"Run away! Run away!" ... what if they don't?

[Valmarius]

Yeah, I don't quite see the need to be that formal, explicit, or absolute about it. The genre has conventions, I see nothing wrong with leaving it at that. Though, again, nothing wrong with formal stake-setting, either.

I think being explicit can help, but I tend to do it before most rolls anyway.
Here's an example from a campaign I ran where the PCs specifically asked for PC's lives to not be at stake:

Very first session, PCs are on an airship that is under attack from hobgoblin sky-pirates (of course).
A number of the pirates head down into the ship and one PC rightly determines that they're headed for the engine room. Bad news for everyone on board.
He states that he's going to heroically swing on a rope, hoping to crash through a lower level window and intercept the pirates.
At this point, I lay out the stakes for him. Theoretically, falling from the rope and to his death could be a possibility, but the players don't want death as stakes.
So I say something along these lines, "Okay, that's going to be an acrobatics check. If you beat a DC of 15 you'll arrive ahead of the badguys, roll lower and you're forced to enter in a different window. You'll have caught up a bit, but you'll be behind them, not in front."

Since there were new players, who instantly reacted to his action with, "That's crazy. You might die!" I spelt out the possible outcomes. Also, it helped to teach them that failure can mean all kinds of things.

 

[Maxperson]

Try being a main-cast character with a deathwish. That'd get really tedious. It didn't seem too unclear: You want to apply a standard (death is final, 'deadly danger' necessarily means an established statistical probability of death that applies equally to anyone exposed to said danger) to figments of imagination, that are intended to evoke other figments that, persistently, are exposed to imaginary danger of imaginary death without every quite being imagined as dying (at least not without a good dramatic reason), and remaining dead permenantly?

Actually one bit is unclear: the problem with using redshirt style NPCs to establish a danger as deadly. If the PCs are not supposed to be aware that they are different from the NPCs, then seeing an NPC succumb to such a danger should establish the deadliness of the danger, shouldn't it?

The threat of death being discussed here isn't to the PC. Yes, the PC is under the threat of death, but the true threat of death only has meaning to the player. The PC is just a sheet of paper. If the player knows you aren't going to kill a PC, then there is no threat of death, even if the sheet of paper is supposed to feel threatened. Unless the DM is willing to kill off a PC, the players aren't going to feel the threat of death and that devalues the game for a lot of us.

 

[Imaculata]

Since there were new players, who instantly reacted to his action with, "That's crazy. You might die!" I spelt out the possible outcomes. Also, it helped to teach them that failure can mean all kinds of things.

I think the meanest thing I ever did as a DM was this:

I had an npc ghost-pirate hold a cutlass against the throat of some poor damsel, and threaten to slit her throat if the players came any closer. One of the players wanted to use a wand with a disarm spell, to disarm the pirate before he could kill her. So I said: "You and the ghost-pirate both roll for initiative. If the ghost-pirate goes first, the girl dies. If you go first, the ghost-pirate gets a save versus the spell, and she might still die if he makes his save."

Then I ended the session, and told them that we would make our rolls at the start of the next session. The players all cried out in agony.

Clearly telling the players the outcome before they take their action, not only gives them an option to reconsider, but it can dramatically up the tension. Especially if the action could end really badly.

(Oh, and in case you were wondering. The player barely rolled higher than the ghost-pirate, and the damsel was saved. )

 

[clearstream]

Yeah, I don't quite see the need to be that formal, explicit, or absolute about it. The genre has conventions, I see nothing wrong with leaving it at that. Though, again, nothing wrong with formal stake-setting, either.

It's worth thinking about what is going on mechanically...

A) the stake (losing a PC) is time played e.g. 1 hour per encounter
B) the odds are risk of death e.g. 0.5% (a half of one percent) per encounter
C) the pot is a mix of character and narrative advancement, e.g. gain 1/6th of a level per encounter

At first, it is easy for a DM to balance the wager. It "costs" a PC about 2 minutes per level. (Cost = risk*time. Rounded to nearest minute.)

10 sessions later, assuming a character survives, their stake has increased to 20 minutes per level. A DM could theoretically increase the pot to match, but that would amount to granting two levels per encounter! It becomes difficult to offer a fair wager to high level characters. DMs and players sense what is going on here, and make accommodations such as declining or removing chance of death.

In terms of the presented design, the D&D 5e approach appears to be to decline the stake by providing means to undo death. I think it is otherwise hard to explain the presence of Gentle Repose, Revivify, Raise Dead, Reincarnate, Resurrection and True Resurrection, and the niceties of their mechanics. Many DMs ignore that to a greater or lesser extent, and instead decline the risk of death. Which is also fine.

Whichever way one does it, my estimate is that chance of permanent death per encounter needs to decline from about 0.4% at tier 1 to about 0.2% at tier 4. My example values are derived from those implicit in the pacing guidelines given by the designers (e.g. in the DMG). Two or three encounters per session, and two or three sessions per level.

My feeling is that a conscious approach to the implicit wager makes death more tolerable, and allows it to take its place in the narrative.

 

[Imaculata]

Whichever way one does it, my estimate is that chance of permanent death per encounter needs to decline from about 0.4% at tier 1 to about 0.2% at tier 4. My example values are derived from those implicit in the pacing guidelines given by the designers (e.g. in the DMG). Two or three encounters per session, and two or three sessions per level.

My feeling is that a conscious approach to the implicit wager makes death more tolerable, and allows it to take its place in the narrative.

So, it seems like you saying that the chance of permanent death should decrease as the players rise in level, rather than increase? Or am I misunderstanding?

 

[clearstream]

So, it seems like you saying that the chance of permanent death should decrease as the players rise in level, rather than increase? Or am I misunderstanding?

Yes, chance of permanent death should decrease as characters rise in level. That is because rewards (relative to level) remain roughly constant, but the wager (time played, i.e. time lost if the character dies) has increased linearly.

One way to use the mechanics provided is to keep the rate at which characters die roughly constant, but allow them access to the revival spells appropriate for their level. So that permanent death becomes less frequent.

 

[Imaculata]

Yes, chance of permanent death should decrease as characters rise in level. That is because rewards (relative to level) remain roughly constant, but the wager (time played, i.e. time lost if the character dies) has increased linearly.

But wouldn't it make more sense to do the exact opposite? To have a learning curve, by starting easy (with a safety net between levels 1-3), and slowly increasing the change of death? After all, the higher level the players are, the more HP they have, and the more abilities they have to prevent death. Isn't that why monsters with insta-kill deathrays are of a high challenge rating?

 

[Oofta]

The threat of death being discussed here isn't to the PC. Yes, the PC is under the threat of death, but the true threat of death only has meaning to the player. The PC is just a sheet of paper. If the player knows you aren't going to kill a PC, then there is no threat of death, even if the sheet of paper is supposed to feel threatened. Unless the DM is willing to kill off a PC, the players aren't going to feel the threat of death and that devalues the game for a lot of us.

Which is something you should discuss in a session 0. Different people play for different reasons.

 

[iserith]

It's worth thinking about what is going on mechanically...

A) the stake (losing a PC) is time played e.g. 1 hour per encounter
B) the odds are risk of death e.g. 0.5% (a half of one percent) per encounter
C) the pot is a mix of character and narrative advancement, e.g. gain 1/6th of a level per encounter

At first, it is easy for a DM to balance the wager. It "costs" a PC about 2 minutes per level. (Cost = risk*time. Rounded to nearest minute.)

10 sessions later, assuming a character survives, their stake has increased to 20 minutes per level. A DM could theoretically increase the pot to match, but that would amount to granting two levels per encounter! It becomes difficult to offer a fair wager to high level characters. DMs and players sense what is going on here, and make accommodations such as declining or removing chance of death.

In terms of the presented design, the D&D 5e approach appears to be to decline the stake by providing means to undo death. I think it is otherwise hard to explain the presence of Gentle Repose, Revivify, Raise Dead, Reincarnate, Resurrection and True Resurrection, and the niceties of their mechanics. Many DMs ignore that to a greater or lesser extent, and instead decline the risk of death. Which is also fine.

Whichever way one does it, my estimate is that chance of permanent death per encounter needs to decline from about 0.4% at tier 1 to about 0.2% at tier 4. My example values are derived from those implicit in the pacing guidelines given by the designers (e.g. in the DMG). Two or three encounters per session, and two or three sessions per level.

My feeling is that a conscious approach to the implicit wager makes death more tolerable, and allows it to take its place in the narrative.

Interesting take!

 

[jasper]

This brings up a lot of internet ink.
1. Is the DM’s campaign railroad or sand box. If sand box and the encounter was fore fifth and sixth shadowed, and the pcs didn’t run. Sucks to be dungeon dressing for the next group of adventurers.
2. Do pcs have plot armour? If so how much?
3. Are the players calm enough for pc death and tpk.
4. Dice as they fall, dm fudge rolls, or be nice to pcs. AKA I not going to nova on low level.
5. Are the pcs playing as team or individuals? I have 37 names on Skully. About 40% are due to people worrying about killing the monster, than throwing a heal at a downed team mate.
6. As long as the players agree to the danger setting, no harm. And Raise Dead is only 1,250 gp away.

I havebeen back into D&D since June 2016. I have two tpk so far. One I will cop to being my fault. The other, well don’t fireball your position. Now I have gotten nicer as I gain more system knowledge but sometimes the dice will go south on the players.
….Oofta The giants simply don't see the group because I'm the DM and it's more interesting for the group to have a holy **** moment than to kill them….. Jasper thinks. Grabs his lucky black d20. “ I only have 5 hit points. I moon the giants Oofta! And insult their mothers!”
Now I have the rep of a killer dm. But this is balanced out as one dm has the rep of going easy on the pcs.