D&D 5E Spell DC vs Magic User Attack Roll

[pemerton]

I don't see a bullette parrying or dodging, or a beholder threatening riposte.

The beholder doesn't riposte, but it can probably dodge, and if the fighter is attacking it from the flank with a magic sword (likely by the time you're dealing with a beholder) it can also "parry" by trying to bring it's anti-magic ray to bear.

As far as the bulette is concerned, if it can't dodge a sword how does it dodge a fireball? In other words, if we assume the wizard can place the ball of fire wherever s/he likes, why does the bulette get a save at all? (Even moreso the giant slug.)

 

[machineelf]

I don't see how it makes less sense than the weak low-level creature rolling a natural 20 and saving automatically.

Do the rules call for natural 20s being automatic success on saving throws? I know they are automatic hits on attack rolls.

 

[Obryn]

Do the rules call for natural 20s being automatic success on saving throws? I know they are automatic hits on attack rolls.

Yes, and 1's are automatic failures.

Skill rolls are the only exceptions to 1 & 20 rules.

 

[Sadrik]

The bell curve/opposed roll vs single-die linear roll isn't going to be solved any time soon. But the most important thing when deigning a game is what they do.

An opposed roll/multiple dire/bell curve roll creates a probability curve that heavily favors results in the middle. The net gain from a +1 bonus varies depending on where it is applied - it changes the probability a lot if the target number was in the middle, but only a little at the extremes.

A flat probability distribution, as with a single d20 against a fixed difficulty, has a static probability change with each +1 added (5% in the case of the d20). This is the same all along the probability spread, up 'till you need a 20 to succeed, then the curve has a discontinuity and suddenly flats out at 0% (or a static 5% if 20 is always a success).

The whole thing is made much more complex because the probability in itself is not always what you're interested in. For example, just as often the proportional increase is interesting, as in how much more damage will I take when a monster needs a 19 to hit me, as opposed to a 20? Edge effects like these is what 5E is trying to avoid with their bounded math.

Which is best? Well, that's what the discussion is about. Personally I prefer a pyramidal probability distribution, like what you get with when rolling 2 dice. I use 2 exploding d6 in my homebrew.

I think the bell curve on dice is fine, so long as it is baked into the system. In this case, we have two ways it is done and they are not equal. One generates -19 to 19 or 39 possible results and the other only 20. If the basic die mechanic is opposed rolls then it works because you can manipulate the system to expect that but not with the vast majority of rolls being d20 vs. a target number.

It is simply better in opposed roll scenarios to have both players go against a DC. Arm wrestling might be a DC 10 STR check for each person each round. Stealth works very good when the person hiding rolls against the DC of the environment they are hiding in and the spotter rolls to spot them in the environment, based on the, you know, environment, not based on how good or bad the hider rolled. It works so much better.

I think this illustrates the opposed roll issue:
1d20 vs DC 10 (55% success)
1d20-1d20 vs DC 0 (~55% success)
As you can see the second one would be wildly more variable. Adding a +1 to one is more predictable in the first and the other is much different.

 

[LostSoul]

The fighter is failing to bypass the defences of a skilled opponent. The wizard isn't failing to cast a spell or hit the immobile patch of ground with their fireball.

This is a distinction I like. Some effects just take place - there's nothing the other side can do about it.

The way I like to look at saves is that they aren't necessarily dodging or anything else; they're a metagame feature. You screwed up, now the game gives you a second chance to save yourself. With fireball, the save isn't "Dodge out of the way-skill"; it's "You are in the middle of a blast of fire, but maybe something can save you and you'll only take 1/2 damage."

I am not sure, but I think TSR-D&D worked like that.

I am playing around with this distinction & spells: some spells need an attack roll (vs. AC, Ref, Fort, or Will - whatever's appropriate), but others just take place. For example, Shocking Grasp needs an attack roll "to hit" so there's no saving throw, but Fireball doesn't - you take damage if you're caught in it, but you can save for half.

(I guess there could be a save any time something bad happens, but I feel like using either an attack/to-hit roll or a save works better.)

 

[pemerton]

This is a distinction I like. Some effects just take place - there's nothing the other side can do about it.

The way I like to look at saves is that they aren't necessarily dodging or anything else; they're a metagame feature. You screwed up, now the game gives you a second chance to save yourself. With fireball, the save isn't "Dodge out of the way-skill"; it's "You are in the middle of a blast of fire, but maybe something can save you and you'll only take 1/2 damage."

I am not sure, but I think TSR-D&D worked like that.

Yes, AD&D and B/X worked like this.

On this model, shouldn't siege weapons, giant's clubs and the attacks of high-level fighters and assassins all be saving throw based, rather than attack roll based?

 

[LostSoul]

Yes, AD&D and B/X worked like this.

On this model, shouldn't siege weapons, giant's clubs and the attacks of high-level fighters and assassins all be saving throw based, rather than attack roll based?

I think so, yes. I'm not sure where the line should be drawn, but I think I'll get into that as we proceed into higher-level play.

 

[Sadrik]

Shocking Grasp needs an attack roll "to hit" so there's no saving throw, but Fireball doesn't - you take damage if you're caught in it, but you can save for half.

I think this is wrong. Shocking Grasp should be a fortitude to resist. It should be automatic to "touch" and then the person that receives the zap should be allowed to resist the damage with a save. I think a Str/Dex vs Dex is not the appropriate thing that should be tested when it comes to a Shocking Grasp spell. It should be a Int vs. Con and then further it should fall in line with the rest of the system and be a save like the vast majority of spells.
I can see an argument where many spells roll to attack, and then the defender also gets a save. This could be applied to weaken a particular spell, giving the opportunity for the spell to fail twice, but I don't like that. Though I can imagine several scenarios where this would apply. A spell that enhances a weapon for a duration and when it deals damage an effect occurs, they would get a save then.

Yes, AD&D and B/X worked like this.

On this model, shouldn't siege weapons, giant's clubs and the attacks of high-level fighters and assassins all be saving throw based, rather than attack roll based?

I don't think so, I think attack rolls is where these should be. None of these should auto-hit. I think high-level fighters virtually auto-hit by virtue of the system but it is still an attack. With an "area" based attack because it is so big, this is a little trickier. I think it is an attack roll still, because it is a function of scale. A giant's club against a giant is not an "area" attack. I think it can be weird to change between those scales. If you do add a rule in the game for large weapon attacks as area attacks it should be a general rule and not something specific to siege weapons or written into the giant monster entry.

 

[Ruzak]

I prefer saving throws for magic for the simple reason that it makes magic feel different from weapons. Whatever is going on when I cast a spell, I want it to work mechanically different from shooting a crossbow.

 

[LostSoul]

I think this is wrong. Shocking Grasp should be a fortitude to resist. It should be automatic to "touch" and then the person that receives the zap should be allowed to resist the damage with a save.

My thinking is that, if you want to touch someone and they don't want you to touch them, you have to roll to resolve that conflict. If the conflict is resolved such that you touch them, then they take the shock. I don't want to have too many rolls so I don't want to have a save here. With fireball, I don't see much of a conflict - you're in the fireball, too bad - so the save is there to help you out.

I guess, logically, every time you take damage in any fashion you could have a save - but I don't feel like the mechanics must necessarily follow their own internal logic. They should deliver a satisfying play experience.

I think a Str/Dex vs Dex is not the appropriate thing that should be tested when it comes to a Shocking Grasp spell. It should be a Int vs. Con and then further it should fall in line with the rest of the system and be a save like the vast majority of spells.

To be clear, I'm talking about my own homebrew 4E-based hack game. I don't really have an opinion on 5E and how it works.