The Forty-Five Theses on the Design and Brokenness of Spells

[Vadskye]

D&D's spell system is one of the most unique things about the game. It allows the game to encompass a great variety of magical phenomena, and it it can make being a magic user uniquely enjoyable, in the right circumstances.

It is also poorly designed and incredibly broken, and Pathfinder did essentially nothing to change that.

Here I present the most important problems with the spell system as currently written. Soon, I will present my proposed Spell Reformation, where I attempt to address all of these problems.

Spells Are Complex, Unintuitive, and Hard to Use

• Spells vary wildly in power and utility, even at the same level.
  Spoiler:

  Sleep is vastly superior to Cause Fear. Daylight does not deserve to be at the same level as Fireball. Crushing Despair is trivial compared to Fear or Slow. Placing Haste at the same level as Rage is ludicrous. There is no shortage of similar examples.

• Spell damage scaling makes no sense.
  Spoiler:

  How much damage does a 5th level spell do at caster level 10 (chosen for easy math)? That depends. Let's take a look. First, AOE spells:
  4d6: Mass Inflict Light Wounds
  10d6: Cone of Cold, Twinned Burning Hands
  15d6: Empowered Fireball

  Now single-target spells:

  7d6: Maximized Inflict Moderate Wounds
  9d6: Maximized Acid Arrow
  10d6: Twinned Magic Missile, Empowered Inflict Serious Wounds, Empowered Searing Light
  14d6: Maximized Scorching Ray
  This is absolutely terrible.

• Many spells are too trivial to be worth including.
  Spoiler:

  Does Helping Hand really need to be a spell? Has anyone actually needed Animal Trance, Calm Animals, Charm Animal, or Hide from Animals?

• Spells are used where class features belong.
  Spoiler:

  What separates a spell from a class feature is that a spell is an optional choice, while a class feature is an inherent part of a class. All of the various "animal control" spells are not strong or useful enough to exist as spells; in order for them to be lowered in level to the point that they would be chosen, they would be overpowered. Instead, they should be redesigned into class features for the druid so that the druid's spells can be interesting and useful.

• Many spells are poorly worded, making it difficult to understand their purpose.
  Spoiler:

  What does Magic Jar do anyway?

• Many spells have overly detailed mechanics, requiring excessive amounts of text to clarify their usage.
  Spoiler:

  Consider the text of Knock:

  SRD wrote:

  The knock spell opens stuck, barred, locked, held, or arcane locked doors. It opens secret doors, as well as locked or trick-opening boxes or chests. It also loosens welds, shackles, or chains (provided they serve to hold closures shut). If used to open a arcane locked door, the spell does not remove the arcane lock but simply suspends its functioning for 10 minutes. In all other cases, the door does not relock itself or become stuck again on its own. Knock does not raise barred gates or similar impediments (such as a portcullis), nor does it affect ropes, vines, and the like. The effect is limited by the area. Each spell can undo as many as two means of preventing egress.

  Consider the following rewording:

  Quote:

  The knock spell opens a locked or otherwise forcibly closed door or container within the area. The spell can undo up to two obstacles on the same object per casting. If used on an arcane locked door, the arcane lock is suppressed for 10 minutes.

  The reworded spell is identical, but far clearer, in 95% of all situations. It gained the ability to open gates and portcullises. Is that minor detail worth adding so much complexity to the spell? I would argue absolutely not. There are many spells like this which would benefit from a rewording and clarification.

• Many spells have trivial effects which are not worth the time investment to track.
  Spoiler:

  The attack bonus from Aid, the miscellanous +1 bonuses from Haste, and so on are trivial. It is unlikely that they will make any difference, but they force players to recalculate attack bonuses and other attributes anyway. This is particularly bad if the bonus is typed, since it forces the player to ensure that he does not have any other bonuses of that specific type - an unnecessarily large amount of mental investment for a +1 bonus.

• There are too many bonus types.
  Spoiler:

  Using such a wide variety of bonus types is unnecessary and makes it much more difficult to keep track of whether a particular effect will apply at full strength or not. Compressing the number of bonus types would make buffing spells much simpler.

• Buffing before combat is time-consuming and unnecessarily complicated.
  Spoiler:

  Tracking durations for a variety of spells and making sure that bonus types match up properly is obnoxious. This slows down the game. The more time you spend buffing before combat, the less time you spend actually enjoying the combat.

• 1 round/level durations scale strangely.
  Spoiler:

  1 round/level durations are terribly inconvenient. They are unusable at low levels (Touch of Fatigue and Summon Monster I being the worst offenders), and not worth the significant bookkeeping to keep track of precise spell durations at any level after about 8th, when they tend to last for a full encounter. Combat duration doesn't get longer as level increases, so why should duration?

• Level-scaling ranges and durations increase complexity substantially for little gain.
  Spoiler:

  Keeping track of precise ranges and durations is time-consuming. In the vast majority of cases, it does not substantially change outcomes or increase enjoyment in any way, but it is still technically necessary. Flat durations and ranges are much easier to use.

• Some spells are terribly designed.
  Spoiler:

  Scare is just a 2nd-level Cause Fear when first acquired. By the time it can affect multiple creatures, any creatures worth affecting are already immune to its effect.

• Spell casting times and components are confusingly formatted.
  Spoiler:

  Including "Casting Time: 1 standard action" and "Components: V, S" on 95% of all spells just makes it difficult to notice when the casting time or components are different from the norm.

• Spell ranges are confusingly formatted on area spells.
  Spoiler:

  Does "Range" refer to the distance away from you that a spell can be cast, or the distance from you that the spell's area extends? It depends! A spell like Bless is fairly specific; it indicates that the area originates from the caster. Bane, however, is completely ambiguous. The area is "All enemies within 50 ft". Within 50 feet of what? Is this a burst, a spread, or something different? The only spells to use a similar format are Circle of Death / Undeath to Death, which affect "Several living creatures within a 40 ft. radius burst". From context - primarily by comparison to Bless - we can determine that Bane is intended to affect a 50 ft. burst centered on the caster, while Circle of Death is intended to affect a burst within the range. However, this is unnecessarily ambiguous. A reasonable and intelligent person might easily read Bane and conclude that it affects a 50 ft. radius centered on a point within the (50 ft.) range.

• Area spells affect arbitrarily chosen and difficult to remember areas.
  Spoiler:

  Quick - how large of an area do Chaos Hammer, Confusion, and Sound Burst affect? If you guessed 20 ft., 15 ft., and 10 ft, congratulations - you're really good at memorizing random numbers. Spells that hit a radius can range from anywhere from a 5 ft. radius to an 80 ft. radius, with no particular patterns. Cones extend out either 15 ft., 30 ft., or 60 ft. from you. Wouldn't spells be so much easier to use if their areas were predictable and easy to remember?

• Spells which affect multiple targets have inconsistent limitations.
  Spoiler:

  Why do most multiple target spells, like Mass Bull's Strength, affect creatures "no two of which can be more than 30 ft. apart", while others, like Holy Aura, affect creatures "within a 20 ft. radius"? Come to think of it, why do any spells use the "no two of which can be more than 30 ft. apart" wording? I don't see any way for that to be easier to work with than "within a 15 ft. radius". Which, of course, raises the question of why we are working with a 15 ft. radius, which is one of the least used values for spell radii.

• Spells which affect cylinders can pass through walls.
  Spoiler:

  A cylinder-shaped spell is explicitly stated to "ignore any obstructions within its area", PHB p. 175.

• Spell descriptions are inconsistently formatted.
  Spoiler:

  Too many cases to list here. Just trust me - they are.

• 1-round casting time spells are terribly designed.
  Spoiler:

  I can't count how many times my players have gotten this particular mechanic wrong. The distinction between a "full-round action" and a "1-round action" is absolutely terribly worded, and makes little sense even when you know which one you are talking about. It feels like getting no actions on the turn you do the casting and two actions on the next turn, and it is incredibly easy to disrupt. There is nothing that a "1-round action" casting time contributes to fluff and enjoyment that couldn't be done better with "Full-round action" casting time.

• Spell schools and subschools are poorly and inconsistently defined.
  Spoiler:

  For more on this, see this entertaining thread. And by "entertaining" I mean "incredibly long and complicated". Though I enjoy it...

Combat Spells are Broken

• Low-level spellcasters have two modes: "useless" and "broken".
  Spoiler:

  A single Sleep or Color Spray spell can end an encounter instantly. However, a low-level spellcaster has only a very small number of those spells, forcing her to do various non-magical tasks like pointlessly firing a crossbow in any situation that does not involve an serious threat. This is not a healthy game dynamic.

• Moderately optimized Pathfinder quickly devolves into "rocket tag".
  Spoiler:

  "Rocket tag" refers to the way high-level D&D/Pathfinder characters can essentially all kill each other instantly. For example, a wizard would die if the fighter got a single full attack (or with some builds, a single charge attack), but the fighter would usually die or be rendered irrelevant if the wizard succesfully affected the fighter with a single spell.

• The "rocket tag" phenomenon is severely detrimental to the game.
  Spoiler:

  When everyone is perpetually one bad roll or poor decision away from death, combat has to be fast, efficient, and brutal. There is no room for finesse or subtlety. This promotes highly technical, tactical play and discourages role-playing.

• Spells can end combat far too quickly, promoting rocket tag.
  Spoiler:

  This is caused by a small army of problems. I will let the subproblems speak for themselves.

• Spells can (virtually) kill opponents instantly.
  Spoiler:

  Save or die spells prevent any sort of sane combat from taking place. Since these effects start at 1st level, with Color Spray and Sleep, this is a problem throughout the game. 10 damage/level is still enough to take just about anyone out of the fight until very high levels, when monster HP skyrockets (but PC and NPC HP doesn't...)

• Spells can render opponents irrelevant instantly.
  Spoiler:

  Total action denial spells like Hold Person and Confusion perform the same function and have the same effect as save-or-die spells.

• Spells do too much damage relative to HP.
  Spoiler:

  Scorching Ray deals roughly 1d6 damage per level. Wizards have a d6 hit die. It is trivially easy for a spellcaster to kill another spellcaster with a single spell - particularly after taking into account Empower and Maximize.

• Buffing before combat yields massive swings in party capability.
  Spoiler:

  A party that chooses to buff before a combat can easily go up multiple ECLs relative to an unbuffed party. This makes it extremely difficult for a DM to plan party-appropriate combat encounters. If the party goes in unbuffed against a monster designed for a buffed party, they can easily be killed or routed. If they go in buffed against a monster designed for an unbuffed party, the encounter will often be trivially easy. Limiting the power and usefulness of precombat buffs would significantly even out gameplay.

• Many buffs are just too good.
  Spoiler:

  Haste is the biggest offender here as a mass spell that gives everyone significant bonuses for a mere 3rd level spell. Enlarge Person is also amazingly strong for a 1st level spell.

• Some buffs can render entire encounters irrelevant.
  Spoiler:

  A dire bear's grappling abilities are nearly unbeatable - unless the fighter has Freedom of Movement, in which case the fight will be a breeze. A vampire can be a terrifying foe - but against a party with Death Ward and Magic Circle against Evil, a vampire is just a pale human without a Con score. (Okay, not literally.) The problem is not merely that these spells exist, but rather that they have such a long duration. This means that it is easy to cast the spell on everyone in the party that might need it, and infeasible for an enemy to wait for the spells to expire.

• Spells can make the subjects effectively invincible to non-spellcasters.
  Spoiler:

  Flight effects and Greater Invisibility are the most prominent offenders here.

• Many multiple target spells that lack single-target versions are vastly more powerful than they should be relative to single-target spells.
  Spoiler:

  What level would Haste or Slow be as a single-target spell? What about Confusion or Fear? These spells have effects that would be roughly level-appropriate as single-target spells, but they affect multiple creatures. This makes them significantly better than they should be.

• Area of effect spells affect too large of an area.
  Spoiler:

  At 5th level, when first aquired, Fireball can deal anywhere from 5d6 to 250d6 damage, depending on how many targets are in the 20 ft. radius. An Empowered Fireball from a 10th level caster can deal anywhere from 15d6 to 750d6 damage. A 20 ft. radius is huge.

• Area of effect and multiple target spells are too powerful relative to single-target spells.
  Spoiler:

  If area spells do equivalent damage to single-target spells, why would I use a single-target spell? Just so that I don't hit my allies? But the large radius of area spells is itself a problem. If that problem is fixed, the supremacy of area spells over targeted spells becomes assured. Buff and debuff spells are (usually) significantly penalized for affecting multiple targets - why shouldn't damage spells be the same?

Noncombat Spells Are Also Broken

• Spells can make make social interactions trivially easy.
  Spoiler:

  Glibness, Charm Person, and Suggestion can all turn a challenging social encounter into a cakewalk. The problem is not that those spells exist, but that they are so easily accessible.

• Spells can make dungeon delving trivially easy.
  Spoiler:

  Find Traps, Knock, and Summon Monster can all deal with traps and obstacles easily. These are particularly problematic because they are generally non-interactive. Disabling a trap or opening a door can be tricky, particularly if there is a battle ongoing. These spells do it faster, better, and more consistently than any mundane alternatives.

• Spells can make stealth and detection irrelevant easily.
  Spoiler:

  Invisibility and Silence are nearly unbeatable together.

• Spellcasters can perform any of these feats with too little investment.
  Spoiler:

  The problem with the above examples is not that magic is theoretically capable of performing these feats. It is magic, after all! That is what magic is for. The problem is that every single example above is done with 3rd level or lower spells. In fact, everything except Suggestion and Glibness can be accomplished by 3rd level!

• Given time to prepare, a spellcaster can perform all of these feats.
  Spoiler:

  It would be bad enough if a spellcaster was capable of overriding a single other area of the game at 3rd level. However, because prepared casters can change spells daily, they can actually override any and all other aspects of the game.

• Spells can exert control over the game world that nothing else can compete with.
  Spoiler:

  A well-placed Dominate Person or Suggestion can turn a city on its head. Scrying and Teleport can dramatically rewrite the whole concept of adventuring and travel when acquired. Spells give almost DM-level control of the game to players - but only to some players. That is not a good system.

Free, permanent duration spells are easily abusable.
Spoiler:
Hello, Explosive Runes. Also, Illusory Wall, though not as many people seem to push that to its logical conclusion.

Other Comments

• Spell resistance is crude and noninteractive.
  Spoiler:

  The default SR for a CR-appropriate monster is designed such that a spellcaster will fail 50% of the time. Futhermore, caster level is one of the statistics that a typical spellcaster is least able to modify, and changes comparatively little over a caster's career. This means that spell resistance, when it applies, simply acts like a flat chance of failure. That is not a healthy balancing mechanism. That's like taking a fully optimized Ubercharger build and calling it "balanced" by slapping a 50% chance to miss onto all of its attacks. Spell resistance should be interactive and more sensitive to character development.

• Spellcasters are unnecessarily penalized for devoting their resources into spells which the whole party requires.
  Spoiler:

  Divine Power and Wall of Stone are a lot more fun to cast - and often more appropriate for a character - then Restoration or Teleport. However, the latter spells can be essential in certain circumstances, so the spellcaster is obligated to spend personal resources to memorize and cast these utility spells.

• Broken spells affect everyone, not just spellcasters.
  Spoiler:

  Magic items are constructed based on spells, and virtually every character has magic items. Poorly designed spells yield poorly designed magic items. In addition, NPCs use spells and monster abilities are often based on spells. If spells are broken, so too is the D&D world as a whole.

[Norgrim Malgus]

Well, I really don't have the time to go through this in a step by step problem/solution format but let me point out a couple of things.

When comparing Divine Spells to Arcane spells, Arcane is supposed to do more damage. The intent is that Arcane casters have more damage spells in their spell lists than Divine casters; at the same time, their spells are intended to inflict more damage. Divine casters have more healing spells and the intent is that they do healing better than Arcane casters, not that they have many in the first place.

Metamagick feats can do some great things to spells, such as your empowered Fireball example. However, an Empowered Fireball is memorized as a 5th level spell and according to Ultimate Magic, 15 dice is appropriate.

Heck, I could quicken, empower, intensify and piercing metamagic a Magic Missile and would wreck a targets day rather successfully. But it is memorized as a 9th level spell and at let's say 17th level, I wouldn't have many slots to spare. The other problem with this amazingly epic grand slam metamagicked MM is all it takes is a Shield spell to force my jaw to hit the floor. Sure, there are ways to tweak it so I can memorize that at a slightly lower spell level but this example works.

Last thing I'll touch on is SR. Can it make affecting said target a pain in the neck? Sure. However, Spell Pen/Greater Spell Pen help me out, the Orange Prism Ioun Stone adds +1 to CL, the Robes of the Arch Magi also adds a +2 to overcome SR, and there are a couple traits/feats that can further push things into your favor. Oh, make your race Elven and you start with +2 to overcome SR.

I would be lying if I said that there are no fixes needed for some things, ambiguous descriptions for certain spells for example. All in all, sometimes what may appear to be a problem, really isn't.

I hope some of this helped out, but as we all know, no such thing as a perfect game system.

[Craig Bonham 141]

Yeah, far too many "examples of brokenness" to go through and argue. Obviously the OP feels very strongly about this and could not be convinced otherwise. But I would say that while not a perfect system the spell system is actually pretty decent. Some of the issues above depend on a group's game-style, the make-up of a party, perception rather than objective reality, and seems locked in on only looking at the Paizo products. A number of excellent third party items out there address some of the issues above.

Sorry it makes you unhappy Vadskye. I would suggest that nothing is going to change since Pathfinder is a love-letter to classic DnD and perhaps you should try another system.

[Wiggz]

I'm a big fan of 4th Ed.'s spell duration of 'encounter'. Seems like an awful lot of spells would be better suited for exactly that.

[Atarlost]

Norgrim Malgus wrote:

Metamagick feats can do some great things to spells, such as your empowered Fireball example. However, an Empowered Fireball is memorized as a 5th level spell and according to Ultimate Magic, 15 dice is appropriate.

I think you should re-read that chapter. A dice cap of 15 is appropriate for a 5th level spell, but at 10th level you should still only be getting 10 dice.

Dice caps are themselves a messed up system. At 10th level a 3rd, 4th, or 5th level spell will do the same damage. At 15th level a 4th, 5th, 6th, or 7th level can do the same damage. High level spells have dice caps higher than the maximum class level, making them essentially wasted.

[Matthew Morris]

You need 54 more before you can nail it to Paizo's door.

[MagiMaster]

If you were going to redesign the spell system, I think you should start, not quite from scratch, but by questioning some of the underlying assumptions like the relation between spell levels and caster levels.

Honestly, I'd like to see casters just have spell slots with no level attached and all spells designed to be cast at any spell level. Casters would still have a caster level that determines how powerful the spells they cast are and spells would still have a spell level that determined how hard they were to learn, but the two wouldn't be related.

That said, at best this would have to wait until PF 2.0 and even then I doubt it will happen just due to the massive amount of work that would be needed to overhaul every spell and magic item that ever existed, or even enough of them to fill out the core book.

[Atarlost]

Spell level probably shouldn't even be a single thing. Some spells should be available early but expensive to cast or available late but cheap to cast. (early SoDs like sleep in the first category, important mobility spells like flight in the second category)

[Da'ath]

I agree with some of your views, disagree with others Vadskye. I'm reserving judgement, however, until I see some of your solutions. I look forward to reading it.

[Norgrim Malgus]

Atarlost wrote:

Norgrim Malgus wrote:

Metamagick feats can do some great things to spells, such as your empowered Fireball example. However, an Empowered Fireball is memorized as a 5th level spell and according to Ultimate Magic, 15 dice is appropriate.

I think you should re-read that chapter. A dice cap of 15 is appropriate for a 5th level spell, but at 10th level you should still only be getting 10 dice.

Dice caps are themselves a messed up system. At 10th level a 3rd, 4th, or 5th level spell will do the same damage. At 15th level a 4th, 5th, 6th, or 7th level can do the same damage. High level spells have dice caps higher than the maximum class level, making them essentially wasted.

I know your level must be appropriate for the extra dice, I was trying to show the OP that 15 dice for a 5th level spell is fine per the dice cap/spell level range in UM.

[WarColonel]

A starting point to use for a redesign could be the Words of Power alternative from Ultimate Magic. Their version of spell seeds can be very useful in conceptualizing a new magic system.

The key to a redesign is whether spells are being rewritten or if spellcasting itself is being redone. The biggest issue with spells I see is in the classes that have weak or no spells of their own. Fighters, paladins, rogues, monks will never be quite as powerful as full casters, both in raw ability and versatility.

That being said, folding spells that seem useless into class abili themselves is only going to increase caster power, thougabilitiesutely. A general change to spells may be the better option, with the decreasing in power of SoD spells and major buffs, the increasing of usefulness of utility spells, and folding spells together. Mass spells, cure/inflict, aid to bless to etc., can be better streamlined into single spells that scale by level or can be cast at higher levels. Metamagic, which can really alter spells, can also be folded in. Schools can be better defined in their roles, and spells with equivalent effects can be destroyed.