Short version: When (if ever) is it "fun" for a player character to die in the game? Should character death be an ever-present threat, or should character death only be a threat for "meaningful" encounters?
Longer version: These questions came to mind after reading some very, ahem, spirited debate about the ongoing UA rules (specifically, whether monsters should be able to crit or not). This post is NOT about the UA, but about what, for you as a player (and DM, I suppose) makes the game fun and what kills the fun? Anecdotally, it feels like a lot of players who picked up D&D with 3rd Edition or later find character death decidedly UNfun, as in: it should be rare, should only be permanent at low levels, and should have some kind of "meaning" (definition left purposefully vague) within the plot or story of the game and campaign. Contrast this with my own (and others) who've been playing D&D much longer, and started with older (much deadlier) editions. Character death is just as likely with a random encounter as with the BBEG; it's simply part of the risk you take when you choose an adventurer's life. More than that: it lends a degree of "reality" and stakes to the game; life and encounters are inherently uncertain, and you're not guaranteed to survive tomorrow, let alone today's encounter with that band of gnolls.
D&D can be played a LOT of ways (though it seems to be designed to be combat-focused). You can have sessions or even campaigns in which you never roll to hit anything, but resolve all encounters and challenges through diplomacy, persuasion, guile, etc. You can have sessions or games which are nonstop meat grinders. And everything in between. None of the options are right or wrong, better or worse (though, again, some will be easier to manage because the game as designed is focused on combat and fighting).
And yes, 5E (and a few earlier editions) do encourage players to spend a good amount of time developing characters - personalities, backstories, allies, rivals, etc. It can be upsetting or sad when a beloved character dies (it's happened to me!) - but does that mean it's not fun?
I guess ultimately I, as a person with roots in old school gaming, have some frustration with how "padded for safety" 5E seems. I have a very hard time, to put it mildly, understanding the mindset of creating a character for the game and being upset if that character dies. Because for me, the bottom line is: adventuring is dangerous. Inherently so. And even though player characters are considered extraordinary in comparison with most commoners, level progression and CR (broken as it is) alone tell us that choosing to pick up a sword or wand and going looking for trouble means you run the risk of dying, be it at the end of a spear wielding by roadside bandits, the traps in an ancient ruin, or the spell cast by the villain they've been chasing for weeks. For me, there IS fun in that because it's what lends stakes to the game: walking into every or even just most encounters being fully certain you're going to survive and succeed becomes boring quickly, and just feels like a power fantasy rather than a series of meaningful challenges.
Even though I have strong opinions on this, please note again: I am NOT, after all, telling anyone how to play the game or how to have fun. I'm interested in dialogue and well intentioned back and forth, but I'm not here to slam anyone or tell you you're having fun the wrong way. (And if I'm contradicting myself here...well, so be it. I'm human, and doing my best to express this stuff and talk with others about it. I'm not claiming to be a bastion of impeccably consistent and logical thought!)
(Note: for the sake of this discussion, please assume a character death was NOT the result of blatantly bad DMing or encounter design. Those are separate topics, and I am very much on the side of a player who is upset because a 2nd level character was killed by a lich they had no chance of avoiding or negotiating with or a 1st level paladin who's forced to fight a drow champion wielding a vorpal sword.)
Character death should be game appropriate, and discussed during session 0.
For newer players, I'd recommend a "Heroic" adventure where death is unlikely, and if it does happen the player can choose whether the plot resurrects them.
For more experienced adventures, I'd opt for a more "Gritty" adventure where death is a legitimate risk, but avoidable with smart play.
The important aspects of death are managing expectations, and ensuring that death isn't the end of fun. As long as the players know what to expect, then all that matter is having everyone on-board.
I would imagine that the greater the risk of death, the less the players will invest in their character's substance. If you want to do a deep lore game, consider tossing in some bumpers.
Character death should be game appropriate, and discussed during session 0.
For newer players, I'd recommend a "Heroic" adventure where death is unlikely, and if it does happen the player can choose whether the plot resurrects them.
For more experienced adventures, I'd opt for a more "Gritty" adventure where death is a legitimate risk, but avoidable with smart play.
The important aspects of death are managing expectations, and ensuring that death isn't the end of fun. As long as the players know what to expect, then all that matter is having everyone on-board.
I would imagine that the greater the risk of death, the less the players will invest in their character's substance. If you want to do a deep lore game, consider tossing in some bumpers.
I pretty much agree. It is all down to the kind of game that your table wants to play. Some people were attracted to the game because it represented a more complex board game or miniature wargame. They might like the strategy, and see paying for their mistakes with rolling a new character as natural as setting up the next battle. Others were drawn to the game as a chance to tell a story as good or better than their favorite books and movies, where they are the heroes. Challenges and tragedy are expected, but they are also expected to have meaning to the plot.
If everyone wants to collectively tell a grand story with deep character growth and a satisfying arc, then death should probably be rare and meaningful. It's hard to get as invested when you're on your third character. And harder to tell that cohesive story. I don't think stakes are only found in death. Characters have a lot to win or lose that isn't just their lives. All kinds of failure can lead to meaningful stakes. If your group wants to make their own private version of a Lord of the Rings style story, then there is a lot of room for stakes that matter, but very little permanent death.
If everyone wants to just roll some dice and go on an old school dungeon crawl, where life is cheap and there is always another adventurer ready to risk it all opening the next door, then death can be commonplace. It can even be part of the fun. It can help tell a different kind of story in a very dangerous world.
There are all kinds of games to be found somewhere in between too. The most important thing is that everyone is on the same page. Because an unexpected death, or even an unexpected rescue by giant eagles, can ruin the fun if you wanted something else. So as long as everyone has talked it out and knows what they're getting into, all kinds of fulfilling stories can be told with DnD..
The issue with a lot of this debate is that "Fun" is not necessarily what people are talking about with character death. "Fun" is not the same thing as immersive. "Fun" is not the same thing as satisfying. A session can be gut-wrenching, horrid, and awful - absolutely no fun at all - and still be immersive and satisfying. Those two things are as important, if not ultimately more important, than "Fun", i.e. an in-the-moment dopamine hit Good Time.
Old Guard players get their immersion and satisfaction from uncertainty and the sense that the universe doesn't care. "Story" is for bohemian losers; Old Guard players absolutely hate "Story" because they want stories. Awesome tales they can tell their buddies about this one time something totally crazy happened in their game and it was wild. The dice fall where they may and if that means Alice dies to an arrow in the eye from a goblin ambush she had no possible way to avoid or protect herself from, then that's just how Adventuring Life is. Suck it up buttercup, get rerolling and we'll find you behind the next closed door - and won't that be a fun story to tell later, about how the dice screwed you in a morbidly hilarious manner? Now you can commiserate with everyone else who lost a PC to something stupid!
New Guys, and Old Guard players who've adopted the newer modes of play in the game, get their immersion from and satisfaction from a strong narrative arc. Uncertainty is important, but it's not as important as Story, and the sense that all the players' actions are building up towards an eventual thrilling conclusion. Character death can absolutely be a part of that, but the death that takes away a story has to offer a new story in its place. Not just the story of a new character, but a new story for the group as a whole that has lost the story they were telling with their original ally. Meaningless nonsense PC assassination from a killer DM takes away the party's story and doesn't offer them a new one in its place, and New Style players will not tolerate that. It breaks their immersion, their satisfaction, and why should they care about a game with no story to it anyways?
That's the big distinction, in my eyes. Capital-S Story versus water-cooler stories, and what drives a player's sense of immersion and satisfaction. Old Guard and New Style are fundamentally opposite and generally need to stay away from each other's tables if the player in question is not flexible in their gaming drives.
I'm a Death all the time type DM. A character might die at any time, for any reason.
I see this as a HUGE draw for RPGs, as it's the ONLY place where this can happen. In all of media, you will never see random character death of main characters.
It's also great for having players pay attention and not act dumb. In the safe game, all too often, players don't pay attention and goof off....as they have nothing to loose: they know the DM will never kill their character.
The player says "whatever, my character just walks past the goblins". The DM rolls for a bunch of arrow shots, but if they roll a bunch of '20's they will just be like "yuk yuk, all the arrows miss your character!" And the player is just like "thanks best buddy DM".
Character death is one of the best parts of the game!
Yurei, I mostly agree with you (except for the kinda mean generalization about how old school players view "Story").
I actually do want story from my games, whether I'm a player or a DM. But I resist, strongly, that D&D is like a novel, or a movie, or a graphic novel. A D&D session and campaign is a story that's constantly writing itself and potentially changing directions - up to and including character permadeath. Outcomes can't - or, at least, IMNSHO, shouldn't - be predetermined or forced to preserve what was assumed to be the story.
The first three books of GRRM's A Song of Ice and Fire are great examples here: characters - MAIN characters - die, permanently. Their deaths matter in the sense that they impact what comes after, the choices that the surviving characters make.
"Character death can absolutely be a part of that, but the death that takes away a story has to offer a new story in its place. Not just the story of a new character, but a new story for the group as a whole that has lost the story they were telling with their original ally."
I think I agree...while also disagreeing? Because the characters, through the players, have agency (or at least should). If one party member were to die, permanently, they still have the ability to react and make choices and (presumably) alter or affect the story. And story-external, the DM should be willing (again, IMNSHO) to work the player back into the game in a way that doesn't feel like someone grabbed the steering wheel and is driving into a ditch.
I want both a story with a capital S and a lot of stories; I want to avoid the extremes of killer DMs who dont' give a sh*t about what's fun for everyone else AND players who can't find enjoyment in the uncertainty and peril that a vicarious adventuring life offers.
Also, I think everyone above is 100% that clear, constructive communications needs to go on when the campaign/session starts, so that expectations and wants are shared, understood, and respected. ("Respected" here doesn't mean, necessarily, that it changes the tone of the game, but that no one's telling someone else their version of fun is wrong.) And it may be that a DM or player isn't right for the group; it happens, and it needs to be OK.
I also think Yurei is spot-on using the word flexibility. D&D is ultimately a collaborative game, and it should be fun for the players AND the DM. As long as their not immovable in what constitutes a fun play style, everyone should be willing to flex and try some things. It might be that a style that's perhaps a bit less deadly than the DM is used but a bit more perilous than the players initially expected could be fun for everyone. Again, communication, respect, and remembering it's a game are paramount here.
Side note: I've never been a big advocate for creating pages/100s or 1000s of words of backstory for characters. I definitely LOVE crafting their personality and sketching in the basics of their history, but I've always felt like the character should be discovered and refined as the game goes along, rather than doing a LOT of work upfront. Then again, that's likely my old school roots showing, since I come from the days when you expected new characters to more likely die than not.
I see this as a HUGE draw for RPGs, as it's the ONLY place where this can happen. In all of media, you will never see random character death of main characters.
Have you ever read Game of Thrones? There's a reason why the joke is, "George RR Martin walks into a bar. Everybody you dies."
... Outcomes can't - or, at least, IMNSHO, shouldn't - be predetermined or forced to preserve what was assumed to be the story. ...
This is the thing that catches a lot of people, in what experiences I've had on this forum.
Some 'Old Guard' style players assume that New Style players have prewritten everything and are unwilling to flex, bend, and follow the game as it progresses. In a sense, they feel like that player isn't really playing D&D but instead using the session as an excuse to show off their Ohh Cee, simply displaying their shiny creation without care or attention to anybody else's character or to the overall plot. After all, for that player they know the overall plot - they already wrote it all down.
That kind of player can safely be termed a Total Nitwit, and it is not what most New Style players are looking for.
Am I capable of generating a character in ten minutes and just playing, the way Old Guard players generally profess is the best way to play? Yes. I've snapkicked for multiple one-shots in the past and I could do it for a campaign. But that character wouldn't fit. It would have ties to the world, it wouldn't have allies or rivals, it wouldn't have interests or motivations. I wouldn't give a single shit about that character or what happened to it. People see that sort of ten-minute snapkick and think "cool! A blank slate that doesn't have any attached baggage I have to worry about!" or "Cool! A character that can be Discovered At The Table!", but that character would be neither of those things. That character wouldn't even be a character. It would be a toon - just a vaguely man-shaped pile of mechanics I drive around doing D&D things with because Rei-the-player agreed to a session and Rei-the-player has an idea what D&D sessions look like.
The character isn't a blank slate because it's not a slate at all - there isn't any story to be told there. This alleged character just popped into existence one day and decided to kill bandits. It doesn't know why, it doesn't care why, it doesn't care what comes afterwards. THis alleged character isn't something that can be "Discovered At The Table" A term I absolutely ******* hate, but that's a story for another day. because there's nothing to discover. Again - this thing popped into existence with no family, no friends, no past, and no future with the sole objective of killing some bandits.
Some people love that. To them, that's all exactly the entire idea, because the story isn't about the characters, it's about what happens to/around the characters. Nobody cares what Gam Samgee did before he adventured, nobody cares about his love life, nobody cares about his development as a person or what his adventures mean to him. They care about whether or not Gam Samgee and his gang will find a way to stop the evil Necrolich from resurrecting the Demon Lord and bathing the world in darkness, and to what lengths they will have to go to do it. Gam Samgee is entirely unimportant to the whole endeavor - literally any batch of 'Legendary Heroes' would work just as well, and did work just as well with Gam Samgee's player's last three characters. Who doesn't matter, only what.
New Style play is asserting that "Who" matters. That the story of Gam Samgee is unique, and Gam Samgee's presence in his band of adventurers is also unique. New Style play asserts that there's value in tailoring a character to the adventure being played and exploring them beyond their ability to fell their foes. That takes work. It takes time, effort, and emotional investment to create a character that fits. And if that time, effort, and emotional investment is repeatedly snubbed without consideration, then the player ends up frustrated and disengaged from the game. Not because they can't handle PC death, but because there's no point in playing a game where their participation doesn't matter. There's no point in being part of a game where literally any five random yaybos will do and the table will spin the exact same story no matter who or what is there.
Which, I suppose, is a long way of saying no - we don't want to control the narrative. We don't want to "write the plot out" ahead of time. We're as keen to see what the winds of fate have in store for us as anyone else. We can handle weird deaths. We can handle ignoble deaths. Hell, I nearly died to evil pumpkin seeds in my PBP game and I was already planning how to deal with Ilyara's increasingly probable demise. But she nearly died because I made choices. She lived because the other players made choices that pulled her back from the brink.
What we don't want is random nonsense, where a DM just rolls a bunch of dice and then kills a player no matter what that player chose to do. That's just ******* disrespectful to a New Style player who sunk real effort into a character and tried to provide the DM plenty of meat for a cool story. If that's the sort of thing a DM wants to do, then they need to ensure their players know to just churn out an endless series of toons and not try to make characters.
Kind of hard to answer but yea if it has meaning and you don't take away pcs afterlife generally if done right it's ok but I had a dm make a op monster did not give us the option to run and forced us to fight it . Why he was doing it the creature said I hate wizards to my character he than proceeded to basically trash us in the end he stood with a grin on his face probably proud of his handy work like he won dnd. Pretty sure he was only one who had fun that day.
Am I capable of generating a character in ten minutes and just playing, the way Old Guard players generally profess is the best way to play? Yes. I've snapkicked for multiple one-shots in the past and I could do it for a campaign. But that character wouldn't fit. It would have ties to the world, it wouldn't have allies or rivals, it wouldn't have interests or motivations. I wouldn't give a single shit about that character or what happened to it. People see that sort of ten-minute snapkick and think "cool! A blank slate that doesn't have any attached baggage I have to worry about!" or "Cool! A character that can be Discovered At The Table!", but that character would be neither of those things. That character wouldn't even be a character. It would be a toon - just a vaguely man-shaped pile of mechanics I drive around doing D&D things with because Rei-the-player agreed to a session and Rei-the-player has an idea what D&D sessions look like.
The character isn't a blank slate because it's not a slate at all - there isn't any story to be told there. This alleged character just popped into existence one day and decided to kill bandits. It doesn't know why, it doesn't care why, it doesn't care what comes afterwards. THis alleged character isn't something that can be "Discovered At The Table" A term I absolutely ****ing hate, but that's a story for another day. because there's nothing to discover. Again - this thing popped into existence with no family, no friends, no past, and no future with the sole objective of killing some bandits.
It's here, I think, we're going to run into some significant disagreements. Characters can be very much discovered at the table. I know, because I've done it/it's happened to me. The character began as a series of dice rolls, but over the course of a session, or multiple sessions, I figure out who they are, what they care about, what drives them. I may not have an elaborate or detailed backstory for them, but they are a character, one I try hard to make distinct from my other characters.
Does this process yield the same kind of results as spending many minutes or hours creating a backstory, family, etc. for a character? Probably not, but I think it all comes down to the player. I'm comfortable in asserting that an engaged, participatory player can bring a "snapkick" character to life, with a bare minimum of information while another player can be given a character with a detailed background, be asked to continue to figure the character out before the first session, and never give that character any sense of personality, individuality, etc.
What we don't want is random nonsense, where a DM just rolls a bunch of dice and then kills a player no matter what that player chose to do. That's just ****ing disrespectful to a New Style player who sunk real effort into a character and tried to provide the DM plenty of meat for a cool story. If that's the sort of thing a DM wants to do, then they need to ensure their players know to just churn out an endless series of toons and not try to make characters.
I'm honestly puzzled here by what you mean by "random nonsense" and the DM just rolling dice and then kills a character no matter what. I don't see anyone advocating for that (again, based on my understanding of what you're writing).
If a group of 2nd level characters encounter a band of hobgoblins as a random encounter on the way somewhere...are you saying the characters shouldn't die? That the random dice rolls (which are foundational to the game) in that encounter should be overridden if it's going to result in a character dying? What if the party has a terrible first round and chooses to not try and negotiate? What if the DM just happens to roll extraordinarily well for initiative, attack, and damage rolls? I'm asking sincerely here, not being rhetorical, because I'm not sure I understand what you mean.
I suppose it would be disappointing if a character were to die in that unplanned encounter, but I still fall back on what seems to be true: adventuring life is extremely dangerous and can be lethal. The foundational choice to go after danger itself could be interpreted as saying even the random encounters aren't all that random, and have some meaning.
It's a two way street, after all. I think most DMs have had campaigns/sessions where the planned Big Bad is captured or killed far more quickly than anticipated, because of dice rolls. (It's certainly happened to me!) The Story IS affected but that's the fun: there's now an unanticipated power vacuum - who or what fills it? What does it mean for the characters to have this sudden success? And so on.
I started playing D&D pre-3rd Edition and I don't ever recall a time when players considered their characters dying to be "fun." It was definitely considered inevitable, especially in campaign settings like Ravenloft and Dark Sun, but it wasn't treated as being fun. The only people I ever met who actually regarded dead PCs as fun were killer GMs who thought that the game was supposed to be GM vs party.
Rollback Post to RevisionRollBack
Find your own truth, choose your enemies carefully, and never deal with a dragon.
"Canon" is what's factual to D&D lore. "Cannon" is what you're going to be shot with if you keep getting the word wrong.
I guess ultimately I, as a person with roots in old school gaming, have some frustration with how "padded for safety" 5E seems. I have a very hard time, to put it mildly, understanding the mindset of creating a character for the game and being upset if that character dies. Because for me, the bottom line is: adventuring is dangerous.
I think this might be where there's some clash.
To me I don't want D&D to be a realistic "Mercenaries who fight monsters, but also never hire as many people as a medieval Free Company would" sim.
I want it mostly to be a sim of a fantasy adventure novel, where deaths are well timed and serve the narrative.
In a grimdark campaign that can mean dying to Ogre #2 in a random encounter and the DM homebrewing out all res magic- But I don't need that to be the default of the system the same way I don't need writers emulating GRRM's. In fact for some fantasy books and TV shows I'm more than happy for absolutely everyone to survive cause it's not that kind of story.
One thing that often gets overlooked in these conversations--a DM need not apply a one size fits all solution even within the same campaign. If some players are more okay with player death than others, you can use monsters whose skills disproportionately harm their characters, different tactics, and other DM tricks to increase the chance of death on those players. That will also increase the perceived chance of death for the other players, which should help them feel more heroic as they keep beating the odds.
Overall, it really does not matter how difficult or deadly the campaign is--so long as folks are having fun, that's fine.
I will note, I think the "5e is too padded and easy" argument is a bad one--there are plenty ways to make 5e plenty deadly if the DM wants. Sure, CR is a mess and if you follow CR at anything except the lowest of levels, your fights will be too easy. But D&D has been doing a bad job at estimating the relative power of monsters and players since 1974, with CR simply being the most recent failing in a long line of mediocre encounter difficulty guidance. It becomes abundantly clear when DMing 5e that CR underestimates player power and encounter difficulty needs to be increased--if the DM chooses to stick sycophantically to CR and not take other actions to make fights more challenging and dynamic, that falls squarely on their shoulders.
I'm a Death all the time type DM. A character might die at any time, for any reason.
I see this as a HUGE draw for RPGs, as it's the ONLY place where this can happen. In all of media, you will never see random character death of main characters.
Clearly you haven't been keeping up with mainstream pop culture in the last decade or so. ;)
People randomly drop dead in real life, why shouldn’t they in a rpg? I don’t try to kill my player’s characters off intentionally but if they do something stupid well it’s on them, decisions come with consequences, and they aren’t always good or wanted ones.
As far as my own characters go, I like to experiment and build unusual characters. Mostly they work out as I have a good grasp of multiclassing but occasionally I build something that just doesn’t fit in well with a party, I play in a single longstanding group so there is rarely conflict but it does sometimes happen. When that happens I have no issue with offering up my character as a sacrifice to the dice gods, so I can use one of the many other character ideas I have living rent free in my head.
Without meaning to derail my own thread, I'm nodding pretty hard along with Short_term_user's post. 5E (and whatever is coming next) is clearly written to make permadeath much rarer than in previous editions, and, assuming the DM isn't just an a**hole and makes an attempt to run the game using the RAW/RAI, character deaths are going to be few and far between.
Some visibility into my own game/style: as I've shared, my gaming roots are in 1E. I'm still learning the differences in approach and while I have some frustrations, I'm overall really happy with 5E and wouldn't go back to a previous edition even if I found a group who wanted to. I've been running a campaign since late 2017. The characters have progressed from level 1 to 8 (they're on the cusp of achieving 9th level). In the 5+ years I've been running this 5E campaign, there have been a grand total of two actual deaths (meaning: death death, not just hitting zero hit points). One of these deaths came about due a very foolish choice by the player, wholly unforced by me; the second came about in a planned encounter due to some very good rolling by me. Both deaths ended up being reversed; in one case, I was able to use the desire of the party to bring the character back to do some worldbuilding and get them on a fun sidequest for the needed material components. In fact: in that particular case, the player made it known to me that they really liked that character, and would appreciate the chance to bring them back. Since the ultimate goal is for everyone to have fun and enjoy the game, I was happy to work that into the campaign. (The second death was more easily reversed due to the characters being at a higher level than they were at the previous death, and they had the magic on hand to deal with it.)
I share this in the interest of giving any readers a very rough idea of my DMing style - because it might be easy to think from my posts that I'm gleefully taking down characters every session. I'm absolutely not doing this - but I'm also not removing the possibility of chance and randomness significantly affecting the characters or the campaign.
I play D&D because I enjoy sitting around the table trying to defeat challenges, overcome obstacles, and effect the outcome of how things play out. Without charactarer deaths, the fact that I overcame those obstacles and beat those challenges doesn’t matter. Why? Because there was no chance that I would fail or die, the obstacles were made to be overcome and no matter how idiotic or how many stupid decisions my character made, they were going to make it to the finish line anyways.
If I wanted to “play” a story with a fixed outcome, then why wouldn’t I have just watched a movie instead? My character shouldn’t be able to fall into the pit of lava and have a nice time slowly swimming back, because there isn’t much of a mechanic for failure if the ultimate failure is taken away.
Though I dislike no character deaths and would never ever play that way, aversion to character death is fine. Why? Because when I spend hours of time and energy into making my character and giving them a real and interesting personality, I don’t want to lose that character right away and be forced to restart due to anything other than my own stupidity or failure to properly deal with the defeatable threats in the game.
If you don’t like character death, don’t play with it. But at my table at least, the possibility of character death is what makes the game worthwhile and fun. Even though the experience of character will likely be neither of those things.
Some people enjoy Game of Thrones, some enjoy Harry Potter, some enjoy both. Some weirdoes even have fun with Attack on Titan or My Little Pony:)
A lot of people are invested too much in the story of their characters and just do not want to see them die, especially if they have plans for them. It is like if the TV show was suddenly cancelled and there was no satisfying conclusion of the story. So for them permadeath of their character is not an option. However, there are people who will enjoy heroic death that will support the cause and some might have fun even with the most silly and easily avoidable deaths.
The whole topic is a very relative, it is like asking if coriander can taste good. It can, to some people.
PS: I agree that (with the exception of lowest levels) 5e feels very safe for the characters and it is (often) hard to experience permanent death of the character. I think that is what majority of players actually prefer, but I agree, that it also robs them of some experience and sense of accomplishment that comes with the early editions that were bit more brutal and unforgiving. I would say that experienced players should not have issues modifying the game to their liking, but for beginners it is better to start with a safer environment (and I would argue that especially the first level can be too deadly). It is really easy for absolute newbies to get themselves killed and if they are attached to their characters, it can ruin the game for them.
I want it mostly to be a sim of a fantasy adventure novel, where deaths are well timed and serve the narrative.
But how do you know when a death serves the narrative or not? What is the criteria for this? (Again, these are genuine questions, not rhetorical.)
My struggle with this approach is that it seems to assume that there's agreed-upon arc or trajectory for the "story" to take, and everyone at the table - DM as well as fellow players (and the dice!) - is going to stick to it. Whereas my decades of experience at the gaming table tells me that while very, VERY broad arcs or trajectories may be able to hold over the course of a campaign, you simply can't dictate where the story goes ahead of time, nor can you really predict where it's going. D&D is meant to be collaborative and, I'd argue, it's meant to have a fair amount of uncertainty (i.e., randomness) to it. Unless the campaign is on rails, allowing for no significant deviation or changes, the narrative often doesn't become clear until near the end of the campaign.
(And this is my problem with so many canon and third party modules and mini-campaigns: they're written like novels, not like games where the characters have full agency. Most of them provide thin or poor hooks/reasons for any sane character to continue on the adventure path.)
It's one thing to carefully craft a character with a detailed backstory, strengths and flaws, and (as a player) have a certain arc or trajectory in mind for the character. That approach, which generally doesn't require specific events at specific times, is good and easily integrated into almost any game. It's another to want the campaign to unfold as a seamless, consistent "narrative" - I guess you get what you want there, as it ends up being more a single person's vision, but D&D, as I've understood, isn't meant to be that. It's a collaborative, constantly evolving and growing "story", told by multiple people, and the end is far from certain or set.
Barring any foolish act from a player character, i think death is most fun during set piece encounters against important enemies than when dying in a random encounter with no importance in the campaign. I think my challenge as a DM resides in striking a good balance between keeping the threat of death omnipresent yet be relatively infrequent, as i don't want my campaigns to never kill PCs nor have them die so frequently.
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
Short version: When (if ever) is it "fun" for a player character to die in the game? Should character death be an ever-present threat, or should character death only be a threat for "meaningful" encounters?
Longer version: These questions came to mind after reading some very, ahem, spirited debate about the ongoing UA rules (specifically, whether monsters should be able to crit or not). This post is NOT about the UA, but about what, for you as a player (and DM, I suppose) makes the game fun and what kills the fun? Anecdotally, it feels like a lot of players who picked up D&D with 3rd Edition or later find character death decidedly UNfun, as in: it should be rare, should only be permanent at low levels, and should have some kind of "meaning" (definition left purposefully vague) within the plot or story of the game and campaign. Contrast this with my own (and others) who've been playing D&D much longer, and started with older (much deadlier) editions. Character death is just as likely with a random encounter as with the BBEG; it's simply part of the risk you take when you choose an adventurer's life. More than that: it lends a degree of "reality" and stakes to the game; life and encounters are inherently uncertain, and you're not guaranteed to survive tomorrow, let alone today's encounter with that band of gnolls.
D&D can be played a LOT of ways (though it seems to be designed to be combat-focused). You can have sessions or even campaigns in which you never roll to hit anything, but resolve all encounters and challenges through diplomacy, persuasion, guile, etc. You can have sessions or games which are nonstop meat grinders. And everything in between. None of the options are right or wrong, better or worse (though, again, some will be easier to manage because the game as designed is focused on combat and fighting).
And yes, 5E (and a few earlier editions) do encourage players to spend a good amount of time developing characters - personalities, backstories, allies, rivals, etc. It can be upsetting or sad when a beloved character dies (it's happened to me!) - but does that mean it's not fun?
I guess ultimately I, as a person with roots in old school gaming, have some frustration with how "padded for safety" 5E seems. I have a very hard time, to put it mildly, understanding the mindset of creating a character for the game and being upset if that character dies. Because for me, the bottom line is: adventuring is dangerous. Inherently so. And even though player characters are considered extraordinary in comparison with most commoners, level progression and CR (broken as it is) alone tell us that choosing to pick up a sword or wand and going looking for trouble means you run the risk of dying, be it at the end of a spear wielding by roadside bandits, the traps in an ancient ruin, or the spell cast by the villain they've been chasing for weeks. For me, there IS fun in that because it's what lends stakes to the game: walking into every or even just most encounters being fully certain you're going to survive and succeed becomes boring quickly, and just feels like a power fantasy rather than a series of meaningful challenges.
Even though I have strong opinions on this, please note again: I am NOT, after all, telling anyone how to play the game or how to have fun. I'm interested in dialogue and well intentioned back and forth, but I'm not here to slam anyone or tell you you're having fun the wrong way. (And if I'm contradicting myself here...well, so be it. I'm human, and doing my best to express this stuff and talk with others about it. I'm not claiming to be a bastion of impeccably consistent and logical thought!)
(Note: for the sake of this discussion, please assume a character death was NOT the result of blatantly bad DMing or encounter design. Those are separate topics, and I am very much on the side of a player who is upset because a 2nd level character was killed by a lich they had no chance of avoiding or negotiating with or a 1st level paladin who's forced to fight a drow champion wielding a vorpal sword.)
Character death should be game appropriate, and discussed during session 0.
For newer players, I'd recommend a "Heroic" adventure where death is unlikely, and if it does happen the player can choose whether the plot resurrects them.
For more experienced adventures, I'd opt for a more "Gritty" adventure where death is a legitimate risk, but avoidable with smart play.
The important aspects of death are managing expectations, and ensuring that death isn't the end of fun. As long as the players know what to expect, then all that matter is having everyone on-board.
I would imagine that the greater the risk of death, the less the players will invest in their character's substance. If you want to do a deep lore game, consider tossing in some bumpers.
I pretty much agree. It is all down to the kind of game that your table wants to play. Some people were attracted to the game because it represented a more complex board game or miniature wargame. They might like the strategy, and see paying for their mistakes with rolling a new character as natural as setting up the next battle. Others were drawn to the game as a chance to tell a story as good or better than their favorite books and movies, where they are the heroes. Challenges and tragedy are expected, but they are also expected to have meaning to the plot.
If everyone wants to collectively tell a grand story with deep character growth and a satisfying arc, then death should probably be rare and meaningful. It's hard to get as invested when you're on your third character. And harder to tell that cohesive story. I don't think stakes are only found in death. Characters have a lot to win or lose that isn't just their lives. All kinds of failure can lead to meaningful stakes. If your group wants to make their own private version of a Lord of the Rings style story, then there is a lot of room for stakes that matter, but very little permanent death.
If everyone wants to just roll some dice and go on an old school dungeon crawl, where life is cheap and there is always another adventurer ready to risk it all opening the next door, then death can be commonplace. It can even be part of the fun. It can help tell a different kind of story in a very dangerous world.
There are all kinds of games to be found somewhere in between too. The most important thing is that everyone is on the same page. Because an unexpected death, or even an unexpected rescue by giant eagles, can ruin the fun if you wanted something else. So as long as everyone has talked it out and knows what they're getting into, all kinds of fulfilling stories can be told with DnD..
The issue with a lot of this debate is that "Fun" is not necessarily what people are talking about with character death. "Fun" is not the same thing as immersive. "Fun" is not the same thing as satisfying. A session can be gut-wrenching, horrid, and awful - absolutely no fun at all - and still be immersive and satisfying. Those two things are as important, if not ultimately more important, than "Fun", i.e. an in-the-moment dopamine hit Good Time.
Old Guard players get their immersion and satisfaction from uncertainty and the sense that the universe doesn't care. "Story" is for bohemian losers; Old Guard players absolutely hate "Story" because they want stories. Awesome tales they can tell their buddies about this one time something totally crazy happened in their game and it was wild. The dice fall where they may and if that means Alice dies to an arrow in the eye from a goblin ambush she had no possible way to avoid or protect herself from, then that's just how Adventuring Life is. Suck it up buttercup, get rerolling and we'll find you behind the next closed door - and won't that be a fun story to tell later, about how the dice screwed you in a morbidly hilarious manner? Now you can commiserate with everyone else who lost a PC to something stupid!
New Guys, and Old Guard players who've adopted the newer modes of play in the game, get their immersion from and satisfaction from a strong narrative arc. Uncertainty is important, but it's not as important as Story, and the sense that all the players' actions are building up towards an eventual thrilling conclusion. Character death can absolutely be a part of that, but the death that takes away a story has to offer a new story in its place. Not just the story of a new character, but a new story for the group as a whole that has lost the story they were telling with their original ally. Meaningless nonsense PC assassination from a killer DM takes away the party's story and doesn't offer them a new one in its place, and New Style players will not tolerate that. It breaks their immersion, their satisfaction, and why should they care about a game with no story to it anyways?
That's the big distinction, in my eyes. Capital-S Story versus water-cooler stories, and what drives a player's sense of immersion and satisfaction. Old Guard and New Style are fundamentally opposite and generally need to stay away from each other's tables if the player in question is not flexible in their gaming drives.
Please do not contact or message me.
I'm a Death all the time type DM. A character might die at any time, for any reason.
I see this as a HUGE draw for RPGs, as it's the ONLY place where this can happen. In all of media, you will never see random character death of main characters.
It's also great for having players pay attention and not act dumb. In the safe game, all too often, players don't pay attention and goof off....as they have nothing to loose: they know the DM will never kill their character.
The player says "whatever, my character just walks past the goblins". The DM rolls for a bunch of arrow shots, but if they roll a bunch of '20's they will just be like "yuk yuk, all the arrows miss your character!" And the player is just like "thanks best buddy DM".
Character death is one of the best parts of the game!
Yurei, I mostly agree with you (except for the kinda mean generalization about how old school players view "Story").
I actually do want story from my games, whether I'm a player or a DM. But I resist, strongly, that D&D is like a novel, or a movie, or a graphic novel. A D&D session and campaign is a story that's constantly writing itself and potentially changing directions - up to and including character permadeath. Outcomes can't - or, at least, IMNSHO, shouldn't - be predetermined or forced to preserve what was assumed to be the story.
The first three books of GRRM's A Song of Ice and Fire are great examples here: characters - MAIN characters - die, permanently. Their deaths matter in the sense that they impact what comes after, the choices that the surviving characters make.
"Character death can absolutely be a part of that, but the death that takes away a story has to offer a new story in its place. Not just the story of a new character, but a new story for the group as a whole that has lost the story they were telling with their original ally."
I think I agree...while also disagreeing? Because the characters, through the players, have agency (or at least should). If one party member were to die, permanently, they still have the ability to react and make choices and (presumably) alter or affect the story. And story-external, the DM should be willing (again, IMNSHO) to work the player back into the game in a way that doesn't feel like someone grabbed the steering wheel and is driving into a ditch.
I want both a story with a capital S and a lot of stories; I want to avoid the extremes of killer DMs who dont' give a sh*t about what's fun for everyone else AND players who can't find enjoyment in the uncertainty and peril that a vicarious adventuring life offers.
Also, I think everyone above is 100% that clear, constructive communications needs to go on when the campaign/session starts, so that expectations and wants are shared, understood, and respected. ("Respected" here doesn't mean, necessarily, that it changes the tone of the game, but that no one's telling someone else their version of fun is wrong.) And it may be that a DM or player isn't right for the group; it happens, and it needs to be OK.
I also think Yurei is spot-on using the word flexibility. D&D is ultimately a collaborative game, and it should be fun for the players AND the DM. As long as their not immovable in what constitutes a fun play style, everyone should be willing to flex and try some things. It might be that a style that's perhaps a bit less deadly than the DM is used but a bit more perilous than the players initially expected could be fun for everyone. Again, communication, respect, and remembering it's a game are paramount here.
Side note: I've never been a big advocate for creating pages/100s or 1000s of words of backstory for characters. I definitely LOVE crafting their personality and sketching in the basics of their history, but I've always felt like the character should be discovered and refined as the game goes along, rather than doing a LOT of work upfront. Then again, that's likely my old school roots showing, since I come from the days when you expected new characters to more likely die than not.
This is the thing that catches a lot of people, in what experiences I've had on this forum.
Some 'Old Guard' style players assume that New Style players have prewritten everything and are unwilling to flex, bend, and follow the game as it progresses. In a sense, they feel like that player isn't really playing D&D but instead using the session as an excuse to show off their Ohh Cee, simply displaying their shiny creation without care or attention to anybody else's character or to the overall plot. After all, for that player they know the overall plot - they already wrote it all down.
That kind of player can safely be termed a Total Nitwit, and it is not what most New Style players are looking for.
Am I capable of generating a character in ten minutes and just playing, the way Old Guard players generally profess is the best way to play? Yes. I've snapkicked for multiple one-shots in the past and I could do it for a campaign. But that character wouldn't fit. It would have ties to the world, it wouldn't have allies or rivals, it wouldn't have interests or motivations. I wouldn't give a single shit about that character or what happened to it. People see that sort of ten-minute snapkick and think "cool! A blank slate that doesn't have any attached baggage I have to worry about!" or "Cool! A character that can be Discovered At The Table!", but that character would be neither of those things. That character wouldn't even be a character. It would be a toon - just a vaguely man-shaped pile of mechanics I drive around doing D&D things with because Rei-the-player agreed to a session and Rei-the-player has an idea what D&D sessions look like.
The character isn't a blank slate because it's not a slate at all - there isn't any story to be told there. This alleged character just popped into existence one day and decided to kill bandits. It doesn't know why, it doesn't care why, it doesn't care what comes afterwards. THis alleged character isn't something that can be "Discovered At The Table"
A term I absolutely ******* hate, but that's a story for another day.because there's nothing to discover. Again - this thing popped into existence with no family, no friends, no past, and no future with the sole objective of killing some bandits.Some people love that. To them, that's all exactly the entire idea, because the story isn't about the characters, it's about what happens to/around the characters. Nobody cares what Gam Samgee did before he adventured, nobody cares about his love life, nobody cares about his development as a person or what his adventures mean to him. They care about whether or not Gam Samgee and his gang will find a way to stop the evil Necrolich from resurrecting the Demon Lord and bathing the world in darkness, and to what lengths they will have to go to do it. Gam Samgee is entirely unimportant to the whole endeavor - literally any batch of 'Legendary Heroes' would work just as well, and did work just as well with Gam Samgee's player's last three characters. Who doesn't matter, only what.
New Style play is asserting that "Who" matters. That the story of Gam Samgee is unique, and Gam Samgee's presence in his band of adventurers is also unique. New Style play asserts that there's value in tailoring a character to the adventure being played and exploring them beyond their ability to fell their foes. That takes work. It takes time, effort, and emotional investment to create a character that fits. And if that time, effort, and emotional investment is repeatedly snubbed without consideration, then the player ends up frustrated and disengaged from the game. Not because they can't handle PC death, but because there's no point in playing a game where their participation doesn't matter. There's no point in being part of a game where literally any five random yaybos will do and the table will spin the exact same story no matter who or what is there.
Which, I suppose, is a long way of saying no - we don't want to control the narrative. We don't want to "write the plot out" ahead of time. We're as keen to see what the winds of fate have in store for us as anyone else. We can handle weird deaths. We can handle ignoble deaths. Hell, I nearly died to evil pumpkin seeds in my PBP game and I was already planning how to deal with Ilyara's increasingly probable demise. But she nearly died because I made choices. She lived because the other players made choices that pulled her back from the brink.
What we don't want is random nonsense, where a DM just rolls a bunch of dice and then kills a player no matter what that player chose to do. That's just ******* disrespectful to a New Style player who sunk real effort into a character and tried to provide the DM plenty of meat for a cool story. If that's the sort of thing a DM wants to do, then they need to ensure their players know to just churn out an endless series of toons and not try to make characters.
Please do not contact or message me.
Kind of hard to answer but yea if it has meaning and you don't take away pcs afterlife generally if done right it's ok but I had a dm make a op monster did not give us the option to run and forced us to fight it . Why he was doing it the creature said I hate wizards to my character he than proceeded to basically trash us in the end he stood with a grin on his face probably proud of his handy work like he won dnd. Pretty sure he was only one who had fun that day.
It's here, I think, we're going to run into some significant disagreements. Characters can be very much discovered at the table. I know, because I've done it/it's happened to me. The character began as a series of dice rolls, but over the course of a session, or multiple sessions, I figure out who they are, what they care about, what drives them. I may not have an elaborate or detailed backstory for them, but they are a character, one I try hard to make distinct from my other characters.
Does this process yield the same kind of results as spending many minutes or hours creating a backstory, family, etc. for a character? Probably not, but I think it all comes down to the player. I'm comfortable in asserting that an engaged, participatory player can bring a "snapkick" character to life, with a bare minimum of information while another player can be given a character with a detailed background, be asked to continue to figure the character out before the first session, and never give that character any sense of personality, individuality, etc.
I'm honestly puzzled here by what you mean by "random nonsense" and the DM just rolling dice and then kills a character no matter what. I don't see anyone advocating for that (again, based on my understanding of what you're writing).
If a group of 2nd level characters encounter a band of hobgoblins as a random encounter on the way somewhere...are you saying the characters shouldn't die? That the random dice rolls (which are foundational to the game) in that encounter should be overridden if it's going to result in a character dying? What if the party has a terrible first round and chooses to not try and negotiate? What if the DM just happens to roll extraordinarily well for initiative, attack, and damage rolls? I'm asking sincerely here, not being rhetorical, because I'm not sure I understand what you mean.
I suppose it would be disappointing if a character were to die in that unplanned encounter, but I still fall back on what seems to be true: adventuring life is extremely dangerous and can be lethal. The foundational choice to go after danger itself could be interpreted as saying even the random encounters aren't all that random, and have some meaning.
It's a two way street, after all. I think most DMs have had campaigns/sessions where the planned Big Bad is captured or killed far more quickly than anticipated, because of dice rolls. (It's certainly happened to me!) The Story IS affected but that's the fun: there's now an unanticipated power vacuum - who or what fills it? What does it mean for the characters to have this sudden success? And so on.
I started playing D&D pre-3rd Edition and I don't ever recall a time when players considered their characters dying to be "fun." It was definitely considered inevitable, especially in campaign settings like Ravenloft and Dark Sun, but it wasn't treated as being fun. The only people I ever met who actually regarded dead PCs as fun were killer GMs who thought that the game was supposed to be GM vs party.
Find your own truth, choose your enemies carefully, and never deal with a dragon.
"Canon" is what's factual to D&D lore. "Cannon" is what you're going to be shot with if you keep getting the word wrong.
I think this might be where there's some clash.
To me I don't want D&D to be a realistic "Mercenaries who fight monsters, but also never hire as many people as a medieval Free Company would" sim.
I want it mostly to be a sim of a fantasy adventure novel, where deaths are well timed and serve the narrative.
In a grimdark campaign that can mean dying to Ogre #2 in a random encounter and the DM homebrewing out all res magic- But I don't need that to be the default of the system the same way I don't need writers emulating GRRM's. In fact for some fantasy books and TV shows I'm more than happy for absolutely everyone to survive cause it's not that kind of story.
One thing that often gets overlooked in these conversations--a DM need not apply a one size fits all solution even within the same campaign. If some players are more okay with player death than others, you can use monsters whose skills disproportionately harm their characters, different tactics, and other DM tricks to increase the chance of death on those players. That will also increase the perceived chance of death for the other players, which should help them feel more heroic as they keep beating the odds.
Overall, it really does not matter how difficult or deadly the campaign is--so long as folks are having fun, that's fine.
I will note, I think the "5e is too padded and easy" argument is a bad one--there are plenty ways to make 5e plenty deadly if the DM wants. Sure, CR is a mess and if you follow CR at anything except the lowest of levels, your fights will be too easy. But D&D has been doing a bad job at estimating the relative power of monsters and players since 1974, with CR simply being the most recent failing in a long line of mediocre encounter difficulty guidance. It becomes abundantly clear when DMing 5e that CR underestimates player power and encounter difficulty needs to be increased--if the DM chooses to stick sycophantically to CR and not take other actions to make fights more challenging and dynamic, that falls squarely on their shoulders.
Clearly you haven't been keeping up with mainstream pop culture in the last decade or so. ;)
People randomly drop dead in real life, why shouldn’t they in a rpg? I don’t try to kill my player’s characters off intentionally but if they do something stupid well it’s on them, decisions come with consequences, and they aren’t always good or wanted ones.
As far as my own characters go, I like to experiment and build unusual characters. Mostly they work out as I have a good grasp of multiclassing but occasionally I build something that just doesn’t fit in well with a party, I play in a single longstanding group so there is rarely conflict but it does sometimes happen. When that happens I have no issue with offering up my character as a sacrifice to the dice gods, so I can use one of the many other character ideas I have living rent free in my head.
Without meaning to derail my own thread, I'm nodding pretty hard along with Short_term_user's post. 5E (and whatever is coming next) is clearly written to make permadeath much rarer than in previous editions, and, assuming the DM isn't just an a**hole and makes an attempt to run the game using the RAW/RAI, character deaths are going to be few and far between.
Some visibility into my own game/style: as I've shared, my gaming roots are in 1E. I'm still learning the differences in approach and while I have some frustrations, I'm overall really happy with 5E and wouldn't go back to a previous edition even if I found a group who wanted to. I've been running a campaign since late 2017. The characters have progressed from level 1 to 8 (they're on the cusp of achieving 9th level). In the 5+ years I've been running this 5E campaign, there have been a grand total of two actual deaths (meaning: death death, not just hitting zero hit points). One of these deaths came about due a very foolish choice by the player, wholly unforced by me; the second came about in a planned encounter due to some very good rolling by me. Both deaths ended up being reversed; in one case, I was able to use the desire of the party to bring the character back to do some worldbuilding and get them on a fun sidequest for the needed material components. In fact: in that particular case, the player made it known to me that they really liked that character, and would appreciate the chance to bring them back. Since the ultimate goal is for everyone to have fun and enjoy the game, I was happy to work that into the campaign. (The second death was more easily reversed due to the characters being at a higher level than they were at the previous death, and they had the magic on hand to deal with it.)
I share this in the interest of giving any readers a very rough idea of my DMing style - because it might be easy to think from my posts that I'm gleefully taking down characters every session. I'm absolutely not doing this - but I'm also not removing the possibility of chance and randomness significantly affecting the characters or the campaign.
I play D&D because I enjoy sitting around the table trying to defeat challenges, overcome obstacles, and effect the outcome of how things play out. Without charactarer deaths, the fact that I overcame those obstacles and beat those challenges doesn’t matter. Why? Because there was no chance that I would fail or die, the obstacles were made to be overcome and no matter how idiotic or how many stupid decisions my character made, they were going to make it to the finish line anyways.
If I wanted to “play” a story with a fixed outcome, then why wouldn’t I have just watched a movie instead? My character shouldn’t be able to fall into the pit of lava and have a nice time slowly swimming back, because there isn’t much of a mechanic for failure if the ultimate failure is taken away.
Though I dislike no character deaths and would never ever play that way, aversion to character death is fine. Why? Because when I spend hours of time and energy into making my character and giving them a real and interesting personality, I don’t want to lose that character right away and be forced to restart due to anything other than my own stupidity or failure to properly deal with the defeatable threats in the game.
If you don’t like character death, don’t play with it. But at my table at least, the possibility of character death is what makes the game worthwhile and fun. Even though the experience of character will likely be neither of those things.
BoringBard's long and tedious posts somehow manage to enrapture audiences. How? Because he used Charm Person, the #1 bard spell!
He/him pronouns. Call me Bard. PROUD NERD!
Ever wanted to talk about your parties' worst mistakes? Do so HERE. What's your favorite class, why? Share & explain
HERE.It can be, for some players and stories.
Some people enjoy Game of Thrones, some enjoy Harry Potter, some enjoy both. Some weirdoes even have fun with Attack on Titan or My Little Pony:)
A lot of people are invested too much in the story of their characters and just do not want to see them die, especially if they have plans for them. It is like if the TV show was suddenly cancelled and there was no satisfying conclusion of the story. So for them permadeath of their character is not an option. However, there are people who will enjoy heroic death that will support the cause and some might have fun even with the most silly and easily avoidable deaths.
The whole topic is a very relative, it is like asking if coriander can taste good. It can, to some people.
PS: I agree that (with the exception of lowest levels) 5e feels very safe for the characters and it is (often) hard to experience permanent death of the character. I think that is what majority of players actually prefer, but I agree, that it also robs them of some experience and sense of accomplishment that comes with the early editions that were bit more brutal and unforgiving. I would say that experienced players should not have issues modifying the game to their liking, but for beginners it is better to start with a safer environment (and I would argue that especially the first level can be too deadly). It is really easy for absolute newbies to get themselves killed and if they are attached to their characters, it can ruin the game for them.
But how do you know when a death serves the narrative or not? What is the criteria for this? (Again, these are genuine questions, not rhetorical.)
My struggle with this approach is that it seems to assume that there's agreed-upon arc or trajectory for the "story" to take, and everyone at the table - DM as well as fellow players (and the dice!) - is going to stick to it. Whereas my decades of experience at the gaming table tells me that while very, VERY broad arcs or trajectories may be able to hold over the course of a campaign, you simply can't dictate where the story goes ahead of time, nor can you really predict where it's going. D&D is meant to be collaborative and, I'd argue, it's meant to have a fair amount of uncertainty (i.e., randomness) to it. Unless the campaign is on rails, allowing for no significant deviation or changes, the narrative often doesn't become clear until near the end of the campaign.
(And this is my problem with so many canon and third party modules and mini-campaigns: they're written like novels, not like games where the characters have full agency. Most of them provide thin or poor hooks/reasons for any sane character to continue on the adventure path.)
It's one thing to carefully craft a character with a detailed backstory, strengths and flaws, and (as a player) have a certain arc or trajectory in mind for the character. That approach, which generally doesn't require specific events at specific times, is good and easily integrated into almost any game. It's another to want the campaign to unfold as a seamless, consistent "narrative" - I guess you get what you want there, as it ends up being more a single person's vision, but D&D, as I've understood, isn't meant to be that. It's a collaborative, constantly evolving and growing "story", told by multiple people, and the end is far from certain or set.
Barring any foolish act from a player character, i think death is most fun during set piece encounters against important enemies than when dying in a random encounter with no importance in the campaign. I think my challenge as a DM resides in striking a good balance between keeping the threat of death omnipresent yet be relatively infrequent, as i don't want my campaigns to never kill PCs nor have them die so frequently.