to get the "best outcome". Others are soft--success or failure merely influences other blockers.

For example, if the task they've set is to persuade the king to send X troops to Farmville to repel the orcs, blockers might be
a) Soft: Lack of information creating uncertainty about how much threat there really is. Failure to resolve this makes resolving other checks harder.
b) Hard: The money guy says there's not enough funds right now to pay for X. Failure to resolve this means you cap out at X/2.
c) Hard: One of the nobles is more worried about the border with Empireland and demands troops for there. Failure to resolve this means you may get no troops.
d) Hard: The king is jealous of the general who would command the troops sent, believing he's getting too popular. Not paranoid, not yet. Failure to resolve this means you may end up with a much lesser-quality general
e) Soft: The king is also very busy with his court favorite concubine, which makes him distracted. Failure to resolve this means the other forces are stronger.

And you can resolve each of those in a bunch of different ways.
a) this might be a matter of recon and presenting evidence. Or a social check.
b) You're going to have to either find more money, prove that the money guy is lying, or forge some documents. May take a heist.
c) Lots of ways to approach this, including outright bribery. Or getting an authority figure like a priest to say that the orcs (being heathens and blasphemers) are more important right now.
d) Sussuasion, deception, or just buttering up the king.
e) You could appeal to the king's sense of duty, you could poison the concubine (non-fatally, I hope), etc.

And some of these have to be done at the same time--the rogue might be slipping into the concubine's room to frame her as cheating while the wizard is working the treasury angle and the cleric is buttering up the high priest. At the same time the fighter and bard are presenting the evidence to the council.

But that's way more than most situations call for.

[Ionathus]

I honestly view this as a conceit of modern game design (especially RPG design) overall. RPGs like Divinity, WoW, Guild Wars 2 and Skyrim might have tutorials, but increasingly, you're just not going to learn all the ins and out of their systems within the game.

You might learn enough to sit down and play, but a fan guide will always have much, much more.

Definitely! I fondly remember the days of N64 instruction manuals and Nintendo Power strategy guides. Of course, even then they were competing with GameFAQs walkthroughs with totally wicked ASCII title art :smallbiggrin:

I think what makes D&D even weirder than checking something like a Skyrim wiki or subreddit or game blog, though, is that in Skyrim you aren't, like, programming and writing the game as somebody else plays it. As the player, you can screw up badly and the game will still continue to be a game with the same functioning mechanics and you can keep making progress. Screw up as the dungeon master in D&D, and things can grind to a halt in an instant...or simply plod along for months at a frustratingly slow pace.


Again, if character 1 can attempt to sus out the motives of Vizier Staback, so can character 2. Obviously the DM can arbitrarily say "no, you can't" but this ultimately just leads to ability checks becoming more irrelevant than they already are.

I disagree in a big way - in fact, I believe limiting "dogpile checks" makes ability checks vastly more relevant than they are in most games. If trying to Insight the Vizier has an associated opportunity cost, then the players are going to care more about that Insight roll. And if the first check fails (and the PCs have the opportunity to communicate this with each other, of course), they'll agonize over whether someone else should give up their chance for focusing on something else to try another go. Maybe that 2nd person has even worse Insight, so now it's a desperation play. That's exciting. If ability checks aren't perpetually a free action, then they carry more weight.


The bolded is literally part of what I recommend though? Pick your action? I agree that everyone gassing up the musician before his big show is cool - but it shouldn't completely overshadow skills of everyone in the party.

Ah, ok. Thank you for clarifying. I thought you were saying buffs made all other ability check decisions irrelevant, and I was offering some reasons that buffs weren't always feasible, but it sounds like we're on the same page there.


The verbal component is something like a quick prayer, and the somatic component is touching a person (or yourself). The idea of a cleric praying constantly, even to the annoyance of the rest of the party, is very normal and in-character. People IRL do this without receiving any tangible benefit!

If you're trying to smooth-talk your way into the rough and tumble Thieves' Guild tavern, and you pause your dialogue with the patrons every 54 seconds to say a prayer and put a hand over your heart or whatever (possibly with a flash of light?), the bouncer is going to throw you out, because that is some major sketchy vibes. If spells were some secret thing you could fire off without anyone noticing, no not even the strangers who are staring directly at you and are highly suspicious of you, we wouldn't have things like the Subtle Spell metamagic.


Heaven forbid the player combatively responds to the DM by obtrusively praying to Ao every minute in IRL time to ensure they always have guidance up.[QUOTE]

Again, I'm not sure if you think that'd be a bad, rude thing or a big brain "gotcha" play, but that is absolutely what I would do with a player who chose to die on the hill of "I always have Guidance and here's my justification."1 I'd set a timer for 54 seconds and require them to do the prayer like clockwork before continuing on with their conversation/stealth/clue hunt/footprint tracking. Guarantee it would be funny for the first 5 minutes, maybe, and then quickly become not worth the hassle. Very likely, they would realize that stopping whatever they're doing every minute for an average bonus of +2 on the die roll is not worth their time, and they would save Guidance for when a single discrete ability check was obviously coming up (locked door, boulder, arm wrestling, guitar solo, etc.).

Almost like that's how it's designed to be used.

[QUOTE]As I said earlier, the only thing keeping these abilities from being abused is that people forget about them. :smalltongue:

At some tables, I guess? At my table, the only thing keeping those abilities from being abused is the players' lack of interest in them. Because they trust the DM and the other players to tell an interesting story no matter how the dice fall, and they're invested more in their (relatively optimized) PC's actions than in the numbers that result from those actions.

I don't mean to turn this into a powergaming vs. roleplaying discussion, but the "guidance 50 times a session" problem simply doesn't exist if your players aren't trying to wring blood from a stone, gaming every possible advantage to maximize their numbers. A little bit of trust goes a long way. My players trust that I'm not going to **** them over just because they didn't mention their d4 buff cantrip, so they don't add 10 minutes to every freaking session by piling it on ad nauseam.

Put simply, the problem you are describing only exists for certain game types. And I swear I'm not trying to say "if you don't like it, then leave" -- but if gaming every possible advantage at the expense of a streamlined play experience is a necessity for you, then yeah, I don't think 5e RAW is the right system to use for the game you're trying to play.


That's only very very rarely a meaningful consequence IMO. Especially if other people are making the same check. A meaningful consequence moves the story along in some direction. "You wasted time and learned nothing" is just status quo ante unless the loss of time is meaningful in and of itself. Which it rarely is except in peculiar circumstances.

Excellently put. I agree wholeheartedly. "You wasted time and learned nothing" parallels pretty well with the "toll booth" checks I described earlier2: there's no uncertainty or doubt about what will happen. It just serves as an arbitrary barrier with a seemingly-arbitrary "cost" that you have to pay by rolling good enough, and until you do, you're stuck here waiting. The only choice is whether to pay with quarters or dimes, which is no choice at all.

And when the ability check is finally passed, everyone breathes a sigh of relief and moves on, glad that that's over with. Doesn't seem terribly fun or compelling to me.

1. Assuming, of course, that the option "talk to them and explain why constantly yelling out 'I cast Guidance on myself' 50 times a session is against the spirit of the spell design, against the spirit of cooperative roleplaying, annoying, and tedious to boot" is off the table
2. Look, I'm going to continue trying to make this analogy happen (https://youtu.be/Pubd-spHN-0)if it kills me

[Tanarii]

Don't do repeated checks if one failure is enough. Because repeated checks almost guarantee failure.
Only if a failed check blocks the way forward entirely.

I mean, generally speaking if you can keep trying the same until you succeed with no time pressure, skip the rolls. Just like if you can make attack rolls until something runs out of hit points, with no attacks back, you might as well skip the rolls. But if a failure means another approach is needed, or there's time pressure, or failure means a setback to overcome, then multiple successes (possibly of different kinds of checks) to succeed entirely work okay.

Skill challenges can be made to work in 5e if handled right.


Then things like group checks... some of the rules there are squiffy. A party of two loud people is likely to be heard, but add the extra noise from four quiet people on top of that and success is easier? Or you get a character that is built for stealth but they never get to use their abilities because everyone else is loud?
Stealth checks to Hide are explicitly individual against all listeners. As are stealth checks to gain Surprise.

That doesn't mean that a group stealth check at a more abstract level to avoid notice might not be appropriate at times though. It'd have to be a situation where the stealthy ones can explicitly help the non stealthy ones to support/cover the success of the group as a whole though.

Personally my favorite example of an apparently non-intuitive group check is group Lore checks, PCs discussing something as a group to decide on who has the right info. That's different from each person checking a know answer / don't know answer, with any one success giving a correct answer. Or even from each person checking and getting right or wrong info based on the result and then the players discussing it. It is an abstraction result for resolution, which is fine depending on the importance of getting a result from the abstraction of character debate vs having the players debate it. Yet for some reason it rubs a lot of folks the wrong way. :smallamused:

[KorvinStarmast]

Seeing the advantage you granted make the difference is one of the best parts of this damned game. So yeah, I don't think there's anything at all wrong with saying "three cheers for teamwork!" and giving the checking PC an auto-succeed and moving on to something that actually challenges the PCs as quickly as possible. It is indeed a team game.


I've played with Guidance-spammers and, without exception, it was because the DM either didn't really know why they were allowing an ability check, or they didn't know how to control the pace of the encounter. Ding Ding Ding. We have a winner. :smallsmile:


WotC presumes to offer strategic advice for running the game with books like the DMG (which is about 10% very solid prescriptive advice for new DMs, plus 25% game lore like magic items and planes of existence, plus wayyy too much info about weird stuff like homebrewing spells that's probably been used by 12 people across 5e's entire run)...It's strange that my "recommended reading" for new DMs is Matt Colville's Running The Game, Keith Amman's The Monsters Know What They're Doing, a select few articles from The Angry GM, and finally, like, 23 pages out of a 300+ page official book that was supposedly written to Guide the Dungeon Master. It means that they gooned it with the DMG when they had a chance to succeed. Guess they rolled a 3 on that attempted (INT) Authoring check.

100% agree with this. Letting players ask for skill checks stifles their creativity too, in my experience, because without that crutch they're free to describe the actual thing they want to do that often falls outside the narrow guidelines of 18 skills. And it's too close to roll playing for my tastes. It's one of the 3.xisms that needs to die in a great burst of hellfire.

That's why this whole discussion is so ironic, the DM is the one who decided out of a multitude of options to let everyone roll and then complains that everyone gets to roll. Like dude, if you don't like the math behind everybody gets a roll don't give everyone a roll, it's your choice pick the method you want. Don't shoot yourself in the foot and then blame the rules, you the DM decides the method, so choose a method you dislike it's that simple. Yep.

As for how a group stealth check can logically work, the stealthy character(s) is using hand signals to silently communicating telling the paladin when to move and when to stop, where to go. That's what gives the paladin advantage (Cancelled out but still) they are partly relying on the stealthy character's skill on where/when to move and are therefore better at moving stealthily during that attempt then during an attempt where they are alone and can't rely on the stealthy player guiding their movements . That's an interesting take on Help.


But really to fix this particular issue the answer is when the first player asks to sus out the Viziers motives don't immediately call for a check, ask every player what they are doing then resolve all the actions at the same time. If multiple people are trying to sus out the Vizier's motives then follow the rules of Working Together which specifically call for one player to make the check, or use the sub-section of group checks. Bravo. Get the whole group of people engaged in the scene. (Of course, my bard is over there lighting up a cigarillo, ignoring the impassioned plea being made by the paladin and the sorcerer to the vizier since she found his hospitality and courtesy lacking ... )

Which is why there are actually two options presented, 1 player makes the check (Possibly at advantage if it's reasonable for help to actually matter), or group checks. For sure there are cases where group checks won't make sense, in those cases you would go with the other option presented in the PHB's Working Together section, "The character whose leading the effort - or the one with the highest ability modifier - can make an ability check with advantage, reflecting the help provided by the other characters ... Moreover, a character can help only when two or more individuals working together would actually be productive."

So setting up camp/pulling on a rope, 1 check at advantage, trying to sus out the Vizier's motives probably not unless they have a creative way to actually work together.
Yep.

It's also worth pointing out any illogicalness to group checks for players wanting to all do the same thing is not an actual issue if the players know in advance that it will be a group check because they will self-regulate and instead focus on trying to find a way to give one person advantage by helping them in someway. Depends on the group, but I've seen that.

{snip the details} If two people have proficiency with thieves tools I allow help on some occasions.
For the record: as a player in a couple of Phoenix's games, I found his approach to work very well in terms of flow and ease of application, and the use of help when it makes sense.

Personally my favorite example of an apparently non-intuitive group check is group Lore checks, PCs discussing something as a group to decide on who has the right info. I use this for Lore/History checks.

[Psyren]

I can't tell if you said that like it's no biggie or it's a horror story, but this sounds like a nightmare slog to me. Can I ask for more context of how it happened 50 times? Just one trigger-happy PC, or a particularly check-heavy encounter design?

Guidance isn't a permanent, unseen buff. It takes an action to cast, it has Verbal and Somatic components, and it requires Concentration. I struggle to think of times that it could be spammed. In social situations, people are going to notice and usually be weirded out by the cleric casting spells on herself mid-conversation. Stealth is out, because of Verbal components. Prolonged checks like searching a room are debatable; sometimes you can justify it, sometimes you can't. Stacking it with something like Enhance Ability is out, because of concentration. Initiative is only possible if you know a fight's about to pop off in the next 54 seconds and can cast it without being noticed and triggering the fight. Guidance is good for picking an occasional lock, or lifting a boulder, but it isn't a Reaction.

I've played with Guidance-spammers and, without exception, it was because the DM either didn't really know why they were allowing an ability check, or they didn't know how to control the pace of the encounter. Usually both, because the problems go hand-in-hand. The problem I always notice is that they try to retroactively say "I'd cast Guidance on myself before I did X" every time they roll for anything, no matter how reactive that check is or whether it's feasible to predict the need for it. The game gets bogged down in making a bunch of checks instead of actually understanding and solving the puzzles in front of you, including their spinoffs like fighting (violence puzzles) or roleplaying (dialogue puzzles). And usually the checks are something totally inconsequential: haggling over price at a tavern, trying to win over a guard, trying to win the pie-baking contest. They could be simple and streamlined and fun, and allowing unrealistic time for buffing in the middle kills that fun for me.

To extend my analogy from before: playing at a table with a Guidance-spammer is like sitting in the passenger seat as the driver of your car scrounges for loose change at the toll booth, insisting that they'll get this right if they can just find 5 more cents. A DM's job with skill checks is to not waste everyone's time counting pennies: DMs should learn to recognize when the scene is dead or dying and wave it through before everyone in the car passes out from boredom.

I missed this earlier but this is fantastic. +100.


To be fair, I also think calling for a group check is rarely a good answer to any D&D situation. If you have a plate clad oaf you need to sneak past an open door, the sneaky player isn't going to feel cool for rolling a 29 if the only thing it achieves is balancing out the paladin's crappy check. That's not fun. That's babysitting (like you said).

In this scenario, a group check should be the least fun and most fail-prone option on purpose because it really is a cartoonishly bad idea. A better idea would be causing a distraction, or using the stealthiest player to sneak ahead solo to neutralize the reinforcements/alarm system, or using magic to disguise yourself, or any number of other options. Or - and this is absolutely a valid choice - saying "to hell with it" and just charging into the room. 9 times out of 10, group checks don't make the game more fun - they feel more like "math" than any other check, and the averaging-out of the rolls really sucks a lot of the character out of each individual roll.

I would much rather say "ok, you'll need a stealth check to get across...or you'll need 2 allies to make an additional check in your favor for you to auto-succeed." Performance to create a distraction, Deception to impersonate the guards and draw them off for a moment, Knowledge(Architecture and Engineering) to identify an acoustic dead zone...something that adds character to the story. That way, the stealthy PC is rewarded for being so dang good at stealth that they get to do more cool stuff and help the team out actively.

Am I missing something? Group checks don't require averaging or a bunch of math. All you do is count the winning dice. It doesn't matter if the Paladin rolled a 10 for a total of 7 or a 1 for a total of -2, they've still only failed once. (Well, unless you as the DM want their value to matter, but the whole point of group checks is a speedy resolution method so why would you?)

An easy way to make the specialist feel special in a group check is to make their big rolls count as multiple successes. Say if they beat the DC by 5 that's 2 successes, by 10 = 3 successes, and a natural 20 means the group makes it automatically, that kind of thing.

[Ionathus]

I missed this earlier but this is fantastic. +100.

Thank you!


Am I missing something? Group checks don't require averaging or a bunch of math. All you do is count the winning dice. It doesn't matter if the Paladin rolled a 10 for a total of 7 or a 1 for a total of -2, they've still only failed once. (Well, unless you as the DM want their value to matter, but the whole point of group checks is a speedy resolution method so why would you?)

An easy way to make the specialist feel special in a group check is to make their big rolls count as multiple successes. Say if they beat the DC by 5 that's 2 successes, by 10 = 3 successes, and a natural 20 means the group makes it automatically, that kind of thing.

Nope, looks like I was missing something in my memory of group checks. RAW they're just counting successes, like you said. It comes up so rarely that I must've seen one of my first DMs do the averaging thing years ago and simply assumed.

"Count the successes" does feel better, though it's still a little game-mechanics-focused for my tastes. If the stealthy character can win the whole group check for the team with a high enough roll, I'd prefer to flavor that as the stealthy PC doing something extra cool individually, like scouting ahead or knocking out a guard. But then, I'm really tableau-driven, so I love spinning three ability checks into an entire scene a la mini skill challenges. We're a story heavy table so we kinda all crave that narrative through-line.

[Psyren]

Thank you!



Nope, looks like I was missing something in my memory of group checks. RAW they're just counting successes, like you said. It comes up so rarely that I must've seen one of my first DMs do the averaging thing years ago and simply assumed.

"Count the successes" does feel better, though it's still a little game-mechanics-focused for my tastes. If the stealthy character can win the whole group check for the team with a high enough roll, I'd prefer to flavor that as the stealthy PC doing something extra cool individually, like scouting ahead or knocking out a guard. But then, I'm really tableau-driven, so I love spinning three ability checks into an entire scene a la mini skill challenges. We're a story heavy table so we kinda all crave that narrative through-line.

Oh for sure, but you can easily plant that narrative through-line retroactively. It's a lot easier to play out and it fits perfectly with 5e.

1) Players state what they want to do ("We want to all sneak past that orc guard post to enter the fort. Robin the Rogue wants to lead the way and help the rest of us who aren't good at sneaking.")

2) DM decides on a resolution method (success is both possible and failure has meaningful consequences, so a die roll is appropriate, and the DM in this case decides on a group check.) They set a DC based on the factors they know (e.g. the orcs in that post are all lazy and decided to get drunk and rowdy because their taskmaster isn't watching them - so he uses their passive perception of 13. If they were instead being actively watchful he might set a higher difficulty instead.)

3) Thanks to expertise and reliable talent, Robin the Rogue absolutely aces his check, scoring multiple successes. The group easily succeeds. Now you can figure out a narrative justification for what happened. For example, Robin perfectly helps the paladin and cleric time the clanking of their armor with the yells from the rowdy orcs after each throw of their gambling game inside the guard post, and it goes unnoticed. He also made them doff and stow their pristine tabards and helmets which would have reflected the moonlight.)

[Danielqueue1]

Repeating what has been said before and adding some of my own. Sounds like uncreative DMs using the wrong tools for the job. Trying to use a club to cut a rope then complaining about the damage rules as it were.

My advice (many have been said before)

1 the DM says when and what to roll.

2 dice come in when things matter
-in combat, a wasted turn is meaningful.
-out of combat (see more later)

3 the DM determines the DC.
-There is nothing in the play materials that says the DC for all skill checks must be eternally set in stone independent of all circumstances. The DC for a character holding on to a rope that somthing heavy is dangling from could be 15, while the DC for two characters could be 10. The DM makes that decision. Ham-fisting it into group check rules is silly. Convincing a priest of your good intentions could be a 5 for a dwarven cleric but much higher for a tiefling assassin. (And a Cleric's success making up for the tiefling's failure totally makes sense as a part of a group check)

4 degrees of success. Different DCs for different results. It doesn't have to be binary.
-stealth check (no one knows you are there/ "could have sworn I heard something, oh well."/ "hey did you hear that? Lets go check it out. " / "Intruders! Sound the alarm!")

-Survival (perfect resting place with a view/ pretty decent protection from the wind, but poor sightlines if something shows up/ OWLBEARS! Roll initiative!)

-insight (he clearly knows exactly what you are talking about, but hasn't decided whether he can trust you yet. / he has that glimmer of recognition in his eye but how much he knows isn't clear just yet. / he has a strange look in his eye, but it's hard to tell if he knows something or just thinks it weird that your friend keeps muttering prayers and poking you every minute of this conversation. / he clearly is giving your cleric the evil eye! He must be opposed to the forces of good!)

5 consequences: if there is no consequences for failure, narrate it and move on.
For picking a lock,
- the lock could be trapped, the trap could be magic, poison gas, a bell on a string, or even a failed trap that just ends up gumming up the lock.
-you could take too long and the assassination target gets up and goes out the front door while you are busy at the back door.
- you could get caught, make too much noise and someone hears the tumblers moving from the other side, a guard patrol could come along, or my favorite "it takes you a few minutes, and you are making good progress, but you only get about half the tumblers set when you hear noises coming your way, you can try again at disadvantage to finish quickly, but if you fail, you will be caught in the open! (Only do this if you already planned to have creatures in the area)

And finally, make it interesting. Have a dynamic world where people do things, go places and have goals. Instead of focusing on purely "fail or success" have enough going on that the distinction isn't black and white. Lets take the example of a group stealth check that's been bounding around. Party sent to kill a bandit lord who seems to be uncatchable to the local guard. Instead of having their hideout be a silent whiteroom dungeon, have it be in a series of tunnels in an abandoned mine next to a waterfall. The waterfall makes lots of noise meaning the DC for the stealth check close to the entrance isn't unreasonable even for someone in heavy armor. Don't ask for a group check, just tell everyone to roll Dex(stealth). turn a group stealth check where the heavy armor guy fails, but everyone else succeeds into one of the guards saying "Droger's rattlin' 'is chains again. I hate when boss has us workin' with that conjurer Blackstaff. Creeps me out he does." Boom you just narrated how a single fail in a group check resulted in a success anyway and set up the boss fight where the Bandit lord has backup with Blackstaff and "Droger." While at the same giving the party a possible avenue of nonviolently convincing the creeped out guard that he has less creepy places to be and if someone else fails a stealth check, a clever player with good deception could turn a fail into a success with a deception check. Bandit around a corner: "hey what was that!" Character imitating previous bandit: "Quit yer yellin' it's just Droger again!" (Roll dice)

[Ionathus]

And finally, make it interesting. Have a dynamic world where people do things, go places and have goals. Instead of focusing on purely "fail or success" have enough going on that the distinction isn't black and white. Lets take the example of a group stealth check that's been bounding around. Party sent to kill a bandit lord who seems to be uncatchable to the local guard. Instead of having their hideout be a silent whiteroom dungeon, have it be in a series of tunnels in an abandoned mine next to a waterfall. The waterfall makes lots of noise meaning the DC for the stealth check close to the entrance isn't unreasonable even for someone in heavy armor. Don't ask for a group check, just tell everyone to roll Dex(stealth). turn a group stealth check where the heavy armor guy fails, but everyone else succeeds into one of the guards saying "Droger's rattlin' 'is chains again. I hate when boss has us workin' with that conjurer Blackstaff. Creeps me out he does." Boom you just narrated how a single fail in a group check resulted in a success anyway and set up the boss fight where the Bandit lord has backup with Blackstaff and "Droger." While at the same giving the party a possible avenue of nonviolently convincing the creeped out guard that he has less creepy places to be and if someone else fails a stealth check, a clever player with good deception could turn a fail into a success with a deception check. Bandit around a corner: "hey what was that!" Character imitating previous bandit: "Quit yer yellin' it's just Droger again!" (Roll dice)

I'm a big fan of "fail forward" game design for D&D. This is a great example where a PC's failed check makes the world richer, rather than more tiresome!

The only change I'd make if I were running the scenario is that the paladin does still suffer a noticeable setback because of their failure (e.g. the guard didn't discover you but is now on higher alert, or the wizard had to burn an illusion spell to cover your failure), and the additional information is only teased and the rest is locked behind further checks. The party can continue sneaking past once the danger of discovery is gone, or they can risk a detour to make Perception/Stealth/other checks to follow up on this "Droger" thread.