My Current Final Answer to Caster/Martial Disparity

[Tanarii]

The 1e Magic User can memorize douzens of spells at level 20 and can learn them by gaining levels. It isn't balanced at all with the fighter.

A 1e Wizard at level 20 either spent 4-5 years getting there, or didn't play with the built in XP rules. It's not really a relevant point of comparison. What matters is levels 1-10.

Now in 3e or later, high level play is a point of comparison. Not a great one, but you still might get there in a year of play or even faster. But that's exactly part of the problem. Spells originally built for "never going to get there" levels of play are now part of the assumed and reachable level span. Even if they're not the normal range of play, they're not effectively out of reach.

[BlacKnight]

A 1e Wizard at level 20 either spent 4-5 years getting there, or didn't play with the built in XP rules. It's not really a relevant point of comparison. What matters is levels 1-10.

Now in 3e or later, high level play is a point of comparison. Not a great one, but you still might get there in a year of play or even faster. But that's exactly part of the problem. Spells originally built for "never going to get there" levels of play are now part of the assumed and reachable level span. Even if they're not the normal range of play, they're not effectively out of reach.

There's no obligation to start at lv 1, and for high level games the assumption is obviously to give each character the same amount of exp.
So if the wizard is lv 20 the other classes would be of even higher level, but let's not kid ourselves that it makes them balanced.

Every option available to the characters in the books is something that should be considered, "it only matters in 1% of the cases" is not a valid defense.
I would also notice that there's no line in the books that say something like "these spells are not meant to be used by players". Which would also beg the question of why the spells are in the players handbooks to start with.

Also, if Companion and Immortal Sets were not meant to be used... would that make those books a fraud? After all TSR would be selling unusable content (yes, I know that those books also have optional content for lower levels, but it's only a part).

[Zarrgon]

Every option available to the characters in the books is something that should be considered, "it only matters in 1% of the cases" is not a valid defense.
I would also notice that there's no line in the books that say something like "these spells are not meant to be used by players". Which would also beg the question of why the spells are in the players handbooks to start with.

Before 3E there was no problematic game rule that said: All spellcasters immediately know all spells in the game. At best, a PC wizard could only create the 'common' spells. Any exotic spell had to be found in gameplay.

Just look at the Spell Compendium. Flip through it and you will find a lot of spells with unique and/or powerful effects. A vast number of those spells come from the 2E Forgotten Realms. But while in 3E all spells are free for anyone to take, in 2E the bulk of those spells were not available to every spellcaster character. The vast majority were only found in one special spellbook that the character could find as treasure.

2E also had at least two rule systems for how rare a spell was. Instead of the 3E where everyone automatically knows every spell, 2E spellcasters only had a chance to even know of a spell. The system also allowed DMs to mark spells as common, uncommon, rare, very rare or unique. The more unknown the spell was, the more of a chance that a character could not get it.

Also, if Companion and Immortal Sets were not meant to be used... would that make those books a fraud?

You might mean Companion and Master.....but you are also crossing the AD&D and D&D streams

[Telok]

Not to anyone, but an idea. Why not have optional cheracter complexity?

Say that at some series of points during character generation and advancement the player gets a choice to opt in (or opt further in) to a set of more complex and advanced abilities. Make it so that characters all get this choice at the same points, maybe they're a feat every level, maybe you spend xp to buy abilities and each ability costs the same amount at particular tiers, maybe its a choice between a basic class ability a basic subclass abiliy and an advanced subclass ability. Whatever it is the character resource comes at the same point for each character.

You need to balance the simple basic options, but you're keeping them simple like basic number boosts or modest at will abilities. This is where you chuck your extra basic attacks, +2 on a skill roll, at will magic that duplicates those things, etc.

The advanced stuff could be your 'with a roll', 'X times per day/encounter', 'has a side effect', 'uses Y amount of resource Z',etc., types of limits with stronger effects like 'knock down or disarm an enemy', 'do bonus damage', 'fireball', 'auto succeed a skill use', 'jump 40 feet', etc. You can put prerequsites on them, make them chains of abilities, whatever.

The approach to balance then is just to figure out and explicitly communicate that each set of abilities available at a particular level/advancement point is as useful in the overall game as the other abilities available at that same level/point. If you decide that +4 damage in melee iscequal to 1/fight trip is equal to 3/day fireball is equal to 3+ on a d6 auto succeed intimidate when you have 9 fights a day and only use intimidate 2/fight thats fine as long as it's communicated to the players and DMs that you have to hit the number of fights.

If you want different forms of game play like 'combat' vs 'social' then you can balance by simply giving the characters their social and combat abilities at the same advancement points. If you want to keep options balanced then you make new advanced abilities for each type in your splat books. You don't put piles of new spells in each splat without putting new maneuvers and skill tricks in too.

Choose a definition of balance, choose a framework that treats all characters equally, and figure out your equivalencies. It sounds a lot like d&d 4e because it is. In combat, by the numbers and character advancement, they got their definition of balance mostly right. There were other problems with d&d 4e, but the character combat power advancement schedule and abilities were successful at their definition of balance.

[BlacKnight]

Before 3E there was no problematic game rule that said: All spellcasters immediately know all spells in the game. At best, a PC wizard could only create the 'common' spells. Any exotic spell had to be found in gameplay.

Just look at the Spell Compendium. Flip through it and you will find a lot of spells with unique and/or powerful effects. A vast number of those spells come from the 2E Forgotten Realms. But while in 3E all spells are free for anyone to take, in 2E the bulk of those spells were not available to every spellcaster character. The vast majority were only found in one special spellbook that the character could find as treasure.

2E also had at least two rule systems for how rare a spell was. Instead of the 3E where everyone automatically knows every spell, 2E spellcasters only had a chance to even know of a spell. The system also allowed DMs to mark spells as common, uncommon, rare, very rare or unique. The more unknown the spell was, the more of a chance that a character could not get it.



You might mean Companion and Master.....but you are also crossing the AD&D and D&D streams

Yeah, I meant the Master Set, but no, I'm not confusing anything with AD&D. The point was about the first edition of D&D, and in that game wizards learn spells by increasing by level.

[KineticDiplomat]

Syllogisms have their uses in bounded sets (and are they really syllogisms in those circumstances?), namely they help serve as a practical check on pure theory. Usually there's a theoretical failing they help highlight even if the exact nature is unknown. Communism mis-defining value in it's original works due to the influence of earlier economists, for example. A fatal flaw that was only really apparent in the theoretical after the practical highlighted it.

Perhaps there could be a theoretical system where resource constraints worked. It has just consistently failed in all the games we've talked about in this thread so far, which are the common reference points for both game design and play across the wider community. No doubt some issue with utility and the social contract contribute are good candidates as to why it fails in these games, but it may be something else. Somewhat irrelevant. The relevant point is that for every practical attempt to implement resource limitations as the major balancing factor in a TTRPG, or even insufficient punishments (looking at you shadowrun), we've seen it fail.

When we talk about D&D, this is particularly egregious. Arguments to the effect of "we should use these D&D techniques for balance" when D&D is a prime example of hideous imbalance is hardly a good indicator that they should work elsewhere when practically applied.

Which brings us back to the real issue: if people advocate techniques known to consistently fail at balancing, even if they theoretically could succeed, then pragmatically they are arguing to not attempt to balance after all. And they may have reasons they don't want to balance - I'd hazard a guess it has to do with TTRPGS, particularly high magic TTRPGS, having a heavy does of power fantasy - but let's not conflate "I want to be unbalanced because that is fun" with "the classes are balanced, because the same things that let me be OP are being used."

And that does go back the OPs point zero - it is ok to be imbalanced if that's what you want. But let's not pretend that false constraints are actually balancing it while we go and enjoy the fruits of being overpowered.

[PairO'Dice Lost]

The 1e Magic User can memorize douzens of spells at level 20 and can learn them by gaining levels. It isn't balanced at all with the fighter.

I didn't say the True Vancian Caster is the 1e Magic-User, just that the fact that the 1e MU was more limited in slots, preparation time, and spell acquisition than the 3e Wizard and was closer to its edition's fighter shows that those sorts of things narrow the gap, in the same way that the 3e Warmage isn't actually a single-school True Evocation Specialist Wizard but its drastically limited school action closes the gap between it and the fighter as well.

ToB classes aren't on the level of full casters anyway, because they lack the variety of tools the latter have.

Again, I didn't say they were, just that they were closer. The point is that Kinetic Diplomat said that resource constraints can't work to balance anything because they haven't in the past, and I said that they can work by showing examples of places where small constraints have made classes a bit more balanced and hypothetical massive constraints could make classes that are much more balanced.

You can make other classes that are on the same level, but then they would just be copies of each other, because the Tier 1 classes can do basically everything so there's no niche left to be filled.

This is an oft-repeated pithy statement, but doesn't really hold when you look at what the classes can actually do. Clerics, druids, wizards, and erudites can all do "summoning" and "area blasting" and "defensive self-buffing" and so on, but the way they do each of those things is quite different--for instance, clerics have costly but willing long-term summons with planar ally, and can get access to short-term utility-focused summons with summon monster; wizards have cheap but coerced long-term summons with planar binding and always have access to short-term utility-focused summons with summon monster, druids only have access to short-term combat-focused summons with summon nature's ally, but have the best feat and spell support for those summons, erudites only have access to short-term combat-focused summons with astral construct and a handful of CPsi powers, but astral constructs are customizable and avoid various anti-summoning wards.

And "basically everything" isn't "actually everything." Druids can't do undead animation, but are generally the best at personal shapeshifting via Wild Shape; wizards can't do healing or resurrection, but are generally the best at "world-changing effects" via a wide variety of permanent/instantaneous Abjurations and Transmutations; clerics can't really do illusions (yes, they have a dozen or so Illusions on their list, but none you care about), but are generally the best at metamagic via Divine Metamagic; erudites can't really do non-personal buffs, but are generally the best at temporal/action shenanigans via time hop/schism/etc.

So really, there's room for at least a few more Tier 1s before you've totally exhausted the design space; heck, a Sorcerer with boosted spell access and a Spirit Shaman with boosted class features almost qualify already, and despite being superficially just "wizard lite" and "druid lite" classes they have their own distinct themes (magical heritage, spirit creatures) to differentiate them from the pack.

I don't think there is a 'side' which is more to blame than the other. People* want to have fighters that are fairly straightforward to use and fits to some mental idea of 'realistic.' Other or the same people also want magic users which have a wide selection of spells which have generally world-changing (or at least party-specific world changing) consequences. They also want the two to be balanced. And, yes, as a general rule, they are aware of the contradictory nature of those impulses.
*Some people, significant enough to shape whether a given edition is a success or not.

I characterize them as a "side" because of all the various arguments for wizards being broken, wizards being balanced, fighters being broken, fighters being balanced, or some position somewhere in the middle, the "keep your magic away from my fighter"/"that's totally unrealistic" faction is the most coherent one and that one that generally comes out of the woodwork to complain in any "how to make the fighter not suck" discussion. There are certainly other arguments on that front, and wizard fanatics who are just as bad, but you don't see e.g. Shrodinger's Wizard Fallacy people swoop into fixed-list caster design threads to complain that an Oracle or Pyromancer class isn't broad enough to "feel magical."

Not to anyone, but an idea. Why not have optional cheracter complexity?

[...]

Choose a definition of balance, choose a framework that treats all characters equally, and figure out your equivalencies. It sounds a lot like d&d 4e because it is. In combat, by the numbers and character advancement, they got their definition of balance mostly right. There were other problems with d&d 4e, but the character combat power advancement schedule and abilities were successful at their definition of balance.

The problem is that those sorts of "equivalences" are easy to do in combat (though not without risking overcorrection and blandifying combat, for which I also cite 4e), but difficult out of combat and difficult both in and out of combat when you take synergies into account, and that's where all the balance problems lie.

Balancing fighters and wizards is utterly trivial if you approach it as a clean room problem where a party including one wizard and one fighter takes part in between [X] and [X+Y] combat encounters per [time period] on a N-by-N-square grid on a featureless plain, but the moment you allow any remotely-interesting combat abilities (e.g. summoned minions, flight), non-combat abilities that can be used as combat abilities (e.g. stone shaping, illusory voices), combat abilities that can be applied before combat and extend into combat (e.g. weapon buffs, damaging auras), then it quickly becomes intractable to try to balance all of those things in any reasonable way. And when you consider that a party can be all fighters, or all wizards, or no fighters or wizards so that basic balancing ideas like "there's a front line/back line distinction to make tanking work" or "someone in the party is going to be able to inflict physical damage" go out the window, it gets even worse.

The one way I can see a "sliding scale of complexity" working is to have multiple classes for each combat that differ in options and complexity but can all contribute the same amount. You could have Wizard and Warblade on the more complex end, Warlock and Barbarian on the more simple end, and Sorcerer/Warmage/Marshal/Fighter somewhere in the middle, but balance them so that a newbie's Barbarian and an expert's Warblade contribute similarly without requiring dumpster-diving on the Barbarian's part or removing interesting abilities on the Warblade's part. Problem is, that only really exists in a 1e-like environment where you pick a class and stick with it, not one where you can go Barbarian X/Warblade Y to get the best of both worlds, and that lack of customization is a dealbreaker for many.

Perhaps there could be a theoretical system where resource constraints worked. It has just consistently failed in all the games we've talked about in this thread so far, which are the common reference points for both game design and play across the wider community. No doubt some issue with utility and the social contract contribute are good candidates as to why it fails in these games, but it may be something else. Somewhat irrelevant. The relevant point is that for every practical attempt to implement resource limitations as the major balancing factor in a TTRPG, or even insufficient punishments (looking at you shadowrun), we've seen it fail.

When we talk about D&D, this is particularly egregious. Arguments to the effect of "we should use these D&D techniques for balance" when D&D is a prime example of hideous imbalance is hardly a good indicator that they should work elsewhere when practically applied.

Which brings us back to the real issue: if people advocate techniques known to consistently fail at balancing, even if they theoretically could succeed, then pragmatically they are arguing to not attempt to balance after all. And they may have reasons they don't want to balance - I'd hazard a guess it has to do with TTRPGS, particularly high magic TTRPGS, having a heavy does of power fantasy - but let's not conflate "I want to be unbalanced because that is fun" with "the classes are balanced, because the same things that let me be OP are being used."

Imagine this thread's title was "My Current Final Answer to the Theurge Problem" and the point of contention that in D&D it's very difficult to build a reasonably balanced character with access to both arcane blasting/control/utility spells and divine healing/buffing/support spells, and you claimed that it obviously isn't possible to ever build such a class because every single prior attempt has been too weak, pointing to the fact that in every single edition of D&D arcane-divine hybrid characters are weaker than their pure-arcane or pure-divine counterparts.

In that scenario, I would point out that of course arcane-divine hybrids are weaker because in every existing edition of D&D it's been a fundamental tenet that arcane casting is its own thing and divine casting is its own thing and you have to pay out the nose if you want access to both, but that doesn't mean you can't make an arcane-divine hybrid class that works by removing that stipulation--the 3e bard is already an arcane caster who heals and buffs like a divine caster and the Divine Magician ACF lets a cleric be a divine caster who can blast or control like an arcane caster, after all--just that it hasn't been done up to this point because that distinction is a sacred cow most D&Ders don't want to get rid of.

Adding resources to the fighter and/or constraining the wizard's resources are the same issue. You can balance beatsticks and casters just fine to within a reasonable tolerance, using the same kinds of balancing mechanisms seen elsewhere in the game, as long as you're willing to make some sacred ground beef.

[Cluedrew]

Not to anyone, but an idea. Why not have optional cheracter complexity?

I think the idea good, the implementation would be tricky of course. I have some recollection of someone suggesting this as a use for 5E sub-classes. Some sub-classes would represent almost the exact same concept just in simpler and more complex implementations. Like the simple fighter might be a juggernaut type - move forward and attack - while the complex fighter might be a canny fighter with dozens of tricks that each only help in a situation or two.

Balance becomes an issue is that the strength of a simple class is (closer to) a point, but for a complex class it is more likely to be a range (and likely a broader range than you expect). If you want to get them balanced than the point has to line up with the place in the range most people end up playing.

Still I like the idea, especially in the context of a system that tries to be many things to many people.

Perhaps there could be a theoretical system where resource constraints worked. It has just consistently failed in all the games we've talked about in this thread so far, which are the common reference points for both game design and play across the wider community. No doubt some issue with utility and the social contract contribute are good candidates as to why it fails in these games, but it may be something else.

Because they system let the constraint be gamed. There are lots of little examples of this but the classic one would be the 15-minute adventuring day, where resources are refreshed so frequently any limits on them don't really matter. I mean that extreme probably didn't happen that often but it kind of highlights the problem: Optimizing your resources took you away from the tension.

And I just realized that I am heading to a slightly different point about how one should try to make the effective path fun, if you let them drift then people might take the effective path instead of the fun one. I guess my only argument for why resource system might be a good balancing tool is kind of proof by contradiction, no matter how many systems have gotten it wrong than if one has gotten it right than it is possible. Of course any example I could use could be debated and I got to go so maybe I will pick this up later.

[Willie the Duck]

And that contradiction is where a lot of these discussions fall apart, in my experience -- there are those who find that contradiction a "world breaker", and those who express outright scathing disdain for that reaction.

If this is related to the what I will call 'verisimilitude' epic threads that you were involved in, I don't know, roughly two years ago, I assure you I have no desire to go back and relitigate those.

I characterize them as a "side" because of all the various arguments for wizards being broken, wizards being balanced, fighters being broken, fighters being balanced, or some position somewhere in the middle, the "keep your magic away from my fighter"/"that's totally unrealistic" faction is the most coherent one and that one that generally comes out of the woodwork to complain in any "how to make the fighter not suck" discussion. There are certainly other arguments on that front, and wizard fanatics who are just as bad, but you don't see e.g. Shrodinger's Wizard Fallacy people swoop into fixed-list caster design threads to complain that an Oracle or Pyromancer class isn't broad enough to "feel magical."

There are definitely arguments that one does and doesn't tend to see pop up. People seeking to create more limited casters are generally seen as doing something self-specific, whereas complex fighter 'fixes' are often seen as 'fixing it wrong.' I just generally disagree with the notion of sides. People have desires, often conflicting, and they voice them. I think the "how to make the fighter not suck" discussions are different from "Oracle or Pyromancer" discussions in that the former are seen as solving a problem* by making a change one might not agree with, while the later seems more like an optional sub-thought to explore.
*the fact that these fixes clearly aren't making it into the official game rules, so who cares if someone else makes a change one disagrees with being somewhat amusing, but par for the course with online discussion.

I think the idea good, the implementation would be tricky of course. I have some recollection of someone suggesting this as a use for 5E sub-classes. Some sub-classes would represent almost the exact same concept just in simpler and more complex implementations. Like the simple fighter might be a juggernaut type - move forward and attack - while the complex fighter might be a canny fighter with dozens of tricks that each only help in a situation or two.

Balance becomes an issue is that the strength of a simple class is (closer to) a point, but for a complex class it is more likely to be a range (and likely a broader range than you expect). If you want to get them balanced than the point has to line up with the place in the range most people end up playing.

The complex classes are always going to be a situation where you try to balance based on some average of the range, based on certain assumptions about the playstyle in which they are used. There are so many other factors to consider, that any complex class is going to run at a different level of ability based on how each campaign plays.

Still I like the idea, especially in the context of a system that tries to be many things to many people.

I think that's the most that can be hoped for, given what D&D means to the gamer base -- if it can be given the option of balance (relative, depending, etc.), if that is something that people consider a high priority.

[Democratus]

I doubt that there is even a single definition of "balance" being used here. Everyone has their own idea of what balance means and what it would look like in a game.

Without having agreement on the concept - how can there ever be consensus?

[Max_Killjoy]

If this is related to the what I will call 'verisimilitude' epic threads that you were involved in, I don't know, roughly two years ago, I assure you I have no desire to go back and relitigate those.

Yeah, there are a lot of things I've wanted to post in this thread that I just haven't.

But, it's also related to the fact that SOMETHING has to be "given up" when it comes to addressing this issue -- could be balance, some characters' power, some players' conceptions of what certain characters "should" be, coherent setting, or something else.

[Max_Killjoy]

I think the idea good, the implementation would be tricky of course. I have some recollection of someone suggesting this as a use for 5E sub-classes. Some sub-classes would represent almost the exact same concept just in simpler and more complex implementations. Like the simple fighter might be a juggernaut type - move forward and attack - while the complex fighter might be a canny fighter with dozens of tricks that each only help in a situation or two.

Balance becomes an issue is that the strength of a simple class is (closer to) a point, but for a complex class it is more likely to be a range (and likely a broader range than you expect). If you want to get them balanced than the point has to line up with the place in the range most people end up playing.

Still I like the idea, especially in the context of a system that tries to be many things to many people.

To me, it adds yet another variable to an already tricky balance situation. Not only do classes have to be balanced across the entire range of levels -- for any and every level, all the classes need to be balanced -- it also adds a complexity axis where the simplest and most complex subclasses all have to be balanced at any and every level.

(Again, going with the thread premise that "balance" is not the thing one decides to give up.)

[Telok]

To me, it adds yet another variable to an already tricky balance situation. Not only do classes have to be balanced across the entire range of levels -- for any and every level, all the classes need to be balanced -- it also adds a complexity axis where the simplest and most complex subclasses all have to be balanced at any and every level.

(Again, going with the thread premise that "balance" is not the thing one decides to give up.)

I'd perceive it more as a method of breaking down the function of balancing into discrete workable chunks. The trick being that all characters would get a build resource at the same time. That of course isn't true in d&d 3e & 5e.

Sort of like saying all classes get one combat thing and one social thing at 5th level. Then you can come to an agreement on the warrior simple options for those being on par with the warrior complex options. Then do the same for the mage options. Then check balance on the warrior vs mage simple options and the warrior vs mage complex options. Eventually you decide level 5 is balanced enough, made easier because you're not trying to compare warrion simple combat options to mage complex social options and both classes got both combat and social things.

You would still want to check cross-level synergy and combos, but hopefully they'll be easier to compare by using the same character building structure and everyone will have some. Synergies and combos aren't a bad thing, they just become a problem when they're too good.

[Pex]

I think the idea good, the implementation would be tricky of course. I have some recollection of someone suggesting this as a use for 5E sub-classes. Some sub-classes would represent almost the exact same concept just in simpler and more complex implementations. Like the simple fighter might be a juggernaut type - move forward and attack - while the complex fighter might be a canny fighter with dozens of tricks that each only help in a situation or two.

Balance becomes an issue is that the strength of a simple class is (closer to) a point, but for a complex class it is more likely to be a range (and likely a broader range than you expect). If you want to get them balanced than the point has to line up with the place in the range most people end up playing.

Still I like the idea, especially in the context of a system that tries to be many things to many people.

For something like this those who want to play the Juggernaught won't mind Canny has his tricks. Juggernaught hits hard and is resilient. They like that, and it is effective. Canny gets his fun with mobility or party tactician or exploits. They approach combat differently. The balance problem, if there is a problem, is what they can do out of combat. They need to be able to contribute in a meaningful way in their own way. Juggernaught can be Strong Man. He can be the guy who stands at the bottom next to a very tall wall everyone climbs on top of each other so Great Perception Little guy can look over and see what needs to be seen. Because of Juggernaut no one else needs to make a check to climb or maintain balance. Canny guy gives the party bonuses to traveling or a tactical bonus to before a combat starts such as initiative or avoid ambushes or maybe he's the party's Face. Spellcasters likely can do these things too, but now they don't have to.

[PairO'Dice Lost]

I'd perceive it more as a method of breaking down the function of balancing into discrete workable chunks. The trick being that all characters would get a build resource at the same time. That of course isn't true in d&d 3e & 5e.

Sort of like saying all classes get one combat thing and one social thing at 5th level. Then you can come to an agreement on the warrior simple options for those being on par with the warrior complex options. Then do the same for the mage options. Then check balance on the warrior vs mage simple options and the warrior vs mage complex options. Eventually you decide level 5 is balanced enough, made easier because you're not trying to compare warrion simple combat options to mage complex social options and both classes got both combat and social things.

You would still want to check cross-level synergy and combos, but hopefully they'll be easier to compare by using the same character building structure and everyone will have some. Synergies and combos aren't a bad thing, they just become a problem when they're too good.

Keeping class structure strict and uniform doesn't actually help much with balancing "simple" and "complex" options if you're allowing relatively free choice at the breakpoints you choose. To make the simple vs. complex balance concept work you really need to allow less build choice to narrow the possible build space.

For example, comparing the Warmage and the Beguiler (two "fixed" caster classes) at various breakpoints is actually pretty easy because you know exactly what they get at every level in terms of class features and spells with the small exception of Advanced Learning spells, but comparing the Beguiler and the Sorcerer (a "variable" caster class) is incredibly difficult because the range of possible sorcerers is so broad, even though they get the same kinds of choices at the same levels from similar lists of largely-overlapping options, even though you can build a sorcerer to be pretty simple or pretty complex based on your preferences, and even if you're deliberately building a sorcerer to be Beguiler-like with mostly Enchantment/Illusion/BFC spells and not making something out there like, say, a half-summoner half-gish. If you want to ensure that the Sorcerer is balanced with the Warmage or Beguiler, you'd have to limit its spell selection mechanism a heck of a lot to even start doing that.

And comparing two similar caster classes in that way is kind of the best-case scenario. Comparing the Monk and the Marshal or the Barbarian and the Swashbuckler (assuming all of those classes were modified to generally not suck and have both simple and complex builds, of course) is much more difficult given that each of those has a totally different role and basically no overlapping mechanics or resource systems. Now imagine actually giving those classes a simple vs. complex slider, where e.g. the Monk and Unarmed Swordsage are supposed to be the simple and complex versions of the same "unarmored skirmisher" class and you're supposed to be able to build something that's 100% Monk, 100% Swordsage, 50%/50% Monk/Swordsage, or anything in between and have those come out vaguely balanced.

In short, unless classes are narrowed to the point that their options are largely fixed and they're closer in breadth to a single build of a 3e class, I don't think parity between simple and complex builds/classes/characters is really an achievable option.

[Telok]

For example, comparing the Warmage and the Beguiler...

Oh, goodness, no. You absolutely can't do it with any version of d&d from the last 20 years. Well, maybe 4e, but it wasn't giving people different options, just variations on standard combat effects. And probably not any version of d&d at all.

Each d&d class is effectively it's own set of abilities on it's own schedule without a real relation to the other classes. About the best you can do in 3.x is compare cleric domains. I mean, beguiler and warmage look similar, but they're getting different types of character resources from the spell lists and class abilities. You're heading back to "is A's combat orange equal to B's social coffee cup" type questions.

[NigelWalmsley]

If you want to ensure that the Sorcerer is balanced with the Warmage or Beguiler, you'd have to limit its spell selection mechanism a heck of a lot to even start doing that.

No you don't. It's balanced with the Beguiler now. People talk about how balancing different classes is really hard, but if you look at the level of balance that actually exists in 3e, it doesn't actually seem to be. You could fill a PHB with classes that are all A) reasonably balanced and B) use different resource management systems. I'm not really sure how much more you could want in terms of "it is in fact possible to balance classes that are different".

The reason imbalance exists is because games are designed by people who are bad at math and don't face strong pressures to make a balanced product. It's not some mystery we can't solve, it's just that if you can balance a system as complicated as D&D, you can get more money from Google or Amazon than WotC would ever imagine paying you.

[Cluedrew]

You could fill a PHB with classes that are all A) reasonably balanced and B) use different resource management systems. I'm not really sure how much more you could want in terms of "it is in fact possible to balance classes that are different".

Do you mean with classes that exist? Because if you do I would really like to see that list. And every other question I have kind of depends on if that list exists or not so I'm going to end here.

[Ignimortis]

Do you mean with classes that exist? Because if you do I would really like to see that list. And every other question I have kind of depends on if that list exists or not so I'm going to end here.

Yes. It's not the best solution, but it might work. Let's see, what do we have in the 3.5 PHB?

Barbarian, Bard, Cleric, Druid, Fighter, Monk, Paladin, Ranger, Rogue, Sorcerer, Wizard. Alright, here goes.

Barbarian, Bard - those are perfectly fine, and Barbarian is a solid option for "lower complexity please" people. Go into a rage, smash people. I never understood why Fighters are often less complex than Barbarians.

Cleric and Paladin get replaced by Healer, Favoured Soul and Crusader. You'd have to beef up Healer's spell-list a bit, but some people do want to play Cleric as a healbot, so that needs to exist.

Druid gets replaced by Wildshape Ranger. Anyone who wants to play a spellcaster type of Druid, gets to play a nature god's Favoured Soul.

Fighter and Monk are now Warblade and Swordsage. People who want to be Monk get an Unarmed Swordsage variant.

Ranger is in a weird spot, because we already have Wildshape Ranger which is generally Ranger but better. Not sure what to do about them, because everything I have is "give them initiating" or "nerf Wildshape Ranger to lose features you actually care about". Perhaps bump up regular Ranger's Animal Companion to their full HD and make the Wildshape variant lose their companion?

Rogue is pretty much fine if you let them take good ACFs. Gestalting them with Fighter as a class might also work.

Sorcerer and Wizard are replaced by Beguiler, Warmage and Dread Necromancer. Needs another one or two classes of that type to finish that off, though - conjuration, transmutation and divination aren't covered by those. Maybe it's for the best, since the former two are the source of most issues with magic, but we do need those archetypes to be covered anyway, in some form.

[NigelWalmsley]

I wasn't necessarily thinking of one-to-one swaps. I don't think that's necessary to prove the point, as I don't think the specific classes in the PHB are essential to the success of the game.

Do you mean with classes that exist? Because if you do I would really like to see that list. And every other question I have kind of depends on if that list exists or not so I'm going to end here.

Just look at the variety of classes that are in the T3/T4 band. T3 alone has twenty, which is enough to discard eight classes for being too similar and still end up with more classes than were in the PHB. And that's from an edition that was notoriously sloppy about balance. If you can get there without trying, I think it's absurd to suggest there's any fundamental problem with having classes that are balanced but mechanically distinct.

[Psyren]

Just look at the variety of classes that are in the T3/T4 band. T3 alone has twenty, which is enough to discard eight classes for being too similar and still end up with more classes than were in the PHB. And that's from an edition that was notoriously sloppy about balance. If you can get there without trying, I think it's absurd to suggest there's any fundamental problem with having classes that are balanced but mechanically distinct.

Even if the community rankings are taken as gospel, there's a lot more variation in that band than you're letting on (any band that includes both Shugenja and Scout, or Wilder and Ninja, has plenty) - and more importantly, it supports rather than disproves the continued existence of caster/martial disparity. The classes with some form of casting are almost universally higher in that band than the classes without, and the classes without innate access to any magic at all like Fighter and Generic Warrior are both lower down and vastly outnumbered. Are the entries on this list better balanced against each other than any ranking that includes druids and samurai, sure - but depending on how much disparity you're trying to get rid of, this list may not be "good enough."

[Cluedrew]

Yeah I have been silent on this because... "Shugenja and Scout": I can guess what the scout is about, because actually scouting works weird in D&D I'm going to assume its more of just a light and fast martial combatant with fewer real scouting abilities. I'm not even entirely sure what language Shugenja comes from let alone what the class actually looks like. So I am going to have to speak in the abstract somewhat.

Sure let's say we start NigelWalmsley's T3/T4 base, use Ignimortis's suggestions to pick and maybe tweak a few of the classes to bring them into a slightly closer band, we are just going to hope its close enough for Psyren's over all balance concern. But even so I am not quite done:

1) This is the Texas Sharpshooter approach to balance*. Create a whole bunch of classes and then pick the balanced ones. Its not a strategy I would recommend for someone actually planning to create more content.

2) Creation strategy aside what can we learn from this? Limited spell lists seem to be the main way of decreasing the power of casters in D&D and martial get powered up by getting more spell like abilities. (Mechanically at least, some of them are flavoured very differently.) Which still seems to be kind of narrow and not quite the 11 different resource systems originally promises. Which I am OK with because that is a lot of rules text.

3) Yes there does seem to be a slant towards more magic = more power even in this band.

* Someone shoots at the side of a barn then draws a target in the middle of the densest grouping of shots. I have no idea what it has to do with Texas.

[NigelWalmsley]

I can guess what the scout is about, because actually scouting works weird in D&D I'm going to assume its more of just a light and fast martial combatant with fewer real scouting abilities. I'm not even entirely sure what language Shugenja comes from let alone what the class actually looks like.

The Scout is basically a nature-themed Rogue. The Shugenja is one of the Oriental Adventures classes that got repackaged in the Completes, it's a caster that gets spells divided by element.

This is the Texas Sharpshooter approach to balance*. Create a whole bunch of classes and then pick the balanced ones. Its not a strategy I would recommend for someone actually planning to create more content.

I wouldn't advocate that as a strategy for creating content, but that's not the point (similarly, it's not a strategy for reducing disparity, so Psyren's complaints about such are entirely beside the point). It's simply an existence proof that balancing classes doesn't require homogenity, and the fact that you can do this well accidentally should suggest that it would be pretty easy to solve the problem if you actually tried. You can write as many classes as have been in any version of the PHB that are reasonably balanced and mechanically distinct, so the notion that balance forces all classes to be the same is simply nonsense.

As far as how you create balanced content, once we've all acknowledged that it's possible, the process is quite simple. It's the same as any other design problem. You define your metrics, then you iteratively refine, test, and redesign until you've hit your target. The idea that this is some unattainable goal, or even particularly hard, is simply FUD.

martial get powered up by getting more spell like abilities.

I think this is misleading, because it suggests that those abilities are inherently magical in some respect, or that we're turning martials into casters. We aren't. It's simply that 3e defined practically every ability worth having as a spell, so any new ability that does interesting things is going to feel like a caster ability, even if it's a martial one by any reasonable standard.

Which still seems to be kind of narrow and not quite the 11 different resource systems originally promises.

Warlock (at-will), Binder (5 round recharge), Wilder (spell points), Totemist (divide points between abilities), Warblade (use abilities or recharge them), Crusader (random ability selection), Swordsage (encounter powers), Healer (prepared slots), Warmage (spontaneous slots), Factotum (spell points, but substantively different from the Wilder in various ways), and Rogue (positioning matters). That's 11.

And, of course, if we look at 4e (the "balance means boring" crowd's favorite talking point), we can cut out whichever three of those seem most similar. Plus that's just confining ourselves to things that have been printed. You could easily imagine classes that had abilities that synergized when used in the right sequence, or inflicted Shadowrun-esque Drain, or any number of other things, and I defy anyone to make a good faith argument that it's impossible to balance those things with existing classes.

Yes there does seem to be a slant towards more magic = more power even in this band.

Sure, but that's not the point. I'm not trying to demonstrate that you can balance magic and martial by doing this, I'm trying to demonstrate that balance doesn't require every class to be identical. If I wanted to prove that you can balance sword guy and spell guy, I'd just point at the Warblade and the Binder or the Crusader and the Warmage or the Swordsage and the Jester or the Scout and the Spellthief. Again, if a game that paid essentially no attention to balance can do that, imagine what you could do if designers actually did their jobs and tested things.

[PairO'Dice Lost]

Oh, goodness, no. You absolutely can't do it with any version of d&d from the last 20 years. Well, maybe 4e, but it wasn't giving people different options, just variations on standard combat effects. And probably not any version of d&d at all.

Each d&d class is effectively it's own set of abilities on it's own schedule without a real relation to the other classes. About the best you can do in 3.x is compare cleric domains. I mean, beguiler and warmage look similar, but they're getting different types of character resources from the spell lists and class abilities. You're heading back to "is A's combat orange equal to B's social coffee cup" type questions.

I wasn't saying that Warmage and Beguiler were "simple" classes according to your proposal, but rather pointing out how much possible variation there is even between two classes with essentially no build choices and basically identical structures simply by varying the specific spells on their respective fixed and unvarying spell lists. As soon as you have different lists of "simple mage combat options" and "complex mage combat options" and "simple mage social options" and "complex mage social options" for a given character to pick from and you try to make all of those possible combinations work together, instead of a single cohesive class you can judge in aggregate, the idea that you can feasibly balance a "simple mage" with a "complex mage" goes completely out the window.

4e is probably the best possible scenario for something like that, and it only came close because it so harshly restricted the possible effects a class could get, and even then it was still, well, 4e.

The other problem with a "one social thing and one combat thing at 5th" setup is that "combat thing" and "social thing" (and "stealth thing" and "downtime thing" and "utility thing" and...) aren't nice distinct categories you can easily split up. Social-oriented mind reading/charms/language spells/etc. can be used in combat, combat-oriented shapeshifting/skill buffs/rerolls/etc. can be used in social contexts, and things like disguises and stat boosts work in both contexts and a half-dozen others besides. To make such divisions meaningful, you'd have to either sharply restrict the scope of effects until you're in 4e "identical hamstrung mechanical effects with flavor that explicitly does nothing" territory or sharply divided usage parameters until you're in 4e "powers everyone uses all the time and rituals no one ever uses" territory, neither of which are really workable.

No you don't. It's balanced with the Beguiler now. People talk about how balancing different classes is really hard, but if you look at the level of balance that actually exists in 3e, it doesn't actually seem to be. You could fill a PHB with classes that are all A) reasonably balanced and B) use different resource management systems. I'm not really sure how much more you could want in terms of "it is in fact possible to balance classes that are different".

The reason imbalance exists is because games are designed by people who are bad at math and don't face strong pressures to make a balanced product. It's not some mystery we can't solve, it's just that if you can balance a system as complicated as D&D, you can get more money from Google or Amazon than WotC would ever imagine paying you.

There are multiple kinds and levels of "balance" you can talk about. There's "balanced" in the sense that you can throw two classes in the same game in place of one another (or in the same party at the same time) and they work out pretty okay because they both have strengths and both have weaknesses and both can contribute and the game's not gonna implode, which you're talking about, and then there's "balanced" in the sense that two classes both have the same structure and same access to similar effects and same utility value in most scenarios and minigames and such and they exist within a narrow band of power relative to one another, which Telok was talking about. I'm personally fine with the former approach to balance, and I think your list of 11 classes would be a great pick for a game that deliberately restricted classes to provide a more balanced pseudo-PHB.

My point with the Sorcerer analogy was that a Sorcerer picking Warmage-like spells and an actual Warmage are like the "simple" and "complex" versions of a single Mage class, and the only way you could reasonably make the Warmage and the Sorcerer roughly evenly balanced with one another in that particular scenario is to pre-select a list for the sorcerer as well, because otherwise you can't simply look at balance between them at a single level (or even multiple level benchmarks) because the power and effectiveness of a given sorcerer's spells known overall is going to vary drastically compared to the warmage's list even if their picks look very similar at certain levels. If you're not trying to make class structures so rigid (and thus more sensitive to outlier abilities) or don't care about the simple version of a class being equivalent to the complex version so long as they're vaguely in the ballpark of one another, that stipulation goes away.

I think this is misleading, because it suggests that those abilities are inherently magical in some respect, or that we're turning martials into casters. We aren't. It's simply that 3e defined practically every ability worth having as a spell, so any new ability that does interesting things is going to feel like a caster ability, even if it's a martial one by any reasonable standard.

It's not necessarily 3e's caster focus that makes people think everything worth having is magical, but that D&D has trained people to think that things formatted like spells are magical even if they're not. As I pointed out in the Great ToB Debates of 2010, this is the Ki Blast feat from PHB2:

KI BLAST
You focus your ki into a ball of energy that you can hurl at an opponent.
Prerequisites: Dex 13, Wis 13, Fiery Fist, Improved Unarmed Strike, Stunning Fist, base attack bonus +8.
Benefit: You can expend two daily uses of your Stunning Fist feat as a move action to create an orb of raw ki energy. You can then throw the seething orb as a standard action with a range of 60 feet. This ranged touch attack deals damage equal to 3d6 points + your Wis modifier. The ki orb is a force effect.
If you fail to throw the orb before the end of your turn, it dissipates harmlessly.
When you take this feat, you gain an additional daily use of Stunning Fist.
Special: A fighter can select Ki Blast as one of his fighter bonus feats. A monk with the Stunning Fist feat can select Ki Blast as her bonus feat at 8th level, as long as she possesses the Fiery Fist feat and a base attack bonus of +6 (other prerequisites can be ignored).

..and here's the Fan the Flames maneuver from ToB:

FAN THE FLAMES
Desert Wind (Strike) [Fire]
Level: Swordsage 3
Prerequisite: One Desert Wind maneuver
lnitiation Action: I standard action
Range: 30ft.
Target: One creature
Flickering flame dances across your blade, then springs toward your target as you sweep your sword through the air.
A skilled Desert Wind adept can gather flame within his weapon and hurl it through the air. When you initiate this maneuver, you launch a fist-sized ball of white-hot fire at a single opponent.
If you make a successful ranged touch attack, your target takes 6d6 points of fire damage.
This maneuver is a supernatural ability.

Both involve a vaguely-Eastern-themed mostly-martial-with-a-bit-of-magical-flavor class using something from some kind of resource pool to chuck fire at people, yet people accepted Ki Blast just fine without complaining that the monk was being turned into a caster while Fan the Flames caused an uproar because it's "too anime" or "giving martial classes spells" or (as Cluedrew put it) "mechanically a spell-like ability while being flavored very differently" because the Fan the Flames text looks like a wizard's spell block while the Ki Blast text looks like a fighter's feat block, even though the flavor is basically identical.

[Ignimortis]

Both involve a vaguely-Eastern-themed mostly-martial-with-a-bit-of-magical-flavor class using something from some kind of resource pool to chuck fire at people, yet people accepted Ki Blast just fine without complaining that the monk was being turned into a caster while Fan the Flames caused an uproar because it's "too anime" or "giving martial classes spells" or (as Cluedrew put it) "mechanically a spell-like ability while being flavored very differently" because the Fan the Flames text looks like a wizard's spell block while the Ki Blast text looks like a fighter's feat block, even though the flavor is basically identical.

Ki Blast also sucks a fair bit more ;)

[PairO'Dice Lost]

Ki Blast also sucks a fair bit more ;)

Like the Quasielemental Plane of Vacuum, yeah. People complained about ToB maneuvers that were weaker than thematically-equivalent feats and class features as well back in the day, I just picked Ki Blast because the similarity to an equivalent maneuver kinda slaps the reader in the face in that way that it doesn't for, say Up the Walls vs. Dance of the Spider or Blazing Berserker vs. Flame's Blessing.

[Cluedrew]

On NigelWalmsley: I don't have much to add to your post besides: yes it does work as an existential proof. Everything else I think either still stands (its not a strategy even if it was not meant to be) or has obviously been corrected (if I say "Is that 11?" and you say "Yes it is." I'll believe you).

On Balance: The balance I aim for is "meaningful contribution". I don't want to make this thread about the definition of balance but it is roughly: often helps, sometimes takes center stage, sometimes fails spectacularly.

On Flavour: The feel of an ability does not only come from flavour text alone. If the ability says "Defeat all non-[story] enemies in this and all adjacent locations. You are defeated." you are probably 80% of the way there in figuring out what this ability is supposed to represent.* By reusing a lot of the underpinnings of the magic system for a new non-magic ability system you have primed people to see it as magic. Plus D&D magic is already pretty weird so using it to represent physical abilities starts getting contrived.

Now one can create a universal ability system that could be used for all sorts of abilities. Or you can create 11 different resources systems if you want. But you can just swap out the flavour text on something that obviously was designed for one use case and get something that fits another.

Spoiler: * if not

Show

Its roughly the text of "I'll See you in Hell", which might be my favorite card in any game. The image shows a man poring out gasoline as monster rush him and the flavour text is: "He made his peace and lit the match."

[Telok]

The other problem with a "one social thing and one combat thing at 5th" setup is that "combat thing" and "social thing" (and "stealth thing" and "downtime thing" and "utility thing" and...) aren't nice distinct categories you can easily split up. Social-oriented mind reading/charms/language spells/etc. can be used in combat, combat-oriented shapeshifting/skill buffs/rerolls/etc. can be used in social contexts, and things like disguises and stat boosts work in both contexts and a half-dozen others besides. To make such divisions meaningful, you'd have to either sharply restrict the scope of effects until you're in 4e "identical hamstrung mechanical effects with flavor that explicitly does nothing" territory or sharply divided usage parameters until you're in 4e "powers everyone uses all the time and rituals no one ever uses" territory, neither of which are really workable.

Most balanced game I ever played had as chatacters a d&d dragon, a 1000 year old drunk werewolf, a cowgirl with six shooters and psychic abilities, a blind circus acrobat with an imaginary friend, and a famous movie actor who could turn into living metal. Everyone could do combat, skills, and social. Nobody was excluded from anything and all could contribute to anything if they wanted to.

Champions with a DM who actually followed the advice in the books, natch.

[NigelWalmsley]

the idea that you can feasibly balance a "simple mage" with a "complex mage" goes completely out the window.

It seems like that that point you've just changed your problem into "balance a classless system", which is reasonably achievable. I actually don't think the hard part of balancing complex and simple options is the doing of it, but doing in it a way that allows people to get what they want out of complex options without those options either feeling like a waste of time, or being overpowered.

There are multiple kinds and levels of "balance" you can talk about. There's "balanced" in the sense that you can throw two classes in the same game in place of one another (or in the same party at the same time) and they work out pretty okay because they both have strengths and both have weaknesses and both can contribute and the game's not gonna implode, which you're talking about, and then there's "balanced" in the sense that two classes both have the same structure and same access to similar effects and same utility value in most scenarios and minigames and such and they exist within a narrow band of power relative to one another, which Telok was talking about.

I don't think the distinction you're trying to make is as compelling as you think it is. If you can solve the problem of getting two different classes to contribute to "the game", then getting two different classes to contribute to any particular minigame is by definition a smaller problem. If you can make a Warmage and a Sorcerer such that you're indifferent between which of the two classes you have in the party, it's difficult for me to believe you can't write some abilities for the Warmage to use in the exploration minigame or whatever.

pre-select a list for the sorcerer as well, because otherwise you can't simply look at balance between them at a single level (or even multiple level benchmarks) because the power and effectiveness of a given sorcerer's spells known overall is going to vary drastically compared to the warmage's list even if their picks look very similar at certain levels.

Again, that just doesn't pass the smell test. There are plenty of different Sorcerer builds that are reasonably balanced with a Warmage, and giving them the ability to pick between different spells isn't inherently going to unbalance things. Look at the Warmage versus the Crusader. One's got a fixed list of options, the other can pick, and they're still reasonably balanced.

the simple version of a class being equivalent to the complex version so long as they're vaguely in the ballpark of one another

That seems like a strawman. No one expects perfect balance, but you can get a much smaller error margin than D&D has typically had without having to give up anything other than the imbalance itself.

By reusing a lot of the underpinnings of the magic system for a new non-magic ability system you have primed people to see it as magic.

Well that gets to the question of what counts as magic. Which is complicated, because the various magic-using classes have a bunch of stuff going on. What makes what a Wizard does "magic" and what a Fighter does "not magic"? Is it mechanical (e.g. spell levels, spell slots)? Thematic (e.g. calling things spells, wearing robes)? Is a Paladin martial or magical? A Crusader? Is the Barbarian's Rage a caster-like ability?

Of course, there's also the question of why we care if our sword guys are using magic. You'll find far less objection to spells like Rage or Heroics or Tenser's Transformation that turn casters into martials. And the source material makes vanishingly little distinction between "sword guy" and "spell guy" to begin with. Gandalf, probably the most iconic fantasy wizard, has a sword, and no one in the setting thinks that's at all weird. The idea that you would have a firm distinction between caster and martial is an invention of D&D, and it's one that causes more problems than it solves.

[martixy]

Here's my final answer: ^{(which I actually arrived at pretty fast)}

Casters and martials don't need to be balanced based on power level.

Casters and martials need to be balanced based on a fun level.

Fun is being able to do things in an encounter and meaningfully contribute. It's having options to choose from and meaningful decisions to make. (Hit it with a sword or a spear is not one of those. Clip its wings so it falls or interrupt its spell is one.)

Then work towards that design goal.