D&D General A puzzle about spell casting in D&D

[Stormonu]

However, there is a tangential casting puzzle across all editions of D&D that has bothered me since day one: an archer has to roll to aim her shot; a person throwing a lit vial of oil has to roll to aim his throw, so why doesn't a caster have to roll to aim her ranged spell?

I've long since fixed this in my own game, but it still bugs me that the core game gives casters this huge (and IMO undeserved) advantage.

That's why there were saving throws - the archer has to roll to hit, the spell victim makes a save. If you make a spellcaster roll to hit then to even things out you'd have to give victims of arrows and weapon blows a saving throw to mitigate the damage too. In older editions, if the spellcaster was required to hit, their THAC0 would cause most attacks to miss. By tying it to saves, the system could be set up so the magic-user had a good chance of affecting the foe. To counter that saves got easier, magic-users spells often added in the effect that a successful save only reduced the damage to half - after all the magic-user's resources were limited, unlike the number of bow shots or sword swings the melee characters could make.

Anyways, when it comes to spells like Fireball and such, it's like hand grenades - "close" is enough.

 

[Umbran]

Why does spell-casting get this sort of benefit of the doubt?

Narratively? Picking a lock differs with every lock, and you have to figure out exactly which motions you are doing as you do them. Spellcasting is much more rote repetition of the same movements.

As a game? Because requiring multiple rolls for one action drastically reduces the chances of success, and means that a specialist in an area that needs them must spread their focus among several stats. If I need Int to know spells and make them potent, Constitution to not drop them, and now Dexterity to deliver them... that's a lot of stats that I need to be good at my job, while the Rogue really only needs Dex. Not a great design, there.

 

[Mannahnin]

Narritively? Picking a lock differs with every lock, and you have to figure out exactly which motions you are doing as you do them. Spellcasting is much more rote repetition of the same movements.

Nitpicking a bit, but if you recognize the kind of lock (or indeed the exact model, for worlds with high enough tech for mass production; maybe Eberron?), it CAN be more of a rote repetition of learned movements.

As a game? Because requiring multiple rolls for one action drastically reduces the chances of success, and means that a specialist in an area that needs them must spread their focus among several stats. If I need Int to know spells and make them potent, Constitution to not drop them, and now Dexterity to deliver them... that's a lot of stats that I need to be good at my job, while the Rogue really only needs Dex. Not a great design, there.

OTOH this is something which can be balanced in design. One flaw some editions of D&D have had, for example, is that a Wizard USUALLY only needs a high Int, whereas a Rogue who wants to be in combat really wants high Con and at least moderate Strength, not just Dex.

It's certainly possible to balance the classes in such a way that they're more equally subject to MAD.

 

[Umbran]

OTOH this is something which can be balanced in design.

Um, the point is that not making the wizard roll twice for a spell is the balance in design. The rogue rolls once to pick a lock, or stab a person. The wizard rolls once (or the target rolls once for a save) to make a spell work.

Rolling dice is fun. Having to repetitively roll multiple times to do one thing, not so much fun, especially when it increases failure rate.

This looks like a case where trying to introduce "realism" is in conflict with gameplay reality.

 

[DND_Reborn]

OTOH this is something which can be balanced in design. One flaw some editions of D&D have had, for example, is that a Wizard USUALLY only needs a high Int, whereas a Rogue who wants to be in combat really wants high Con and at least moderate Strength, not just Dex.

Considering how few HP Magic-Users (and later Wizards) have in D&D, CON (which in 5E is also used for Concentration checks) is important for them (as much, if not more so than Rogues need it IMO), and they also want a good DEX since they don't typically wear armor.

With the relative importance of DEX and CON in 5E, I don't consider any class in 5E really SAD anymore.

Anyway, as others have pointed out, wizards are rolling either for attacks or their targets are rolling saves. Many spells don't require either, but those often don't affect anyone hostile to the caster.

IMO, good design results in any task being able to have the outcome determined by a single die roll. (Longer situations might require multiple rolls of the same sort, but not different sources.)

 

[Cap'n Kobold]

I've just been re-reading Gygax's description of spell-casting in his PHB and DMG. It talks about various components in spell casting, including the need for the somatic components to "be begun and completed without interruption in a clean, smooth motion" (DMG p 65) and the fact that "gestures must be exact and movements free and as prescribed" (PHB, p 100).

There are other class abilities in D&D that require exact hand motions to be completed cleanly - picking pockets, disarming traps, some weapon attacks - and since the earliest days these abilities have required a dice roll to determine whether or not the character successfully performs the motions in question. But generally it's never been required to make a roll to see if the somatic components can be properly performed. Why does spell-casting get this sort of benefit of the doubt?

Those other class abilities all have a measure of opposition: No pocket-picking is the same and you're trying to do it without the victim seeing you. When disarming a trap, you're pitting your mind and/or finesse against the designer and situation. When trying to hit someone, they are almost always trying to defend themselves.

Gygax's spellcasting is different: there i no opposition to successfully completing the movements to cast a spell, so it works. In most cases where there would be opposition, such as being struck or grappled while trying to do the movements - you just automatically fail.

 

[Mannahnin]

Um, the point is that not making the wizard roll twice for a spell is the balance in design. The rogue rolls once to pick a lock, or stab a person. The wizard rolls once (or the target rolls once for a save) to make a spell work.

Rolling dice is fun. Having to repetitively roll multiple times to do one thing, not so much fun, especially when it increases failure rate.

Yes. Someone else already made that exact point up-thread, and I wrote a post agreeing with them and expanding on that point with more details and examples.

I was referring to the multiple ability dependency issue, not the multiple rolls for spells issue. Sorry if that wasn't clear. I'll try to edit better to make more clear which part I'm responding to.

 

[Ruin Explorer]

I've just been re-reading Gygax's description of spell-casting in his PHB and DMG. It talks about various components in spell casting, including the need for the somatic components to "be begun and completed without interruption in a clean, smooth motion" (DMG p 65) and the fact that "gestures must be exact and movements free and as prescribed" (PHB, p 100).

There are other class abilities in D&D that require exact hand motions to be completed cleanly - picking pockets, disarming traps, some weapon attacks - and since the earliest days these abilities have required a dice roll to determine whether or not the character successfully performs the motions in question. But generally it's never been required to make a roll to see if the somatic components can be properly performed. Why does spell-casting get this sort of benefit of the doubt?

The simple answer, Doyle-ist answer is that it's because spells burn a resource, but picking a pocket etc. does not.

If you could cast spells all day, you know there'd be some kind of check. It's almost universally true that in systems where spellcasting doesn't have a direct, mandatory resource cost, there will be a check, and usually it's kind of harsh one.

 

[Lanefan]

TBF, 5E does have casters make attack rolls for ranged spells (though not area effect ones). And as noted 4E has them roll for every spell.

I don't remember seeing anything in the 4e rules regarding having to aim a fireball or lightning bolt; though if this came in after the initial round of DMG-PH-MM releases I'd have missed it as those were the only 4e rulebooks I bought.

Like Paul mentioned, games which make casters check for every spell to see if it works typically don't limit them with spell slots or memorization restrictions. I've been running 5 Torches Deep for most of the last year, and in that game casters know a limited number of spells, but can keep casting them as long as they keep passing checks. But once they fail they lose access for the day.

As Psi pointed out, spells in D&D also allow saves, so if we require caster checks AND allow saves, we get two points of failure and make spells terribly unreliable. 5E design takes this into account, making spells almost all require EITHER an attack roll OR a save. Spells which require both generally have particularly potent effects, or the save is against some additional bonus nasty effect. My new wizard in RotFM has Ray of Sickness for one of his 1st level spells. I have to hit, and then the enemy gets a save to reduce the damage to half and to avoid the Poisoned condition. But Poisoned is quite nasty, so I'm happy to take the gamble on the spell.

That's 5e. There's any number of spells in 1e - which is what the OP was asking about - where there is no save. (Slow, I'm looking right at you here)

My own beef is that by RAW a caster can always exactly position a fireball (for example) such as to hit the enemy but not the allies; this is mitigated in 1e by fireballs expanding to fill a volume and associated possible misjudgments of available space, but even that goes away in 3e and later. But then you go on to say...

One thing I started toying with in the last 5E game I was running was having casters for offensive area effect spells roll a targeting check to see if it lands exactly where they want or not (perhaps a bit long or short). This creates more excitement with Fireballs, for example, and makes them act a bit less like laser-guided munitions which can be dropped into a melee and run no risk of hitting friendlies.

...this; and now you're on to it! There should be an aiming roll just like any other ranged attack. (and note they can also miss a bit left, right, up or down; rather than just long or short)

There IMO should also be a small risk (1/d20 maybe, if under duress) of missing everything e.g. your lightning bolt sails over everyone's heads, or even of fumbling with an aimed spell (best spell fumble ever: caster wearing a fireproof cloak casts fireball, fumbles, fireball goes off inside cloak...)

 

[Lanefan]

Most premade adventures will specify a 10' ceiling, thus making the height of the ceiling a moot issue for melee combat.

A standing joke in our crew is the simple question "How high is the ceiling?".

This came about due to one particularly unwise caster who loved chucking fireballs into combat after asking that question. Now when any caster asks it there's often either a round of laughter (if it's safe) or a resounding "No!" if the space is confined.