5E How to skill check (and why 5e got stealth wrong)

FrogReaver

Adventurer
I want to provide an alternative framework to the goal and approach method that's gotten so much attention the last few days. I'm going to call it the situation - action method.

Skill checks are inherently about resolving uncertainty. That's why when there isn't uncertain there's no point in rolling. That said, skill checks are not about resolving uncertainty relating to your goal, but rather uncertainty brought about through whatever action you took to achieve your goal. A simple example might help.

Goal: You want to the Lord Mayor of the Town to not increase hostilities with their neighboring town. Action: You attempt to persuade him. Resolution of Action: The DM determines there's no chance you can persuade him, Consequences of Action: but there is uncertainty about whether your attempt to persuade him will anger him. So he calls for a persuasion roll to determine whether the Lord Mayor is angry at you.

When should the DM demand an approach to go along with the action? When the approach might change the outcome or the DC he will set.

Approach: In the above example, the action to attempt to persuade the Lord Mayor could have been met with the DM asking the player to clarify "what he says or does to persuade him", because the DM knows that if you mention something about Blithe the sorcerer desiring the increased tensions that the Lord Mayor has a real chance to heed your advice. In this case the approach becomes important even though it isn't always.

So far I've explained the Action Framework. What do I mean about situation? Well this part is simple as well. Currently in D&D we pit players against individual NPC's. For example, consider a rogue attempting to sneak up to a camp of orcs we measure the pc vs each member's passive perception in the camp of orcs (assuming it's uncertain in the first place). That resolution method doesn't make sense other than under a heavy simulationist framework. The current method would be to resolve the Rogue's Action of sneaking up to the camp of orcs by comparing the pc's stealth check to each individual orc's passive perception. The better method would be for the DM to set a DC that takes into account the whole situation. In this case a camp of orcs on the lookout might be assigned a DC 20.

Going back to the Lord Mayor persuasion example, the DC may depend on who else is present for the persuasion attempt, or if you chose a particularly busy time to bring this before the Lord Mayor etc.

Anyways, this is my preferred resolution methodology. Determine if a check needs made by considering both the action and the consequences of that action, and only asking for additional approach information as needed and then if either of them auto succeed or fail and what consequences the action may have for failure or success. If a check gets made assign a DC that takes into account the whole situation.

Hopefully this helps put some thought and definition into the more traditional styles.
 
Last edited:

ad_hoc

Adventurer
The better method would be for the DM to set a DC that takes into account the whole situation. In this case a camp of orcs on the lookout might be assigned a DC 20.
I think the DM is already empowered to do that.

The quick and easy method given by the rules is to have advantage/disadvantage.

I think few people would say that being more granular, in this case giving the orcs a +9 rather than advantage's +5, is against the spirit of the rules.
 

FrogReaver

Adventurer
I think the DM is already empowered to do that.

The quick and easy method given by the rules is to have advantage/disadvantage.

I think few people would say that being more granular, in this case giving the orcs a +9 rather than advantage's +5, is against the spirit of the rules.
The DM is empowered to do anything.

By the way, I'm not suggesting we give the Orcs + anything. I'm suggesting the whole situation be taken into account and a DC assigned.
 

Oofta

Title? I don't need no stinkin' title.
I just make multiple opposed rolls based on how many orcs I think could reasonably be observing the area. Some of them may have advantage or disadvantage based on normal environmental variables.
 

FrogReaver

Adventurer
I just make multiple opposed rolls based on how many orcs I think could reasonably be observing the area. Some of them may have advantage or disadvantage based on normal environmental variables.
But you don't actually need to know which orcs spotted the rogue in this situation, or even how many. The rogue is just wanting to sneak up to the encampment. The question is whether he gets spotted. Once he's spotted you can take that part of the story wherever is most interesting or most fun or most realistic - Whatever your sensibilities desire.
 
Last edited:

Charlaquin

Goblin Queen
I want to provide an alternative framework to the goal and approach method that's gotten so much attention the last few days. I'm going to call it the situation - action method.

Skill checks are inherently about resolving uncertainty. That's why when there isn't uncertain there's no point in rolling. That said, skill checks are not about resolving uncertainty relating to your goal, but rather uncertainty brought about through whatever action you took to achieve your goal. A simple example might help.

Goal: You want to the Lord Mayor of the Town to not increase hostilities with their neighboring town. Action: You attempt to persuade him. Resolution of Action: The DM determines there's no chance you can persuade him, Consequences of Action: but there is uncertainty about whether your attempt to persuade him will anger him. So he calls for a persuasion roll to determine whether the Lord Mayor is angry at you.

When should the DM demand an approach to go along with the action? When the approach might change the outcome or the DC he will set.

Approach: In the above example, the action to attempt to persuade the Lord Mayor could have been met with the DM asking the player to clarify "what he says or does to persuade him", because the DM knows that if you mention something about Blithe the sorcerer desiring the increased tensions that the Lord Mayor has a real chance to heed your advice. In this case the approach becomes important even though it isn't always.
I don’t think I disagree with any of this. However, I do question how the DM can know whether or not the approach might change the outcome or the DC without hearing the approach first. Or, to be blunt, I don’t believe they can. Certainly, I would not be confident in my own assumption that the approach can’t change the outcome or DC without having heard the approach.

So far I've explained the Action Framework. What do I mean about situation? Well this part is simple as well. Currently in D&D we pit players against individual NPC's. For example, consider a rogue attempting to sneak up to a camp of orcs we measure the pc vs each member's passive perception in the camp of orcs (assuming it's uncertain in the first place). That resolution method doesn't make sense other than under a heavy simulationist framework. The current method would be to resolve the Rogue's Action of sneaking up to the camp of orcs by comparing the pc's stealth check to each individual orc's passive perception.
Who is “we,” exactly? I don’t think that’s how I’d resolve such an action.

The better method would be for the DM to set a DC that takes into account the whole situation. In this case a camp of orcs on the lookout might be assigned a DC 20.

Going back to the Lord Mayor persuasion example, the DC may depend on who else is present for the persuasion attempt, or if you chose a particularly busy time to bring this before the Lord Mayor etc.

Anyways, this is my preferred resolution methodology. Determine if a check needs made by considering both the action and the consequences of that action, and only asking for additional approach information as needed and then if either of them auto succeed or fail and what consequences the action may have for failure or success. If a check gets made assign a DC that takes into account the whole situation.

Hopefully this helps put some thought and definition into the more traditional styles.
Again, I agree with you, I think this is an excellent action resolution method. But I propose that additional approach information is always needed. I always need it, at any rate.
 

Oofta

Title? I don't need no stinkin' title.
But you don't actually need to know which orcs spotted the rogue in this situation. The rogue is just wanting to sneak up to the encampment. The question is whether he gets spotted. Once he's spotted you can take that part of the story wherever is most interesting or most fun or most realistic - Whatever your sensibilities desire.
But to me the "whole situation" is basically taking into consideration the number of times the rogue might be spotted, right? You can do that with a set DC, use passive perception for the orc, possibly have the rogue roll stealth multiple times as he crosses different sections of the camp.

I'm saying that because as a DM, taking the whole scenario into account is kind of meaningless. Or at least it is to me. I guess what I'm saying is that I don't know how I would set the DC ... what factors would I take into consideration? What if it's not just orcs (perception +0) but they've captured a black dragon wyrmling (perception +4)?

If I break it up into a couple of scenes then I know what to do. The outskirts of the encampment the rogue may have advantage or even automatic. Sneaking through the heart of the camp? Goes from being not all that much more difficult to very dangerous with the rogue rolling with disadvantage and orcs having advantage (+5 to passive if I go that way).

For social encounters I'm much more likely to set a DC and adjust based on approach, context, points raised and other details.
 

MarkB

Hero
The current method would be to resolve the Rogue's Action of sneaking up to the camp of orcs by comparing the pc's stealth check to each individual orc's passive perception. The better method would be for the DM to set a DC that takes into account the whole situation. In this case a camp of orcs on the lookout might be assigned a DC 20.
I notice that you didn't suggest asking for an approach when determining how to resolve the whole "rogue sneaking into a camp of orcs" situation. This certainly seems like something that may be modified significantly based upon the chosen approach.

On the other hand, I've never really liked the "multiple stealth rolls" approach to sneaking either, because - barring something like Reliable Talent - once you're requiring more than 2-3 rolls, the player's luck will almost inevitably run out at some point. It's basically imposing super-disadvantage on the check.
 

FrogReaver

Adventurer
I notice that you didn't suggest asking for an approach when determining how to resolve the whole "rogue sneaking into a camp of orcs" situation. This certainly seems like something that may be modified significantly based upon the chosen approach.

On the other hand, I've never really liked the "multiple stealth rolls" approach to sneaking either, because - barring something like Reliable Talent - once you're requiring more than 2-3 rolls, the player's luck will almost inevitably run out at some point. It's basically imposing super-disadvantage on the check.
1. It was just an example

2. If you think it would matter then call for one.. Depending on the details I probably would. But in the context of a encampment with a clearing surrounded by the forest then I wouldn't require an approach, though a well placed volunteered approach could help, up to the player (provided the rogue was wanting to get right up to the edge of the camp.)
 

FrogReaver

Adventurer
On the other hand, I've never really liked the "multiple stealth rolls" approach to sneaking either, because - barring something like Reliable Talent - once you're requiring more than 2-3 rolls, the player's luck will almost inevitably run out at some point. It's basically imposing super-disadvantage on the check.
Yep. that's the part 5e did wrong with stealth. They treated it like a 1pc v many npc skill, but it really should 1 pc vs 1 situation, like all other skills
 

FrogReaver

Adventurer
I don’t think I disagree with any of this. However, I do question how the DM can know whether or not the approach might change the outcome or the DC without hearing the approach first. Or, to be blunt, I don’t believe they can. Certainly, I would not be confident in my own assumption that the approach can’t change the outcome or DC without having heard the approach.
The DM can reasonably judge the typical set of approaches that a player means when he doesn't elaborate on an action. I sneak up to the encampment is a reasonable action declaration in that sense. The player is trying to be sneaky while getting up to the encampment. Unless there's a lot more granular information or detail going on in that scene I don't know why you would need anything more than the action declaration. If there is then that falls under the, if the approach would make a real difference then you ask for it in more detail.


Who is “we,” exactly? I don’t think that’s how I’d resolve such an action.
The D&D 5e community as a whole. To my knowledge the RAW prescribes stealth vs passive perception to resolve stealth.


Again, I agree with you, I think this is an excellent action resolution method. But I propose that additional approach information is always needed. I always need it, at any rate.
I don't think you need It nearly as much as you think you do.
 

5ekyu

Adventurer
I think the DM is already empowered to do that.

The quick and easy method given by the rules is to have advantage/disadvantage.

I think few people would say that being more granular, in this case giving the orcs a +9 rather than advantage's +5, is against the spirit of the rules.
I resolve sneak past groups as a "help" action advantaged observer. More than one person looking, its against Pper+5.

I resolve sneaking against a camp or settlement setup to be on watch as against the basic aptitude of their commander - 10-15-20 DMG possible guidelines plus or minus 5 for exceptional resources, time, morale etc.

But regardless, there may be choke points that straight up stealth fails st, no matter the roll.

Afaik its all within RAW.
 

FrogReaver

Adventurer
But to me the "whole situation" is basically taking into consideration the number of times the rogue might be spotted, right? You can do that with a set DC, use passive perception for the orc, possibly have the rogue roll stealth multiple times as he crosses different sections of the camp.
Or you just roll one time to see if his action is successful. Stretching out the basic stealth approach scene just isn't interesting unless there's a lot more going on in the scene than what we have currently described.

I'm saying that because as a DM, taking the whole scenario into account is kind of meaningless. Or at least it is to me. I guess what I'm saying is that I don't know how I would set the DC ... what factors would I take into consideration? What if it's not just orcs (perception +0) but they've captured a black dragon wyrmling (perception +4)?
Your the DM. It's your SUBJECTIVE call on what to set the DC at.

If I break it up into a couple of scenes then I know what to do. The outskirts of the encampment the rogue may have advantage or even automatic. Sneaking through the heart of the camp? Goes from being not all that much more difficult to very dangerous with the rogue rolling with disadvantage and orcs having advantage (+5 to passive if I go that way).
If getting to the different levels is important and yields interesting and important information or events then by all means break it apart into sections.

But the resolution method would still be the same. Set a DC for each level. In this case you aren't setting the DC for an individual orc, you are setting the DC for a specific orc encampment at a specific time. If it' a generic orc encampment in a more or less generic location without any obviously sight enhancing features then maybe the standard easy, medium, hard suffices.

For social encounters I'm much more likely to set a DC and adjust based on approach, context, points raised and other details.
If you can do it for social I'm confident u can do it for exploration.
 

Charlaquin

Goblin Queen
The DM can reasonably judge the typical set of approaches that a player means when he doesn't elaborate on an action. I sneak up to the encampment is a reasonable action declaration in that sense. The player is trying to be sneaky while getting up to the encampment. Unless there's a lot more granular information or detail going on in that scene I don't know why you would need anything more than the action declaration. If there is then that falls under the, if the approach would make a real difference then you ask for it in more detail.
I don’t know why I would need more specificity either, which is exactly why it’s important. I can’t know if the approach would make a difference unless the player tells me what their approach is.

The D&D 5e community as a whole. To my knowledge the RAW prescribes stealth vs passive perception to resolve stealth.
RAW says to “Make a Dexterity (Stealth) check when you attempt to conceal yourself from enemies, slink past guards, slip away without being noticed, or sneak up on someone without being seen or heard,” which I believe covers your suggestion of setting a DC to account for the scenario as a whole.

It also says that, “Until you are discovered or you stop hiding, that check's total is contested by the Wisdom (Perception) check of any creature that actively searches for signs of your presence.” So, if one of the orcs searches for the character, that’s the total they have to beat to find them. This might happen, for instance, if the player fails the initial check and sets the camp on alert as a consequence.

Additionally, it says that, “When you hide, there's a chance someone will notice you even if they aren't searching. To determine whether such a creature notices you, the DM compares your Dexterity (Stealth) check with that creature's passive Wisdom (Perception) score. which equals 10 + the creature's Wisdom modifier, as well as any other bonuses or penalties. If the creature has advantage, add 5. For disadvantage, subtract 5.” This might be a good way to set the DC for that initial check. This might also be relevant if there are certain creatures in the encampment with better Wisdom (Perception) than the orcs; maybe their chief has higher Wisdom, or maybe they have guard dogs that might have advantage due to Keen Smell, putting their passive Perception much higher than the orcs’.

Personally, I do t run Stealth 100% RAW, in part because I don’t run passive checks 100% RAW.

I don't think you need It nearly as much as you think you do.
Well that’s nice, but unconvincing.
 

Fanaelialae

Adventurer
I don’t know why I would need more specificity either, which is exactly why it’s important. I can’t know if the approach would make a difference unless the player tells me what their approach is.


RAW says to “Make a Dexterity (Stealth) check when you attempt to conceal yourself from enemies, slink past guards, slip away without being noticed, or sneak up on someone without being seen or heard,” which I believe covers your suggestion of setting a DC to account for the scenario as a whole.

It also says that, “Until you are discovered or you stop hiding, that check's total is contested by the Wisdom (Perception) check of any creature that actively searches for signs of your presence.” So, if one of the orcs searches for the character, that’s the total they have to beat to find them. This might happen, for instance, if the player fails the initial check and sets the camp on alert as a consequence.

Additionally, it says that, “When you hide, there's a chance someone will notice you even if they aren't searching. To determine whether such a creature notices you, the DM compares your Dexterity (Stealth) check with that creature's passive Wisdom (Perception) score. which equals 10 + the creature's Wisdom modifier, as well as any other bonuses or penalties. If the creature has advantage, add 5. For disadvantage, subtract 5.” This might be a good way to set the DC for that initial check. This might also be relevant if there are certain creatures in the encampment with better Wisdom (Perception) than the orcs; maybe their chief has higher Wisdom, or maybe they have guard dogs that might have advantage due to Keen Smell, putting their passive Perception much higher than the orcs’.

Personally, I do t run Stealth 100% RAW, in part because I don’t run passive checks 100% RAW.


Well that’s nice, but unconvincing.
While I do think that the approach can oftentimes be useful or even necessary, I also think that sometimes it isn't. Particularly when it involves specialized knowledge that the character would know far better than the player or DM would.

For example, picking a lock. Let's assume, for the sake of simplicity, that the character has already determined it isn't trapped. I only have an amateur understanding of lock picking, and even then only for modern locks (as opposed to medieval designs). Most of my players know even less than I do on the subject. But the character with expertise in Thieves Tools is a pro. So no approach is needed. Assuming a basic lock, unlimited time and the right tools, the character just succeeds. If it's a unique lock that the character might not be able to figure out, or I want to see how long it takes to open (because a wandering monster might show up) then we roll for it. An approach might be warranted under special circumstances (the character left his tools at home and needs to improvise) but that's its own thing.

Another case (IMO) is a knowledge check. While a player is free to expound on why their character is likely to know something, if we just want to know how much the character knows about the McGuffin of McGuffininess (which the player knows nothing about and therefore has little to no idea how their character might know) I say just let the dice decide. Assuming the character is knowledgeable, if they roll well they know about the item, and if not they have an idea of where they might be able to look for more information.
 

Oofta

Title? I don't need no stinkin' title.
While I do think that the approach can oftentimes be useful or even necessary, I also think that sometimes it isn't. Particularly when it involves specialized knowledge that the character would know far better than the player or DM would.

For example, picking a lock. Let's assume, for the sake of simplicity, that the character has already determined it isn't trapped. I only have an amateur understanding of lock picking, and even then only for modern locks (as opposed to medieval designs). Most of my players know even less than I do on the subject. But the character with expertise in Thieves Tools is a pro. So no approach is needed. Assuming a basic lock, unlimited time and the right tools, the character just succeeds. If it's a unique lock that the character might not be able to figure out, or I want to see how long it takes to open (because a wandering monster might show up) then we roll for it. An approach might be warranted under special circumstances (the character left his tools at home and needs to improvise) but that's its own thing.

Another case (IMO) is a knowledge check. While a player is free to expound on why their character is likely to know something, if we just want to know how much the character knows about the McGuffin of McGuffininess (which the player knows nothing about and therefore has little to no idea how their character might know) I say just let the dice decide. Assuming the character is knowledgeable, if they roll well they know about the item, and if not they have an idea of where they might be able to look for more information.
It seems like a lot of people conflate a gaming philosophy like goal and approach with the role of the dice. I disagree with GAA for the reasons you point out, I'm not a locksmith and have no clue. As far as I know all locks can be opened with a Fonzie fist bump.

But one style of play (page 236 DMG) is to only use dice as rarely as possible. Which is fine. It does come with some pluses and minuses that the book points out including the fact that as much as we might like to think otherwise, no DM is completely neutral.

In other words, different style? Great. Claim RAW SAYS ... umm ... nope.
 

Charlaquin

Goblin Queen
While I do think that the approach can oftentimes be useful or even necessary, I also think that sometimes it isn't. Particularly when it involves specialized knowledge that the character would know far better than the player or DM would.

For example, picking a lock. Let's assume, for the sake of simplicity, that the character has already determined it isn't trapped. I only have an amateur understanding of lock picking, and even then only for modern locks (as opposed to medieval designs). Most of my players know even less than I do on the subject. But the character with expertise in Thieves Tools is a pro. So no approach is needed. Assuming a basic lock, unlimited time and the right tools, the character just succeeds. If it's a unique lock that the character might not be able to figure out, or I want to see how long it takes to open (because a wandering monster might show up) then we roll for it. An approach might be warranted under special circumstances (the character left his tools at home and needs to improvise) but that's its own thing.
Picking a lock is an approach to the goal of opening a locked door. Breaking it down is another.

Another case (IMO) is a knowledge check. While a player is free to expound on why their character is likely to know something, if we just want to know how much the character knows about the McGuffin of McGuffininess (which the player knows nothing about and therefore has little to no idea how their character might know) I say just let the dice decide. Assuming the character is knowledgeable, if they roll well they know about the item, and if not they have an idea of where they might be able to look for more information.
In my opinion, the idea of a “knowledge check” is kind of silly. There’s no real action being taken there (again, in my opinion), so a check isn’t the appropriate way to resolve it. The Intelligence-related skills are great for resolving attempts to analyze, identify, study or otherwise attempt to learn something you didn’t before. But for recalling information already known, I don’t think a check is the best way to handle that. I’d rather just tell the player “yes, you know that” or “no, you don’t” based on their character’s background and Proficiencies.
 

Fanaelialae

Adventurer
In my opinion, the idea of a “knowledge check” is kind of silly. There’s no real action being taken there (again, in my opinion), so a check isn’t the appropriate way to resolve it. The Intelligence-related skills are great for resolving attempts to analyze, identify, study or otherwise attempt to learn something you didn’t before. But for recalling information already known, I don’t think a check is the best way to handle that. I’d rather just tell the player “yes, you know that” or “no, you don’t” based on their character’s background and Proficiencies.
That's a perfectly legitimate way to run it, but it certainly isn't how everyone does it.

There are times when the answer is obvious, and no roll is necessary. Your character knows his name unless suffering amnesia, but he almost certainly doesn't know the true name of a Demon Lord unless possession of that knowledge had been pre established.

However, sometimes it isn't so obvious, particularly for a DM who believes their role to be that of an impartial arbiter. In that situation, such a DM will find it useful to allow the dice to decide how much to reveal to the players.
 

Beleriphon

Totally Awesome Pirate Brain
Yep. that's the part 5e did wrong with stealth. They treated it like a 1pc v many npc skill, but it really should 1 pc vs 1 situation, like all other skills
For Stealth I'm pretty sure the implication of the rules, if not the outright rules, are roll once and compare to passive Preception until something about being stealthy changes dramatically enough to need a new roll. That isn't that different than your suggestion of one roll vs the entire camp, since at a practical level probably every single orc is going to have the same score, so if the rogue can beat one they can beat them all.
 

Charlaquin

Goblin Queen
That's a perfectly legitimate way to run it, but it certainly isn't how everyone does it.

There are times when the answer is obvious, and no roll is necessary. Your character knows his name unless suffering amnesia, but he almost certainly doesn't know the true name of a Demon Lord unless possession of that knowledge had been pre established.

However, sometimes it isn't so obvious, particularly for a DM who believes their role to be that of an impartial arbiter. In that situation, such a DM will find it useful to allow the dice to decide how much to reveal to the players.
Which is also perfectly reasonable and valid. Just not the way I like to do things.
 

Advertisement

Top