Surprise round question

[Ovinomancer]

Well, yeah, but what I specifically meant to ask was how you, personally, use your discretion as a DM to determine under what conditions to call for a Perception check from a sleeping creature, and under what conditions the answer to the question of whether a sleeping creature does or doesn't notice some element of the environment is certain.

An example was given up-thread of a situation in which it was uncertain whether a sleeping person would wake up when another person snuck into the room and removed a book from a tall stack of books in order to borrow it. I said I might resolve the removal of the book by calling for a DEX (Sleight of Hand) check with a DC set depending on how difficult conditions make it to complete the task without making some sort of comotion that wakes the sleeping person. I would only call for such a check provided there's a meaningful consequence for failing the check, however, e.g. the sleeping person might not allow the book to be borrowed if awakened, and the book would have to be acquired in some other way. A similar approach could be used with a DEX (Stealth) check to see if the borrower's movement through the room awakens the sleeper, but again a meaningful consequence would need to be in play.

What I wouldn't do is contest either of those rolls with a WIS (Perception) check from the sleeping person because an ability check resolves the outcome of some effort on the part of the creature making the check. A sleeping creature is making no effort to notice elements of its environment, so in my book it shouldn't be asked for a check to resolve such an effort.



I agree, and as a DM, I like to keep things simple for myself. The game gives me two pieces of evidence relating to how aware sleeping creatures are of their environment. First, as I've already mentioned, is the Elven trait trance which strongly implies that non-elves are less conscious than semiconscious when they sleep, a state that I would regard as unconscious, and for which I would impose the Unconscious condition, which specifies that an unconscious creature is unaware of its surroundings.

Second is the sleep spell. One of its effects is to impose the Unconscious condition. Now, one might make the argument that the condition is imposed only in the case of magical sleep, but I would think that what is particularly magical about the spell isn't the quality, or depth, of the sleep imposed, but that it puts you to sleep when you otherwise wouldn't have been, and that its effects resemble mundane sleep in most, if not all, respects, the main difference being that it can impose sleep during combat, which seems to be intended.



Yes, I'd throw perception in there too. An unconscious creature is unaware of its surroundings, and a creature's WIS (Perception) check measures its "general awareness" of its surroundings. So if the creature is unconscious there's no need to check its perception with respect to its surroundings. We already know what the outcome is.



It's a bit odd to call part of the game's rules text a "ruling", and I'm not trying to get into a debate about what the rules should be, but you are of course free to rule these things as you wish at your table regardless of what the actual rule-book says and in accordance with what makes sense to you. The way it all makes sense to me straight out of the box is that a creature only has one awareness with which to perceive its environment, just like it has only one WIS score. It can choose to occupy that awareness in a variety of ways, but once occupied for an exploration round, or whatever interval of time we're concerned with, it cannot apply its awareness to any other task that requires it. So creatures who have occupied their awareness with exploration tasks other than keeping watch for hidden threats (which are often resolved with a WIS (Survival) check, btw) don't have the option to also use their awareness to keep watch for hidden threats.

A trancing elf, on the other hand, by virtue of remaining semiconscious, has a partial awareness that he or she can apply to keeping watch, albeit at disadvantage, provided the elf's partial awareness is not otherwise occupied. For example, the DM might allow elves to trance while walking and performing other sorts of exploration tasks as well, if need be, because the meditative trance only takes up part of the elves' awareness. It's one of the things that makes elves special.

But what ends the unconscious condition? The rolling of initiative? Surely it's not the loudness of combat, as they're unaware of such things in their surroundings. If the sleep spell imposes normal sleep, then it surely takes an action to awaken from sleep if attack, yes?

In reality, how you play your game is your business, and I certainly wish you all the funs. But there's a serious logical disconnect in your presentation here.

 

[Maxperson]

Well, yeah, but what I specifically meant to ask was how you, personally, use your discretion as a DM to determine under what conditions to call for a Perception check from a sleeping creature, and under what conditions the answer to the question of whether a sleeping creature does or doesn't notice some element of the environment is certain.

An example was given up-thread of a situation in which it was uncertain whether a sleeping person would wake up when another person snuck into the room and removed a book from a tall stack of books in order to borrow it. I said I might resolve the removal of the book by calling for a DEX (Sleight of Hand) check with a DC set depending on how difficult conditions make it to complete the task without making some sort of comotion that wakes the sleeping person. I would only call for such a check provided there's a meaningful consequence for failing the check, however, e.g. the sleeping person might not allow the book to be borrowed if awakened, and the book would have to be acquired in some other way. A similar approach could be used with a DEX (Stealth) check to see if the borrower's movement through the room awakens the sleeper, but again a meaningful consequence would need to be in play.

What I wouldn't do is contest either of those rolls with a WIS (Perception) check from the sleeping person because an ability check resolves the outcome of some effort on the part of the creature making the check. A sleeping creature is making no effort to notice elements of its environment, so in my book it shouldn't be asked for a check to resolve such an effort.

I don't have a codified system. However, like you above, if someone were to sneak into a library where a PC was sleeping to take a book, I'd roll stealth to see how quiet the person was, with a dex check like you mentioned, and if an appropriate level of noise was made, I'd check to see if the PC was awakened. The sleeping passive perception would likely be used, but for very important rolls, I'd probably use an opposed check with the PC being penalized appropriately. That way the player would be more involved with the important roll.

I'd check whether it was important or not, though. More than once I've thought something was or was not important to a player/PC and was wrong. Also, even when not important, the roleplaying that comes from the encounter is usually very good and worth it to the group to have the check happen.

I agree, and as a DM, I like to keep things simple for myself. The game gives me two pieces of evidence relating to how aware sleeping creatures are of their environment. First, as I've already mentioned, is the Elven trait trance which strongly implies that non-elves are less conscious than semiconscious when they sleep, a state that I would regard as unconscious, and for which I would impose the Unconscious condition, which specifies that an unconscious creature is unaware of its surroundings.

Second is the sleep spell. One of its effects is to impose the Unconscious condition. Now, one might make the argument that the condition is imposed only in the case of magical sleep, but I would think that what is particularly magical about the spell isn't the quality, or depth, of the sleep imposed, but that it puts you to sleep when you otherwise wouldn't have been, and that its effects resemble mundane sleep in most, if not all, respects, the main difference being that it can impose sleep during combat, which seems to be intended.'

I disagree here to a degree. First, unconscious which is what happens when you are knocked out is much more severe than sleep. You can't make a noise or nudge someone out of unconsciousness. You can do that with sleep. That means to me that there are levels of unconsciousness within the game, even if the unconscious rules don't spell that out.

Yes, I'd throw perception in there too. An unconscious creature is unaware of its surroundings, and a creature's WIS (Perception) check measures its "general awareness" of its surroundings. So if the creature is unconscious there's no need to check its perception with respect to its surroundings. We already know what the outcome is.

Sleeping creatures have a level of awareness, though. Many things in the real world make it into dreams through that awareness. Others wake the sleeper up due to that awareness. If you are truly unaware of something, it cannot wake you up.

It's a bit odd to call part of the game's rules text a "ruling", and I'm not trying to get into a debate about what the rules should be, but you are of course free to rule these things as you wish at your table regardless of what the actual rule-book says and in accordance with what makes sense to you. The way it all makes sense to me straight out of the box is that a creature only has one awareness with which to perceive its environment, just like it has only one WIS score. It can choose to occupy that awareness in a variety of ways, but once occupied for an exploration round, or whatever interval of time we're concerned with, it cannot apply its awareness to any other task that requires it. So creatures who have occupied their awareness with exploration tasks other than keeping watch for hidden threats (which are often resolved with a WIS (Survival) check, btw) don't have the option to also use their awareness to keep watch for hidden threats.

All a rule set is, is an official compilation of rulings and mechanics. 5e has official mechanics for perception. The rules regarding tracking, etc. are clearly not mechanics, so they are just official rulings that tie into those official perception mechanics. I view the official rulings as wrong for the reasons stated and won't use them.

 

[Arial Black]

An example was given up-thread of a situation in which it was uncertain whether a sleeping person would wake up when another person snuck into the room and removed a book from a tall stack of books in order to borrow it. I said I might resolve the removal of the book by calling for a DEX (Sleight of Hand) check with a DC set depending on how difficult conditions make it to complete the task without making some sort of comotion that wakes the sleeping person. I would only call for such a check provided there's a meaningful consequence for failing the check, however, e.g. the sleeping person might not allow the book to be borrowed if awakened, and the book would have to be acquired in some other way. A similar approach could be used with a DEX (Stealth) check to see if the borrower's movement through the room awakens the sleeper, but again a meaningful consequence would need to be in play.

What I wouldn't do is contest either of those rolls with a WIS (Perception) check from the sleeping person because an ability check resolves the outcome of some effort on the part of the creature making the check. A sleeping creature is making no effort to notice elements of its environment, so in my book it shouldn't be asked for a check to resolve such an effort.

First (and once again), you don't need to 'use an action' to hear things! Neither in-game (you just roll Perception, no Action required) or in real life (you don't need to consciously make an 'effort' in order to hear or see things; your senses are passive).

Second, the logical disconnect. In the book-swiping example, in the question of whether or not the small sound of the book being removed from the pile and whether that wakes the sleeper or not, how can it fail to involve the senses of the sleeper?

If the sleeper wakes then it is because the sleeper sensed/heard the book sliding out of the stack! If the sleeper was deaf then it could not have woken him. In-game, deaf creatures auto-fail Perception checks involving hearing.

The way you run it, how good the hearing of the sleeper has no impact. This is the logical disconnect!

Somehow, you acknowledge that loud sounds auto awaken sleepers while at the same time denying the fact that sleepers can hear anything on the grounds that they cannot make a conscious choice to hear anything!

 

[Hriston]

How? What, in world, wakes them? The rattle of Initiative dice being rolled?

There must be some perceived stimulous that wakes them. 'The DM deciding to arrange the action using combat rounds and initiative' is a meta-game construct that cannot be known or sensed by the creatures in the world.

I imagine your players often find themselves wondering why they're rolling initiative and then, if they win, wasting their turn wondering what to do because nothing has happened yet. Is that an accurate assessment of the type of thing that goes on in your games in the event of a surprise attack?

When combat begins in my games, I describe some element of the action or movement that triggers combat to let the players know combat is beginning. Usually it's something that's reflected in the system like, "The Orc swings his club at you." But it could just as easily be color entirely of the DM's invention. For example, going back to the situation where a hidden assassin begins combat by attacking with a dagger from melee range, it was discussed that the sound of the dagger being drawn from its sheath might alert the target to the imminent attack, but alternatively it might be that the blade catches a beam of moonlight as the assassin raises it to strike, and it flashes momentarily into the target's eyes. Either one could be said to alert the target to the attack, and both work just as well whether the target is awake or asleep.

You are objectively wrong on this point.

In 5E, 'surprised' is not a mere natural language synonym for 'mildly astonished'. It is a jargon word for a specific rules effect that makes you unable to act or react straight away. It in no way implies or requires the victim to 'know' they are under attack! Sleeping creatures are not immune to surprise, especially on the grounds that they don't know they are being attacked! Nor does the fact that someone is about to attack them cause them to automatically wake up and be aware of it.

Not "about to attack them", they are being attacked in the very moment they awaken.

Also, if it isn't the attack that surprises them, what in-world stimulus causes them to be surprised? Maybe they're surprised by the DM calling for initiative?

Because in this situation the assassin is definitely not shouting a battle cry, beating spears on shields or doing anything else noisy. He is specifically and deliberately staying quiet, and you are ruling that the target automatically wakes up before any noise is made, on the frankly ludicrous grounds that 'combat is usually noisy'!

Not "before", but at the same time! And it isn't necessarily noise. It could be a beam of moonlight glinting off the exposed blade and shining in the target's eyes that wakes him up.

You are ruling that a meta-game cause of 'DM decides to roll initiative' automatically wakes in-world creatures.

Rubbish! Combat doesn't start because the DM decides to roll initiative. Combat starts because one of the players at the table (including the DM) decides to have their character attack, leading to the in-world event of the character actually making an attack.

You don't need to make an 'effort' in order for your senses to provide information.

I agree. Furthermore, if it doesn't require an effort, it shouldn't require an ability check. The information should simply be provided if it's relevant.

You can deliberately stop doing other things in order to concentrate on looking for something or hearing something (in game terms using your action to make a Wisdom(perception) check), but you can still see and hear, still perceive things, even without deliberately choosing to. In game terms, the DM says, "You open the door, you see a well-appointed drawing room. Make a Perception check". You don't need to spend an Action to make a Perception check, because your senses may be sharp enough to detect the thug with the club standing behind the door.

You don't need to "spend an action" to do anything outside of combat. All you need to do is tell the DM if you're engaging your awareness in some other exploration task or it's assumed you're keeping watch for hidden threats. The DM in your example is making such an assumption. Personally, however, I wouldn't ask for a roll and would just check your passive Perception instead.

Sure. At my ear, the traffic noise was louder than the noise of the broom on the kitchen ceiling. But the broom woke me and the traffic didn't.

I think you're comparing apples and oranges here. The constant white noise of the traffic was something you could tune out and sleep through, but if it was actually louder in your room than the thumping noise coming from your floor then it would have drowned it out. Low frequency vibrations carry very well through solid objects like your floor and, depending on how close to the floor you were sleeping, may have carried through your body as well. All in all, this situation is too complicated to be useful for discussion and is in need of determination from the DM as to what factors were responsible for you not being able to remain asleep, as I assume you were trying to do.

This is surreal. You don't have to 'initiate' hearing something.

My point exactly! If you don't have to "try" to do something, then it doesn't call for an ability check! That way, a WIS (Perception) check to notice something by hearing resolves whether you notice that thing when you otherwise wouldn't have if you hadn't tried. If you can notice it without trying, then you just do. Why roll?

I think you understand 'perception' to mean something different than the rest of us. If your senses pick it up, you 'perceive' it. It may, or may not, wake you up.

I can only speak for myself, but I think it's obvious I'm using "perception" here to mean what it does in the game, that is that aspect of WIS that allows you to notice and detect something that is otherwise hidden. If your senses just "pick it up" involuntarily, then it isn't hidden, and you just notice it. You don't have to make a WIS (Perception) check to hear what anyone present can hear. You just hear it, no check required.

Also, I don't see why being good at noticing things would make you a light sleeper.

Oh, you can be unfair to NPCs too!

So what you consider "fair" is that your own preconceived expectation of the currency between skill selection at char gen and agency over the in-game situation is respected at all times. Sorry, but in my games, your ability to min-max is not guaranteed.

'Unrealistic' is the idea that sleepers choose to wake up at all!

Well, what would you call it if someone notices something in their sleep and wakes up as a result? Why does noticing something wake you up?

By using my wisdom and experience of living life to understand the types of things that might or might not awaken a sleeper, as opposed to things that automatically would (explosion nearby) or would not (an ant breaths in).

I do the same. I simply prefer to narrow the range of uncertainty down to nil because, as you say, sleeping creatures choosing to wake up seems "unrealistic" and, for me, cheapens the use of die rolls, so I simply inform the players when their characters wake up.

Again, what in-world stimulous wakes them?

Do you consider it within the DM's purview to imagine a flash of moonlight that awakens a sleeping creature and invite the players to imagine it too? Or is that somehow overstepping the DM's role?

 

[Maxperson]

When combat begins in my games, I describe some element of the action or movement that triggers combat to let the players know combat is beginning. Usually it's something that's reflected in the system like, "The Orc swings his club at you." But it could just as easily be color entirely of the DM's invention. For example, going back to the situation where a hidden assassin begins combat by attacking with a dagger from melee range, it was discussed that the sound of the dagger being drawn from its sheath might alert the target to the imminent attack, but alternatively it might be that the blade catches a beam of moonlight as the assassin raises it to strike, and it flashes momentarily into the target's eyes. Either one could be said to alert the target to the attack, and both work just as well whether the target is awake or asleep.

Why are PCs/NPCs in your game so incompetent that they are literally unable to attack an unaware target? Why do they have to flub it so that the target sees/senses them somehow before the attack hits? It makes no sense that fate makes it impossible for an assassin to do his job properly by always causing moonlight to glint off of a blade, him to make noise, or whatever in order to alert the target. It's really okay for the blade coming out of the PC's chest to be the first clue that an attack happened and surprise him.

 

[Arial Black]

I imagine your players often find themselves wondering why they're rolling initiative and then, if they win, wasting their turn wondering what to do because nothing has happened yet. Is that an accurate assessment of the type of thing that goes on in your games in the event of a surprise attack?

There is a difference between 'player' and 'character'. The 'player' knows that the DM told them to roll initiative, and it may make the 'player' wonder why. But the 'character' doesn't know that initiative has been rolled because the 'character' doesn't realise that he is an avatar in a 5E game of D&D. Whether or not the 'character' realises he is in combat, or that combat is imminent, depends on that 'character' being aware of clues in his environment, and that is determined in-game by the 'players' making die rolls and using abilities.

When combat begins in my games, I describe some element of the action or movement that triggers combat to let the players know combat is beginning. Usually it's something that's reflected in the system like, "The Orc swings his club at you." But it could just as easily be color entirely of the DM's invention. For example, going back to the situation where a hidden assassin begins combat by attacking with a dagger from melee range, it was discussed that the sound of the dagger being drawn from its sheath might alert the target to the imminent attack, but alternatively it might be that the blade catches a beam of moonlight as the assassin raises it to strike, and it flashes momentarily into the target's eyes. Either one could be said to alert the target to the attack, and both work just as well whether the target is awake or asleep.

Wait, you mean that the creature.....perceives....the clues that combat is imminent? Hears the blade being drawn or sees the shaft of moonlight on his closed eyelids?

That sounds a lot like Perception to me! If only there was something in the game that simulated Perception, like a skill or something!

Here, you freely admit that sleeping creatures can perceive sensory stimulous, AND admit that sleepers who sense such stimulous awaken. Just like we've been saying all along and just like you've been denying whenever you say that sleepers are not aware while saying that noises wake sleepers.

The problem, the logical disconnect, is that you have the assassin automatically fail! No matter how stealthy, he will always give the game away by whatever means you make up on the fly: sound of blade being drawn or shaft of moonlight or whatever. It's as if the world somehow conspires to wake the sleeper.

Also, if it isn't the attack that surprises them, what in-world stimulus causes them to be surprised? Maybe they're surprised by the DM calling for initiative?

Actually, it's the lack of stimulous, or the sleeper failing to react to the stimulous quickly enough, that causes 'surprise'.

Rubbish! Combat doesn't start because the DM decides to roll initiative. Combat starts because one of the players at the table (including the DM) decides to have their character attack, leading to the in-world event of the character actually making an attack.

Agreed. But it is the in-game events that may (or may not) be perceived by the creature, not 'initiative being rolled' that wakens the creature or 'combat time starts'.

If it doesn't require an effort, it shouldn't require an ability check. The information should simply be provided if it's relevant.

No.

Although we may judge that a particular sight or sound is automatically sensed, or automatically not be sensed, by the creatures in question, there remains a huge swathe of things that may or may not be sensed depending on how good the creature's senses are (Perception modifier) and on unpredictable chance (the d20).

Although we may judge that something a stealthy assassin does cannot be perceived by a sleeper or that will automatically wake a sleeper, there remains a huge swathe of things that the assassin may do that may or may not awaken the sleeper. More skillful and experienced assassins are less likely to give themselves away in this fashion than less skillful or experienced ones. This is reflected in Stealth skill modifiers and other special abilities.

This being a game, where the outcome is not certain (if it was it would be 'Magical Story Time' instead of a Role-Playing Game), the exciting bits are resolved through the dice, modified by various abilities, in such a way that being better at something makes you more likely to succeed. This helps us create characters with different strengths and weaknesses.

Those wanting to create assassins should be rewarded for being good at Stealth (by being more likely to succeed at Stealth than if they hadn't bothered being proficient in Stealth), but your way means that no matter what they do or how skilled they are then the sleeper automatically wakes.

You do the same with Perception. You avoid having the sleeper make Perception checks (on the grounds that they are asleep and cannot 'choose' to hear!) and have them wake up anyway, no matter how poor their Perception.

You don't need to "spend an action" to do anything outside of combat. All you need to do is tell the DM if you're engaging your awareness in some other exploration task or it's assumed you're keeping watch for hidden threats. The DM in your example is making such an assumption. Personally, however, I wouldn't ask for a roll and would just check your passive Perception instead.

You don't need to tell the DM that you are walking around with your eyes open! You don't need to state in advance that you can hear noises.

I've no problem with using passive Perception BTW; it keeps your investment in the Perception skill relevant.

My point exactly! If you don't have to "try" to do something, then it doesn't call for an ability check! That way, a WIS (Perception) check to notice something by hearing resolves whether you notice that thing when you otherwise wouldn't have if you hadn't tried. If you can notice it without trying, then you just do. Why roll?

Because that's how senses work! In a group of real people there may be sounds that none of them hear (ant breathes in), sounds that all of them hear (explosion), and sounds that some of them hear but some don't (leopard concealed in the trees treads on some dry leaves). How do we work out which people hear the threat and which don't? Well, those with better Perception are more likely to hear the threat than those who are not so Perceptive.

I can only speak for myself, but I think it's obvious I'm using "perception" here to mean what it does in the game, that is that aspect of WIS that allows you to notice and detect something that is otherwise hidden. If your senses just "pick it up" involuntarily, then it isn't hidden, and you just notice it. You don't have to make a WIS (Perception) check to hear what anyone present can hear. You just hear it, no check required.

Some things are hidden in regard to some creatures while not being hidden to others, and one possible reason is that some creatures managed to Perceive a threat while others didn't because of a combination of chance and keen senses. This is what the check is for, and the game engine is built for this. If it did not expect us to roll for this stuff then it wouldn't bother with Stealth or Perception skills!

Also, I don't see why being good at noticing things would make you a light sleeper.

It's called Survival of the Fittest. The creatures that didn't hear the approaching predator and wake up in time were much less likely to survive to have such dozy offspring.

So what you consider "fair" is that your own preconceived expectation of the currency between skill selection at char gen and agency over the in-game situation is respected at all times. Sorry, but in my games, your ability to min-max is not guaranteed.

I'm using 'fair' in the sense that creatures have a chance to be stealthy and that being more stealthy means being more likely to remain undetected. And that creatures that are more Perceptive have a better chance of perceiving something than those that are less perceptive. It's about the logic of the situation, not about character building.

Well, what would you call it if someone notices something in their sleep and wakes up as a result? Why does noticing something wake you up?

It's a survival mechanism. All the creatures who can't be woken like this have already been eaten!

I do the same. I simply prefer to narrow the range of uncertainty down to nil because, as you say, sleeping creatures choosing to wake up seems "unrealistic" and, for me, cheapens the use of die rolls, so I simply inform the players when their characters wake up.

Your certainty in every single case is what is unrealistic. Especially in a game, where such uncertainty combined with skill is the heart of a game where situations are resolved through rolling dice!

Do you consider it within the DM's purview to imagine a flash of moonlight that awakens a sleeping creature and invite the players to imagine it too? Or is that somehow overstepping the DM's role?

I would find it absurd if a skilled assassin stealthily drawing his blade was the direct cause of a flash of moonlight that automatically woke every creature within a 30 foot radius!

 

[Hriston]

But what ends the unconscious condition?

Some element of the action (or movement) that begins combat wakes them up. Once they are no longer asleep, they no longer have the condition.

Surely it's not the loudness of combat, as they're unaware of such things in their surroundings.

This line of reasoning runs counter to my experience of sleep and awareness. When I'm asleep, although my eyes are nearly always shut, my ears do not close up. They pick up any noise around me just as if I had been awake and send information to my brain about those noises. But I'm aware of none of this. If you asked me later what sort of noises I heard during the night I wouldn't be able to tell you because I would have no recollection of having heard them. My subconscious brain is keeping track, however, and if it picks up anything it considers alarming, it wakes me up, and only then do I become aware of what's going on around me.

If the sleep spell imposes normal sleep, then it surely takes an action to awaken from sleep if attack, yes?

The spell explicitly imposes magical sleep. Combat has already started when creatures are affected, however, because the casting of the spell itself triggers combat if it hasn't started already. I would attribute that the sleep can only be ended for the spell's duration by the enchanted creatures taking damage or being shaken or slapped awake to the sleep's magical nature, but I would certainly allow either of those events to awaken those affected by mundane sleep as well.

 

[Arial Black]

Here's how we handle it:-

Assassin sneaks into room where the sleeper is in bed (Stealth check versus passive Perception, -5 for disadvantage).

Assuming success, the assassin moves to melee range, draws his knife and strikes!

This counts as combat(!) so the rules are applied: determine surprise, roll initiative. Both the assassin and the sleeper roll initiative, and we know that the assassin is not surprised.

Is the sleeper surprised? Probably; it's a skilled assassin after all, and he's caught his target sleeping.

Roll an opposed Stealth/Perception check, with disadvantage on the Perception because of disadvantage for being asleep.

Why not passive Perception?

"Such a check can represent the average result for a task done repeatedly, such as searching for secret doors over and over again"

...which doesn't apply here...

"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"

...because rolling a check may give the game away, but the assassin is going to reveal himself in a nanosecond anyway by way of stabbing the sleeper.

Given that the assassin is almost certainly going to be skilled in Stealth, the sleeper (proficient or not), disadvantaged by sleep, is probably going to suffer the game effects of surprise.

But, does the sleeper wake up? If so, why?

There must be some in-game reason to wake up. Meta-game reasons (we've just rolled initiative) cannot be the reason.

Don't worry, a legitimate reason is coming toward your heart right about now!

Without any sleep being involved, the combat would play out as the attacker striking first because even if the sleeper rolled higher initiative he cannot move or act on his first turn. With the sleep involved, as soon as the attack hits and does damage (almost certainly a crit due to the assassin ability) this is sufficient sensory input to awaken a creature from normal sleep. Although he might be dying.

However, the sleeper had a fair chance. He had a roll, it's just that he failed.

But he did not automatically fail! He might have succeeded, against the odds. He might have an extraordinarily high Perception modifier, advantage to cancel the disadvantage, even a Weapon of Warning. It would be wrong to dismiss those hard-won advantages by ruling that the sleeper cannot roll, just as it would be wrong to have the sleeper automatically awake and render all of the assassin's training pointless.

So what if the sleeper's Perception check beats the assassin's Stealth, as unlikely as that seems?

He wakes up! Something woke him, and the DM can narrate the fluff of it how he likes. It may very well be the sound of the knife being drawn, or a shaft of moonlight reflecting from the blade, or the shadow passing over the sleeping eyelids, or the sudden lack of night-time noises caused by the assassin's silence spell, or the scent of the assassin's oiled beard of villainy. The fluff can be whatever sounds reasonable, but the crunch is simply that he woke up because he won the opposed Stealth/Perception check.

The game is literally made so that dramatic situations are resolved by rolling dice, modified by the abilities of the creatures involved. It goes against both the game and against the point of playing it to just decide what happens without any rolls; at that point it's just Magic Story Time.

 

[Hriston]

Why are PCs/NPCs in your game so incompetent that they are literally unable to attack an unaware target? Why do they have to flub it so that the target sees/senses them somehow before the attack hits? It makes no sense that fate makes it impossible for an assassin to do his job properly by always causing moonlight to glint off of a blade, him to make noise, or whatever in order to alert the target. It's really okay for the blade coming out of the PC's chest to be the first clue that an attack happened and surprise him.

You seem to misunderstand me. Creatures (whether PC, NPC, or monster) in my games can attack an unaware target, and if they were stealthy, the target will be surprised by the attack and won't notice them until the attack hits or misses. The target, if it was asleep, will become alert and aware at the start of combat with the knowledge that an attack is now underway, however, but this doesn't interfere with the assassin's use of its signature ability. A surprised target will remain surprised until the end of its first turn, giving the assassin an opportunity to auto-crit if it wins initiative, so far from it being "impossible for the assassin to do his job properly", the assassin in my games can do his job, and the only difference I can see is that I don't check the WIS (Perception) of the unaware target, which makes it easier for the assassin, not harder.

 

[Maxperson]

You seem to misunderstand me. Creatures (whether PC, NPC, or monster) in my games can attack an unaware target, and if they were stealthy, the target will be surprised by the attack and won't notice them until the attack hits or misses. The target, if it was asleep, will become alert and aware at the start of combat with the knowledge that an attack is now underway, however, but this doesn't interfere with the assassin's use of its signature ability. A surprised target will remain surprised until the end of its first turn, giving the assassin an opportunity to auto-crit if it wins initiative, so far from it being "impossible for the assassin to do his job properly", the assassin in my games can do his job, and the only difference I can see is that I don't check the WIS (Perception) of the unaware target, which makes it easier for the assassin, not harder.

Okay. It seemed like you were saying that there would always be something to make the target aware of the attack just prior to it happening in order to allow them to be surprised. That didn't make any sense to me.