OneWinged4ngel
2007-09-05, 10:21 PM
((My popular article, reposted from the WotC boards / Dreamscarred Press. Enjoy))
Many people go about creating new classes all the time, but a lot of them aren't particularly good. Here, I'm going to examine a few of the elements of what I view to be good class design, and hopefully help a few people out (since they seem to keep asking me for help in this regard). I would love to get some feedback here to make this guide more complete and better able to communicate its points.
Ultimately, this guide should help people to create fun and balanced classes, while avoiding common pitfalls.
Is there a reason to create a new class?
This is one that people miss out on a lot. I can't count the number of classes I've seen on these boards that actually served very little purpouse, because oftentimes the exact same concept could be managed by existing classes (and done with similar or sometimes better mechanical elegance).
In order for a base class to be a solid new addition to the roster, it has to do at least one of two things:
It must either
A) Fill a new niche. For example, the Artificer fills the item creation and use niche as a class, where previously that was never the focus of any base class.
or B) Be mechanically original. For example, while the Warblade fills the "Fighter" niche, it does so in a new and mechanically unique and elegant way.
If it doesn't do either of these things, it may as well not exist.
This post from Tempest Stormwind helps to illustrate my point:
I agree with this 100%.
For me, a good base class has to fit two criteria: It must be mechanically original and it must fill a niche no other base class can do. I'm willing to be lax somewhat on one if the other's particularly strong, though. The society mind (http://boards1.wizards.com/showthread.php?p=7311764) and blue mage (http://boards1.wizards.com/showthread.php?t=137596) do both of these quite nicely (summary of the two: The blue mage is the Final Fantasy archetype of learning enemy spells by experiencing them firsthand, and the society mind is a psionic support class that makes the team operate as if they were parts of a single greater being), while the gemini dancer doesn't fit the niche criterion at all and gets a ton of press as if it did. A lot of it looks like mechanics for the sake of mechanics, actually -- one of the major reasons I dislike Sztany's Ultimate series, in fact.
(On an unrelated note, by the way, if it wasn't clear before, the society mind was published in Untapped Potential, meaning several DMs who require print sources may now peruse it at their leisure. Dreamscarred hasn't forgotten about it, either, and it will be used in later books.)
Those two criteria can help you in designing other base classes. Step 1 is ALWAYS to ask yourself if your concept REALLY needs a new base class instead of just some imagination. From there, it's just a matter of mechanics.
For Example...
Untapped Potential's other base class, the Marksman, was designed as follows: We noted that there was no full-base-attack manifesting class, and when we looked at all the published full-base-attack classes, every one of them except the ranger (and even the ranger, to some extent) had a heavy, heavy melee focus. Thus, a ranged full-base-attack class would be mechanically original, and would be a good place to go. From there, we looked at places like archetypal "small-scale psionics" (i.e. Starcraft Ghosts, The Shadow) and archetypal ranged combatants (there's a very powerful ability there drawn from classic Western films, for instance), and came up with interesting ideas mechanically (such as the ability to develop a signature weapon style). Finally, we put it all together and pared it down until it looked reasonably balanced. To illustrate further, here's some WotC examples.
Duskblade: Fills a niche no other single class can (fighter/mage, a popular niche; the psychic warrior comes close but feels completely different since it can't blast), sort of mechanically original (not so much, but what mechanics are there are solid). Conclusion: Good class.
Warlock: Fills a niche no other class can (and that niche can vary somewhat depending on what you choose, but always having magic at your fingertips has a certain appeal), mechanically original (Self-evident!). On the plus side, it's dripping in unique flavor (which was enough to lure my brother, who NEVER plays anything *but* The Half Orc Barbarian, into trying something new). Conclusion: Good class.
Samurai: Not mechanically original (all of its abilities are fighter bonus feats, essentially), fills no new niche (a Lawful fighter could do this). Conclusion: Bad class.
Swashbuckler: Doesn't fill a new niche (warrior rogue, light fighter), not mechanically original at all. Conclusion: Bad class. But insightful strike's good, so it's worth considering as a dip.
Beguiler: Doesn't fill a new niche (sneaky spellcaster; illusionist comes close), but executes it with mechanical elegance and brilliance. Conclusion: Good class. You'll note that this is the reverse of the duskblade (which fills the niche but isn't mechanically original), yet both are good because their better side is as good as it is.
All of these work without introducing a new system either. If I wanted to, I could laud the Expanded Psionics Handbook, Pact Magic section of the Tome of Magic, the Totemist from Magic of Incarnum, and the Tome of Battle classes as "good" (they all fill various niches and do so with mechanical elegance) and the rest of the Tome of Magic and Magic of Incarnum as "poor" for similar reasons (MoI doesn't fill any new niche except for the totemist, Truenaming fails on so many mechanical levels, and shadow magic is weak on both fronts but shows promise). Are you seeing how to design a good class from this by now? I'd hope I'm being helpful.
If we're including variants of existing base classes, then my money currently lies on (in no particular order) Seerow's Fighter* (http://boards1.wizards.com/showthread.php?t=728493), OneWinged4ngel's Paladin (http://boards1.wizards.com/showthread.php?t=761045), my own Marshal (http://boards1.wizards.com/showthread.php?t=752414), BlaineTog's Soulknife (http://boards1.wizards.com/showthread.php?t=310984), and RadicalTaoist's Ranger (http://boards1.wizards.com/showthread.php?t=738077) (even if the latter may need a bit of tweaking; it's getting intensive playtesting in our group now, right alongside Blaine's soulknife). All of these except the ranger are intended to "fix" an existing class, replacing it altogether (Seerow's fighter is the result of many, many long discussions of what the figher's lacking as a class, OW4's paladin is a reimagining from the ground up of what a PALADIN means, Blaine's soulknife fixes many weak mechanical flaws while giving the soulknife a niche of its own isntead of a bastard niche between stealth and tanking, and my marshal makes a dip class into something worth taking as a defining class). The ranger is meant as a variant that can be used in addition to normal rangers, trading mystical abilities for Tome of Battle maneuvers.
Hope that helps.
* I haven't yet read Otto's classes beyond a cursory skimming, but I like what I see there, particularly the hexblade.
^--It also covers my next point a bit.
Mechanics for mechanics' sake is bad. Seriously. People add little +1s and -1s and tiny little details that don't actually affect the character much based on this and that, and bog it down (and actually can limit the versatility of the character concepts it can fill), when you actually don't need any new mechanics. Some good examples of "mechanics for mechanics sake" are actually some classes that have received high acclaim from some, but criticism from board veterans like myself and Tempest Stormwind, such as Szatany's classes and Frasmage's Gemini Dancer. This is largely because the presentation looks good, but the underlying mechanics are actually rather unnecessary.
Basically, mechanics should have some meaning and significance. They should do something you care about, instead of bogging down the system.
Dead levels suck.
Make sure the player gets something new every level. The reason for this is simple and straightforward: No one wants to take a level where there is no benefit, and moreover dead levels are boring and just plain not fun. Even if it's something minor, you should at least put *something* to fill the void. Ideally, you should be giving a fairly even progression of cool class features.
It is worth noting that gaining a level of spells or something is *NOT* a dead level, because a level of spells actually represents an array of new class features. In fact, it often represents a larger array of new class features at the new level than many other classes get. Don't confuse "class features" with "entries in the special column," because it's just not true. It's also worth noting that if you have a lot of abilities that scale by level, it can also be acceptable to have a few seemingly "dead" levels, since you're actually getting something nifty at those levels even if you don't see it right there in the special column. A bad dead level is one where you're basically the same guy as you were at the last level, except with a bit higher numbers.
You should feel more powerful, more versatile, and generally cooler at every level. A player should not be doing just the same old thing, except with slightly higher numbers.
Tempest helps to clarify, here:
I should add that if you have a lot of class abilities that scale by level, you can also get by with this.
The classic "bad" dead level is Fighter 5. You play EXACTLY the same way you did as Fighter 4, except with slightly different numbers.
This is, at its core, what the no-dead-level approach is meant to avoid. You don't need flashy new abilities at each level, but the character should feel more powerful, more versatile, or similar at every level. He should NOT feel "just like I was, but more so".
(Note a quick comparison in general design philosophy with core melee and ToB characters, by the way -- core melee use the same tactics at level 20 that they do at level 1, more or less: charge, full attack. They have some new features (spells, rages, unique feats, etc) to make the delivery of these tactics possible, and to make the numbers bigger... but it's still the same tactics. Meanwhile, a warblade at level 20 plays significantly different from one at level 1, even if fundamentally they still involve move-in-and-attack. It's because they get options -- and they get them at every level, so it feels good to be a warblade every time you level up. Make your class something you'd want to feel proud to be.)
Uneven progressions suck. There are two common sides of the coin here. There's toploading, and there's the "suck now, but own later" mentality. Both of these ideas generally suck. Toploading is bad because it means that most of your class's progression is actually useless (for instance, the Swashbuckler is often considered a 3-level long class. The other 17 levels are wasted). And "pay for it now for power later" and similar such uneven progressions *really* don't actually work that way in play. Sure, many PrCed up gish builds will be pretty lame at low level, and killer at higher level, but the reality is that most campaigns aren't actually played from levels 1-20. They'll be more like "5-12" or "1-14" or "12-18" or whatever. So that "evening out cost and benefit over levels" doesn't really exist. Making a class good at one level and crappy at another is a bad thing. Ideally, a class progression should be as even as possible and a class should contribute to the party in a level-appropriate way at *every* level. No more, no less. It doesn't actually have to be perfect... but it should be a fairly even progression of cool class features.
I think this honorable Crane puts it fairly well here.
"Suck Now Rule Later" and it's reverse are poor points of game balance and design. Even assuming that all games run from levels 1 to 20 (wildly untrue), and that the characters remain the same throughout all 20 levels (often untrue) that still means that only half your players are having fun at any given point in your campaign, and the others are only hanging on in hopes they'll get a moment to shine at some point via DM intervention. Your players should enjoy every fight they meet and consider every challenge interesting, not singlehandedly win the first 3 fights of the day and stand around useless for fight number 4.
Make the class for everyone who's going to use it, not just you. Basically, this follows a principle that when designing something for public consumption, you want to make it adaptable to everyone's needs. A class shouldn't look like "your specific character's build choices." It should be able to embody a variety of concepts. Writing extensive fluff on the history of some order and the exact way a certain character fights and so forth doesn't actually make for a better class in any way. We've all seen these classes that look like one guy's character, instead of a real base class that can be adapted to a variety of concepts.
Keep it flexible in build. Building on the last point, a class should provide many "viable" build options, allowing it to embody a variety of concepts. With a look at the Wizard, we can see that you can make a tricky illusionist, a war wizard that makes buildings explode, a calculating seer, or a thousand other concepts. Where possible, you shouldn't be restricting the sort of concepts you can use with the class.
Give it options in play. Using the same tactic over and over is boring. If you're a trip fighter with that one trick (trip, trip, trip) then your gameplay is going to become more monotonous. By contrast, the Warblade introduces more versatility and options into every battle.
This is notably distinct from versatility in build. Versatility in build refers to the ability of a Fighter to be built in many different ways, but versatility in play refers to have many options of actions available to you during play.
Plot writing abilities SUCK. Just don't do it. This is a no-no. When I say "plot writing" abilities, I mean stuff like the HORRIBLE Thunder Guide class in the Explorer's Handbook where you get abilities like "Serial Hero: At 8th level, famed Korranberg Chronicle reporter Kole Naerrin writes a serialized account of your adventures appearing over the course of thirteen weeks. You earn 1000 gp per point of your charisma bonus for the rights to your story (minimum 1000gp)." Seriously, WTF? "A guy writes a book about you" isn't a class ability. A class ability is supposed to be some ability that your character has, not something that happens in the plot.
This PrC from the Dungeonomicon parodies the plot-writing abilities and "The class is actually just my specific character put into a progression" problems that we see *alarmingly* often, which is just stupid.
Elothar Warrior of Bladereach
"My name is Elothar. Your name is unimportant, for you shall soon be dead."
The city of Bladereach sits at the mouth of the Typhon River that flows from the Bane Mires into Ferrin's Bay. The elves of Celentian's caravan come every year to trade with the largely human inhabitants of Bladereach and sometimes they leave more than the wares of the Black Orchard Hills when they leave. The results of these dalliances find that they never fit in amongst the people of Bladereach, and are taught the hard secrets of battle that the children of Bladereach have to offer. Often, these half-elven warriors turn to adventuring.
Prerequisites:
Skills: 9 ranks in Use Rope
Race: Half-elf.
Region: Must be from Bladereach.
Special: Name must be Elothar.
Hit Die: d8
Class Skills: The Elothar Warrior of Bladereach's class skills (and the key ability for each skill) are Balance (Dex), Climb (Str), Concentration (Con), Craft (Int), Handle Animal (Cha), Heal (Wis), Hide (Dex), Jump (Str), Knowledge (all skills taken individually) (Int), Listen (Wis), Move Silently (Dex), Profession (Wis), Ride (Dex), Search (Int), Spot (Wis), Survival (Wis), Swim (Str), and Use Rope (Dex).
Skills/Level: 4 + Intelligence Bonus
BAB: Good (1/1), Saves: Fort: Poor; Reflex: Poor; Will: Good
Level, Benefit
1 Way of Two Swords
2 Tommy, Legacy of the Water Stone
3 Magic Swords, Immunity to Petrification
4 I've got that!
5 Double Riposte, Fistful of Rubies
6 Der'renya the Ruby Sorceress
7 Ways and Paths
8 Name of the First Eagle
9 Blessing of the Gnome King
10 Flying Ship, Your Money is No Good Here
11 Demesne of Tralathon
12 Mark of Ruin
13 Sword of Kas, Dwarf Friend
14 Happily Ever After, Khadrimarh
All of the following are Class Features of the Elothar Warrior of Bladereach class:
Weapon and Armor Proficiency: An Elothar Warrior of Bladereach gains proficiency with the Nerra Shard Sword, the Kaorti Ribbon Dagger, and the Shuriken.
Way of Two Swords (Ex): With a single standard action, an Elothar Warrior of Bladreach may attack with a one-handed or light weapon in each hand at no penalties to-hit or damage for the weapon in his primary or off-hand.
Tommy: At 2nd level, an Elothar Warrior of Bladereach is joined in his adventures by Tommy, a 5th level Halfling Rogue from Figmountain. Tommy is a loyal cohort and gains levels when the Elothar Warrior of Bladereach does. Other Halflings will be impressed by Tommy's apparent loyalty and the Elothar Warrior of Bladereach gains a +3 bonus to his Diplomacy checks when dealing with Halflings if Tommy is present.
Legacy of the Water Stone (Sp): An Elothar Warrior of Bladereach of 2nd level has touched the fabled Water Stone, and gleaned a portion of its powers thereby. He may cast create water as a spell-like ability at will. The caster level for this ability is 5.
Magic Swords (Su): Any sword a 3rd level Elothar Warrior of Bladereach holds has an enhancement bonus equal to 1/3 of his character level (round down, no maximum). The enhancement bonus fades one round after the Elothar Warrior of Bladereach stops touching the weapons.
Immunity to Petrification (Ex): At 3rd level, an Elothar Warrior of Bladereach cannot be petrified.
I've Got That! (Sp): At 4th level, an Elothar Warrior of Bladereach can mimic the effects of a drawmij's instant summons at will. The Elothar Warrior of Bladereach does not need an arcane mark on the item, nor does he need a sapphire to call the item in question.
Double Riposte (Ex): If an opponent provokes an attack of opportunity from a 5th level Elothar Warrior of Bladereach, the Elothar Warrior of Bladereach may attack with a weapon in each hand at no penalty. This is considered a single attack of opportunity for purposes of how many attacks of opportunity the Elothar Warrior of Bladereach is allowed in a turn.
Fistful of Rubies: At 5th level, an Elothar Warrior of Bladereach finds 10,000 gp worth of rubies.
Der'renya the Ruby Sorceress: At 6th level, an Elothar Warrior of Bladereach is joined in his travels by Der'renya the Ruby Sorceress, a beautiful Drow magician. She is a Wizard 6/ Seeker of the Lost Wizard Traditions 4, and gains levels when he does. Other dark elves will be angered by Der'renya's betrayal, and will be if anything even less friendly with the Elothar Warrior of Bladereach if encountered with her.
Ways and Paths (Su): At 7th level, an Elothar Warrior of Bladereach can make his way back to any plane he's ever been to. By wandering around in the wilderness for three days, he can make a Survival check (DC 25) to shift himself and anyone traveling with him to another plane.
Name of the First Eagle (Sp): At 8th level, an Elothar Warrior of Bladereach can speak the name of the first eagle, which summons a powerful giant eagle that has the attributes of a Roc (though it is only large sized). The eagle appears for one hour, and may be summoned once per day.
Blessing of the Gnome King (Su): At 9th level, an Elothar Warrior of Bladereach has pleased the king of the Gnomes so thoroughly that he is granted a portion of the gnomish power. The Elothar Warrior of Bladereach can speak with burrowing animals and sees through illusions as if he had true seeing cast upon him by a 20th level Sorcerer.
Flying Ship: At 10th level, an Elothar Warrior of Bladereach finds a Flying Ship from the Eberron setting. And can pilot it around.
Your Money is no Good Here: An Elothar Warrior of Bladereach of 10th level gets free drinks and food at The Wandering Eye, a tavern in Sigil.
Demesne of Tralathon: At 11th level, an Elothar Warrior of Bladereach gains sole control of Tralathon, a small demiplane that appears to be an abandoned Githyanki outpost. Tralathon has several one-way portals that exit onto places on the Astral Plane, the Prime Material, and Limbo. The Elothar Warrior of Bladereach may planeshift to Tralathon at will as a spell-like ability.
Mark of Ruin (Su): At 12th level, an Elothar Warrior of Bladereach is permanently marked with the Mark of Ruin, which causes all of his melee attacks to ignore hardness and damage reduction.
Sword of Kas: An Elothar Warrior of Bladereach finds the Sword of Kas at level 13.
Dwarf Friend (Ex): The deeds of an Elothar Warrior of Bladereach are well remembered by Dwarves when he reaches level 13. Dwarves he encounters are treated as Friendly.
Happily Ever After: At 14th level, an Elothar Warrior of Bladereach becomes king of Bladereach with Der'renya as his queen. The castle of Halan Shador, that used to belong to the Lichking Hadrach is his to rule from.
Khadrimarh: A 14th level, an Elothar Warrior of Bladereach has a young adult white dragon named Khadrimarh as a pet.
Elothar Warriors of Bladereach in your campaign: You may want to adapt this prestige class to the specifics of your campaign. In other campaign worlds, the race, region, and name requirements of this class may need to be changed to fit with the overall narrative.
MAD isn't a bad thing.
Now, before you say "What? zomgwtf? MAD makes you weaker!" Well, yes, it does. However, SAD is actually a problem, while MAD is probably a good thing. Allowing a person to excel in different ways by excelling in different stats increases his customizability and versatility in build, which is a good thing. A world where everyone invests in and only cares about the same stat is *not* such a good thing. This goes back to the "make the class versatile in build" point.
Allowing for decent multiclassing is a good thing, too.
You know what's annoying about a lot of full spellcasters? They kinda just feel cheated when you multiclass 'em most of the time. This isn't fun... it impedes on the "make it versatile in build" principle. By contrast, we see a more elegant multiclassing mechanic in Tome of Battle classes, where your maneuvering abilities don't just become completely obsolete if you decide to take a few levels at level 7 or 8 (As opposed to getting magic missile at level 9 with a caster level of 1).
Thankfully, they've got a few PrCs and a feat or two (like Practiced Spellcaster) to mitigate the multiclassing issue of some classes, but it would be better if they didn't have to. However, this is a fairly advanced concern, and people aren't going to mind *terribly* if it doesn't multiclass well, just because multiclassing generally sucks across most of the board in D&D.
This quote from Tempest helps to clarify this point a bit.
Just a minor clarification, then: A class should multiclass well, and reward players who stick with them all the way (note: capstones alone aren't the way to do this, though they do sweeten the deal a bit). The trick is to balance these two competing forces and do it well.
The best example I can think of currently is the warblade. It has simple build elegance in the single class, progressing given abilities, granting new ones, and having no dead levels (every level past 4 has either a maneuver swap or a new maneuver, plus the new Battle X abilities), and a capstone to sweeten level 20. However, it also gains a fair bit from multiclassing, both INTO and OUT OF warblade, due to the interesting way ToB multiclassing works. A non-warblade considering to multiclass has a more compelling reason every level to consider a warblade level. A high-level warblade doesn't lose out too much by taking other classes, especially if all he needs to get more or less "back on track" is simply take a warblade level later on. The pull to multiclass is almost exactly countered by what the class provides.
(Add to this that warblades are diverse in their builds, diverse in their play, capable in their niche, and FUN, *while* being relatively simple, and you get a spectacular class in general, but I wanted to highlight the multi/singleclass balance present.)
Choose a paradigm for balance. This is to say, don't just shoot in the dark, then post on the boards saying "is this balanced?" Decide, from the beginning, what you consider to be your standard for balance. Many people will recommend the Rogue or Psychic Warrior as a "middle-of-the-road" point for balance. Frank uses the single-classed transmuter Wizard. It's ultimately up to you. But the point is... know what power level you're shooting for and go for that.
Keep your conceptual goals in mind. You want to know where you're headed. You find a new niche to fill, or you think of a mechanically original way to handle something. From there, keep that goal in mind, and work towards realizing the concept of the class, and moreover keep in mind *how it will work in play.* Synergy matters. If you're making a paladin, don't just throw in a bunch of holy warrior-y abilities for 20 levels... think of the cohesive whole and how the whole thing works together to create an even class progression that fills a useful and fun role.
Capstones are cool abilities, but don't really change the class's playstyle. Basically, a capstone should be something and cool and shiny that says "congrats, you just hit 20th level." However, it should *not* be something that significantly alters the style of play, such as, say, a Duskblade's Arcane Channelling, which is a "meat of the class" ability. You want to get those sort of abilities when you can use them for more than 1 level. Instead, a capstone is something like "you turn into an outsider type," which is cool and all, but doesn't really revolutionize the way the class is played.
Mind the CR system. You want your base class to be balanced, so measure it up against encounters of the appropriate level (including monsters, other characters, encounter traps, and non-combat obstacles and encounters). Look up the CR system, and know what it's supposed to mean. If a character cannot contribute in a way appropriate to his or her level against encounters appropriate for their level, or if they can completely floor all of those encounters, you don't have something balanced on your hands. What you DO NOT want to do is just eyeball it and say "Hey, that looks balanced." Examine it. Scrutinize it. Make comparisons. Playtest it. Get other people to playtest it if possible, so that you can get away from your own biased opinion.
Balance your options. That is to say, each build option should be good in its own way, with no clear "best" or "worst" choice. When you can feel the indecision, that's balance, right there. Pretty straightforward, but worth mentioning. You don't want "Cat's Grace vs. Bite of the Wererat." You want "Invisibility vs. Silence."
Present your class clearly. The last thing is that you want your class write-up to look nice. This doesn't actually have so much to do with class design itself, but it's an important point when designing classes. You want some flavorful stuff to entice the reader to pay attention, like a quote from a character of the class or a picture. You want to have a clear table, clear ability entries, and something that's legible instead of all just kinda blending together in a great blob of text. You might even want to link up your spell list table to stuff in the SRD or something. Whatever. The most important part of this is making the rules clean and concise, to avoid misinterpretations and generally make everything go down smooth. A badly explained ability entry has led to more than one long, heated, pointless argument on these boards. Don't make it happen to your class.
That's all I got off the top of my head. May edit more in as it comes to me. Please have the courtesy to reply maturely, like it was a Regdar's thread. The last thing I want to see is frivolous tangents and opinion bumps corrupting an otherwise valuable resource.
-Signed,
http://i169.photobucket.com/albums/u211/SatinFist/Sephiroth.gif
Many people go about creating new classes all the time, but a lot of them aren't particularly good. Here, I'm going to examine a few of the elements of what I view to be good class design, and hopefully help a few people out (since they seem to keep asking me for help in this regard). I would love to get some feedback here to make this guide more complete and better able to communicate its points.
Ultimately, this guide should help people to create fun and balanced classes, while avoiding common pitfalls.
Is there a reason to create a new class?
This is one that people miss out on a lot. I can't count the number of classes I've seen on these boards that actually served very little purpouse, because oftentimes the exact same concept could be managed by existing classes (and done with similar or sometimes better mechanical elegance).
In order for a base class to be a solid new addition to the roster, it has to do at least one of two things:
It must either
A) Fill a new niche. For example, the Artificer fills the item creation and use niche as a class, where previously that was never the focus of any base class.
or B) Be mechanically original. For example, while the Warblade fills the "Fighter" niche, it does so in a new and mechanically unique and elegant way.
If it doesn't do either of these things, it may as well not exist.
This post from Tempest Stormwind helps to illustrate my point:
I agree with this 100%.
For me, a good base class has to fit two criteria: It must be mechanically original and it must fill a niche no other base class can do. I'm willing to be lax somewhat on one if the other's particularly strong, though. The society mind (http://boards1.wizards.com/showthread.php?p=7311764) and blue mage (http://boards1.wizards.com/showthread.php?t=137596) do both of these quite nicely (summary of the two: The blue mage is the Final Fantasy archetype of learning enemy spells by experiencing them firsthand, and the society mind is a psionic support class that makes the team operate as if they were parts of a single greater being), while the gemini dancer doesn't fit the niche criterion at all and gets a ton of press as if it did. A lot of it looks like mechanics for the sake of mechanics, actually -- one of the major reasons I dislike Sztany's Ultimate series, in fact.
(On an unrelated note, by the way, if it wasn't clear before, the society mind was published in Untapped Potential, meaning several DMs who require print sources may now peruse it at their leisure. Dreamscarred hasn't forgotten about it, either, and it will be used in later books.)
Those two criteria can help you in designing other base classes. Step 1 is ALWAYS to ask yourself if your concept REALLY needs a new base class instead of just some imagination. From there, it's just a matter of mechanics.
For Example...
Untapped Potential's other base class, the Marksman, was designed as follows: We noted that there was no full-base-attack manifesting class, and when we looked at all the published full-base-attack classes, every one of them except the ranger (and even the ranger, to some extent) had a heavy, heavy melee focus. Thus, a ranged full-base-attack class would be mechanically original, and would be a good place to go. From there, we looked at places like archetypal "small-scale psionics" (i.e. Starcraft Ghosts, The Shadow) and archetypal ranged combatants (there's a very powerful ability there drawn from classic Western films, for instance), and came up with interesting ideas mechanically (such as the ability to develop a signature weapon style). Finally, we put it all together and pared it down until it looked reasonably balanced. To illustrate further, here's some WotC examples.
Duskblade: Fills a niche no other single class can (fighter/mage, a popular niche; the psychic warrior comes close but feels completely different since it can't blast), sort of mechanically original (not so much, but what mechanics are there are solid). Conclusion: Good class.
Warlock: Fills a niche no other class can (and that niche can vary somewhat depending on what you choose, but always having magic at your fingertips has a certain appeal), mechanically original (Self-evident!). On the plus side, it's dripping in unique flavor (which was enough to lure my brother, who NEVER plays anything *but* The Half Orc Barbarian, into trying something new). Conclusion: Good class.
Samurai: Not mechanically original (all of its abilities are fighter bonus feats, essentially), fills no new niche (a Lawful fighter could do this). Conclusion: Bad class.
Swashbuckler: Doesn't fill a new niche (warrior rogue, light fighter), not mechanically original at all. Conclusion: Bad class. But insightful strike's good, so it's worth considering as a dip.
Beguiler: Doesn't fill a new niche (sneaky spellcaster; illusionist comes close), but executes it with mechanical elegance and brilliance. Conclusion: Good class. You'll note that this is the reverse of the duskblade (which fills the niche but isn't mechanically original), yet both are good because their better side is as good as it is.
All of these work without introducing a new system either. If I wanted to, I could laud the Expanded Psionics Handbook, Pact Magic section of the Tome of Magic, the Totemist from Magic of Incarnum, and the Tome of Battle classes as "good" (they all fill various niches and do so with mechanical elegance) and the rest of the Tome of Magic and Magic of Incarnum as "poor" for similar reasons (MoI doesn't fill any new niche except for the totemist, Truenaming fails on so many mechanical levels, and shadow magic is weak on both fronts but shows promise). Are you seeing how to design a good class from this by now? I'd hope I'm being helpful.
If we're including variants of existing base classes, then my money currently lies on (in no particular order) Seerow's Fighter* (http://boards1.wizards.com/showthread.php?t=728493), OneWinged4ngel's Paladin (http://boards1.wizards.com/showthread.php?t=761045), my own Marshal (http://boards1.wizards.com/showthread.php?t=752414), BlaineTog's Soulknife (http://boards1.wizards.com/showthread.php?t=310984), and RadicalTaoist's Ranger (http://boards1.wizards.com/showthread.php?t=738077) (even if the latter may need a bit of tweaking; it's getting intensive playtesting in our group now, right alongside Blaine's soulknife). All of these except the ranger are intended to "fix" an existing class, replacing it altogether (Seerow's fighter is the result of many, many long discussions of what the figher's lacking as a class, OW4's paladin is a reimagining from the ground up of what a PALADIN means, Blaine's soulknife fixes many weak mechanical flaws while giving the soulknife a niche of its own isntead of a bastard niche between stealth and tanking, and my marshal makes a dip class into something worth taking as a defining class). The ranger is meant as a variant that can be used in addition to normal rangers, trading mystical abilities for Tome of Battle maneuvers.
Hope that helps.
* I haven't yet read Otto's classes beyond a cursory skimming, but I like what I see there, particularly the hexblade.
^--It also covers my next point a bit.
Mechanics for mechanics' sake is bad. Seriously. People add little +1s and -1s and tiny little details that don't actually affect the character much based on this and that, and bog it down (and actually can limit the versatility of the character concepts it can fill), when you actually don't need any new mechanics. Some good examples of "mechanics for mechanics sake" are actually some classes that have received high acclaim from some, but criticism from board veterans like myself and Tempest Stormwind, such as Szatany's classes and Frasmage's Gemini Dancer. This is largely because the presentation looks good, but the underlying mechanics are actually rather unnecessary.
Basically, mechanics should have some meaning and significance. They should do something you care about, instead of bogging down the system.
Dead levels suck.
Make sure the player gets something new every level. The reason for this is simple and straightforward: No one wants to take a level where there is no benefit, and moreover dead levels are boring and just plain not fun. Even if it's something minor, you should at least put *something* to fill the void. Ideally, you should be giving a fairly even progression of cool class features.
It is worth noting that gaining a level of spells or something is *NOT* a dead level, because a level of spells actually represents an array of new class features. In fact, it often represents a larger array of new class features at the new level than many other classes get. Don't confuse "class features" with "entries in the special column," because it's just not true. It's also worth noting that if you have a lot of abilities that scale by level, it can also be acceptable to have a few seemingly "dead" levels, since you're actually getting something nifty at those levels even if you don't see it right there in the special column. A bad dead level is one where you're basically the same guy as you were at the last level, except with a bit higher numbers.
You should feel more powerful, more versatile, and generally cooler at every level. A player should not be doing just the same old thing, except with slightly higher numbers.
Tempest helps to clarify, here:
I should add that if you have a lot of class abilities that scale by level, you can also get by with this.
The classic "bad" dead level is Fighter 5. You play EXACTLY the same way you did as Fighter 4, except with slightly different numbers.
This is, at its core, what the no-dead-level approach is meant to avoid. You don't need flashy new abilities at each level, but the character should feel more powerful, more versatile, or similar at every level. He should NOT feel "just like I was, but more so".
(Note a quick comparison in general design philosophy with core melee and ToB characters, by the way -- core melee use the same tactics at level 20 that they do at level 1, more or less: charge, full attack. They have some new features (spells, rages, unique feats, etc) to make the delivery of these tactics possible, and to make the numbers bigger... but it's still the same tactics. Meanwhile, a warblade at level 20 plays significantly different from one at level 1, even if fundamentally they still involve move-in-and-attack. It's because they get options -- and they get them at every level, so it feels good to be a warblade every time you level up. Make your class something you'd want to feel proud to be.)
Uneven progressions suck. There are two common sides of the coin here. There's toploading, and there's the "suck now, but own later" mentality. Both of these ideas generally suck. Toploading is bad because it means that most of your class's progression is actually useless (for instance, the Swashbuckler is often considered a 3-level long class. The other 17 levels are wasted). And "pay for it now for power later" and similar such uneven progressions *really* don't actually work that way in play. Sure, many PrCed up gish builds will be pretty lame at low level, and killer at higher level, but the reality is that most campaigns aren't actually played from levels 1-20. They'll be more like "5-12" or "1-14" or "12-18" or whatever. So that "evening out cost and benefit over levels" doesn't really exist. Making a class good at one level and crappy at another is a bad thing. Ideally, a class progression should be as even as possible and a class should contribute to the party in a level-appropriate way at *every* level. No more, no less. It doesn't actually have to be perfect... but it should be a fairly even progression of cool class features.
I think this honorable Crane puts it fairly well here.
"Suck Now Rule Later" and it's reverse are poor points of game balance and design. Even assuming that all games run from levels 1 to 20 (wildly untrue), and that the characters remain the same throughout all 20 levels (often untrue) that still means that only half your players are having fun at any given point in your campaign, and the others are only hanging on in hopes they'll get a moment to shine at some point via DM intervention. Your players should enjoy every fight they meet and consider every challenge interesting, not singlehandedly win the first 3 fights of the day and stand around useless for fight number 4.
Make the class for everyone who's going to use it, not just you. Basically, this follows a principle that when designing something for public consumption, you want to make it adaptable to everyone's needs. A class shouldn't look like "your specific character's build choices." It should be able to embody a variety of concepts. Writing extensive fluff on the history of some order and the exact way a certain character fights and so forth doesn't actually make for a better class in any way. We've all seen these classes that look like one guy's character, instead of a real base class that can be adapted to a variety of concepts.
Keep it flexible in build. Building on the last point, a class should provide many "viable" build options, allowing it to embody a variety of concepts. With a look at the Wizard, we can see that you can make a tricky illusionist, a war wizard that makes buildings explode, a calculating seer, or a thousand other concepts. Where possible, you shouldn't be restricting the sort of concepts you can use with the class.
Give it options in play. Using the same tactic over and over is boring. If you're a trip fighter with that one trick (trip, trip, trip) then your gameplay is going to become more monotonous. By contrast, the Warblade introduces more versatility and options into every battle.
This is notably distinct from versatility in build. Versatility in build refers to the ability of a Fighter to be built in many different ways, but versatility in play refers to have many options of actions available to you during play.
Plot writing abilities SUCK. Just don't do it. This is a no-no. When I say "plot writing" abilities, I mean stuff like the HORRIBLE Thunder Guide class in the Explorer's Handbook where you get abilities like "Serial Hero: At 8th level, famed Korranberg Chronicle reporter Kole Naerrin writes a serialized account of your adventures appearing over the course of thirteen weeks. You earn 1000 gp per point of your charisma bonus for the rights to your story (minimum 1000gp)." Seriously, WTF? "A guy writes a book about you" isn't a class ability. A class ability is supposed to be some ability that your character has, not something that happens in the plot.
This PrC from the Dungeonomicon parodies the plot-writing abilities and "The class is actually just my specific character put into a progression" problems that we see *alarmingly* often, which is just stupid.
Elothar Warrior of Bladereach
"My name is Elothar. Your name is unimportant, for you shall soon be dead."
The city of Bladereach sits at the mouth of the Typhon River that flows from the Bane Mires into Ferrin's Bay. The elves of Celentian's caravan come every year to trade with the largely human inhabitants of Bladereach and sometimes they leave more than the wares of the Black Orchard Hills when they leave. The results of these dalliances find that they never fit in amongst the people of Bladereach, and are taught the hard secrets of battle that the children of Bladereach have to offer. Often, these half-elven warriors turn to adventuring.
Prerequisites:
Skills: 9 ranks in Use Rope
Race: Half-elf.
Region: Must be from Bladereach.
Special: Name must be Elothar.
Hit Die: d8
Class Skills: The Elothar Warrior of Bladereach's class skills (and the key ability for each skill) are Balance (Dex), Climb (Str), Concentration (Con), Craft (Int), Handle Animal (Cha), Heal (Wis), Hide (Dex), Jump (Str), Knowledge (all skills taken individually) (Int), Listen (Wis), Move Silently (Dex), Profession (Wis), Ride (Dex), Search (Int), Spot (Wis), Survival (Wis), Swim (Str), and Use Rope (Dex).
Skills/Level: 4 + Intelligence Bonus
BAB: Good (1/1), Saves: Fort: Poor; Reflex: Poor; Will: Good
Level, Benefit
1 Way of Two Swords
2 Tommy, Legacy of the Water Stone
3 Magic Swords, Immunity to Petrification
4 I've got that!
5 Double Riposte, Fistful of Rubies
6 Der'renya the Ruby Sorceress
7 Ways and Paths
8 Name of the First Eagle
9 Blessing of the Gnome King
10 Flying Ship, Your Money is No Good Here
11 Demesne of Tralathon
12 Mark of Ruin
13 Sword of Kas, Dwarf Friend
14 Happily Ever After, Khadrimarh
All of the following are Class Features of the Elothar Warrior of Bladereach class:
Weapon and Armor Proficiency: An Elothar Warrior of Bladereach gains proficiency with the Nerra Shard Sword, the Kaorti Ribbon Dagger, and the Shuriken.
Way of Two Swords (Ex): With a single standard action, an Elothar Warrior of Bladreach may attack with a one-handed or light weapon in each hand at no penalties to-hit or damage for the weapon in his primary or off-hand.
Tommy: At 2nd level, an Elothar Warrior of Bladereach is joined in his adventures by Tommy, a 5th level Halfling Rogue from Figmountain. Tommy is a loyal cohort and gains levels when the Elothar Warrior of Bladereach does. Other Halflings will be impressed by Tommy's apparent loyalty and the Elothar Warrior of Bladereach gains a +3 bonus to his Diplomacy checks when dealing with Halflings if Tommy is present.
Legacy of the Water Stone (Sp): An Elothar Warrior of Bladereach of 2nd level has touched the fabled Water Stone, and gleaned a portion of its powers thereby. He may cast create water as a spell-like ability at will. The caster level for this ability is 5.
Magic Swords (Su): Any sword a 3rd level Elothar Warrior of Bladereach holds has an enhancement bonus equal to 1/3 of his character level (round down, no maximum). The enhancement bonus fades one round after the Elothar Warrior of Bladereach stops touching the weapons.
Immunity to Petrification (Ex): At 3rd level, an Elothar Warrior of Bladereach cannot be petrified.
I've Got That! (Sp): At 4th level, an Elothar Warrior of Bladereach can mimic the effects of a drawmij's instant summons at will. The Elothar Warrior of Bladereach does not need an arcane mark on the item, nor does he need a sapphire to call the item in question.
Double Riposte (Ex): If an opponent provokes an attack of opportunity from a 5th level Elothar Warrior of Bladereach, the Elothar Warrior of Bladereach may attack with a weapon in each hand at no penalty. This is considered a single attack of opportunity for purposes of how many attacks of opportunity the Elothar Warrior of Bladereach is allowed in a turn.
Fistful of Rubies: At 5th level, an Elothar Warrior of Bladereach finds 10,000 gp worth of rubies.
Der'renya the Ruby Sorceress: At 6th level, an Elothar Warrior of Bladereach is joined in his travels by Der'renya the Ruby Sorceress, a beautiful Drow magician. She is a Wizard 6/ Seeker of the Lost Wizard Traditions 4, and gains levels when he does. Other dark elves will be angered by Der'renya's betrayal, and will be if anything even less friendly with the Elothar Warrior of Bladereach if encountered with her.
Ways and Paths (Su): At 7th level, an Elothar Warrior of Bladereach can make his way back to any plane he's ever been to. By wandering around in the wilderness for three days, he can make a Survival check (DC 25) to shift himself and anyone traveling with him to another plane.
Name of the First Eagle (Sp): At 8th level, an Elothar Warrior of Bladereach can speak the name of the first eagle, which summons a powerful giant eagle that has the attributes of a Roc (though it is only large sized). The eagle appears for one hour, and may be summoned once per day.
Blessing of the Gnome King (Su): At 9th level, an Elothar Warrior of Bladereach has pleased the king of the Gnomes so thoroughly that he is granted a portion of the gnomish power. The Elothar Warrior of Bladereach can speak with burrowing animals and sees through illusions as if he had true seeing cast upon him by a 20th level Sorcerer.
Flying Ship: At 10th level, an Elothar Warrior of Bladereach finds a Flying Ship from the Eberron setting. And can pilot it around.
Your Money is no Good Here: An Elothar Warrior of Bladereach of 10th level gets free drinks and food at The Wandering Eye, a tavern in Sigil.
Demesne of Tralathon: At 11th level, an Elothar Warrior of Bladereach gains sole control of Tralathon, a small demiplane that appears to be an abandoned Githyanki outpost. Tralathon has several one-way portals that exit onto places on the Astral Plane, the Prime Material, and Limbo. The Elothar Warrior of Bladereach may planeshift to Tralathon at will as a spell-like ability.
Mark of Ruin (Su): At 12th level, an Elothar Warrior of Bladereach is permanently marked with the Mark of Ruin, which causes all of his melee attacks to ignore hardness and damage reduction.
Sword of Kas: An Elothar Warrior of Bladereach finds the Sword of Kas at level 13.
Dwarf Friend (Ex): The deeds of an Elothar Warrior of Bladereach are well remembered by Dwarves when he reaches level 13. Dwarves he encounters are treated as Friendly.
Happily Ever After: At 14th level, an Elothar Warrior of Bladereach becomes king of Bladereach with Der'renya as his queen. The castle of Halan Shador, that used to belong to the Lichking Hadrach is his to rule from.
Khadrimarh: A 14th level, an Elothar Warrior of Bladereach has a young adult white dragon named Khadrimarh as a pet.
Elothar Warriors of Bladereach in your campaign: You may want to adapt this prestige class to the specifics of your campaign. In other campaign worlds, the race, region, and name requirements of this class may need to be changed to fit with the overall narrative.
MAD isn't a bad thing.
Now, before you say "What? zomgwtf? MAD makes you weaker!" Well, yes, it does. However, SAD is actually a problem, while MAD is probably a good thing. Allowing a person to excel in different ways by excelling in different stats increases his customizability and versatility in build, which is a good thing. A world where everyone invests in and only cares about the same stat is *not* such a good thing. This goes back to the "make the class versatile in build" point.
Allowing for decent multiclassing is a good thing, too.
You know what's annoying about a lot of full spellcasters? They kinda just feel cheated when you multiclass 'em most of the time. This isn't fun... it impedes on the "make it versatile in build" principle. By contrast, we see a more elegant multiclassing mechanic in Tome of Battle classes, where your maneuvering abilities don't just become completely obsolete if you decide to take a few levels at level 7 or 8 (As opposed to getting magic missile at level 9 with a caster level of 1).
Thankfully, they've got a few PrCs and a feat or two (like Practiced Spellcaster) to mitigate the multiclassing issue of some classes, but it would be better if they didn't have to. However, this is a fairly advanced concern, and people aren't going to mind *terribly* if it doesn't multiclass well, just because multiclassing generally sucks across most of the board in D&D.
This quote from Tempest helps to clarify this point a bit.
Just a minor clarification, then: A class should multiclass well, and reward players who stick with them all the way (note: capstones alone aren't the way to do this, though they do sweeten the deal a bit). The trick is to balance these two competing forces and do it well.
The best example I can think of currently is the warblade. It has simple build elegance in the single class, progressing given abilities, granting new ones, and having no dead levels (every level past 4 has either a maneuver swap or a new maneuver, plus the new Battle X abilities), and a capstone to sweeten level 20. However, it also gains a fair bit from multiclassing, both INTO and OUT OF warblade, due to the interesting way ToB multiclassing works. A non-warblade considering to multiclass has a more compelling reason every level to consider a warblade level. A high-level warblade doesn't lose out too much by taking other classes, especially if all he needs to get more or less "back on track" is simply take a warblade level later on. The pull to multiclass is almost exactly countered by what the class provides.
(Add to this that warblades are diverse in their builds, diverse in their play, capable in their niche, and FUN, *while* being relatively simple, and you get a spectacular class in general, but I wanted to highlight the multi/singleclass balance present.)
Choose a paradigm for balance. This is to say, don't just shoot in the dark, then post on the boards saying "is this balanced?" Decide, from the beginning, what you consider to be your standard for balance. Many people will recommend the Rogue or Psychic Warrior as a "middle-of-the-road" point for balance. Frank uses the single-classed transmuter Wizard. It's ultimately up to you. But the point is... know what power level you're shooting for and go for that.
Keep your conceptual goals in mind. You want to know where you're headed. You find a new niche to fill, or you think of a mechanically original way to handle something. From there, keep that goal in mind, and work towards realizing the concept of the class, and moreover keep in mind *how it will work in play.* Synergy matters. If you're making a paladin, don't just throw in a bunch of holy warrior-y abilities for 20 levels... think of the cohesive whole and how the whole thing works together to create an even class progression that fills a useful and fun role.
Capstones are cool abilities, but don't really change the class's playstyle. Basically, a capstone should be something and cool and shiny that says "congrats, you just hit 20th level." However, it should *not* be something that significantly alters the style of play, such as, say, a Duskblade's Arcane Channelling, which is a "meat of the class" ability. You want to get those sort of abilities when you can use them for more than 1 level. Instead, a capstone is something like "you turn into an outsider type," which is cool and all, but doesn't really revolutionize the way the class is played.
Mind the CR system. You want your base class to be balanced, so measure it up against encounters of the appropriate level (including monsters, other characters, encounter traps, and non-combat obstacles and encounters). Look up the CR system, and know what it's supposed to mean. If a character cannot contribute in a way appropriate to his or her level against encounters appropriate for their level, or if they can completely floor all of those encounters, you don't have something balanced on your hands. What you DO NOT want to do is just eyeball it and say "Hey, that looks balanced." Examine it. Scrutinize it. Make comparisons. Playtest it. Get other people to playtest it if possible, so that you can get away from your own biased opinion.
Balance your options. That is to say, each build option should be good in its own way, with no clear "best" or "worst" choice. When you can feel the indecision, that's balance, right there. Pretty straightforward, but worth mentioning. You don't want "Cat's Grace vs. Bite of the Wererat." You want "Invisibility vs. Silence."
Present your class clearly. The last thing is that you want your class write-up to look nice. This doesn't actually have so much to do with class design itself, but it's an important point when designing classes. You want some flavorful stuff to entice the reader to pay attention, like a quote from a character of the class or a picture. You want to have a clear table, clear ability entries, and something that's legible instead of all just kinda blending together in a great blob of text. You might even want to link up your spell list table to stuff in the SRD or something. Whatever. The most important part of this is making the rules clean and concise, to avoid misinterpretations and generally make everything go down smooth. A badly explained ability entry has led to more than one long, heated, pointless argument on these boards. Don't make it happen to your class.
That's all I got off the top of my head. May edit more in as it comes to me. Please have the courtesy to reply maturely, like it was a Regdar's thread. The last thing I want to see is frivolous tangents and opinion bumps corrupting an otherwise valuable resource.
-Signed,
http://i169.photobucket.com/albums/u211/SatinFist/Sephiroth.gif