And martials miss their attack rolls just as well. Or, they roll crappy damage that doesn't down the enemy. Magic Missile (along with most spells) doesn't miss
To slightly extend this: spells often do half damage, or have some minimum effect, even if the target makes their save.
Outside of Graze, which is fixed <attack ability modifier> damage, non-caster characters rarely if ever achieve the same thing. If their efforts fail, they simply
fail, no effect. It doesn't take much for even a high-level Fighter (3-4 attacks per round) to go a whole round without achieving anything.
The casters slots just have to outlast the most damaged PCs hp. I’d suggest this is a big part of why in practice many of us often see slots outlast hp.
Well-said. Just
one party member being on, say, 1/3 HP, with no hit dice left, dipping into the communal coffers just to stay alive? That's already a powerful argument for "hey guys, we should long rest."
Yes, any character "can" outdamage any other character in play. This is part of the problem with these type threads, they way oversimplify things and don't take into account numerous variable in actual play.
In an extreme example of this, a sword and board Wizard with an 8 strength, no proficiency in shields or the longsword she is swinging with disadvantage "can" outdamage an optimized GWM/PAM fighter if that Wizard rolls high enough to hit and the fighter doesn't. Roll enough times and it is not even in doubt, the Wizard WILL outdamage that fighter at some point.
And this, too, is part of the problem with these types of threads: people pretend that extraordinarily unlikely circumstances, or massive outright defiance of the explicit text, or the nebulous never-defined "situation" or "context" etc., somehow make probability and numbers totally 100% irrelevant and easily dismissed. Usually with BS non-arguments like "white room" etc.
Quite obviously FrogReaver is talking about consistency. Do you think spellcasters can
consistently, reliably, testably, outdamage sword and board characters?
One-in-a-million chances happen about once in a million attempts. You don't hinge plans on one-in-a-million chances unless you have no other choice. You hinge plans around what is most likely, most expected, most probable. The nature of probability is that in most cases, near-the-center results happen,
especially when you average across many attempts. That is quite literally what "regression to the mean"
means; over many trials, the
observed average result will hug quite close to the
theoretical average most of the time.
Hence why my math above went out of its way to presume few, short (but perhaps not quite short
enough), weak combats, and
still found the claim of 32 potions for the whole group to fall not just short, but
significantly short. Even if I over-estimated numbers by
double, you'd be burning through that entire stack in 3, perhaps 4 adventuring days--and that means needing another 1600 GP
every single time you go to town. Assuming the town even
has 32 healing potions, which again is far from guaranteed.
