D&D 5E Invsibility vs Cloak of Elvenkind

TheCosmicKid · Mar 19, 2017

Hriston said:
Agreed about the confusion, but again the positive visual evidence provided by the cloak seems stronger to me than the negative evidence of not seeing an invisible creature. Thus the advantage.

Seeing thin air is "positive sensory evidence". It's evidence of absence. What your eyes are telling you is affirmatively inconsistent with what your ears are telling you. Negative evidence, absence of evidence, would be a situation where you simply can't see: it's dark, or the noise is behind you, or it's coming from another room. Your eyes aren't confirming what your ears are telling you, but they're not contradicting it either. See the difference?

Hriston said:
No, if you can locate someone with your eyes they can't hide from you at all.

You now appear to be arguing that the Perception skill is purely nonvisual. This seems implausible. You can hide from someone with standard cover or concealment. You don't need full cover or concealment. The normal presumption is that if you're hiding in shadows or behind a tree, and the other creature perceives you, they have located you at least partially with your eyes. 3rd Edition even called this out as a "Spot check" as opposed to a "Listen check".

Hriston said:
Yeah, but that's a failed Stealth check on the part of the cloak wearer. We were discussing situations where an invisible creature would fail while a cloak wearer wouldn't, and why that would be the case.

And I'm arguing that there are no such situations. Whatever way you formulate the cloak as providing advantage on Stealth checks, outright invisibility ought to provide at least the same advantage. Seeing a subtle color shift is harder than seeing an uncloaked creature; seeing a ripple in the air is harder than seeing a visible creature. Concluding that a stump is actually a creature requires setting aside an assumption about physics; concluding a patch of thin air is actually a creature requires setting aside an assumption about physics. All the arguments you've given for advantage run in parallel this way.

TheCosmicKid · Mar 19, 2017

Ovinomancer said:
3) You actually narrate both results of a check with advantage? How does that work for you?

I sometimes do this. Well, not narrate both results, but if the player rolls a 10 and a 15, it's often not difficult to construct a narrative where the character would have failed but the advantageous situation gives them the edge they need to succeed. It's not really any different than doing the same with a flat bonus. "This attack would have hit you, but you can deflect it because you've got a shield." That sort of thing.

ClearlyTough69 · Mar 20, 2017

Springheel said:
The problem, as I see it, is that the rules generally (but not always) seem to distinguish between noticing someone by sight or noticing someone by sound.

"A given area might be lightly or heavily obscured. In a lightly obscured area, such as dim light, patchy fog, or moderate foliage, creatures have disadvantage on Wisdom (Perception) checks that rely on sight." (PHB 183)

But there are no guidelines anywhere for when DMs should allow detection of a hiding creature based on sight and when to use sound. In the majority of cases, both would apply to varying degrees.

I think this gets to the heart of much of the confusion about hiding and perception; this, and what you know about a creature's location if you detect it but can't see it. I've tussled with these problems for months.

My eureka moment was to break down the analysis of a Perception check into its component senses. I use one Perception check result for all senses, but analyse each sense in turn. I'll illustrate by example:

1. A rogue is trying to sneak past a ranger out in the open at twilight (which lightly obscures normal vision). Her Stealth check result is 13.

2. Can the ranger see the rogue in the twilight? No, because his Perception check suffers disadvantage because the rogue is lightly obscured, giving a result of 10. OK, now we know the rogue is unseen, so whatever else happens, the rules for unseen attackers and targets apply (advantage and disadvantage on attack rolls).

3. Can the ranger hear the rogue? Yes, because the obscurement of twilight does not affect hearing, so there's no disadvantage applied to the Perception check, giving a result of 15. Now we know that the ranger is aware of the rogue (so can't be surprised), and knows her exact position so she is no longer hidden (by RAW). The rogue remains unseen.

You might think this approach makes hearing a more useful sense than sight, but that is to forget the power of the unseen attacker/ target mechanic. My approach does make it harder to be surprised than the approach of automatically applying disadvantage to Perception checks that rely on hearing, but it works for me.

PS I said 'by RAW' earlier because I think it's too strong that hearing - or smell for that matter - can pinpoint a creature's exact position to an unlimited range: they just don't have the precision of sight. At my table the limits are 30 and 10 feet respectively; beyond that you can determine a creature's general whereabouts but not its exact position (so you have to guess at which square on the grid to attack, for example).

EDITED to make clear I am illustrating by example

Springheel · Mar 20, 2017

ClearlyTough69 said:
1. Can the ranger see the rogue in the twilight? No, because his Perception check suffers disadvantage because of the light obscurement of twilight. OK, now we know the rogue is unseen, so whatever else happens the rules for unseen attackers and targets apply (advantage and disadvantage on attack rolls).

I don't quite follow...with light obscurement, the ranger has the ability to see the rogue, they just roll with disadvantage to do so. How did you establish that the rogue was unseen?

2. Can the ranger hear the rogue? Yes, because we ignore the obscurement caused by the twilight, so there's no disadvantage applied to the Perception check. Now we know that the ranger is aware of the rogue (so can't be surprised). By RAW, the ranger knows the rogue's exact position so she is no longer hidden, but the rogue remains unseen.

So even though the rogue is in dim light, you wouldn't apply disadvantage to the ranger's Perception check?

PS I said 'by RAW' earlier because I think it's too strong that hearing - or smell for that matter - can pinpoint a creature's exact position to an unlimited range: they just don't have the precision of sight. At my table the limits are 30 and 10 feet respectively; beyond that you can determine a creature's general whereabouts but not its exact position (so you have to guess at which square on the grid to attack, for example).

I agree that range should play a part in stealth.

ClearlyTough69 · Mar 21, 2017

Springheel said:
I don't quite follow...with light obscurement, the ranger has the ability to see the rogue, they just roll with disadvantage to do so. How did you establish that the rogue was unseen?

I was illustrating by example but it clearly didn't work. I have edited my post to make this explicit. So I do still call for a contested check. And I agree, the ranger might see the rogue, but actually didn't because his Perception check was too low.

Springheel said:
So even though the rogue is in dim light, you wouldn't apply disadvantage to the ranger's Perception check?

Here, I'm ignoring sight so dim light is irrelevant. I'm looking at hearing instead, which I'm treating as a standard Perception check, without disadvantage. This time the Perception check exceeds the Stealth check, so the ranger detects the rogue's position. However, the ranger still can't see the rogue, so the rules for seen/ unseen still apply, ie the rogue still has advantage on attack rolls against the ranger.

Springheel · Mar 23, 2017

ClearlyTough69 said:
Here, I'm ignoring sight so dim light is irrelevant. I'm looking at hearing instead, which I'm treating as a standard Perception check, without disadvantage. This time the Perception check exceeds the Stealth check, so the ranger detects the rogue's position. However, the ranger still can't see the rogue, so the rules for seen/ unseen still apply, ie the rogue still has advantage on attack rolls against the ranger.

I'm curious why you decided to ignore the dim light? It seems like the rogue lost an advantage that they should have had from being in the shadows. While at the same time, telling the ranger that he beat the rogue's stealth score, but still can't see him, strikes me as odd. Would you have run the scenario any differently if the rogue was in complete darkness (heavily obscured)?

ClearlyTough69 · Mar 23, 2017

[DELETED BY POSTER]

ClearlyTough69 · Mar 23, 2017

Springheel said:
I'm curious why you decided to ignore the dim light? It seems like the rogue lost an advantage that they should have had from being in the shadows. While at the same time, telling the ranger that he beat the rogue's stealth score, but still can't see him, strikes me as odd. Would you have run the scenario any differently if the rogue was in complete darkness (heavily obscured)?

I'll preface my reply by saying this is a house rule - one that I've seen nobody else use, but which for me feels entirely congruent with RAW, and expressly acknowledges all those places in the rules that talk of Perception checks relying on a particular sense (for example the rule you quoted up-thread for lightly obscured areas, or all the beast with Keen Hearing, Keen Sight, or Keen Smell).

I'm ignoring dim light because I've already established that the ranger's Perception check relying on sight was too low to see the rogue: he literally can't see her in the dim light. However, I then go on to assess the Perception check as one relying on hearing. The ranger can hear the rogue just as well whether she's in blazing sunlight or in pitch darkness. So there's no disadvantage to his Perception check relying on hearing, which is now good enough to beat the rogue's Stealth check. By RAW that means the ranger now knows the rogue's exact position. He still can't see her (remember those shadows), but he can target her with disadvantage (following the rules for unseen targets).

The only advantage the rogue has lost is that of surprise; she can still attack the ranger with advantage (because she's unseen by him), meaning she can still benefit from sneak attack too.

Had the rogue been in total darkness, I would not even have bothered with a Perception check relying on sight, because, as the rules say, you are effectively blinded when looking into a heavily obscured area (this was not in the first printing of the PHB, but has been clarified in errata since). So any chance of perceiving the rogue relies on hearing (and I guess other senses) and nowhere in the rules does it say that such a Perception check suffers disadvantage owing to obscurement.

epithet · Mar 24, 2017

ClearlyTough69 said:
I'll preface my reply by saying this is a house rule - one that I've seen nobody else use, but which for me feels entirely congruent with RAW, and expressly acknowledges all those places in the rules that talk of Perception checks relying on a particular sense (for example the rule you quoted up-thread for lightly obscured areas, or all the beast with Keen Hearing, Keen Sight, or Keen Smell).
...

Unless the ranger were actively searching for the rogue in this scenario (which would typically require spending an action to do so,) the rogue's dexterity (Stealth) roll would be opposed by the ranger's passive perception. As you point out, this gives disadvantage on perception based on sight, but since this is probably using passive perception it wouldn't require a roll. Disadvantage would drop the ranger's passive perception by 5, so if the rogue's stealth is less than 5 better than the disadvantaged passive score it would be reasonable to rule that the ranger detects the rogue but the rogue remains unseen. If the rogue is an assassin, he now hates you.

If the ranger is using an action to search for the rogue, then the check would regularly be made with disadvantage. Applying your rule, though, you would instead call for the ranger to make two perception checks (still rolling the same number of dice.) One success means he hears the rogue, two means he sees the rogue.

I like it. I think I'll start doing it this way.

ClearlyTough69 · Mar 25, 2017

epithet said:
Unless the ranger were actively searching for the rogue in this scenario (which would typically require spending an action to do so,) the rogue's dexterity (Stealth) roll would be opposed by the ranger's passive perception. As you point out, this gives disadvantage on perception based on sight, but since this is probably using passive perception it wouldn't require a roll. Disadvantage would drop the ranger's passive perception by 5, so if the rogue's stealth is less than 5 better than the disadvantaged passive score it would be reasonable to rule that the ranger detects the rogue but the rogue remains unseen. If the rogue is an assassin, he now hates you.

If the ranger is using an action to search for the rogue, then the check would regularly be made with disadvantage. Applying your rule, though, you would instead call for the ranger to make two perception checks (still rolling the same number of dice.) One success means he hears the rogue, two means he sees the rogue.

I like it. I think I'll start doing it this way.

Yes, you've described exactly what I do: for a passive Perception check, apply disadvantage (-5) to see the rogue, and straight passive Perception to hear her; for an active check, roll two dice and apply Perception modifier, and if one beats the rogue's Stealth score, she is heard, and if both beat the Stealth score, she is seen and heard.

I'll add to that: I think it's unreasonable for hearing to automatically pinpoint a creature's exact position beyond a range of 30 feet, so if you hear a creature that is further away, but can't see it, you only know its general whereabouts (in a 90 degree arc). If your Perception score beats the stealth check by a wide margin, you know its whereabouts with greater precision. I can provide the detail it anyone's interested.

I also have maximum ranges for different senses, with active Perception checks having greater range than passive ones.