PDA

View Full Version : Emergent Casters



SpiderBrigade
2007-04-03, 01:47 PM
This is something I was thinking about while following the Forcecage Fix Idea (http://www.giantitp.com/forums/showthread.php?t=39157) thread, and decided to start up a new thread rather than (further) derail that one. My basic idea is that it may be almost impossible to really balance the spell-caster classes without a complete rewrite of the magic system, and I'd like some opinions on/explorations of that view.

The core of the issue has to do with emergent properties. Basically, there are some situations where you take thing A and thing B, and when you combine them you get something that has not only the qualities of A and B, but new properties as well. So A+B=ABC, or similar.

Wizards (and also clerics, and to some extent druids) have a bad case of this, and it is the source of their strength/overpoweredness. Looking at the spells individually, most of them are fine (with some exceptions like polymorph, a few others). However, when you add together combinations of spells, or feats and class abilities that let you use spells differently, you can achieve much more powerful effects.

Now, this is also true to some extent for the melee (or fighter-type) classes. Put Power Attack together with some charging feats, for instance, and you ramp up your damage very quickly. But in this case you rely only on class abilities and feats. Even using only the core books, there are many more spells than feats, and therefore many more potential combinations. Furthermore, a fighter-type build will be forced to choose one particular "combination," while a caster will know enough spells to have access to many. This also makes casters much more flexible.

This is what makes balancing spells so hard. Each one must be, individually, equivalent in power to other spells of its level, as well as roughly comparable to the class abilities other characters get at that point in their progression. However, the spell must also not become "too powerful" when combined with any of the other spells available to that class. Given the sheer number of spells out there, the problem becomes impossibly complex.

Some people look at the Tome of Battle as a vision of what could be done with meleers to bring them into line with casters. This is feasible as long as the goal is to make fighter-type classes as breakable as magic-users. The more maneuvers are published, the greater the range of emergent techniques a martial adept will have access to. Hell, there are already many builds on the WotC optimization boards that do incredible things with that book.

Of course, one of the things that ToB does do better is to limit the flexibility of the adepts. That's the other problem with casters: the wizard is famously Batman, and even clerics and druids have a lot of versatility. A charge-specialized fighter (for instance) will come across lots of situations where his abilities can't be used. Meanwhile, the caster will simply switch to a different powerful combo move, one appropriate to the situation. He will know enough spells to handle a wide variety of challenges.

The conclusion that I come to from this is that, unless the very nature of how wizards and other casters work is changed, there's not a lot that can be done to bring their power down, rather than bringing the ability of the other classes up. There are too many spells to remove the possibility of powerful synergistic combinations while keeping the individual spells balanced against each other, and casters know so many spells that they WILL have one or two strong combos.

The solution may be to drastically reduce the number of spells any caster can know. The sorceror is already much less powerful than the wizard, although still very strong. He is more akin to the fighter in that situations where he cannot use his specialization will be more difficult for him, and as a result is likely to keep his options flexible at the cost of forgoing some of the "win-button" tactics.

I should just say to wrap this up, that I'm not posting this to somehow get an answer to "how to balance casters." I don't think there's a simple houserule that will fix the problem, which is kind of the point of the thread. However, if you think you HAVE such a houserule, I'd be interested to hear it! Most of the ones I've seen in the past are essentially a from-scratch rewrite of the magic system, either by rebuilding the caster classes, or by fussing with spell descriptions and spell levels. Neither one of those is IMO practical because you're now dealing with a 75% homebrew system.

storybookknight
2007-04-03, 01:53 PM
Interestingly, a few of the articles over at Wizards.com (in any situation where people are making premade characters for quickplay) have stayed away from any casting class that doesn't have a limited set of "spells known" - or in polls to create a balanced party, "prepared" spellcasters were more or less disallowed. When "prepared" casters were allowed, the spells were picked beforehand, with no chance of players switching them.

Ikkitosen
2007-04-03, 02:15 PM
For wizards, I like the idea of only using specialists (diviners lose 2 schools) and replace Scribe Scroll with Focused Specialist. That way each wizard loses 3 schools and is forced to choose many of their spells from their specialist school.

This is not a total fix, just a step in a direction.

storybookknight
2007-04-03, 02:22 PM
For wizards, I like the idea of only using specialists (diviners lose 2 schools) and replace Scribe Scroll with Focused Specialist. That way each wizard loses 3 schools and is forced to choose many of their spells from their specialist school.

This is not a total fix, just a step in a direction.

Isn't the point//flavor of the wizard its versatility, in comparison to the sorceror's raw power?

Now, if instead you gave wizards 6 spells known per level and 4 spells per day, whereas the sorcerer has 4 spells known and 6 per day, you might achieve a happier balance.

Ikkitosen
2007-04-03, 02:23 PM
Isn't the point//flavor of the wizard its versatility, in comparison to the sorceror's raw power?

Now, if instead you gave wizards 6 spells known per level and 4 spells per day, whereas the sorcerer has 4 spells known and 6 per day, you might achieve a happier balance.

5/8 of all spells ever is still FAR more versatility than a sorc.

storybookknight
2007-04-03, 02:37 PM
5/8 of all spells ever is still FAR more versatility than a sorc.

And 5/8s of all spells ever is still probably not enough of a "hit" to rebalance the game.

Ikkitosen
2007-04-03, 03:00 PM
No, but it is a step in the right direction! The OP states they're not expecting to fix the problem, just have a discussion. I think this is a good, flavoursome houserule myself.

Person_Man
2007-04-03, 03:19 PM
I think the main answer is found in the PHBII with the Duskblade and Beguiler. Instead of having a huge array of spells, casters should only have a limited list. With more discrete powers, it becomes much easier to balance casters against non-casters.

Furthermore, no PrC should have more than 4/5 caster progression, and the first level of the PrC should always be a missed level of caster progression. That way there would be a real tradeoff for caster PrC.

I would also say that all spells should have a Saving Throw, and that passing the Save should mean no effect on the target.


Another alternative is to up-power all of D&D, and work ToB rules into core. Melee and Skill Monkey classes would progress on manuevers and stances of various segregated types, and casters would progress on spells.

SpiderBrigade
2007-04-03, 08:58 PM
I think the main answer is found in the PHBII with the Duskblade and Beguiler. Instead of having a huge array of spells, casters should only have a limited list. With more discrete powers, it becomes much easier to balance casters against non-casters.Yes. Limit the spell list and you curtail the synergy any one class can get. Plus you can have members of what is ostensibly the same class having very different "sets" of abilities, as opposed to what you have now where every even-slightly-optimized wizard is going to have certain spells. Of course, it's still very possible to break these newer caster classes...but you have to work harder.


Furthermore, no PrC should have more than 4/5 caster progression, and the first level of the PrC should always be a missed level of caster progression. That way there would be a real tradeoff for caster PrC.

I would also say that all spells should have a Saving Throw, and that passing the Save should mean no effect on the target. Agree with the former, as there are waaay too many PrCs that give you these awesome new abilities, AND you progress in casting. I mean, for a wizard...oh no, you just lost out on some metamagic feats and familiar bonuses. Not crippling.

The latter is getting into the territory of "completely rewriting the magic system, for me. Which is probably necessary, so fair enough.


Another alternative is to up-power all of D&D, and work ToB rules into core. Melee and Skill Monkey classes would progress on manuevers and stances of various segregated types, and casters would progress on spells.I think this is what a lot of folks are hoping will happen if/when 4.0 comes around. But there are also a lot of people who think that "blade magic," while exciting, is not what Fighters should be about. So that's tricky.

Merlin the Tuna
2007-04-03, 09:13 PM
Furthermore, no PrC should have more than 4/5 caster progression, and the first level of the PrC should always be a missed level of caster progression. That way there would be a real tradeoff for caster PrC.This is a good general rule, but it's not one that necessarily should always be the case. In 95% of cases, spellcasting should be getting lost somewhere, but it doesn't necessarily need to be in the PrC. The Spellwarp Sniper (CScoundrel) is a good example of this -- it's got full spellcasting progression, but it also requires a sneak attack die.

storybookknight
2007-04-04, 01:04 AM
Granted; simplify, simplify, simplify is the key.

How much more balanced would a Warmage be if they were the only ones who could cast the Energy Orb spells?

I like the notion of limited spell lists - maybe the way Psionics does with its powers, so that specialist wizards have access to a set of spells noone else does, or you have to specialize in Illusion to get Phantasmal Killer... stuff like that.

Especially if you said "specialist wizards get to know all the spells on their specialist list, but only a few that aren't" - make it a real difference from the Sorcerer.

Zincorium
2007-04-04, 03:40 AM
Granted; simplify, simplify, simplify is the key.

How much more balanced would a Warmage be if they were the only ones who could cast the Energy Orb spells?

I like the notion of limited spell lists - maybe the way Psionics does with its powers, so that specialist wizards have access to a set of spells noone else does, or you have to specialize in Illusion to get Phantasmal Killer... stuff like that.

Especially if you said "specialist wizards get to know all the spells on their specialist list, but only a few that aren't" - make it a real difference from the Sorcerer.

Personally, I really like the way that the Dread Necromancer, Warmage, Beguiler and Duskblade spell lists work. You have a set number of spells, that the DM knows about in advance and can alter if necessary without having to search through 20+ possible rulebooks to see if things from those need to be off limits.

You can cast any spell on your list. No agonizing over which spell to take at any particular level, far easier for beginning players and lazy experienced ones like me.

Granted, all of those classes get some other benefit for the decreased casting, higher HD, special abilities, etc. I think this is a good thing, as it reduces the eggshell factor of beginning spellcasters and gives them something worthwhile they can do even if they run out of spells to cast.

Shiny, Bearer of the Pokystick
2007-04-04, 06:30 PM
As regards published materials: I think the Duskblade and Beguiler both present excellent alternatives to the 'emergent/ultraflexible' nature of core prepared casters; so, agreement there.

I do not, sadly, enjoy the Tome of Battle at all; I like some of the rules refinements in it, and I really do think it's cool- hell, awesome- but I do not want the melee classes to suffer from the emergent problem we've just described. Agreement there as well.

As regards the dreaded Homebrew Wastes, where a thousand failed rewrites and revised spell lists languish in torment eternal, I have been working on something, which may be crazy, but I'd be interested in opinions on the basic principles thereof.
I'll hide it in spoiler because it's long, and I can summarize it at the end if you don't feel like reading.

Firstly, my proposed new arcanist is as chained to the dice as anyone else; similar to the Truenamer in Tome of Magic, his casting is skill-based.

Secondly, the effects the arcanist is able to produce are sharply limited; each category of effects essentially forms a separate class. Rather than the traditional 'schools' of magic, magical effects are separated by their basic nature- thus, an arcanist who specializes in manipulating life would have effects resembling both necromancy and conjuration(healing), but only effects of that type- perhaps a few otherwise, but using them is extremely difficult.
To somewhat counteract this significant reduction in versatility, the exact effect of the arcanist's magic, for instance, its duration, range, or target, is somewhat mutable- though more difficult than the standard version, the arcanist could, for instance, cast a fireball-like spell that wove through his allies like a tentacle of flame, or a semi-mobile orb of fire.
Manipulating a spell in this way, however, is more difficult, and likely decreases the spell's total power; similar to how a psion could likely augment one aspect of a power, but not two.

To my mind, this gives the arcanist a 'versatile' or at least 'creative' spin, without the emergent problem.

Thirdly, the vast majority of the arcanist's effects (except, of course, those that have the 'created but not inherently magical' property) allow saving throws.

Fourthly, to compensate for these changes further, the arcanist's skill list is expanded, their skill points upgraded, and their hit die increased to d6.

With this system, the locus of choice shifts.
The 'which spells do I take' model is what core assumes- an easily abusable system, the source of the emergent problem, and above all, not a terribly fun thing; one has to consider a mind-boggling array of options at the start of every adventure, indeed, every day. Using a wizard's abilities is fun; starting your wizardly day with the equivalent of filling out your day planner is, to my mind, not; thus, from a player's perspective, not entirely lovely.

From a DM's perspective, the 'spell choice' model is also a nightmare- the DM is expected to know the abilities of every class, all the time, and consider their likely synergies when designing encounters. The number of possible ramifications from even a single spell (for instance, fly) may be large enough to necessitate the redesign of massive sections of dungeon, or the creation of entirely new design principles; expanding this issue a thousand thousand times, for every spell produced, then a thousand thousand times more, for all their possible combinations, renders the task nigh-impossible when even the most aware DM comes up against even a moderately optimized player.

But I digress.

With my system, the locus of choice isn't an essentially out-of-character 'Spell Selection'. Instead, it's something in-character, based on tactical decision and role-playing choice; to whit, an 'Effect Selection' system.

Essentially, I feel that granting everyone the same basic effects available eliminates the emergent problem- but granting flexibility in how those effects are employed keeps the fun available for players.

Rather than taking table time agonizing over whether to take Fireball or Dispel Magic (or, worse, immediately moving to the most 'optimized' choice regardless of character), they instead make quicker, more immediate, more in-character decisions: do I try to pick off the archers by increasing the range of my fire effect, or do I try to increase the damage and attempt to take down their front-liners- or should I manipulate sonic energy to damage the building itself, instead? Do I bolster my attempt to beguile the barmaid completely, or try to gain the attention of the entire tavern?

I've seen classes and prestige classes sort of groping towards this model of choice for some time now, and I think psionics does a good job with it...but I'm hoping that keeping the 'magic flavor' and applying it to an effect-decision system will help things.

...I've ranted.

Fax Celestis
2007-04-04, 06:44 PM
For wizards, I like the idea of only using specialists (diviners lose 2 schools) and replace Scribe Scroll with Focused Specialist. That way each wizard loses 3 schools and is forced to choose many of their spells from their specialist school.

This is not a total fix, just a step in a direction.

Something I've done in my games is "reschool" all the spells in existence (a task in and of itself) into four schools: Creation, Alteration, Divination, and Destruction. Wizards are forced to specialize, and in doing so, lose access to a full half of the spells available. Ultraspecialists like the Red Wizards of Thay are really out on a limb as they only get spells from one school. This also makes the sorceror a more viable option, since he doesn't have to specialize.

Matthew
2007-04-15, 04:31 PM
Some good points here SpiderBrigade and I am in full agreement with your conclusion that there appears to be no easy fix for this problem. Bringing Spell Casters under control is a difficult process.

Ikkitosen
2007-04-15, 04:38 PM
Something I've done in my games is "reschool" all the spells in existence (a task in and of itself) into four schools: Creation, Alteration, Divination, and Destruction. Wizards are forced to specialize, and in doing so, lose access to a full half of the spells available. Ultraspecialists like the Red Wizards of Thay are really out on a limb as they only get spells from one school. This also makes the sorceror a more viable option, since he doesn't have to specialize.

Sounds grand - how long have you spent re-schooling? In doing so have you limited the books available?

Kyace
2007-04-15, 05:53 PM
Emergent: simples rules that can model/create more complex systems than one would seem.
combination grouping: The raw number of possible groups of x items uniquely picked from a total possible selection of x_max items.

If we say that x is the number of different spells a spellcaster knows and x_max is the number of possible spells that the spellcaster could learn, then you can use (x_max!)/(x!*(x_max-x)!) to figure out how many possible spellbooks filled with x spells from a x_max available spells. If you were limited to spellbooks of only 5 spells and only 100 possible spells, there would over 75 million spellbook combinations possible. Since x_max! grows very fast (faster than even all polynomials and exponential functions), the likelyhood that a possible spellbook contains a gamebreaking combination of spells increases very quickly.

What you all are talking about is basic combinations, not emergent theory.

You want emergent theory? Fine: Lets make a mock spell system with only four spells and only a couple rules.
Spell 1: Fire! Target 5 foot wide square has it's background magic value set to 52.
Spell 2: Ice! Target 5 foot wide square has it's background magic value set to 32.
Spell 3: Void! Target 5 foot wide square has it's background magic value set to -64.
Spell 4: Wall! Target 5 foot wide square has it's BMV set to 1 for one day. If the square's value was already 1, set it to 0.

Rule 1: At the start of a round, apply the following rules to all squares at once (all squares follow the rules based on the values at the start of the round, not based on any changes since then).
Rule 2: If a square has a background magic value other than 1 and one of it's neighbor's doesn't have a BMV (or has a BMV of 0), then half the value of the BMV square's value then set all neighboring squares without a BMV to the same half value.
Rule 3: If a square has a BMV of 13, then deal 1d6 fire damage to anything in that square and set that square's BMV to 0. If the BMV equals 8, deal 1d6 cold damage and set square's BMV to 0. If the BMV equals -16, then deal 1d6 untyped damage and set square's BMV to 0. If the BMV equals -1, deal 1d6 of both fire and cold and set square's BMV to 1.
Rule 4: If a square with a negative value and a square with a positive value neighbor each other, negate the value of the positive.
Rule 5: If a square has a value of -8, then all neighboring squares without BMVs of 0 gain values of 32. If a square has a value of -13, then all neighboring squares with BMVs of 0 gain values of -64.

Form a large enough box of Wall!s, maybe with some Wall! mazes or something in it, cast Fire!, Ice! and Void! within 4 or so squares and watch the result.

PinkysBrain
2007-04-15, 07:05 PM
This is what makes balancing spells so hard. Each one must be, individually, equivalent in power to other spells of its level, as well as roughly comparable to the class abilities other characters get at that point in their progression.
The problem with magic has always been it's combination of flexibility and potency more than combos. So no, spells should not be equivalent in power to the class abilities of other characters ... unless those characters have the same level of flexibility IMO.