5E A Reliable Talent for Expert Stealth

DMZ2112

Chaotic Looseleaf
I'm curious to hear from other D&D5 dungeon masters how you challenge high-level rogues in terms of going about undetected.

Shorthand: With Reliable Talent and Expertise, an 11th level rogue can easily have a Stealth check result range of 23-33, which is far beyond the capacity of any passive Perception in the Monster Manual and really beyond the capacity of most Perception checks -- if the dungeon master permits a roll -- representing a 'nearly impossible' DC /on average/.

To be clear, /negating/ this advantage would be easy. Denying the rogue a place to hide is not a solution. I'm specifically interested in hearing about scenarios where palatable, credible circumstances were engineered to challenge a rogue with a Stealth check result in the high 20s.
 

Ovinomancer

No flips for you!
I'm curious to hear from other D&D5 dungeon masters how you challenge high-level rogues in terms of going about undetected.

Shorthand: With Reliable Talent and Expertise, an 11th level rogue can easily have a Stealth check result range of 23-33, which is far beyond the capacity of any passive Perception in the Monster Manual and really beyond the capacity of most Perception checks -- if the dungeon master permits a roll -- representing a 'nearly impossible' DC /on average/.

To be clear, /negating/ this advantage would be easy. Denying the rogue a place to hide is not a solution. I'm specifically interested in hearing about scenarios where palatable, credible circumstances were engineered to challenge a rogue with a Stealth check result in the high 20s.
1) don't confuse DCs with contested rolls -- rolling a 33 is not a Nearly Impossible challenge, even if it's pretty much a de facto one.

2) You don't challenge a rogue with stealth challenges at this point, except on rare occasion and then well telegraphed. They are really, really good at sneaking. If you apply the stealth rules reasonably, this is just very awesome and not an "I win button." As you note, you have to have the right conditions to hide, so it's not an all-the-time thing or should even be assumed -- and I'm generous with hiding opportunities.

3) Why do you want to challenge to rogue straight at his skills? D&D in general encourages this (you go after the fighter with fighting, for example), but it's a bit of a dead end in challenge design. Let the rogue be awesome where they've invested their build choices. Instead of making stealthing the challenge, make the challenge one that stealth helps, but can't solve.


That said, rogue skill design is one of my biggest pet peeves about 5e. If you're going to do something as interesting as bounded accuracy, don't blow it up that badly and that quickly. Still, it's a minor problem -- you just look to challenge rogues in other ways than going straight at their skills.
 

jaelis

Explorer
You can arrange difficult circumstances and say they give a bonus to observer's perception checks. For instance, when trying to move quietly through dry leaves, you can give observers a +10 bonus to hear. With that mechanic, you can tune the difficulty as you like. It's not hard to come up with situational modifiers like that. Basically it gives the rogue a chance to hide in situations when an ordinary character would have no hope.
 

DMZ2112

Chaotic Looseleaf
1) don't confuse DCs with contested rolls -- rolling a 33 is not a Nearly Impossible challenge, even if it's pretty much a de facto one.
Maybe I'm not understanding you, but the situation seems /worse/ with contested rolls. Most monsters can't beat a 20, and the rogue will roll a 23 more than half the time and higher the other half.

2) You don't challenge a rogue with stealth challenges at this point, except on rare occasion and then well telegraphed. They are really, really good at sneaking. If you apply the stealth rules reasonably, this is just very awesome and not an "I win button." As you note, you have to have the right conditions to hide, so it's not an all-the-time thing or should even be assumed -- and I'm generous with hiding opportunities.
While I do on occasion design encounters in which hiding is unhelpful or impossible, generally speaking I come down the other way, on this: it /should/ be assumed. Denying the opportunity to hide has to be pretty carefully gauged, at least in combat. The rogue is clearly designed with the expectation that they will at least have the chance to get sneak attack every round. Usually this is a matter of keeping enemies engaged, but not always. Negating the rogue's combat potency is a kick in the fork to the player's fun.

Not having to roll is also not fun, in my experience, which is where I'm coming from on this. I'd just like the rogue to be /capable/ of failure under normal circumstances, even if that failure is a rare occurrance.

3) Why do you want to challenge to rogue straight at his skills?
I would have asked a different question if I had a problem challenging a rogue otherwise. :)

That said, rogue skill design is one of my biggest pet peeves about 5e. If you're going to do something as interesting as bounded accuracy, don't blow it up that badly and that quickly. Still, it's a minor problem -- you just look to challenge rogues in other ways than going straight at their skills.
This is where my head's at just now. The rogue just... ignores bounded accuracy, after a point.

You can arrange difficult circumstances and say they give a bonus to observer's perception checks. For instance, when trying to move quietly through dry leaves, you can give observers a +10 bonus to hear. With that mechanic, you can tune the difficulty as you like. It's not hard to come up with situational modifiers like that. Basically it gives the rogue a chance to hide in situations when an ordinary character would have no hope.
Those are some loud leaves, but I take your point. However, even with a massive +10 to the roll your average CR 11 monster has a check result range of 11 to 31 and is still /guaranteed/ to fail more than half the time -- regardless of the rogue's result. I'm terrible at probability math but I think the monster's actual odds are worse than 3:1 against.
 

jaelis

Explorer
Those are some loud leaves, but I take your point. However, even with a massive +10 to the roll your average CR 11 monster has a check result range of 11 to 31 and is still /guaranteed/ to fail more than half the time -- regardless of the rogue's result. I'm terrible at probability math but I think the monster's actual odds are worse than 3:1 against.
75% success against a garden-variety monster, under adverse conditions, seems pretty reasonable to me. If you want to have a particularly alert guard, you can give it proficiency, or even expertise, in perception.
 

BookBarbarian

Expert Long Rester
At high levels I rarely challenge the party with the same kind of challenges that I would at low levels.

At the point where reliable talent comes into play the test will not be "Can the rogue sneak in there?" but "What will they do when they get the secret-info/cursed-macguffin/item-that-is-way-bigger-than-expected/etc?
 

Sabathius42

Explorer
In the games I have ran, the "payment" the rogue pays for always being hidden is that the rest of the party gets beat down more than their fair share of the time. The party pays a price for a "never attacked because they are always hidden" character by having monsters instead attack wizards/sorcerers and other lightly defended characters when there is no rogue in their way. Not using the rogues ability of taking half damage once-per-round on attacks is leaving even more damage on the table to go towards other more squishier comrades.

If you have a second or third character in the party who leaves all the attacks to someone else (by flying/invisibility/etc) the damage on the characters left to sustain it can quickly grow out of control.

DS
 

Blue

Double sized Hobbit
I'm curious to hear from other D&D5 dungeon masters how you challenge high-level rogues in terms of going about undetected.

Shorthand: With Reliable Talent and Expertise, an 11th level rogue can easily have a Stealth check result range of 23-33, which is far beyond the capacity of any passive Perception in the Monster Manual and really beyond the capacity of most Perception checks -- if the dungeon master permits a roll -- representing a 'nearly impossible' DC /on average/.

To be clear, /negating/ this advantage would be easy. Denying the rogue a place to hide is not a solution. I'm specifically interested in hearing about scenarios where palatable, credible circumstances were engineered to challenge a rogue with a Stealth check result in the high 20s.
Hmm, let's see. A hero of our story has spent class features and a feat on becoming the absolute best at stealth that they can be. They now can do legendary feats of stealth that mundane creatures just won't notice and only the best prepared or most lucky will catch.

Woo, sounds good!
 

iserith

Magic Wordsmith
In a practical sense, this means that the rogue will almost always surprise monsters (unless he's traveling with other, less stealthy people) and will almost always have advantage on the attack roll if there's a place to hide in combat. If the rogue is on his or her own, it will also mean that scouting around without being detected will almost always succeed.

Personally, if a rogue tries to hide in the same place twice in order to get advantage on an attack roll, I impart disadvantage on the Stealth check. That does not increase the challenge (which remains static) - it just increases the difficulty which may still be quite low, comparatively speaking. If the rogue's player thinks there's a risk of failure, he or she will try to find another place to hide which may or may not be possible, depending on the situation.

Monsters with keen senses may be of some use here, as this effectively increases the difficulty, so that could be tacked onto some monsters to give them an edge. Monsters with blindsight or tremorsense or the like are also something to consider.

You could also implement something like the Mark of Finding from the Eberron supplement. The way that works is that the NPC/monster imprints the rogue in its memory (range 30 feet, sight), then the following benefit is gained: "When your quarry is within 60 feet of you, you have a sense of its location: it can't be hidden from you, gains no benefit from invisibility, and your attacks against it ignore half cover." So you stick some monsters in your adventure with that ability, appropriately renamed, telegraph that they exist and play on. The challenge for the rogue is identifying and taking those monsters out before they can imprint and ruin his or her ability to hide. Increase the difficulty by having it such that if one monster of this kind imprints, they all imprint until the original one is taken out (or something like that). Then have that one monster dodge and play keep-away while the others pound the rogue.
 

aco175

Adventurer
I tend to cheat a bit and make a few more monsters show up if the encounter is not as difficult as expected. I do this only sometimes to make the fight more fun and not be a wipe out each time things go not as expected from what was designed. I also tend to let the player have the PC with the cool trick like this. I'm not sure how to directly counter the stealth problem.

You could try an adventure with more of a direct stealth mission. This may mean the fighter needs to be helped or keeps tipping off the guards. You could have a counterstealth adventure like the sniper movies where a master stealth tracks him down to be the backstabee instead of being the backstabber.

In the end, you have a player who designed a PC to be something cool to him and you should let him play it mostly, while still challenging him once in a while.
 

Shiroiken

Adventurer
A player in my first 5E campaign did this. I simply looked at it and unless he was obviously visible to the monster (basically violating the rules for hiding), I pretty much just let him sneak around as much as he liked. The only time he was really challenged on it was when he was trying to hide in places where he might be accidentally visible, such as moving between columns. I think he was spotted maybe once between levels 11 and 18, and it didn't bother me, because he spent character resources to be that good and he never tried to abused the idea of being a super ninja.
 

jgsugden

Adventurer
The underlying thought here is that the DM needs a way to counterct the incredibly high stealth.

They do not.

If your PCs are nigh undetectedable due to insane stealth, celebrate that awesome ability. If they enter every combat stealthed, and hide during combat often, and are rarely, if ever, found... cheer. That is your party doing what they're supposed to do.

Now, that does not mean that an NPC that is aware of their abilities might not look for countermeasures... but as a DM, you do not need to worry about challenging that ability independent of that NPC natural story incentive.
 

Esker

Explorer
Yeah, a rogue that puts expertise in stealth and stays single classed long enough to get reliable talent has earned the ability to be undetected pretty much anywhere they can stay out of line of sight. Now if the player is finding it unfun to be able to do this at will too often, that's another story, and then the DM might want to try to come up with circumstances where there is a reasonable chance of failure, such as posting elite guards with, say, +7 in perception or so who are on high alert (and therefore making active checks). In that case they have about a 15% chance of spotting the rogue with a +13 in stealth and reliable talent. It's low enough that they're still pretty likely to feel awesome, but meaningfully high enough that it feels like a real risk to try.

Of course (as others have said), even if they can stay undetected, that's only part of the battle most of the time. If they're trying to pull off a heist, say, they need to be able to pick locks, perceive threats or objectives, deduce solutions to problems, maybe even disguise themselves for the times when hiding isn't available, and lie their way out of trouble. If they have expertise in stealth, perception, investigation, thieves' tools and deception, well then that sounds like an epic tailor-made challenge that they have a real chance of succeeding at, while a character without reliable talent probably has almost no chance of success simply due to the exponential failure rate inherent in a sequence of d20 rolls.
 

Ovinomancer

No flips for you!
Maybe I'm not understanding you, but the situation seems /worse/ with contested rolls. Most monsters can't beat a 20, and the rogue will roll a 23 more than half the time and higher the other half.



While I do on occasion design encounters in which hiding is unhelpful or impossible, generally speaking I come down the other way, on this: it /should/ be assumed. Denying the opportunity to hide has to be pretty carefully gauged, at least in combat. The rogue is clearly designed with the expectation that they will at least have the chance to get sneak attack every round. Usually this is a matter of keeping enemies engaged, but not always. Negating the rogue's combat potency is a kick in the fork to the player's fun.

Not having to roll is also not fun, in my experience, which is where I'm coming from on this. I'd just like the rogue to be /capable/ of failure under normal circumstances, even if that failure is a rare occurrance.



I would have asked a different question if I had a problem challenging a rogue otherwise. :)



This is where my head's at just now. The rogue just... ignores bounded accuracy, after a point.



Those are some loud leaves, but I take your point. However, even with a massive +10 to the roll your average CR 11 monster has a check result range of 11 to 31 and is still /guaranteed/ to fail more than half the time -- regardless of the rogue's result. I'm terrible at probability math but I think the monster's actual odds are worse than 3:1 against.
Seems you have a good handle on it and don't need my advice. Enjoy it!
 

77IM

Explorer!!!
Denying the rogue a place to hide is not a solution.
Would it be OK to make it difficult to get into, or stay in, the places to hide?

I think we've all seen a movie where the hero detects enemies approaching. The hero looks around and there's no hiding place! The enemies come around the corner, and the hallway is strangely empty, and they just walk on by. Then the camera pans up, and the hero is on the ceiling, holding themselves aloft by pressing hands and feet against opposite walls. (Presumably some kind of Athletics or Acrobatics check -- not that I want to restart that tired old debate.)

OR the classic: the hero is in a maze of shipping containers down by the docks. Enemies with submachine guns are searching for him or her. But the hero picks them off one by one! (The challenge here is for the rogue to detect the enemies using Perception; if they fail to do so, the enemy comes around a corner and just spots them.)

OR how about: The hero needs to cross open ground, with scant cover. Can they dive and roll behind various crates and barrels? Wait until the guards' backs are turned or the spotlight is pointed elsewhere? Time their movement to coincide with a passing cart or passing shadow? (Even if you consider these tasks to be Stealth-based, depending on the circumstances the DC might be incredibly high. Like, an ordinary person might have no chance of hiding behind a tree that narrow, but our ninja PC can hit the required DC 30... usually.)

OR the hero can create a distraction! (Possibly some kind of Deception or Sleight-of-Hand trick to make the guards actually abandon their post and check out that sound.) How about some teamwork -- another character distracts the guards so that the hero can slip in behind them!

OR difficult terrain can be an issue. Maybe the only way to stay behind cover involves a difficult jump, a walk along a balance beam, or squeezing through a hole? (Athletics, Acrobatics, maybe thieves' tools if a door is locked.)

OR what if some other thing is making sound? The stolen MacGuffin might be noisy and clanky, or be an incompetent NPC. We know the hero can sneak up behind that guard, but can he or she kill the guard silently? What if the guard tries to shout? (Maybe you use grappling rules for this. Maybe slitting a throat silently is more Sleight of Hand than Stealth.



-----

Now, if all you are talking about is combat, then it's fine for the rogue to hide every single round and get Sneak Attack every round. That's kind of by design. In some environments the enemies can move to flank and get line-of-sight on the rogue, but usually the PC can just Cunning Action Dash their way into a better hiding spot.

The only tricky issue here is Blindsight, and whether or not mundane Stealth skill works on it. Generally you can't Hide from someone who can "see" you unless you are "obscured," so I think it's fair to rule that a creature which "sees" by hearing, temperature, vibrations, scent, etc. isn't going to be fooled by shadows, foliage, a conveniently-placed barrel, etc. The bigger question is, what counts as "obscurement" to Blindsight? Can Blindsight see right through solid objects? A wall may block light but not sound, for example. (Recall the scenes depicting Daredevil's senses in season 1 of that show, where he can pinpoint people blocks away.) Would white noise block sound-based Blindsight? Would wind block scent-based Blindsight? Would soft, absorbent clothing block Echolocation? How does ooze Blindsight even work? I have no answers, but I can tell you from experience that every time the rogue gets spotted by Blindsight, that their player is going to be asking these questions.
 

Xeviat

Explorer
The high-level rogue doesn't have to sneak through a dark dungeon past a sleepy guard. The high-level rogue has to sneak past a dragon, in broad daylight, at a full tilt run.

I was actually just thinking about starting a post about whether or not expertise ruins the game. Higher numbers just become auto success, without pushing the DCs past the point that non-specialists can't make it. Other systems have features that let you take a penalty to your check, or voluntarily raise the difficulty to do something they couldn't otherwise do before. Taking a -5 penalty to stealth at full speed, or a -10 penalty to stealth while dashing and moving. A penalty to stealth in light obscurement/dim light. An even higher penalty to stealth in bright light. This would push the specialist to have to try hard, don't put them up against things a non-specialist can achieve.
 

Fenris-77

Explorer
The high-level rogue doesn't have to sneak through a dark dungeon past a sleepy guard. The high-level rogue has to sneak past a dragon, in broad daylight, at a full tilt run.

I was actually just thinking about starting a post about whether or not expertise ruins the game. Higher numbers just become auto success, without pushing the DCs past the point that non-specialists can't make it. Other systems have features that let you take a penalty to your check, or voluntarily raise the difficulty to do something they couldn't otherwise do before. Taking a -5 penalty to stealth at full speed, or a -10 penalty to stealth while dashing and moving. A penalty to stealth in light obscurement/dim light. An even higher penalty to stealth in bright light. This would push the specialist to have to try hard, don't put them up against things a non-specialist can achieve.
I think 5E could really benefit from a little more thinking like this. Expertise in particular suffers from being a really effective but boring high level ability. Adding effect in exchange for penalties to the roll is a great way to add some flash back in. That said, I also think 5E could use a mechanism for partial success too...
 

Xeviat

Explorer
A player in my first 5E campaign did this. I simply looked at it and unless he was obviously visible to the monster (basically violating the rules for hiding), I pretty much just let him sneak around as much as he liked. The only time he was really challenged on it was when he was trying to hide in places where he might be accidentally visible, such as moving between columns. I think he was spotted maybe once between levels 11 and 18, and it didn't bother me, because he spent character resources to be that good and he never tried to abused the idea of being a super ninja.
A PC with 20 Dex, Proficiency/Expertise in Stealth, the Skulker feat, and Reliable Talent is Batman. They get to do Batman things. Batman can still be challenged. You're right, let them be Batman. That's what they wanted to make.

The wizard can go invisible and use silence to be super stealthy. Let your Rogue vanish when someone looks away from them, that's cool.
 

77IM

Explorer!!!
I was actually just thinking about starting a post about whether or not expertise ruins the game.
You should create this thread. I hate Expertise, and while I would not go so far as to say it "ruins the game," it's definitely a design mistake.

My preferred fix would require D&D to have some sort of formal degree-of-success mechanism for non-combat rolls. As an analogy, fighter with a greatsword is doing substantially more damage than a cleric with a mace even if the fighter's attack bonus is only slightly higher. I'd prefer have the rogue with Expertise in Stealth get a modest bonus to their check (maybe a flat +2) but somehow their degree of success is greater. It's similar to your idea of -5 penalty to hide while moving full speed, except I'd just let the expert do that without a trade-off (because their bonus is not so ridiculously high). But, the default assumption of ability checks in 5E is a binary result (success or fail), with "Succeed by 5 or more" stuff being very special-case. I'm not sure how to generalize such degrees-of-success.
 

Xeviat

Explorer
You should create this thread. I hate Expertise, and while I would not go so far as to say it "ruins the game," it's definitely a design mistake.

My preferred fix would require D&D to have some sort of formal degree-of-success mechanism for non-combat rolls. As an analogy, fighter with a greatsword is doing substantially more damage than a cleric with a mace even if the fighter's attack bonus is only slightly higher. I'd prefer have the rogue with Expertise in Stealth get a modest bonus to their check (maybe a flat +2) but somehow their degree of success is greater. It's similar to your idea of -5 penalty to hide while moving full speed, except I'd just let the expert do that without a trade-off (because their bonus is not so ridiculously high). But, the default assumption of ability checks in 5E is a binary result (success or fail), with "Succeed by 5 or more" stuff being very special-case. I'm not sure how to generalize such degrees-of-success.
Alright! I'll figure out something tomorrow.

I've been working on an expanded skill system, and Expertise kind of throws a wrench in things. Expertise is needed to reach the vaunted "Impossible DC 30" with any reliability, but it also puts skills outside the reach of the saving throw system. Thus, you can't have skill actions like "demoralize" or "feint" as uses of intimidation or deception, as skills are on a different track. Look at how bards and rogues make the best grapplers ... it's just weird.
 

Advertisement

Top