D&D 5E Attacking from Stealth. When you can / cant Hide - A thorough breakdown

[Mistwell]

For the hiding behind ally halfling trick, Mearls said he would apply disadvantage to the hide checks that happen after you've already attacked from that spot.

 

[Rod Staffwand]

The 5e hiding rules are just horribly written. I get the "rulings over rules" ethos and agree with it, but the hiding rules are such a jumble of vague concepts and poorly worded corner-cases and inconsistent language that they do next to nothing to help a DM or player make a ruling.

The primary issue is that they equate visibility with awareness, which common sense says is not the same thing. Another issue is that they don't do enough to delineate the differences between:
1. Your presence being completely unknown to your enemies.
2. Your presence being known, but your location being unknown.
3. Your presence and general location being known, but your specific location being unknown or obscured from view.

"You can't hide from a creature that can see you..." This is a poor sentence. A much more effective sentence would have been: "You can't hide from a creature that sees you." This seems to be the intent. Obviously you can hide from a creature that can see you--but hasn't yet. If a creature has noticed you must lose line of sight before attempting to hide. Extending this means that, if you are hidden, you can remain so from creatures that can potentially see you (but haven't yet).

The rules go on to explore various caveats, such as being seen or making noise. The effects of these are left to the DM to determine, especially in regards to combat. You might remain hidden to move up to backstab an enemy, but showing yourself openly would cause you to be seen.

So the rules mostly make sense, but the poor writing hides them behind a wall of confusion.

 

[Authweight]

The stealth rules are pretty hopelessly fubared IMO. My solution is to rewrite them entirely:

1) Out of a fight, you can attempt to stealth anywhere, anytime you like. You just need to provide a convincing reason for why/how this could work. The target number will be the enemy's passive perception with situational modifiers. Trying to hide in plain sight is possible, but extremely difficult. If you are hidden and a fight begins, you remain hidden until something happens to reveal you.

2) In a fight, you can attempt to hide using the hide action. For this to work, you must be heavily obscured or in 3/4 cover. Anything else will automatically fail unless there is a really good reason otherwise. This targets the enemy's passive perception. While you are hidden, enemies are assumed to know your general location, but not precisely enough to target you with an attack. They can definitely fling a fireball in your general direction though.

3) Once you are hidden, you remain hidden until something happens that reveals you. Enemies can use their action to actively look for you. If one enemy sees you, they can use their action to aid another. You can move and remain hidden as long as you remain at least lightly obscured or stay within some level of cover. Attacking or moving out of cover/concealment reveals you immediately. You can attempt to move from one area of cover to another while remaining hidden, but doing so will require a new stealth check with situational modifiers based on the difficulty of the maneuver.

My logic for this approach is that the hide action represents you ducking out of sight. At that point, if you succeed, then your enemies can't follow your movement, letting you stay hidden in places you couldn't hide in directly. If enemies really want to try and find you, they can look for you, or just move into a place where they can see you.

By my rules, you can hide behind a chair before the fight starts, but not while the fight is going on. However, you can hide behind a wall, scurry across the room behind cover, and then roll behind a chair and stay hidden. To me, that feels about right and matches the tone I want. YMMV though.

 

[Melkor]

It bothers me that the ambiguity of the rules as written requires this level of effort to try and make sense of.

 

[Joe Liker]

It bothers me that the ambiguity of the rules as written requires this level of effort to try and make sense of.

Nah, the rules as written make sense, and they work fine.

It only becomes a problem when DMs think their only means of shutting down a blatantly exploitative player is to read rules at them.

The 5e rules are not written with punctilious rigor in mind, so such a rigorous analysis as we see in this thread is bound to collapse. In 5e, the DM is supposed to adjudicate stealth pretty much every time. Mearls has indicated this in nearly those exact terms, so I'm a little mystified why the discussion continues as if a closer reading of the rules can eventually result in perfect, DM-proof, algorithmic clarity.

 

[Authweight]

Nah, the rules as written make sense, and they work fine.

It only becomes a problem when DMs think their only means of shutting down a blatantly exploitative player is to read rules at them.

The 5e rules are not written with punctilious rigor in mind, so such a rigorous analysis as we see in this thread is bound to collapse. In 5e, the DM is supposed to adjudicate stealth pretty much every time. Mearls has indicated this in nearly those exact terms, so I'm a little mystified why the discussion continues as if a closer reading of the rules can eventually result in perfect, DM-proof, algorithmic clarity.

The problem is that if they wanted it to be a DM adjudication, they should have just said so in the book. They gave a bunch of rules that seem like they're supposed to tell you how stealth works, but those rules don't actually end up making much sense when you try and parse them out.

The other issue is that if they wanted it to be DM adjudicated, they should have told us at least the general role they want stealth to have in the game. This is particularly important because hiding is a key rogue class feature. It would be great if they said something like, "hiding is always up to the DM. However, we assumed that in most cases a character can attempt to hide if they have some cover. If you want to make hiding more difficult, consider giving rogues or other stealthy characters [some other thing] to compensate." That would be clear, effective writing. Instead, the book leaves it very unclear how they imagined stealth working in the game.

 

[Joe Liker]

The problem is that if they wanted it to be a DM adjudication, they should have just said so in the book. They gave a bunch of rules that seem like they're supposed to tell you how stealth works, but those rules don't actually end up making much sense when you try and parse them out.

The other issue is that if they wanted it to be DM adjudicated, they should have told us at least the general role they want stealth to have in the game. This is particularly important because hiding is a key rogue class feature. It would be great if they said something like, "hiding is always up to the DM. However, we assumed that in most cases a character can attempt to hide if they have some cover. If you want to make hiding more difficult, consider giving rogues or other stealthy characters [some other thing] to compensate." That would be clear, effective writing. Instead, the book leaves it very unclear how they imagined stealth working in the game.

I do agree that the rules would be better with the addition of your passage.

 

[sgaon]

The stealth rules are pretty hopelessly fubared IMO. My solution is to rewrite them entirely:

1) Out of a fight, you can attempt to stealth anywhere, anytime you like. You just need to provide a convincing reason for why/how this could work. The target number will be the enemy's passive perception with situational modifiers. Trying to hide in plain sight is possible, but extremely difficult. If you are hidden and a fight begins, you remain hidden until something happens to reveal you.

2) In a fight, you can attempt to hide using the hide action. For this to work, you must be heavily obscured or in 3/4 cover. Anything else will automatically fail unless there is a really good reason otherwise. This targets the enemy's passive perception. While you are hidden, enemies are assumed to know your general location, but not precisely enough to target you with an attack. They can definitely fling a fireball in your general direction though.

3) Once you are hidden, you remain hidden until something happens that reveals you. Enemies can use their action to actively look for you. If one enemy sees you, they can use their action to aid another. You can move and remain hidden as long as you remain at least lightly obscured or stay within some level of cover. Attacking or moving out of cover/concealment reveals you immediately. You can attempt to move from one area of cover to another while remaining hidden, but doing so will require a new stealth check with situational modifiers based on the difficulty of the maneuver.

My logic for this approach is that the hide action represents you ducking out of sight. At that point, if you succeed, then your enemies can't follow your movement, letting you stay hidden in places you couldn't hide in directly. If enemies really want to try and find you, they can look for you, or just move into a place where they can see you.

By my rules, you can hide behind a chair before the fight starts, but not while the fight is going on. However, you can hide behind a wall, scurry across the room behind cover, and then roll behind a chair and stay hidden. To me, that feels about right and matches the tone I want. YMMV though.

I really like this; I was working on something similar myself. The thing I would change would be coming out concealment reveals you automatically--I want to allow rogues to gain advantage on attacks from hiding (otherwise, what's the point?), so I'd say they're revealed after an attack from hiding, (allowing a jump from concealment to backstab or whatever), just as though they were hiding before combat started. Do you think this OP's them? Maybe an exception if they pop out from the same 3/4 cover or obscurity they hid behind/in, to accommodate for common sense?

Also, as regards being able to hide in combat with at least 3/4 cover or heavy obscurity, this would be what is affected by Skulker/wood elf/lightfoot halfling.

 

[Authweight]

I really like this; I was working on something similar myself. The thing I would change would be coming out concealment reveals you automatically--I want to allow rogues to gain advantage on attacks from hiding (otherwise, what's the point?), so I'd say they're revealed after an attack from hiding, (allowing a jump from concealment to backstab or whatever), just as though they were hiding before combat started. Do you think this OP's them? Maybe an exception if they pop out from the same 3/4 cover or obscurity they hid behind/in, to accommodate for common sense?

Also, as regards being able to hide in combat with at least 3/4 cover or heavy obscurity, this would be what is affected by Skulker/wood elf/lightfoot halfling.

Rogues can always range attack from hiding with a thrown dagger or a crossbow, so they can benefit that way. It wouldn't break the game though to let the rogue run out from hiding and gank someone, although if they're dual-wielding then they're getting two attacks with advantage on both which essentially means four chances to get off their sneak attack dice. That's very strong, but it rewards clever play.

Other possibilities I've considered for letting rogues hide and then run out to gank someone:

1) You can remain hidden while using the dash action, but not while moving normally. So a rogue can cunning action dash, then immediately attack while still hidden. Doing anything in between would mean you're revealed.

2) You can attempt to run out and gank someone, but it requires a successful stealth check against their passive perception.

 

[Eric V]

The problem is that if they wanted it to be a DM adjudication, they should have just said so in the book. They gave a bunch of rules that seem like they're supposed to tell you how stealth works, but those rules don't actually end up making much sense when you try and parse them out.

The other issue is that if they wanted it to be DM adjudicated, they should have told us at least the general role they want stealth to have in the game. This is particularly important because hiding is a key rogue class feature. It would be great if they said something like, "hiding is always up to the DM. However, we assumed that in most cases a character can attempt to hide if they have some cover. If you want to make hiding more difficult, consider giving rogues or other stealthy characters [some other thing] to compensate." That would be clear, effective writing. Instead, the book leaves it very unclear how they imagined stealth working in the game.

This is precisely on point. When I read the rulings in different backgrounds, it's CLEAR they expect DMs at individual tables to run with it as they see fit. For stealth, there is actually a lot written on it (in different locations, which is poor organization, but whatever), giving the idea that something specific was in mind when they wrote that section of the rules.