What's the Problem with Save-or-Die?

[Hassassin]

Want an effect that can kill instantly?

Make it do enough damage to kill instantly.

So what's the advantage of dealing 1000d6 damage instead of "you die"?

Or do you mean it should deal damage that may not be enough to kill instantly?

 

[Hassassin]

One die roll.
Soda reverts the game back to level 1.

If Finger of Death was:
A melee touch attack to make contact
Then a Will/Wisdom save to bypass the target's soul defending itself against death magic
Then a Fortitude/Constitution save to kill the target's body

Then it'll be okay. That's three chances to save yourself, much like damage based combat.

What should the chances be?

Is it OK if it has at least a 5% chance of killing you (e.g. 50% on touch attack, 40% on will save, 25% on fort save)? If so, why is that better than one d20 where 1 fails? Is it just psychological?

 

[Li Shenron]

In the hands of the players, a save-or-die spell will easily be cast on the first round against the BBEG (and with as many buffs on the caster as possible to maximize the chance of success), spoiling an otherwise potentially challenging and memorable encounter.

The DM is always free to use it more sparingly, such as for instance not casting it on the first round but let the BBEG save it for when he's losing the fight. Still, it does deliver a sense of not really being in control of your PC's destiny, if one damned roll is all that separates you from death.

But I agree that at least some save-or-die effects must be in the game... one thing to notice tho, is that most iconic monsters attacks with save-or-die effect might not really be save-or-die after all! Petrification is not death, there are ways to revert the effect that are not resurrection.

Also, if coup-de-grace offends you, it would be enough to just rule out that it cannot be used during a fight. The reason why coup-de-grace was put in the game was for assassinations and executions, so that being a 100hp character does not save you from a stab in the heart while sleeping or a throat slash while completely bound.

 

[Li Shenron]

I would also give rogues a SoD on critical sneak attack.

In 3ed this can actually happen, but it's not common. A high level Rogue may trigger the 50hp massive damage save-or-die effect. It's not easy... even at 20th level the average sneak attack damage is only 35 (plus your normal damage), and lots of extra damage dice means the results cluster around the average, plus the ST DC is an easy, static 15 Also, it's minimally dependent on whether it's critical since the sneak attack damage is not multiplied. Just to say that however at least the possibility exists...

 

[delericho]

You left out easy character creation.

Partly true. New PCs still need to be written in, though.

In this case, I think it's easy to just select another spell if you don't like all or nothing.

Agreed, but so many of the high-level spells were SoDs. Plus, when making use of pre-gen modules, it's not so easy.

This is not a bad idea, although it doesn't work with all effects.

I would also give rogues a SoD on critical sneak attack.

I think giving it on a standard sneak attack is probably too much. However, there's certainly scope for an advanced sneak attack power (or the assassin's death attack) which could be a SoD. And, indeed, there's ample room to expand those "critical sneak attacks" - attacks that hamstring opponents, or blind them (temporarily or permanently), or whatever.

So what's the advantage of dealing 1000d6 damage instead of "you die"?

I suppose it depends on how the game scales. If there's an upper limit to levels and/or hit points, then you reach a point where it's just cleaner and quicker to go with SoD. On the other hand, if there is no upper limit, a character/monster might theoretically survive 1000d6 damage...

(After all, a fireball is almost a SoD at 1st level, but really isn't at 10th.)

Practically speaking, and barring a miniscule number of exceptions, D&D campaigns have always had a limited number of levels/hit points. So we do indeed reach a point (quite quickly, actually), where we should just go SoD.

Or do you mean it should deal damage that may not be enough to kill instantly?

I guess it comes down to probabilities. If you want X% chance of killing, and have a feel for how many hit points the expected target should have, you could work out the damage code appropriately.

Personally, I feel that a spell like disintegrate loses its sting if it doesn't at least have the possibility of killing outright, regardless of hit points. YMMV, of course.

 

[Minigiant]

What should the chances be?

Is it OK if it has at least a 5% chance of killing you (e.g. 50% on touch attack, 40% on will save, 25% on fort save)? If so, why is that better than one d20 where 1 fails? Is it just psychological?

It is partially psychological. You'll feel less hose from insta death if you had more than one chance to save yourself.

As for chances, I'd prefer the caster or defender to have to on average roll over 10 if they get 1 roll or over 15 if 2 rolls. That will range from ~3% but increase dramatically as the caster levels or targets the right save.

 

[Ratskinner]

I don't really "hate" SoD, but I think they need some tweaking. I voted other because its their overall effect on gameplay that bothers me more than any one aspect, but I could easily have checked off several boxes.

My biggest problem is that, IME, don't help anything, and end up making more work for the DM.

If the PCs have the SoD advantage, the fights are uninteresting. If the bad guys have the SoD advantage, TPKs or near TPKs become much more likely, while half the players sit and watch bored and helpless. It also increases the likelihood of a few survivors abandoning their comrades. Particularly in older editions, bizarre poison needle traps seemed to exist solely for punishing rogue/thief players for making that "bad" choice, in spite of their being necessary for the party. They can add some tension when aimed at the party, but it seems to come at too high a cost overall.

Additionally, they can really mess up a story-based game. When an important villain or hero gets insta-gacked... it can really mess up the plot. Not even so much from the major characters as from minor characters with important information. PCs rarely seem to question knowledgable-but-dead NPCs.

All of that leads the DM (me) to build in workarounds, like sneaking in some Remove Paralysis scrolls before an encounter with ghouls or something. (Of course that can still be thwarted when the party sells them off....)

YMMV, of course.

 

[Fede]

What should the chances be?

Is it OK if it has at least a 5% chance of killing you (e.g. 50% on touch attack, 40% on will save, 25% on fort save)? If so, why is that better than one d20 where 1 fails? Is it just psychological?

One die roll.
Soda reverts the game back to level 1.

If Finger of Death was:
A melee touch attack to make contact
Then a Will/Wisdom save to bypass the target's soul defending itself against death magic
Then a Fortitude/Constitution save to kill the target's body

Then it'll be okay. That's three chances to save yourself, much like damage based combat.

I disagree with Minigiant, and my reason is also the answer to Hassassin's question.
I don't consider Minigiant's proposal a solution to the SoD problem, because having to roll three saving throws in a row is only a way of obfuscating the probability of killing(and making it depend on more than one score, but that's orthogonal to the problem).

It's different than having saving throws that are on subsequent turns, like 4E's turn into stone medusa, because in that case you have some time to react (e.g. a cleric casting a spell to heal you, or drink a potion to save yourself, etc.) and it's not a "gotcha!" one turn kill.

 

[Minigiant]

I disagree with Minigiant, and my reason is also the answer to Hassassin's question.
I don't consider Minigiant's proposal a solution to the SoD problem, because having to roll three saving throws in a row is only a way of obfuscating the probability of killing(and making it depend on more than one score, but that's orthogonal to the problem).

It's different than having saving throws that are on subsequent turns, like 4E's turn into stone medusa, because in that case you have some time to react (e.g. a cleric casting a spell to heal you, or drink a potion to save yourself, etc.) and it's not a "gotcha!" one turn kill.

The point of my multiple save solution is to simulate combat.

The fighter engages the orc hits 3 times. The orc dies.
The wizard casts death spell on the orc and hits 3 times. The orc dies.

The first turn argument doesn't work. I have seen archer PCs Swiss Cheese a major enemy away and I've seen charging brutes smash a foe first turn with lucky rolls on initiative.

 

[ExploderWizard]

The saving throw was designed to give an adventurer some chance to avoid at least the worst effects of terrible magical and mythic powers.

If these effects are removed from the game then what is there to separate the fantastic from the mundane? If myth and magic can accomplish nothing but dealing damage do we really need it at all?

If magic isn't going to have a distinctly different feel from the everyday then get rid of it.