The Horriffic Emptiness of Dungeons

[Kindling]

I've just finished reading R Scott Bakker's The Judging Eye, and, while I don't want to give too many spoilers, I was particularly struck by an extended sequence in which a group of characters enter an abandoned underground city built by an ancient pre-human civilisation. Almost a classic dungeoncrawl, then, and it was superbly written - the author was able to create an incredible sense of atmosphere and of how utterly unnerving such an experience would be.

I thought one of the most interesting things about it, though, was that for the majority of the time they spent in the dungeon, the characters just explored, bickered, and got scared. Eventually they did encounter monsters at the climax of the novel, and it was hugely dramatic when they did, but the battle served as the release after a long time building up tension. For most of their way into the depths it was the environment itself that seemed to be what they were up against, the oppressive dark and silence of ancient subterranean space, haunted by the memories of the atrocities that caused its abandonment.

By contrast there doesn't seem to be anywhere near as much dead space in gaming dungeons. Typically more rooms will include traps, puzzles and/or monsters than won't. I know you can find a similar kind of slow build in some published works like Death Frost Doom or in the St Severine's Skull stuff that Steerpike posted on his blog a while back, but it seems to be the exception, part of what made DFD stand out to people as unique.

One thing that's been bothering me recently is how little I've ever actually used dungeon environments in games I've run, and how much I may have missed out on by doing that. The Judging Eye seems like the perfect example of how I'd like a dungeon I ran to feel, but I wonder if its slow build and "nothing happens... nothing happens... nothing happens... nothing happens... SUDDENLY MONSTERS" format might just not translate that well from novel to tabletop.

I like the idea of the environment and its exploration being the focus, and of storytelling through dungeon design - things for players to interact with that aren't traps or enemies. I like the idea of running the dungeoncrawl as something approaching a horror scenario - because although it takes a hero to brave the depths, they're probably not gonna feel like much of a hero while they're creeping around down there all alone in the dark. I like the idea of a sense of alien scale to the place - either physical scale or in terms of time and just how fucking ancient everything down there is. I like the idea of the dungeon being haunted - maybe not only by literal ghosts that might come through the walls clanking their chains but in a more oppressive, all-encompassing sense.

I don't like the idea of kicking down the door and fighting the orcs.

Would anyone like to talk to me about the way they run dungeons and possibly give me some advice? I want to use emptiness atmospherically, but I'm worried that emptiness could just be boring. I'd be interested to hear if anyone has tried anything similar and how it went - particularly, perhaps, from Steerpike about how St Severine's skull went in actual play, but in fact from anyone with a lot of experience with dungeons and how they like to handle pacing in them and why.

[Hibou]

I've just finished reading R Scott Bakker's The Judging Eye, and, while I don't want to give too many spoilers, I was particularly struck by an extended sequence in which a group of characters enter an abandoned underground city built by an ancient pre-human civilisation. Almost a classic dungeoncrawl, then, and it was superbly written - the author was able to create an incredible sense of atmosphere and of how utterly unnerving such an experience would be.

I thought one of the most interesting things about it, though, was that for the majority of the time they spent in the dungeon, the characters just explored, bickered, and got scared. Eventually they did encounter monsters at the climax of the novel, and it was hugely dramatic when they did, but the battle served as the release after a long time building up tension. For most of their way into the depths it was the environment itself that seemed to be what they were up against, the oppressive dark and silence of ancient subterranean space, haunted by the memories of the atrocities that caused its abandonment.

By contrast there doesn't seem to be anywhere near as much dead space in gaming dungeons. Typically more rooms will include traps, puzzles and/or monsters than won't. I know you can find a similar kind of slow build in some published works like Death Frost Doom or in the St Severine's Skull stuff that Steerpike posted on his blog a while back, but it seems to be the exception, part of what made DFD stand out to people as unique.

One thing that's been bothering me recently is how little I've ever actually used dungeon environments in games I've run, and how much I may have missed out on by doing that. The Judging Eye seems like the perfect example of how I'd like a dungeon I ran to feel, but I wonder if its slow build and "nothing happens... nothing happens... nothing happens... nothing happens... SUDDENLY MONSTERS" format might just not translate that well from novel to tabletop.

I like the idea of the environment and its exploration being the focus, and of storytelling through dungeon design - things for players to interact with that aren't traps or enemies. I like the idea of running the dungeoncrawl as something approaching a horror scenario - because although it takes a hero to brave the depths, they're probably not gonna feel like much of a hero while they're creeping around down there all alone in the dark. I like the idea of a sense of alien scale to the place - either physical scale or in terms of time and just how fucking ancient everything down there is. I like the idea of the dungeon being haunted - maybe not only by literal ghosts that might come through the walls clanking their chains but in a more oppressive, all-encompassing sense.

I don't like the idea of kicking down the door and fighting the orcs.

Would anyone like to talk to me about the way they run dungeons and possibly give me some advice? I want to use emptiness atmospherically, but I'm worried that emptiness could just be boring. I'd be interested to hear if anyone has tried anything similar and how it went - particularly, perhaps, from Steerpike about how St Severine's skull went in actual play, but in fact from anyone with a lot of experience with dungeons and how they like to handle pacing in them and why.

That book sounds really cool, and I'm going to have to check it out.

I agree with you in that I think this is the way dungeons should be, but most tabletops don't really support that level of non-combat tension development, where most of the interaction probably requires heavy player investment to be on a familiarity level such that they can speak to the situation without, say, having to roll dozens of Knowledge checks. There are some kinds of dungeon experiences that lend themselves more naturally to the encounters you're referring to, such as chasing an enemy or monster who has retreated into such a place, or happening upon one while doing something else and letting the players get curious.

I think the more action-y stuff we are accustomed to (e.g. your kicking in the door example) is fit for castles, battlegrounds, mountain pass ambushes, foot chases through cities, etc., while the dungeon should really provide the eerie and oppressive contrast to those settings that you are referring to. The common theme with all of those places is that they're close to sunlight and accessible by mount (mostly). Dungeons are stale, cramped, dark, and generally limit movement by either intent (e.g. a prison) or practicality (e.g. a mine), which I think should be your main focus to invoke the feeling you want.

I think one of the first options you have is to make it apparent somehow that literally nobody except the players even know that it's there, or maybe that if someone does they're keeping it secret effectively. This doesn't mean it has to be completely unknown; it could just be so remote that no one knows much or cares to find out, or that the players discovered a much larger portion of what has long been believed to be a tiny crypt or reservoir or something.

With this kind of adventure you will probably want to take time to develop tension about the place before they even enter it. Maybe if it's an ancient underground complex, some of its ceiling has collapsed and a dangerous, dark hole has opened up on a horse trail, or it was flooded partially by a river. Thinking about those kinds of environments creeps me out but also makes me want to go find out more about them, so that might be a good feeling to try to invoke. You should also look up pictures of abandoned sites, urban decay, flooded ruins, etc. to get yourself a feel for how to describe these places when you develop your own.

In music I often hear people use variants of the old saying that absence of sound can be as significant as an abundance of it. If the dungeon you have developed is mostly empty, crumbling, or filled with useless and rusted doodads, the ones that you place with significance will be all the more notable when they are witnessed. I can think of an early example in my DMing experience where I was a bit too over-the-top with a dungeon where people were tortured - lots of bodies, equipment, etc. However, I think if I had changed a few details and made it a more subtle scene with a couple of very old bodies and black blood stains that looked as if they were trying to escape from the room whose door had slammed shut on them, it would've made for a more compelling scene. I doubt any of this is new to you, but it is important to try and remember the KISS principle when you start trying to do too much.

Edit: sorry, on mobile and accidentally hit submit when my post was incomplete

[Steerpike]

I want to use emptiness atmospherically, but I'm worried that emptiness could just be boring. I'd be interested to hear if anyone has tried anything similar and how it went - particularly, perhaps, from Steerpike about how St Severine's skull went in actual play, but in fact from anyone with a lot of experience with dungeons and how they like to handle pacing in them and why.

I have, as you might imagine, thoughts on this!

My current real-life 5E game also plays with this idea - it's set atop a gigantic megadungeon - but I definitely did more of this in St. Serevine's Skull, and have used similar techniques in Ravenloft and Lamentations of the Flame Princess Games. I ran an adventure in my main campaign about a year back that was basically a classic haunted house thing that worked along similar lines - there were 4 levels to it, with precisely 1 monster per level, but with 5-10 rooms per level (so, roughly 10-15% occupied, so to speak), and some of the monsters moved around quite a bit. Although I haven't run it, The God That Crawls I think would pull this off really well.

Here's a collection of assorted observations:

-This sort of thing generally tends to be more effective the fewer players you have. The more players you have, the safer everyone feels, and also the more people argue and discuss where to go next, and the less agency people have over the exploration. This leads to more tedium and less suspense. But if you have 2-4 players, I think the "mostly empty" dungeon can work really well. It can still work with a bigger group but you need to be more on the ball with keeping things moving.

- If your players are expecting balanced, level-appropriate encounters with commensurate treasure and "fair" challenges, mostly-empty dungeons are drained of much of their suspense. If they're not sure what to expect and are aware that they may potentially be confronted by things they'll need to run from in order to save themselves, there's a lot more tension.

- On the other hand, making things too deadly too often can lead to players becoming so paranoid the go into what I think of as "SWAT mode" and comb through every room with maximum caution, and/or they just will be too terrified to touch anything. Both of these reactions tend to be pretty boring, so what needs to happen is the idea that something is down in the dungeon with them, or that danger is nearby, but that they are still most likely going to be okay most of the time. There needs to be a balance between safe and scary.

- Emptiness plays best when the players can still find information, clues, and objects - when the environment is telling a story. Graffiti, dead bodies, barricades, bloodstains, journals, carvings, murals, tapestries, paintings, rooms which served a clear function, rooms where something dramatic has occured. Ideally these things should in some sense connect to the reason the players are present, even indirectly.

- These sorts of details are also a great way to reveal a monster's presence without actually showing the monster, which is so absolutely key to successful horror build-up. Good horror needs foreshadowing. It's like a striptease - you just see glimpses at first, you don't just spring the monster on someone right away. Slime trails. Suggestive bones. Gougings on a wall. Strange sounds. Blood-stained scraps of diary describing previous encounters with the thing, penned by survivors. This is what helps mitigate the "nothing happens... nothing happens... nothing happens... nothing happens... SUDDENLY MONSTERS" thing. I agree with Hoers that this can go over-the-top into gruesome territory too quickly, so these things need pacing, too.

- Give players escape routes when the monster shows up. Chases are much scarier than fights. Being stalked and hunted and getting lost in a giant maze of empty rooms is a lot scarier than walking through a bunch of empty spaces before being cornered and fighting.

- In a similar vein, make at least some of the very few monsters mobile.

- Here's the structure I would avoid: Characters enter empty rooms filled with random details -> find monster -> kill monster -> characters explore more empty rooms with random details -> players find additional monster(s) -> kill monsters -> treasure -> leave dungeon.

- Here's more of a structure that I think works better: Characters enter empty rooms -> find vague evidence of monster -> more empty rooms -> brief glimpse of monster (or what could be monster) from afar, or evidence that monster was recently present -> more information gleaned about monster and/or treasure or history of the place -> encounter monster, know they're outmatched because they've obtained information about it, narrowly escape, perhaps not unscathed -> find more evidence of monster, hinting at weaknesses -> find weapons/tools to defeat monster while being stalked and harassed by monster -> hunt monster down and kill it with cleverly executed plan.

I'll think of more later. This is totally my jam, though.

[Hibou]

I want to use emptiness atmospherically, but I'm worried that emptiness could just be boring. I'd be interested to hear if anyone has tried anything similar and how it went - particularly, perhaps, from Steerpike about how St Severine's skull went in actual play, but in fact from anyone with a lot of experience with dungeons and how they like to handle pacing in them and why.

I have, as you might imagine, thoughts on this!

My current real-life 5E game also plays with this idea - it's set atop a gigantic megadungeon - but I definitely did more of this in St. Serevine's Skull, and have used similar techniques in Ravenloft and Lamentations of the Flame Princess Games. I ran an adventure in my main campaign about a year back that was basically a classic haunted house thing that worked along similar lines - there were 4 levels to it, with precisely 1 monster per level, but with 5-10 rooms per level (so, roughly 10-15% occupied, so to speak), and some of the monsters moved around quite a bit. Although I haven't run it, The God That Crawls I think would pull this off really well.

Here's a collection of assorted observations:

-This sort of thing generally tends to be more effective the fewer players you have. The more players you have, the safer everyone feels, and also the more people argue and discuss where to go next, and the less agency people have over the exploration. This leads to more tedium and less suspense. But if you have 2-4 players, I think the "mostly empty" dungeon can work really well. It can still work with a bigger group but you need to be more on the ball with keeping things moving.

- If your players are expecting balanced, level-appropriate encounters with commensurate treasure and "fair" challenges, mostly-empty dungeons are drained of much of their suspense. If they're not sure what to expect and are aware that they may potentially be confronted by things they'll need to run from in order to save themselves, there's a lot more tension.

- On the other hand, making things too deadly too often can lead to players becoming so paranoid the go into what I think of as "SWAT mode" and comb through every room with maximum caution, and/or they just will be too terrified to touch anything. Both of these reactions tend to be pretty boring, so what needs to happen is the idea that something is down with them, or that danger is nearby, but that they are still most likely going to be okay most of the time. There needs to be a balance between safe and scary.

- Emptiness plays best when the players can still find information, clues, and objects - when the environment is telling a story. Graffiti, dead bodies, barricades, bloodstains, journals. Ideally these things should in some sense connect to the reason the players are present, even indirectly.

- These sorts of details are also a great way to reveal a monster's presence without actually showing the monster, which is so absolutely key to successful horror build-up. Good horror needs foreshadowing. It's like a striptease - you just see glimpses at first, you don't just spring the monster on someone right away. Slime trails. Suggestive bones. Gougings on a wall. Strange sounds. Blood-stained scraps of diary describing previous encounters with the thing, penned by survivors. This is what helps mitigate the "nothing happens... nothing happens... nothing happens... nothing happens... SUDDENLY MONSTERS" thing. I agree with Hoers that this can go over-the-top into gruesome territory too quickly, so these things need pacing, too.

- Give players escape routes when the monster shows up. Chases are much scarier than fights. Being stalked and hunted and getting lost in a giant maze of empty rooms is a lot scarier than walking through a bunch of empty spaces before being cornered and fighting.

- In a similar vein, make at least some of the very few monsters mobile.

- Here's the structure I would avoid: Characters enter empty rooms filled with random details -> find monster -> kill monster -> characters explore more empty rooms with random details -> players find additional monster(s) -> kill monsters -> treasure -> leave dungeon.

- Here's more of a structure that I think works better: Characters enter empty rooms -> find vague evidence of monster -> more empty rooms -> brief glimpse of monster (or what could be monster) from afar, or evidence that monster was recently present -> more information gleaned about monster and/or treasure or history of the place -> encounter monster, know they're outmatched because they've obtained information about it, narrowly escape, perhaps not unscathed -> find more evidence of monster, hinting at weaknesses -> find weapons/tools to defeat monster while being stalked and harassed by monster -> hunt monster down and kill it with cleverly executed plan.

I'll think of more later. This is totally my jam, though.

You hit the nail on the head with the second structure.

There's a pretty key example of it that I'll give from the movie Alien (spoilers below):

[spoiler=Alien]
It's a particularly good example because it speaks to a lot of Steerpike's points. In the movie the characters encounter a dungeon like this to begin the movie, and in a sense go through the entire sequence - abandoned, decrepit place that is rather inhospitable, with hints of its origins and a growing dread, followed by a long, slow exposure to the room with the eggs. John Hurt's character eventually encounters and is incapacitated by the monster of this dungeon: the face-hugger. This prompts the crew to book it back to the ship to save him (and themselves).

What's particularly chilling about that first act is that it's all just a setup for the real dungeon - which is their own ship. The thing about Alien is that not only does it do all of the tension-building stuff mentioned above, but it turns a place that should be relatively familiar and safe to the crew into an even more hostile place. Suddenly the tight-but-efficient corridors, monitoring equipment, and life support systems themselves become unfamiliar, frightening places where the absence of noise or movement adds to the tension, rather than reassuring them. The xenomorph uses the ship's ducts, maintenance shafts, and other spaces to move around as the crew frantically tries to come up with a solution, and most of the glimpses they get until later on are either encounters where they outright die (perhaps some of them ought to be considered NPCs), or it's by proxy as their fellow crew are killed. I don't know that I need to say a whole lot more since anyone reading this has probably seen it anyway, but you get the idea.
[/spoiler]

As a recap, it has:

-A small group, which dwindles as the encounters with the monster take place
-A foe they're surprised by that poses a problem for their quick solutions (the face-hugger and early xenomorph), followed by one they just have to run from mostly
-Even though the ship is unsafe, they have that sense of partial safety as they regroup and stay in safe areas
-No shortage of the environment playing a role as the empty passageways and such become frightening, and their comm systems and whatnot give them clues of the monster's whereabouts
-Room to escape, at least for some
-Monster is extremely mobile, and claustrophobic corridors with lots of dead ends make for a terrifying chase

------------

The problem that seems to exist, assuming you're using a pretty standard game system like D&D, C&S, etc. is that it's hard to create a lot of these atmospheric conditions because of a multitude of problems, most notably:

1) Some players are going to love this theme, while others aren't up for it at all; this kind of campaign is probably going to have a much higher death count than a typical game, and you'll have to get players on board who are okay with having these kind of run-or-die threats be regular (so far as regular means "sometimes or often the climactic encounter of an adventure")

2) It's really easy to metagame and lose that kind of suspense, at least for people like myself, once you speculate on bounds for the monster's abilities

How would you guys handle these problems?

[Steerpike]

The problem that seems to exist, assuming you're using a pretty standard game system like D&D, C&S, etc. is that it's hard to create a lot of these atmospheric conditions because of a multitude of problems, most notably:

1) Some players are going to love this theme, while others aren't up for it at all; this kind of campaign is probably going to have a much higher death count than a typical game, and you'll have to get players on board who are okay with having these kind of run-or-die threats be regular (so far as regular means "sometimes or often the climactic encounter of an adventure")

2) It's really easy to metagame and lose that kind of suspense, at least for people like myself, once you speculate on bounds for the monster's abilities

How would you guys handle these problems?

These are both pretty big problems, I totally agree.

Problem 1 is tough, because there really are some players who just really sincerely don't like the idea of even theoretically being outmatched or having their characters be in "real" danger. I think it can be mitigated to a certain degree by being as generous as possible in terms of letting players get away. Drawing attention to escape routes is important here, but also making it clear that the threat is genuine by describing it properly, maybe even giving characters some sort of check to size it up. System makes a genuine difference here. If you're playing LotFP or 5E or a retroclone or Call of Cthulhu or Fear Itself or whatever, it's pretty much baked into the expectations of the game that this sort of power disparity will occur. If you're playing 4th edition or to a lesser extent Pathfinder, or if you're playing something more like Fate or another narrative-heavy/storyelling game, it's not going to be as obvious unless you strain against the system to a degree. This is, incidentally, one of the reasons I really gravitated away from Pathfinder and towards 5th.

Problem 2, I think, is a bit easier to solve if you use a scratch-made monster and play it smartly. If you see your players having an OOC discussion about monster abilities, as a DM I'd suggest nudging them to have an in-character discussion instead. Something like this deleted scene from Alien (I also agree that Alien is a near-perfect example of the Horrifying Emptiness of Dungeons done well in film form; The Descent also comes to mind). Honestly it's scenes like that that are some of my favourite scenes in horror movies and in D&D - watching people carefully plan and think creatively to solve a problem.

[sparkletwist]

Honestly it's scenes like that that are some of my favourite scenes in horror movies and in D&D - watching people carefully plan and think creatively to solve a problem.

Scenes like that are great in fiction, but I don't really see the circumstances existing to create them in most RPG's, particularly if you're playing where there is "real" danger. The big thing that makes the scene from Alien so good (and probably why you prefer IC planning to OOC planning) is that it isn't just a group of people tactically planning, but rather the characters genuinely reacting to the threat. They're scared, they're confused, they're in disagreement, and it makes for great drama. However, this isn't usually how it works in games. Players tend to plan tactically to the best of their ability out of character, leading to problem #2 cited by Hoers. During key moments like this they tend not to focus particularly much on their character's feelings or flaws-- and this goes double if you're playing the sort of game where losing the fight means you're dead so you don't get to play your character any more.

You said that using Fate you'd have to "strain against the system" but (perhaps not surprisingly) I don't see a problem here, and in fact I think that Fate is actually set up to be pretty good at delivering the sort of scene that you seem to want to create. The use of compels against a character's aspects for fate points can provide a mechanical reward to acting in character even when what the character would do isn't the smartest option, so characters will be more prone to acting scared and confused, because (tactically speaking) there's actually an incentive to do so. Fate's use of softer failure states can mean that players can concede a conflict by realizing they are outmatched and fleeing, which is exactly the sort of imagery you want. There isn't the same "real" danger, but putting a character into "real" danger doesn't inherently create any sense of IC tension. It creates OOC tension, but the player will often not translate that into IC tension. Instead, the game will slow down and good RP will be brushed aside as everyone goes into tactics mode and tries to make the best plan possible so as to not die. And rightfully so! You're playing for keeps, so they are too.

[Steerpike]

Scenes like that are great in fiction, but I don't really see the circumstances existing to create them in most RPG's, particularly if you're playing where there is "real" danger... During key moments like this they tend not to focus particularly much on their character's feelings or flaws-- and this goes double if you're playing the sort of game where losing the fight means you're dead so you don't get to play your character any more.

I enjoy the planning and tactical thinking-through of these scenes as well as/in addition to the drama of dealing with a threat, so it could be we're enjoying different things about that scene. That said, I just flat-out disagree that the possibility of a character dying (which need not always mean the character becomes unplayable) inherently discourages drama. This goes to the crux of what I think is a fairly long-standing difference of opinion or experience between us. You're suggesting that "real" danger to a character doesn't create IC tension, and that might be true for some players, but I simply disagree that it's true for all. I've seen players have very genuine emotional reactions when faced with "real" threats - dread, anxiety, paranoia, low-key panic. I've seen hands trembling in these moments, players practically shouting "SHUT THE DOOR!" at one another after an escape, groaning with dread on hearing a monster moving in an adjacent room. You don't need to provide incentives in the form of mechanical rewards for roleplaying a desire for self-preservation and fear of a threat when the players are already experiencing a version of the exact same state - at least, not for me, anyway. This isn't to disparage those sort of incentives generally, I just don't think they're the only answer to Hoers' problem. The goal here, at least for me, isn't (just) to get characters to act a certain way, it's to elicit actual emotional responses form players themselves.

I don't disagree that Fate could produce a scene where the players perform their characters' fear or anxiety very well, and it might even be able to conjure some of those feelings in players, but I also don't think that making "a plan not to die" necessitates stripping away the emotion from a scene. I do think Fate could handle this sort of thing fine, and possibly better for certain types of players.

For, me I think Hoers' problem #2 isn't caused by the "realness" the threat, it's caused by the break in immersion that comes from remembering that we're playing a game with mechanics, and that the monster has particular statistics and mechanical attributes. In other words, it comes from remembering the unrealness of the threat. This sort of thinking disenchants the monster, strips it of its mystery and otherworldly terror and loathsomeness, because it's reduced to a set of numbers to defeat. Seen this way, at least for me, there's a way in which mixing in mechanical incentives for roleplaying in moments like the one being described can actually hinder the production of real emotions in players, since those incentives can form part of the reminder of the "mechanical" nature of the game being played. Hence, keeping discussions as in-character as possible obfuscates the mechanical dimension of the monster (since characters can't talk about statistics) and tries to sustain the collapse between character and player so that the character is never stepping outside of the character's head-space.

I know we're going to disagree on this, though. I just don't think we approach the player-character relationship the same way. Which is interesting, because I think somehow we still DM well for one another. I'm going to probably use some of the techniques I described above in Alptraum, so you'll have to tell me how it goes, though obviously it will be Fate, so perhaps it'll offer an interesting "test case" for some of our contentions.

[sparkletwist]

which need not always mean the character becomes unplayable

You're right, so I should clarify, I was talking about permadeath. In games where death-with-resurrection makes sense in the world, that is a softer failure state, and I have less issues with it. I still have some, but that's a tangent and also something we've been over.

You're suggesting that "real" danger to a character doesn't create IC tension, and that might be true for some players, but I simply disagree that it's true for all. I've seen players have very genuine emotional reactions when faced with "real" threats - dread, anxiety, paranoia, low-key panic.

Actually, what I said was that putting a character into "real" danger doesn't inherently create any IC tension, but it does create OOC tension. The things you've cited are all OOC tension. You're making the players react but this does not necessarily say anything about the characters, and it does not really automatically lead to any roleplay. They still have to respond in character in order for anything they're thinking or feeling to be reflected in the character's actions.

You don't need to provide incentives in the form of mechanical rewards for roleplaying a desire for self-preservation and fear of a threat when the players are already experiencing a version of the exact same state - at least, not for me, anyway.

For players to be roleplaying effectively, they still have to respond how the character would, though. If you just make the player feel some desire for self-preservation, this won't necessarily manifest itself in any IC way. Since the tension belongs to the player, that player may instead try to seek player-oriented (rather than character-oriented) solutions to the problem. In other words, "gamism," not "simulationism."

For, me I think Hoers' problem #2 isn't caused by the "realness" the threat, it's caused by the break in immersion that comes from remembering that we're playing a game with mechanics, and that the monster has particular statistics and mechanical attributes.

I saw this firsthand with a player who was new to Asura, and, really, new to the whole "Fate way of doing things." The party had gotten into a fight with a powerful brute who could dish out some really impressive damage, and one of the other player characters had just taken a critical hit and was in pretty bad shape. Immediately after that, it was the new player's turn, and he did not react in character to this at all-- it made him tense, sure, but his focus was immediately on how low the other character's hit points were and what mechanical abilities his own character had to help out so that another player wouldn't lose his character. Interestingly, in this case, the new player only dropped back into character and started roleplaying again when we explained some of how the game works and the mentality behind it, which meant that there was no "real danger." Situations like this are largely what shape my thinking in this matter and why I prefer to do things the way I do.

I'm going to probably use some of the techniques I described above in Alptraum, so you'll have to tell me how it goes, though obviously it will be Fate, so perhaps it'll offer an interesting "test case" for some of our contentions.

Definitely, especially since I'll be reacting to things in a "Fate-oriented" way. I'm particularly interested in how you'll use compels.

[Steerpike]

Actually, what I said was that putting a character into "real" danger doesn't inherently create any IC tension, but it does create OOC tension. The things you've cited are all OOC tension. You're making the players react but this does not necessarily say anything about the characters, and it does not really automatically lead to any roleplay. They still have to respond in character in order for anything they're thinking or feeling to be reflected in the character's actions.

I guess what I'm calling "in character" tension is tension where the emotional states of the players organically matches or mirrors that of the characters. They are "synched" to characters in this moment so that they share in a similar anxiety. This could be contrasted with a mismatched tension, where the players are calm (or maybe even laughing, making fun of their characters' plight) but having their characters freak out and panic - still effectively roleplaying, in a certain sense, but by making their characters perform, rather than experiencing emotional reactions themselves. I think of this sometimes as 1st person versus 3rd person roleplaying - I'm definitely a 1st person roleplayer.

For players to be roleplaying effectively, they still have to respond how the character would, though. If you just make the player feel some desire for self-preservation, this won't necessarily manifest itself in any IC way. Since the tension belongs to the player, that player may instead try to seek player-oriented (rather than character-oriented) solutions to the problem. In other words, "gamism," not "simulationism."

I agree with the gameism/simulationism distinction here, but this is why as a general practice I try to minimize obtrusive mechanics or meta-level considerations in horror games of this type. Ideally, for me, the action the player takes for self-preservation should match the action the character takes for self-preservation as much as possible. If, as the DM, you've genuinely unnerved the player and given them the appropriate feeling of menace and threat, they should be deep in the character's head, not thinking about meta-level rewards or incentives. I don't think it's impossible to mix some of those in to an extent, but the more the game's mechanics intrude, from my perspective, the less "synched" I feel with a character. If I know for a fact my character is not in any real danger, I can be a lot more cavalier with them, and I might really need a lot of meta-level incentives to have them act in a way that still makes sense in-game. If my character is in real danger and I want to protect the character, then the character and I are sharing headspace and motivations, trying to problem-solve.

So in your example, for instance, I don't think it's that crazy that upon seeing a party-member get severely injured, you think about what you can do to try and get them to safety or heal them or shield them or something. If they're not in real danger and you know, as the player, they're not in real danger, and their player also knows they're not in real danger and thus is not desperate for help, then everyone's mind-set is totally at odds with how the actual people in that situation would react knowing that one of their party is in danger of dying, isn't it?

[Steerpike]

Here's an example from another medium. Take this scene in the brilliant horror video game Outlast (or this similar one from the sequel). This game is one where there is a permadeath mode where you have to start over if you die.

In this scene, the character is in "real" danger, and the player knows it. The player must direct the character to run from his assailants, barricade doors, sneak through vents, leap obstacles, and generally scamper away from enemies. Player and character are synched, because the player by this point is freaked out (my heart hammers every time I play this scene) and so is their character. The actions they take make sense in-universe. When permadeath is enabled the tension of this scene is nuts even though I've probably played it 10, 15 times.

If, on the other hand, the player has enabled god-mode and knows that they cannot really be harmed - if they feel safe and assured that even if the bad guys catch up to them, their character will be fine - the scene is instantly a lot less terrifying. Now there's no real reason for the player to hurry or barricade doors, even though presumably the character would be freaking out and trying to get away. The player and character are now out of synch. The player might still take those actions, but their motivations for taking them are going to be very different from the character's motivations. There would need to be some other reason for them to barricade doors and sprint like a rabbit down corridors, like maybe they get bonus experience if they get out unscathed, or unlock an achievement for making the run in a certain time, or something. Behaviour that used to just make sense because of shared headspace now needs an extrinsic motivation because that headspace is compromised.

This scene would also be a lot less extreme if gamey elements like hit-point bars, hovering power-ups, damage totals, and other numbers and extradiagetic features were included. Those, to me, are the equivalent to rules discussions and mechanical interventions popping up, Hoers' problem #2 rearing its head.

That's an extreme example, but I hope it illustrates what I'm trying to say to some extent. Obviously roleplaying games are a bit different from videogames, but I think most of the principles of identification between characters and players are pretty broadly similar.

[sparkletwist]

I think of this sometimes as 1st person versus 3rd person roleplaying - I'm definitely a 1st person roleplayer.

That makes sense, and by that metric I'm definitely a 3rd person roleplayer.

Ideally, for me, the action the player takes for self-preservation should match the action the character takes for self-preservation as much as possible.

I honestly don't quite understand what you're saying here, but my guess is that what you're saying is that the action the character takes should be what the player thinks the character would "realistically" do, unfettered by game mechanics thinking. That's a worthwhile goal, but unfortunately no RPG mechanics are really good enough to allow doing that and not running into nasty surprises from time to time. Yes, this is absolutely a flaw in the mechanics, but that doesn't change the fact that it exists; if you're playing with permadeath and want to not die, when the going gets tough, it's almost always going to be a healthy amount of system mastery that keeps you alive.

If I know for a fact my character is not in any real danger, I can be a lot more cavalier with them, and I might really need a lot of meta-level incentives to have them act in a way that still makes sense in-game.

Why is it so hard to roleplay a character realistically facing mortal peril without the prospect of having the character being taken away from you hanging over your head? We roleplay greedy characters even though having gp gains nothing in the real world, we roleplay altruistic characters even though saving NPCs doesn't actually help anyone real, and we roleplay stupid characters even though we're smart people who know a lot about how the game works. So why is this any different?

(That said, Asura is based around superhumans carrying out gonzo over-the-top heroics, so in its case characters acting like they can't die is probably not unreasonable anyway...)

So in your example, for instance, I don't think it's that crazy that upon seeing a party-member get severely injured, you think about what you can do to try and get them to safety or heal them or shield them or something.

That's reasonable, but my point was that it seemed pretty obvious the player's thinking wasn't so much "what can I (as the character) do to save him?" but rather "what can I (as the player) do mechanically within this system to keep his hp from dropping below 0?" As I mentioned before, RPG systems are not perfect, and different genres are going to require different rules anyway-- how you stay alive in an action movie RPG is probably going to be very different than how you stay alive in a gothic horror RPG-- so ultimately, by creating a situation that is "perilous" (on some level) for the player, you are going to encourage thinking in terms of game-mechanical tactics rather than character-level tactics, because it's ultimately the game mechanics that determine whether the character lives or dies.

If, on the other hand, the player has enabled god-mode and knows that they cannot really be harmed - if they feel safe and assured that even if the bad guys catch up to them, their character will be fine - the scene is instantly a lot less terrifying.

I think this is kind of a strawman, considering you are comparing RPG play styles with using an actual cheat code in a video game. Yeah, if you put in a cheat code, you're probably not going to be playing like you're in peril...

[Steerpike]

I could be entirely off-base here, but I thought of what might be a pretty decent analogy, I think, for our different ways of seeing the player-character relationship. I am basically Stanislavskian whereas I think you are perhaps more sympathetic to Diderot's paradox of the actor, what Stanislavski would call the "art of representation."

For the Stanislavskian, the goal is to "experience the role," to experience the genuine emotions of the character so that you and the character almost entirely fuse - "the actor immerses him or herself in the circumstances of the fictional reality experienced by the character in the play. In a state of absorption, he or she responds 'naturally' and 'organically' to that situation and the events that proceed from it."

In contrast, the art of representation is all about nailing a part, standing in a sense at arm's length from the character, "the object of [the actor's] acting," and maintaining a distinction between the two the better to craft a performance. For Diderot the character is a model that the actor imitates, and in order to make the right sorts of calculations to portray the character as effectively as possible, the actor has to maintain a distance between themselves and the character, to ask themselves first "how would this character act?" and then "how can I achieve a representation of that action?"

Maybe? I could be crazy.

Yes, this is absolutely a flaw in the mechanics, but that doesn't change the fact that it exists; if you're playing with permadeath and want to not die, when the going gets tough, it's almost always going to be a healthy amount of system mastery that keeps you alive.

It's never going to be perfect, for sure, but the question for me is basically how to minimize this problem.

Why is it so hard to roleplay a character realistically facing mortal peril without the prospect of having the character being taken away from you hanging over your head? We roleplay greedy characters even though having gp gains nothing in the real world, we roleplay altruistic characters even though saving NPCs doesn't actually help anyone real, and we roleplay stupid characters even though we're smart people who know a lot about how the game works. So why is this any different?

I'm not necessarily defending full-on permadeath here (my own games don't usually even use it) so much as the the presence of intentionally unbalanced encounters where there are clear fail conditions which may include death, with or without the perma- in front of it.

But I don't think mortal peril is unique here. When I present players with moral conundrums, for instance, those problems are meant to elicit genuine feelings not of terror but pity and empathy, so that when they act altruistically they are also "synched" with their character because the world feels real enough that they care about the moral implications of their actions and they feel good about helping people and guilty about hurting them. I definitely think that greed acts similarly - when my players get cool magic items, their eyes light up much in the same way their characters' would, because now they have a cool new toy to play with as part of a power fantasy. In some recent sessions I had two players get into a very intense (and fun) moral argument with a lot of what felt like quite genuine-feeling hand-wringing over the completely-not-real-but-still-real-feeling consequences of their actions. I've also had players feel pretty guilty over collateral damage their characters did, to the point where they went out of their way to make amends and set right what went wrong.

...you are going to encourage thinking in terms of game-mechanical tactics rather than character-level tactics, because it's ultimately the game mechanics that determine whether the character lives or dies.

Maybe. Maybe not. A good system (for this kind of play) tries to align game and character tactics, right? But I can think of examples that problematize this.

Let's say the characters enter a room in a nearly-empty dungeon and hear something slithering around behind the next door. They know from reading some dusty scrap of journal that some sort of mutant snake-woman thing of terrible deadliness lives in this place. They've found huge masses of moulted skin in some of the abandoned chambers, as well as a corridor full of crushed bones from a dozen warriors, as from some monstrous anaconda's coils. In another room, they found the body of another adventurer, something of a local legend, who came here to slay the beast; he's dead, his face black with poison, his body partially eaten, his magical sword snapped in two. Murals they've glimpsed indicate that this thing has incredibly lethal poison. Somewhere in the complex there's supposed to be an amulet that protects against this venom, but they haven't found it yet. There are also hints that the snake-woman dislikes fire.

They know absolutely nothing about its AC, HP, attack modifier, or poison DC. They have no clue what its challenge rating is, none at all. They don't know how many attacks the thing gets or what its perception roll is or anything. But they know that their DM is not one to provide only "level-appropriate" encounters.

The door, ajar, begins to creak open. They glimpse yellow, slitted eyes, an image the journal mentioned. The characters, realizing the snake-woman is upon them, flee, with the fighter sticking around just long enough to desperately hurl a flask of alchemist's fire at the sinuous shape that fills the doorway. They hear an angry hissing as they turn tail and slam the door behind them, jamming it shut. The wizard casts arcane lock to seal the door. The characters live, but hear the monster thrashing around inside, a smell of burning scales. The thing is still alive, just temporarily deterred.

I'd argue that the above involves almost zero real system mastery. The characters survived not because someone knew when to use Mechanical Ability X but because they had paid attention to the dungeon and acted prudently. Eventually when they actually fight the thing system mastery might come in to a certain unavoidable extent, but the "Aliens" style in-character conversation they're going to have when they get to a safe location probably isn't going to be wild guesses about hit point totals but rather how to find the amulet and then how to lure the snake-woman into a trap they're going to set using the oil they found on an abandoned sub-level... or whatever, you get my drift. The conversation is about using in-universe resources and the environment and skills the characters possess, not numbers.

I'm making up a story here, but this is not far off from encounters I've run.

I think this is kind of a strawman, considering you are comparing RPG play styles with using an actual cheat code in a video game. Yeah, if you put in a cheat code, you're probably not going to be playing like you're in peril...

That's fair, I was picking a very extreme example to try and illustrate the point. On the other hand if you have 100% assurances from the DM that death is completely off the table for your character unless you give prior permission - which is a way some people for sure actually play - isn't that effectively a kind of low-key god-mode?

[Ghostman]

An extreme approach to this kind of horror gaming would be narrowing the scope of game mechanics so that they cover only direct confrontation, leaving everything else to be handled via freeform RP. If those rules were sufficiently biased against PCs, then there might emerge an incentive to avoid actions that invoke game mechanics and instead search for solutions in the unmechanized space. That would be in stark contrast to most RPGs, where working within the frame of game mechanics is almost always preferable because those mechanics represent a known territory, a kind of roadmap that can be used to navigate toward predefined victory conditions, while the space that falls outside the mechanics is an unknown territory completely subject to the game master's whims. Thus the rules that cover direct confrontation would need to be so obviously rigged that players find it preferable to risk diving headlong into the unknown.

[LoA]

To be frank, I don't think this sense of dread is worth trying to convey in any of the RPG's mentioned so far. I'm wondering if there are any games that were tailored made for this sort of thing though.

[Steerpike]

An extreme approach to this kind of horror gaming would be narrowing the scope of game mechanics so that they cover only direct confrontation, leaving everything else to be handled via freeform RP. If those rules were sufficiently biased against PCs, then there might emerge an incentive to avoid actions that invoke game mechanics and instead search for solutions in the unmechanized space. That would be in stark contrast to most RPGs, where working within the frame of game mechanics is almost always preferable because those mechanics represent a known territory, a kind of roadmap that can be used to navigate toward predefined victory conditions, while the space that falls outside the mechanics is an unknown territory completely subject to the game master's whims. Thus the rules that cover direct confrontation would need to be so obviously rigged that players find it preferable to risk diving headlong into the unknown.

To be frank, I don't think this sense of dread is worth trying to convey in any of the RPG's mentioned so far. I'm wondering if there are any games that were tailored made for this sort of thing though.

Lamentations of the Flame Princess essentially exploits the rules-minimalism of older editions to accomplish something not far off from what you're describing, Ghostman, and I'd put it forth as the closest to D&D that seems purpose-built from cultivating a sense of dread. LotFP consciously aims to "flatten" the power level and progression of characters to rig combat against them. For example, characters gain drastically fewer Hit Points per level after a certain point in the game. There are a few mechanics for things other than fighting, but really not that many.

Also, "unmechanized space" is a really good phrase for what I think a lot of old school play (horror-oriented or otherwise) aims to produce.