Making Balanced Spellcasters is Hard.

[Username17]

This is a thread about how to generate game mechanics for spellcasters, and as such it's going to be really long and take a while to get to the point. For that, I apologize.

First things first, there is the basic design decision of Point Based vs. Level Based. The more point based your system is, the greater versatility it has. However, it also has more min/maxxing. A level based system has more character breadth but has less customizability in character design.

Level Based design differs from Point Based design on the fundamental concept that essentially people are forced to spend their points on a diverse set of abilities that they otherwise would not purchase. Another way to look at this is that in a Level Based system, you get your main power at some rate, and then you get a variety of tangential abilities and defenses for free. The trade-off is customizability vs. diversity of individual characters.

Secondly, the non-magical characters. Unlike magical characters, the non-magical character can be designed in a truly genre-independent fashion. Whether you are going class-based or point based or some combination, the non-magical characters can be designed in a truly archetypical fashion.

That is, there are relatively few things that warriors do, in the broad scheme of things, at leas so far as combat is concerned. They can:

Dish Out Damage
Affect Difficult Targets
Impede Enemies
and of course Take Punishment

That's really it. As such, you can have a point system where people can purchase abilities which correspond to those basic concepts, or you could have a level based system in which people advance one or more of these concepts each level depending upon class - or even a hybrid where characters advanced one or more of these concepts each level at their whim while still gaining various minimum other abilities at a base minimum. Under this over simplification, having an ability to take damage less often is considered to be the same basic ability as taking less damage from each attack - as in the long term it is (think of it as gambling for having a higher damage resistance - double or nothing, as long as the odds are fair it doesn't really make any difference from a game balance standpoint). More unusually, note that having Ghost Touch Weapons, or the ability to penetrate Damage Reduction is essentially just another facet of the ability to have a Ranged Attack - in either case you are affecting enemies you would otherwise be unable to hurt.

---

Now, once you've set the power level for Warriors, you have a variety of choices with what to do with Magic. Unfortunately, here your choices are not iconic. When you write game mechanics for spellcasting in one completely balanced way, it will feel completely different from writing a different set of balanced game mechanics.

Your basic choice for Magic is whether to make it:

More powerful than Swords (or whatever weapons people are using)
Less powerful than Swords
or Exactly as powerful as Swords

Now, if you make spells just as powerful as Swords, you are in a situation like Champions strives for - one in which characters who are "magical" pretty much feel the same as non-magical characters. The only real difference might be that a third of the enemies are magic resistant and a third of the enemies physical resistant, for example. This is very easy to balance, just invert all the physical effects - you don't even need new game mechanics. Spells in this system don't need an endurance bar unless swords do as well.

Contrarywise, if you make spells More Powerful than Swords, as is the case in many (probably most) game systems, you have the immediate problem of limiting the spellcasting in some fashion. The basic methods are these:

Charge Casting: In Charge Casting, there is a limited amount of spell juice that can be used in a period (day, resting period, month, whatever), or even a limited amount which can be used ever. This includes spell slots, spell points, expendable components, limited ability uses, etc.

Guess and Check Casting: In Guess and Check Casting, the player chooses available magical effects ahead of time under the assumption that sometimes your effects are going to be inappropriate to the situation. Ideally, for every time your flame blast is twice as effective as a sword attack you should encounter a fire beast who is flat immune to your magical might.

Master Plan Casting: In Master Plan Casting, spellcasting takes longer to accomplish than other activities. When your spell takes effect, it will presumably be larger, but in the mean-time enemies can run away or attempt to stop you. In a sense you are investing your actions into a much larger action at the end.

Skills and Failures Casting: In Skills and Failures Casting, your spells have a chance of not going off at all. This means that spellcasting is like having more dodge ability - you get a bigger effect for a lower reliability of your effect happening at all. You are, in essence, gambling your actions for a greater reward for success vs. a smaller chance of getting that success.

Recharge Delay Casting: In Recharge DelayCasting, you can't use your magic for a period after you use it. This can be done with backlash, drain, or just a good old fashion timer. In any case, you are essentially borrowing from your future actions for a greater effect right now.

Note that there is really very little reason to combine these limitations - although virtually all game systems do. While a simple Skills and Failures system of magic limitation can afford spells which are twice as effective as normal actions but half as likely to work - a Skills and Failures + Guess and Check system would be much harder pressed to do that. For example, to support a spell effect that was twice as good, you might have a 75% chance of getting your spell off, and run into 1/3 of your enemies against whom all of your spells were completely ineffective.

The flip side:

Making Magic Less Powerful:

While a more powerful magic set has to have things to make it not function in some circumstances to compensate, when magical effects are less powerful, then magic has to be compensated with greater accessability. Intriguingly, the methods of doing that are precisely the opposite of the limitations to restrain a more powerful magic system:

Unfaltering: The opposite of a charge caster, the Unfaltering Caster is able to use magic more than the Warrior can swing a sword. This usually requires some kind of endurance limit to be in place for swords in the first place (D&D does not have one, for example, but Champions does).

Right Tool, Right Job Casting: The opposite of a Guess and Check Caster, the Right Tool, Right Job Caster always has an appropriate magical effect. In this system, while a fire bolt is less impressive than a volley of arrows - the spellcaster can substitute in an ice dart, a bolt of lightning, or a mental thrust as desired. This makes the spellcaster able to confront any enemy - though not as well as a Knight could confront an enemy who played fair.

Preemptive Strike Casting: The opposite of the Master Plan Caster, the Preemptive Strike Caster actually gets his magic off before sword blows can happen. This generally requires a system which has weapon speed or the like already in place, or a phase system for simultaneous play (magic going off in the magic phase, weapon attacks going off in the combat phase, for example).

Reliability Casting: The opposite of the Skills and Failures Caster, the Reliability Caster takes comfort from the fact that while swords and arrows miss - spells don't. While a lightning bolt in this system is less hurtful than a halberd - its unerring accuracy makes it very similar over the long haul.

Autofire Casting: The opposite of the Recharge Delay Caster, the Autofire Caster is able to cast spells more often than an archer can pump out arrows. More, smaller spells can add up to a similar effect over all.

Once again, while you could combine multiple benefitsto compensate for weaker magic - there is very little reason to do this. When a spell is more likely to hit and more likely to be the correct tool for the job, this makes the math somewhat intractable - which leads to game imbalance.

[Username17]

Effects on game balance:

Here's a walkthrough of the difficulties of balncing some of the types of magic - we'll start with the early ones:

Charge Casting: The biggest problem with this casting type is that it is very front loaded. That is, players are going to run through their power at the beginning of the fight, not the end. This means that to make this balanced, battling has to regularly go to the point of characters running out of charges. As such, the game balance here is largely in terms of the setting - not the game mechanics. It puts the burden of game balancing the system squarely on the head of the person setting up the opposition.

Guess and Check Casting: For this to be a meaningful limitation, you actually need to go up against enemies on whom your magic does not function sometimes. Since the DM/GM/Storyteller/Whatever probably knows what magical effects you've committed yourself to before it comes up, this can put the players in a very antagonistic situation with the DM - who is, after all, deliberately putting in enemies with the explicit intent of screwing over some of the players.

More later.

-Username17

[RandomCasualty]

A really interesting post, I like it.

Though you forgot one major limitation for casters, perhaps one of the most classic.

Fragility: Casters trade their ability to take damage with their ability to dish out damage. This requires casters to require protection from other characters in order to be successful.

Though I think you should also analyze the out of combat uses of spellcasters too. Spells like teleport, commune and raise dead do stuff that fighters can't. Necessary stuff.

While looking at things from a straight combat perspectve, mixing limitations doesn't seem like a neccessary thing, but I think once you start adding in noncombat casting you really have to.

You need some kind of charge casting if you want to introduce any kind of long term magic, or you must place a limitation on how many spells you can have active at once.

Of course, charge casting sucks badly for combat balancing because to balance it, you have to prevent the wizard from resting when he wants. This is tough to do and in the worst case it looks like the DM is just being a jerk. As far as NPCs are concerned, they could hardly care they have charges because they probably won't live that long.

Now, another big design question is how strong you want the counters to be.

The way I see it, you can run a tactical emphasis setup where certain classes/monsters counter other classes/monsters. So for instance rogue > wizard > meleer > archer, for instance. In this setup you'd be able to beat stuff that's better than you (more CR) by being an appopriate counter class. Really, I don't think this is more of a strategy game setup as opposed to an RPG. Basically this amounts to balancing wizards in a rock, paper, scissors manner. They'll dominate certain stuff, but other monster types, like the rusher, will beat them. Master plan casting produces this effect, where archers can easily disrupt the wizard's spells and take him down. Similarly a creature could have a massive HP/AC total but be totally vulnerable to magic. Also if you use this system you can apply different limitations to each type of spellcaster. Wizards use master plan casting for instance and clerics do something else.

Now, the other option is the egalitarian level based design. The counters that exist are soft counters and situational based ones. An archer won't always have an advantage over a wizard, but he will if the wizard chooses a certain tactic (read as "decides to cast a certain spell"). This is where mixing limitations becomes really important because it allows the wizard to counter various things. Counters don't really assure you of victory in this scenario, they just give very slight numerical edges.

Now because wizards have extra non-combat abilities over the fighter they actually have to struggle in this design to be formidable combatants.

So under this system, the perfectly played wizard will be better than a fighter, the poorly played wizard will absolutely suck and the mediocre played wizard will probably weigh in at a bit worse than a fighter, at least in the combat arena. The wizard becomes a class thats entirely based on playing it well in terms of good spell selection and such. In fact, it rather sucks if you don't play it well.

I think that's almost what you have to do. The wizard has to for the most part suck, and rely on player creativity to give him an edge. If he's not constantly working to get every last drop of power out of his spells, then generally he'll lag behind the fighters in combat, because after all, combat is all the fighter does, he has to be good at it.

I'll come back to this topic later tonight probably when I have more time.

[Username17]

Fragility:

Oddly enough, Fragility usually manifests itself as a form of Master Plan Casting. Since a spellcaster is easier to drop with that mechanic, it is easier to stop them before they get their spell off than to stop a warrior from getting their sword attack - so the game balancing act here is actually the same as if they had a delay on their spells.

Now, here's the game balance skinny on Master Plan Casting and Recharge Delay Casting:

When you are investing your time (that is, taking perhaps more actions, or simply taking the risk that your spell won't go off before the battle ends or you get a face full of fist), you need to be getting interest. When you are borrowing from your time (that is, taking larger actions now at the expense of actions taken later), you need to be paying interest on the effects of your spells.

That is to say, a spell that might never get around to being cast at all (either because it takes too long or the spellcaster is too fragile) needs to be more powerful than whatever the Wizard could be doing right now with a different character set-up. If a spell used now is going to be done so at the expense of your ability to do things later (either because it slaps you with charge limits or simply won't let you cast again for a while) - it needs to be less powerful than what you would otherwise be doing over the course of that period combined.

Actions now are worth more than potential actions in the future, and actions potential in the future are worth less than actual actions you are necessarily taking.

This means that Fragility doesn't really make a lot of sense when combined with Charges or Cast Delay - because it has opposite effects in the big picture.

-Username17

[RandomCasualty]

1081103947[/unixtime]]
Oddly enough, Fragility usually manifests itself as a form of Master Plan Casting. Since a spellcaster is easier to drop with that mechanic, it is easier to stop them before they get their spell off than to stop a warrior from getting their sword attack - so the game balancing act here is actually the same as if they had a delay on their spells.

Well, somewhat, however fragility is a different thing in that if your initiative is high enough you can fire off your spell and lose basically nothing, where master plan casting is going to open you up to attack either way, whether you gain surprise, win init or whatnot.

Where as master plan casting is basically based around giving up time and risking spell failure through disruption by spending more time, fragility is about giving up the ability to soak damage and defend yourself in exchange for more power.

In many cases master plan casting is much worse than simply being fragile. Take a time stop combo in 3.0 in which you blow off a bunch of spells and nuke the other guy. If you are simply fragile, then if you win initiative, you win that battle. If time stop was a 1 round spell however, you're pretty much screwed trying to solo anything. Even if you win initiative, it just means you start casting it earlier. They'll have disrupted your spell long before you get it off, unless the encounter starts at extreme range. From a tactical standpoint, that's pretty significant.

Basically fragility allows for wizards going solo and makes who acts first super important. Fragility really emphasises randomness in combat, because combats are super deadly. When you've traded away most of your defensive ability for offensive ability, basically if they go first, you die, if you go first, they die.

Master plan casting doesnt care as much who acts first, so long as when the wizard acts he has some means of lasting a round or more to get his spell off. Master plan casting pretty much requires wizards have someone or something to delay the enemy so they can cast. The solo wizard is going to get owned with a master plan casting system.

While MP casting and fragility are somewhat close conceptually, they produce very different effects on game balance. There's no good counter for fragility beyond simply acting first and this actually stresses out of combat actions, like hiding and invisibility spells. This also tends to leave fighters in the dust. On the other hand, good movement speed or ranged weapons are a nice counter for a MP caster, and fighters always have their place defending the caster while he gets off his attack spell. Overall MP casting is a much more stable tool for balance.

The main reason I bring fragility up is mostly as an example of a bad balancing tool.

IMO, one of the main problems in 3rd edition is that they use the fragility concept far too much as an excuse to give out excessive amounts of offensive power to wizards, and eventually it just gets to be a simple roll of the initiative dice that determines the winner. I think wizards need to be made less fragile and rely more on some of the other balancing systems, especially MP casting.

The 1 round casting time is one of the most underused mechanics in 3E IMO. Lots more spells should have it I think.

[rapanui]

I'm working on a system that has caster with spells that are probably often more powerful than weapons. They don't have limited casting (not in any real sense... although somethimes they can summon even more powerful effects that do cost them a certain resource) and they don't HAVE to be fragile... although most of them will likely have a d8 HD at most (I should mention the system is classless).

The key to this game then, is that there is a system that starts manisfesting itself significantly when spellcasters start getting the "uber" spells... you know: the ones that drop characters in a single round. This system allows for even characters with weak targetted saves to survive a couple of assults by the mage, and come up with counter tactics.

And there will be counter-tactics for every character, and I've tried to structure it so that all characters will have a set of tactics available to attack weak points... the problem is they have to find those weak points first.

I could extrapolate further on this system if you want, or you can wait until its finished. Or you can ignore me. I suppose it all goes.

[virgil]

I've been thinking about Master Plan Casting, and thought of some possible concerns. There is a certain limit to having long casting times that would still be applicable to combat, since disruption of casting is likely to remain and standing in one place with a timer over your head is just asking for trouble. If you retain mobility while casting, then you can get away with at least a slightly longer casting time.

One thing it does do is make surprise attacks, or any situation where you get 'buff time', exceedingly useful for the Master Plan caster.

Another option I thought of is instead of casting the spell for several rounds solid, you take as much time to cast as you would swing a sword, but the spell itself has a timer. You cast the spell on someone, two rounds later, they get hurt. A similar situation would be what MMOs call the DOT (damage over time), where it creates a minor effect multiple times over its duration.

[angelfromanotherpin]

The fundamental quality of magic, that it is not science, means that the same inputs do not necessarily produce the same effects. So in terms of making magic feel like magic, I have to go with Guess and Check.

That has the advantage of synergy with the level system, since breadth of ability is how you become more reliably awesome.

[Orion]

How is Charge Magic different from Recharge Magic?

[RiotGearEpsilon]

Charge Magic implies you charge /immediately/ before you cast the spell, I think.

[angelfromanotherpin]

There was an old GURPS variant system, which worked pretty well for my taste. It was basically a mana-cost system, except for two things:

1) Mana recharged slowly.
2) You could spend more mana than you had, but going over the limit meant there was a chance of magical backfire. The further you went over the limit, the greater the risk and the worse the backfire.

This had the effect of making wizards unwilling to use their magic for anything that could be achieved another way, but let even low-grade casters go out in spectacular blazes of glory. The backlash roll made it feel like you were tampering with forces best left undisturbed. It was pretty cool.

[virgil]

Charge magic means you have a mana bar, which refreshes wholly after a certain interval, but you can cast any combination of magic over that interval equals the value of the mana bar.

Recharge magic means either something else important to you acts as a mana bar (such as health) or you just have a cooldown period immediately after casting a spell.

I've never seen GURPS magic in action, but having the option of killing yourself can create bad things. You can end up creating throwaway, kamikaze wizards. From the sounds of it, GURPS wizards started off with Charge magic, and once they ran out, they went into Recharge mode.

[angelfromanotherpin]

Normal GURPS Magic (as distinct from what I mentioned above) is just a fatigue-cost system, where you spend the same fatigue you'd spend on physical exertion to cast spells, and as you get low you get tired, and eventually pass out. If you try cast a spell that costs more fatigue than you have, you take the difference in damage, and that damage is a penalty to your casting skill. So you can throw off a last spell at the cost of bleeding at the nose, but if it's too big, it'll cripple/kill you and you'll probably fail anyway.

Kamikaze throwaway characters are always an option, you don't need magic to do it. Most people take a really dim view of that sort of behavior, though.

[virgil]

Ah, I see. Are there any forms of preparation, because that's sounding like the Recharge system whole cloth.

[Psychic Robot]

Are there any non-GURPS examples of charge magic?

[SphereOfFeetMan]

Are there any non-GURPS examples of charge magic?

...D&D.

[virgil]

D&D is a conglomerate of many...

Charge: Limited spells per day
Guess & Check: Specific spell memorization (sans sorcerer)
Master Plan: Inherent fragility in arcane casters
Reliability: A more realistic result of guess & check in D&D

[virgil]

One thing I've been considering very recently is the idea of something that is inherently Master Plan in casting. Instead of doing nothing but chanting for X actions before unleashing the spell, what if you take a penalty to actions you take for X actions and the next action is then unleashed with the primary spell goal. A similar effect would be that the spell could have a snowball effect, the first round being comparatively minor, the second round sees the effect increase, and the culmination of power is attained on the third round (an action being spent to maintain each).

[JonSetanta]

I'm considering Charged (resting period) as a universal mechanism for all 'class' abilities of a new d20-like RPG.
Some less effective power 'modes' will be provided as backup for when a character's universal per-battle powers run out but as a whole the Charged method is what gets a battle over quickly.

I suppose Recharge Delay would also fit since the rest period varies depending on the battle time scale.
If it's a small scale personal battle, rest is mere minutes (undecided so far, maybe just 1d10 minus something) but if a battle scales up to army vs. army it could be up to a matter of hours.

There will be rituals/incantations/whatever as well for the more durable effects, and those will be usable effectively at-will but with longer cast time, but the actual combat stuff will be "until end of combat" or instant.
Some battle powers will have lasting effects beyond end of combat but this requires reaching a certain threshold for affecting a foe, similar to failing multiple saving throws in a row.
Undecided as to how this will work but I have a few options.

I hate per-day. Talisman knows exactly how much I hate per-day.
So, I plan to eliminate that crap from the start.

[Ice9]

I think there may be one more balancing type, one that's very hard to balance but still gets used a lot:

Separate Domain Casting
Casting is more powerful in certain parts of gameplay and less powerful in others. For instance, magic might be less powerful in combat, but more powerful in obstacle avoidance, travel, and social encounters. The problem with this is the Charge problem but moreso; the DM has to shape the campaign to include a balanced ratio of these different tasks.

A modified form of this would be that magic is more/less powerful at specific subcomponents of combat - for instance, magic may be worse at actually killing people, but better at battlefield control and defense/healing. This is a probably more practical to balance, but does limit characters to certain combat roles.

[Ice9]

Something else I thought of, which is somewhat related to this thread - how to have mind-control and damage be cooperative, without making mind-control do damage.


The 3E Situation:
Mind control and damage aren't really cumulative in any meaningful way. In fact, they're almost anti-cumulative, because if you have a Wizard trying to mind-control something while a Fighter kills it, it's quite possible to have the mind-control succeed when it's almost dead, giving you a mortally injured minion. Even if the mind-control is more the Sleep type, it's still either useless or invalidates the others' efforts.

The 4E Solution:
Mind control only lasts for a round, and it does damage. This is completely cumulative with damage, because it is damage. However, it makes things rather bland when nearly every power ends up working the same way - and it means you can never do something like Jedi-Mind-Trick a guard to let you into the royal court - you'll either injure him slightly and start a fight, or you'll blast him unconcious.


So, my solution:
Mind-Control is a cumulative effect, pretty much like the TNE wound system - you roll a level/stat based check against their willpower, with a low-variability roll like 3d6. If you beat their willpower by a lot (as in, being many levels higher than them), then you go straight to complete control. More likely, you slow their mind somewhat, giving them penalties to things and putting future Mind-Control attempts closer to complete success.

The important thing is that one of the penalties accrued this way is something that helps people trying to kill that foe - lower AC, lower damage threshold, can't use certain defensive moves, etc. And vice versa, dealing damage also boosts mental control - let's say you take a -X penalty to your willpower at 1/4 injured, 1/2 injured, 3/4 injured, and dying.

The end result would hopefully be that while it's possible to mind control someone without injuring them at all, or kill someone without mentally affecting them at all, the two activities are cumulative against the same enemy - especially if it's a tough one - without being just the same thing under different names.

[CatharzGodfoot]

The neat thing about TNE-style mind control is that when you first put it into effect, you've got a person who's near-catatonic. She's suffering from (possibly multiple) tier 1-3 confusion effects as well as the tier 4 domination. You actually have to let her rest and heal up the lesser afflictions before she become a useful ally.

[Ice9]

I guess that's a good point - against even-level foes anyway, but I would hope that against a significantly weaker foe you could go straight to useful control without stunning them or blasting them in an obvious way - for instance, the previously mentioned "Jedi Mind Trick" on a guard.