DM advice: How do you NOT kill your party?

[Caliban]

The point that tends to get overlooked, whenever someone brings up that fallacy, is that most people in the world actually aren't Scotsmen. Likewise, most games are not First-Person Shooters, or ball-games, or tabletop RPGs.

If you're not role-playing, then it's not an RPG. Claiming that FATE (for example) is an RPG would be like claiming that ice hockey is a ball-game because a puck is kind of like a ball. I can see where you're coming from, and the historical roots are evident, but it clearly doesn't pass by the obvious definition.

In your opinion, using your definition. Good enough for the games you play, but that's about it.

There are many correct ways to role-play. Meta-gaming, specifically, is not one of those ways.

In your opinion, using your definition. Good enough for the games you play, but that's about it.

Which is fine. For you and your games. You cross the line when you start telling other people, whose games you have no direct knowledge of or involvement in, that they are somehow playing D&D wrong, or their game isn't "really" D&D. You don't have that authority.

 

[Tony Vargas]

Not at all. I'm referring to both the stakes for individual rolls and for challenges overall.

Some systems present some mechanics that necessarily have certain stakes attached - like anything that does damage can kill you if you're 1hp away from death at the time. You could make your dungeons into padded cells where no one can get hurt, thus never invoke a mechanic that puts death even theoretically in some tail instance, on the table.

I wouldn't try it if I were claiming to run D&D, though.

I think this is irrelevant. If you don't want death as a possibility, just take it off the table up front. There are other possible outcomes to losing a combat challenge.

Remember, a freak character death is a possible outcome of /winning/ a combat challenge, too. So if you set the stakes - win the fight, gain some treasure; lose the fight be taken prisoner - the party could win or lose, with those outcomes. The four out of five of them not killed by a freak damage spike, that is. Or you could fudge the freak damage spike and return to the intended stakes.

It might be implausible to others that only certain deaths are possible and others ("ridiculous" ones) are not. I'm less concerned about implausibility, which is a matter of preference, than I am of where the technical problem is occurring such that fudging is seen as the solution.

Technical problems occur in the stats given to potential foes, the combat rules, the resolution mechanics, and the class designs, among others. It's a very complex system with many potential points of failure.

Again, if you don't want death as a possibility, change the stakes.

Meh, you could decide the 3.0 orcs in question want to take the party prisoner, even the elves, and arm them with saps to that end, instead of their customary greataxes, I suppose. I question whether it'd go over well, especially as a matter of course. But you didn't have had to go to that extreme in 4e, nor should you often need to in a 3rd+ level 5e game - you could have relatively easy fights without a meaningful risk of unintended/lame freak character deaths or other oddities. Of course, you might, conversely, have issues with 'deadly' encounters not exactly living up to the label, either, even when you mean for them to...
...and, you could always adjust that from behind the screen, too.

To have a discussion about whether fudging is a viable solution to a particular problem, I would need to have a concrete definition of what you mean by "immersion."

What I mean by "immersion" is "bizarre, nearly-impossible-to-articulate, problem some players have with some mechanics and events in-game, but not others, without being able to satisfactorily explain the distinction that makes one set intolerably immersion-shattering and the other just fine." And, yeah, that can leave you needing to make ad-hoc adjustments after the panacea 'stake setting' portion of the challenge.



I have to admit, though, you have hit on a completely effective answer to the OP's question: just never present the party with challenges that include potentially deadly dangers.

 

[5ekyu]

Almost all my dangerous combats are layered, with multiple waves of possibles.

Almost all my groups are non-drones, so there are internal divisions and political nuances.

Those two make for highly fluid situations where all sorts of betrayal opportunities exist.

 

[Satyrn]

Remember, a freak character death is a possible outcome of /winning/ a combat challenge, too. So if you set the stakes - win the fight, gain some treasure; lose the fight be taken prisoner - the party could win or lose, with those outcomes. The four out of five of them not killed by a freak damage spike, that is. Or you could fudge the freak damage spike and return to the intended stakes.

But what iserith is saying is that if character death isn't one of the intended stakes, the DM is better off not rolling dice that will result in character death than fudging that die roll.

Like, in your scenario you've set failure as resulting in being taken prisoner. By what iserith has been saying, the DM should have his monsters act in accordance with those stakes, making attacks that would render the PCs unconcious - use melee attacks instead of ranged so you can invoke the rule that lets you choose to make the final blow a knockout.

And when it comes to spellcasters, the DM should be using spells like Hold Person instead of Power Word Kill.

 

[Tony Vargas]

But what iserith is saying is that if character death isn't one of the intended stakes, the DM is better off not rolling dice that will result in character death than fudging that die roll.

Not will, but 'could,' remotely, possibly only in combination with some other roll or some bad player decision...
So, refrain from making attacks, for instance.

Yeah, I get it. I just don't think it's a viable approach in the context (5e D&D), let alone the 3e example.

 

[Satyrn]

Not will, but 'could,' remotely, possibly only in combination with some other roll or some bad player decision...
So, refrain from making attacks, for instance.

Yeah, I get it. I just don't think it's a viable approach in the context (5e D&D), let alone the 3e example.

Not viable all the time, no. But then I want the stakes to include death.

Also, yeah. 3e's monster manual orc was one of those cases where the CR was definitely wrong.

 

[TheSword]

How about death is still on the table but not TPK.

So if i my fireball by some statistical fluke rolls eight sixes, every character fails their save and 48 is enough to kill every character I instead fudge the roll so it is enough to kill all but one. Or even better fudge the DC so that a character passes.

I would prefer not to use the fireball at all in that situation but you can’t anticipate 4 failed saves and a max dice roll.

If one character survives badly burnt and then decides to throw himself back into combat, then by all means TPK the fool.

 

[Inchoroi]

If you are planning a major combat, (such as a boss fight) one that has potential to be extremely lethal, what steps do you follow to keep the party alive?
What tactics do you use to prevent a party wipe without the party knowing you helped them out a little.

Don't get me wrong. I'm not an easy DM. I've killed a few PC's in my day, and I'm good with that if it happens here. But this is a boss fight. I want them to win. But I also want it to be dramatic. I want them to survive by the skin of their teeth.

How do you guys create that illusion if you sense things are going very very badly for the party?

I have learned to use the encounter building system (Through Kobold Fight Club), but never to rely on it. I've built encounters I thought were going to crush my party and they wiped the floor with my baddies. And I have built encounters I thought would be a breeze and almost TPK'ed. I'm sure we have all experienced this.

What steps do I take to keep the party alive?

I don't take any.

I put monsters and give those monsters abilities where and when it makes sense based on the story. If they steamroll an encounter, more power to them! They should feel happy that they were badass! If, however, they go into a fight expecting it to be easy and find that it really is not, well, that's on them, too.

This happened just a couple sessions ago, actually. The party went whole-hog on a group of cultists, whom they knew for a fact had a lot of worshipers and magic. They still went for it. They killed a lot of them...but it got them down to the point where they had no choice but to retreat. They ran, hid in the dungeon and took a short rest. Since the bad guys were clever, but knew also that they had limited resources, they gambled: the head cultists started a ritual that would bring a very powerful creature to this plane of existence, into which the party walked mid-ritual. Fight for survival ensued.

All seemed to be lost, as the head priest bade his cultists continue the ritual and fought the party on his own. He was beating the party senseless, until a very lucky critical hit managed to sever the creature's shield arm. This single moment of luck changed the entire encounter, and it was glorious! Without that, the high priest would have eventually wore them down, until the ritual was complete and the extraplanar creature came through the gate they had created and, well, destroyed the party.

I didn't tell them this, but they had 3 rounds left when they stopped the ritual. Three rounds before a group of level 7 characters had to fight a CR 17.

 

[TheSword]

To me fudging as a DM is like a small pressure release valve on a boiler system. Sure you could just have unweildy double thickness pipes and run at much lower temperature. Instead have your pressure valve to help prevent the boiler blowing up when happenstance causes the pressure to go much greater than intended.

The pressure release of that DM fudge let’s me run my campaign hotter and leaner, more efficiently, whilst avoiding the whole system melting down in a TPK.

 

[iserith]

Some systems present some mechanics that necessarily have certain stakes attached - like anything that does damage can kill you if you're 1hp away from death at the time. You could make your dungeons into padded cells where no one can get hurt, thus never invoke a mechanic that puts death even theoretically in some tail instance, on the table.

I wouldn't try it if I were claiming to run D&D, though.

Eh, it wouldn't be my cup of tea, but I wouldn't necessarily say it's not D&D.

Remember, a freak character death is a possible outcome of /winning/ a combat challenge, too. So if you set the stakes - win the fight, gain some treasure; lose the fight be taken prisoner - the party could win or lose, with those outcomes. The four out of five of them not killed by a freak damage spike, that is. Or you could fudge the freak damage spike and return to the intended stakes.

[MENTION=6801204]Satyrn[/MENTION] got this.

Technical problems occur in the stats given to potential foes, the combat rules, the resolution mechanics, and the class designs, among others. It's a very complex system with many potential points of failure.

By "technical," I'm referring to the techniques the DM uses since how the play experience turns out is so largely dependent on the DM in D&D 5e.

Meh, you could decide the 3.0 orcs in question want to take the party prisoner, even the elves, and arm them with saps to that end, instead of their customary greataxes, I suppose. I question whether it'd go over well, especially as a matter of course. But you didn't have had to go to that extreme in 4e, nor should you often need to in a 3rd+ level 5e game - you could have relatively easy fights without a meaningful risk of unintended/lame freak character deaths or other oddities. Of course, you might, conversely, have issues with 'deadly' encounters not exactly living up to the label, either, even when you mean for them to...
...and, you could always adjust that from behind the screen, too.

I look at intended difficulty and actual difficulty as different things since difficulty is modified by player choice and randomness. An encounter intended to be easy might end up being deadly based on what the players do and the result of the dice. That should always be considered a possibility in my view. And if you don't want PCs dying to encounters that are intended to be easy (for example), then making death a possibility, however remote, seems counterproductive.

What I mean by "immersion" is "bizarre, nearly-impossible-to-articulate, problem some players have with some mechanics and events in-game, but not others, without being able to satisfactorily explain the distinction that makes one set intolerably immersion-shattering and the other just fine." And, yeah, that can leave you needing to make ad-hoc adjustments after the panacea 'stake setting' portion of the challenge.

My position is that if they can't explain it so that I can figure out a solution, then they need to suck it up.

I have to admit, though, you have hit on a completely effective answer to the OP's question: just never present the party with challenges that include potentially deadly dangers.

Yeah. I guess if that's not desirable it leads us back around to the beginning: If you're not going to fudge and you're not going to take death off the table, how do you prevent PCs from dying? A rather strange proposition.

For my part, you just accept that death happens from time to time. Prepare for it by solving the iteration time issue and set the proper expectations. As long as you're not selling merch and will soon find yourself sitting on 500 T-shirts related to a particular character who just fell into a pit trap and died, you're probably going to be okay.