PDA

View Full Version : [3.5/PF] Class Design, Dead Levels versus Power Creep



Newwby
2014-01-27, 09:37 AM
I've been having three recurring issues for a while with the various classes I've designed for 3.5/PF and I'm looking for some playgrounder input on each subject.

Dead Levels versus Power Creep :smallannoyed:
The first is my aversion to dead levels. In an effort to fill every level I often end up creating features which, whilst minor in effect/power on their own, add up to a more powerful class than I originally envisioned. Sometimes this power is overarching and dangerously/game-breakingly wide in scope - but more often it's so narrow it handicaps the class in other respects (think 'this is the sole purpose of the class, to use it for anything else is suboptimal'). Are dead levels sometimes useful to space out class features and provide optimal 'stepping off' points for prestige classes?. Should I try follow a similar class feature as the 3.5 core rulebook classes with copious dead levels - or as the core PF rulebook classes or WotC dead (https://www.wizards.com/default.asp?x=dnd/cwc/20061013a) level articles (https://www.wizards.com/default.asp?x=dnd/cwc/20070227x) with flavorful but ultimately nigh-powerless features?

Options versus Too Many Roles :smallcool:
The second problem is that I always try to design classes with a multitude of paths/options for progression and with at least a slight capability for social capabilities or skillmonkey functions too. Sometimes this works out and I end up with a class that *could* fill any role needed (but not all at once) but it often ends up with a one-man party of a class. Should I endeavour to keep classes more traditionally focused (e.g. 'this is a melee class it hits things' and 'this is a skillmonkey it handles social/exploration roles) or is it preferable to allow for a multitude of playstyles at the risk of creating a one-man band?

Saving Throws :smallfrown:
The third and least problem is to do with saving throws. I usually try to follow the standard chassis options presented by classes (High BAB, High Fort, 2+Int sp as a warrior class for example) but often I find myself allowing a class two high saving throws because it feels more appropriate to the theme than a single one. Is preserving the 'feel' of the class or keeping the saving throws in line with a traditional class more important? How common exactly is it for a class to have two, or even three, high saving throws? Do prestige classes tend toward multiple high saving throws more often than base classes or am I imagining that?

I welcome any and all input, including any tangents discussing key aspects, or linking to helpful resources, of homebrew class design that I should consider.

Adam1949
2014-01-27, 10:55 AM
The first is my aversion to dead levels. In an effort to fill every level I often end up creating features which, whilst minor in effect/power on their own, add up to a more powerful class than I originally envisioned. Sometimes this power is overarching and dangerously/game-breakingly wide in scope - but more often it's so narrow it handicaps the class in other respects (think 'this is the sole purpose of the class, to use it for anything else is suboptimal'). Are dead levels sometimes useful to space out class features and provide optimal 'stepping off' points for prestige classes?.
I'd like to start by saying that I am exactly like that; seeing that a level is empty makes me unhappy, even if it frankly needs that slot empty. In general, the higher in power you go, the less class features a class needs filling up the levels; classes with a full 9 level spell-list are a good example of this. That being said, in regards to Prestige jump-out points, I prescribe instead to the theory that there should be good abilities spread fairly evenly throughout the class which make good benchmarks to jump from (Dread Necromancer has levels 8 and 12, for example, while Fighter is only good for 1, 2, or 4 levels depending on how many feats/BAB you need).



The second problem is that I always try to design classes with a multitude of paths/options for progression and with at least a slight capability for social capabilities or skillmonkey functions too... Should I endeavour to keep classes more traditionally focused (e.g. 'this is a melee class it hits things' and 'this is a skillmonkey it handles social/exploration roles) or is it preferable to allow for a multitude of playstyles at the risk of creating a one-man band?
This one's actually a lot easier; the reason that classes like Cleric and Wizard are so powerful is because they can truly fill any role that they want. A good class that can do 'anything' is one that still tries to focus on other things first and foremost; Factotum is the Ur-Example here, but Binder is another one that fills that role. In both cases, they both still have a focus, but that focus isn't strictly required to make a decent character. Generally, though, it's best to make a class with the assumption that they'll fill a role first, and fill in for a role second.


The third and least problem is to do with saving throws. I usually try to follow the standard chassis options presented by classes (High BAB, High Fort, 2+Int sp as a warrior class for example) but often I find myself allowing a class two high saving throws because it feels more appropriate to the theme than a single one. Is preserving the 'feel' of the class or keeping the saving throws in line with a traditional class more important? How common exactly is it for a class to have two, or even three, high saving throws? Do prestige classes tend toward multiple high saving throws more often than base classes or am I imagining that?
In 3..5 the only class I can think of that has 3 good saving throws is the Monk. After that, it's frankly up to you to decide what a class should have in saving throws; frontliners generally have good Fortitude and/or Will (depending on their theme), characters meant to be moving around a lot generally have good Reflex, spellcasters generally have good Will. A class is OK with having 1 good save, but that also means they have 2 bad saves, making its weaknesses obvious; filling those in with, say Evasion or Mettle helps limit that.
In regards to skill points, there's a good rule of thumb I've found; unless the class is INT-focused or INT-dependent, they should never have a skill list of 2+INT Mod; it's too limiting, no matter what their skill list is.

Tanuki Tales
2014-01-27, 11:18 AM
First off, let me say that Dead Levels are always a bad thing. A class should always be getting something new and interesting at every single level so that the player never feels like they're being an idiot for not just multiclassing out. Let me also say that "flavorful, but ultimately useless" class features are even worse than dead levels. One of the hallmarks of why the Monk class is terrible is because it has a menu of such abilities and none of them really have synergy with each other.

Now, I find myself in a similar place as you most the time, but it's usually not because of power creep, but because of feature glut. I rarely ever make something higher than Tier 3, but I always end up with way too many class features. So, a trick I've learned for getting rid of feature glut and not having dead levels is to build the class so it's very modular or it's abilities are front loaded but scale all the way to level 20. Thus, the class is getting something new at practically every level, but you don't need to invent something new to fill those levels.


Second off, Tier 3 is considered the "sweet spot" for game design:


Capable of doing one thing quite well, while still being useful when that one thing is inappropriate, or capable of doing all things, but not as well as classes that specialize in that area. Occasionally has a mechanical ability that can solve an encounter, but this is relatively rare and easy to deal with. Challenging such a character takes some thought from the DM, but isn't too difficult. Will outshine any Tier 5s in the party much of the time.

So your class being able to fill multiple roles but only filling one role the best is what you want to aim for. You want to be making the Bard, the Binder or the Factotum, not the Fighter, Rogue or Monk.

Third off, it's a pet peeve of mine, but I highly suggest against giving non-Int focused SAD classes 2+Int mod skills. That's just not enough skills to really make a functioning character. I always make the floor for skills being 4+Int mod.

Fourth off, I wouldn't worry too much about saves. Only the Monk and Paladin get all good saves, true, but a lot of them do get two good saves. Even having all good saves doesn't really add that much to a class' power unless they already have a ton of other stuff going for them.

Fifth off, I've noticed that you mention wanting to design around what's "traditional" or is "standard" by the game design for 3.5 and Pathfinder. Well...don't. Those game designers barely knew a single thing about the games they've written and are not idols to mold your work in the shadow of. These homebrew forums could create a game far superior to any of the muck they forced down our throats for exuberant sums of money and that's because we care about what's actually fun and more reasonably playable (and sometimes balanced) than whatever moth's fart they call their "design process".

Kensen
2014-01-27, 11:38 AM
Dead levels are just very disappointing, especially for people who don't want to multiclass. So in my opinion, it's better to add minor abilities that are thematically appropriate, just as long as they don't unnecessarily add to the number of things you have to keep track of.

The PF fighter's bravery is both a good and bad example. It's a small bonus that affects a relatively small subset of spells, so it's not unbalancing. It's thematically appropriate - warrior types are not supposed to easily scared! However, I personally dislike bonuses that only work on Wednesdays or if you're wearing green. :D I never remember to add the bonus and I also have to constantly remind the GM that if it's a fear effect, I get +x.

As for party roles, you can give the class a class feature that lets you choose a few extra class skills. The sorcerer does this with bloodlines, for example. I love it how they can pick up that one skill you really want them to have. This way it is possible to be good at something the class isn't specifically intended for.

As pointed out, you generally should pick either one or two good saves... (I mean if you want it to be similar to official classes) unless the class is somehow obsessed with perfecting oneself (such as the monk, or the 3.5 favored soul). Note that a number of PF classes have only one good save, but they get situational bonuses such as the aforementioned bravery, or the ability of the swashbuckler (playtest) advanced class that lets you add your Cha modifier on any saving throw x times per day. Maybe this way you can avoid the temptation of giving the class a second good save if you're not sure it deserves it.

I hope that helps! :)

Newwby
2014-01-28, 03:02 AM
-snip- In general, the higher in power you go, the less class features a class needs filling up the levels; classes with a full 9 level spell-list are a good example of this.
I like this, considering the overall power level of the class before filling in with other features and ending up with what Tanuki called 'power glut'.


That being said, in regards to Prestige jump-out points, I prescribe instead to the theory that there should be good abilities spread fairly evenly throughout the class which make good benchmarks to jump from (Dread Necromancer has levels 8 and 12, for example, while Fighter is only good for 1, 2, or 4 levels depending on how many feats/BAB you need).

I read as many homebrew design sources as I can get my hand on and someone else (from pensandpencils) suggested having a class feature that improves over time improve in predictable level skips (e.g. every third level, every fourth level) - if the improvements are more than purely numerical that accomplishes that goal hopefully? Would you consider brand new class features more valuable for the purposes of putting in x levels before prestige-ing out?


This one's actually a lot easier; the reason that classes like Cleric and Wizard are so powerful is because they can truly fill any role that they want. A good class that can do 'anything' is one that still tries to focus on other things first and foremost; Factotum is the Ur-Example here, but Binder is another one that fills that role. In both cases, they both still have a focus, but that focus isn't strictly required to make a decent character. Generally, though, it's best to make a class with the assumption that they'll fill a role first, and fill in for a role second.
That's good advice, I'll try to stray away from 'can do everything' and towards to 'could do anything' - perhaps customisation for class features is a good solution for my problem here? E.g. you gain a new class feature this level, pick option A and you gain search flavored abilities, pick option B and you gain social flavored abilities. Compared to that, how do fluid class features stack up? E.g. a class that would gain the same choice but is capable of changing it every day (like a wizard choosing spells). I realise that obviously it's more powerful but *how* much more powerful?


Let me also say that "flavorful, but ultimately useless" class features are even worse than dead levels. One of the hallmarks of why the Monk class is terrible is because it has a menu of such abilities and none of them really have synergy with each other.

Now, I find myself in a similar place as you most the time, but it's usually not because of power creep, but because of feature glut. I rarely ever make something higher than Tier 3, but I always end up with way too many class features. So, a trick I've learned for getting rid of feature glut and not having dead levels is to build the class so it's very modular or it's abilities are front loaded but scale all the way to level 20. Thus, the class is getting something new at practically every level, but you don't need to invent something new to fill those levels.
What if the class features are flavorful and well within the theme of the class, but also useful to a small degree? E.g. I have a 'conjurer general' class which is basically a conjuration-focused sorcerer that sacrifices other casting for class features that improve their minions - one of the early 'flavorful' abilities I added was a simple +2 bonus for their minions against any fear and mind-affecting saves - is a dull +number bonus a good idea to fill in dead levels?

I like the idea of front-loading class features and then having them scale. I realise now that's probably the thing I like most about the 'conjurer general' class I designed at the moment. I recently did a test where I took away all the abilities that were simply improvements of an earlier ability from the class table, so I could see where unique class features cropped up, and it seemed that all the class features bar one appeared in the first eight levels, then scaled.


Second off, Tier 3 is considered the "sweet spot" for game design:
I've read as much as I could find on the tier system and I find I can understand 'why' a class might fall in to a tier when explained to me, but I struggle to place a class I'm designing within a tier intentionally - perhaps due to the aforementioned power creep. My intention is, on publishing whatever homebrew design finishes first, on the GitP forums that I'll ask fellow playgrounders to evaluate the tier of the class and I'll adjust from there.


In regards to skill points, there's a good rule of thumb I've found; unless the class is INT-focused or INT-dependent, they should never have a skill list of 2+INT Mod; it's too limiting, no matter what their skill list is.

Third off, it's a pet peeve of mine, but I highly suggest against giving non-Int focused SAD classes 2+Int mod skills. That's just not enough skills to really make a functioning character. I always make the floor for skills being 4+Int mod.
I'm seeing a lot of '2+INT, oh god no!' so hopefully it will relieve you all to know that although some of my frontliner class designs started with that few in skill points, only one still retains that few to this day and it is, as you all suggested, an INT-based class.


than whatever moth's fart they call their "design process".
:smallbiggrin: I will devote more time to the homebrew of these forums (and brilliant gameologists) than DnD Tools/Pathfinder then - homebrew always seems far more interesting.


As for party roles, you can give the class a class feature that lets you choose a few extra class skills. The sorcerer does this with bloodlines, for example. I love it how they can pick up that one skill you really want them to have. This way it is possible to be good at something the class isn't specifically intended for.
My crowning glory (and closest to publishing on the forums once I finish another rewrite) is a class (the 'Antecedent') that derives it's skill list and known spells from class features it chooses every other level. Maybe it's coming to tabletop game design after video game design but I love customisation and the idea that two people playing the same class could play entirely different characters.


As pointed out, you generally should pick either one or two good saves... (I mean if you want it to be similar to official classes) unless the class is somehow obsessed with perfecting oneself (such as the monk, or the 3.5 favored soul). Note that a number of PF classes have only one good save, but they get situational bonuses such as the aforementioned bravery, or the ability of the swashbuckler (playtest) advanced class that lets you add your Cha modifier on any saving throw x times per day. Maybe this way you can avoid the temptation of giving the class a second good save if you're not sure it deserves it.

I hope that helps! :)

That's a really good idea for resolving the issue I'm having with the saving throws actually! Two always seems unfair in comparison to other classes but 1.5x (effectively) works in nicely. I've, on the aforementioned 'Antecedent', given it one high saving throw (Will) then allowed the player to choose either Fortitude or Reflex as an additional high saving throw (improving the progression) at 2nd level. The addition of 'evasion', 'grit' or 'mettle' gives a high saving throw a great deal more worth doesn't it? Should I be careful about adding those features to a class or does every class deserve one?

Thanks to you all for addressing my questions in such depth - the fact that you all mostly agreed on what I should do gives me a great grounding in what approach to take to the next round of rewrites. I'm cursed with an abysmal attention span so I tend to work on a plethora of homebrew designs at once rather than focusing on one until it's finish - I've got seven class designs on the go at the moment. I operate on a progressing scale of rewrites of 'notes' then 'draft writeup' (both in notepad), then a full word-processor writeup which I leave for a time to gather dust (creating more notepad notes as I think of things) before the final rewrite. Most of my classes are currently in the word-processor writeup stage but I'm midway through the final rewrite with one. :smallsmile:

Now apologies :smalltongue: but I've come up with two additional questions -
Chassis versus Class Features
Firstly should I scale the class's chassis on it's power level? Using my previously mentioned 'conjurer general' homebrew class, I originally give it medium bab, a decent hd, skills and two good saving throws because it didn't have anything that let it operate alone - everything it achieved or could achieve was done through proxy of minions. Later I ended up adding a feature which allowed it to temporarily take on aspects of it's minions to accomplish things they could do - a feature which drastically increased it's solo survivability. Should I reduce it's chassis to compensate for the increased power, scale back other class features or just go back to the older design? At the moment I'm thinking of converting it to a prestige class with strict requirements or labeling it a 'DM's should allow it for the big bad only' class since it would quite nicely fill the role of a world-conquering caster.

Pools versus Spell Slots
Secondly, if I utilise a 'pool' (e.g. you have x amounts of points in this pool to power your class features, they return when you rest), is it messy to also use spell slots? Should I pick one and drop the other? Are Vancian spell slots easier to play with or is a pool easier to work with?

[Apologies for the essay-length reply, it really got away from me]

HammeredWharf
2014-01-28, 03:36 AM
Chassis versus Class Features

The chassis is overrated too often in D&D design IMO. Sure, having a poor BAB is worse than having a good BAB, but simply having a good chassis means very little. For example, Monk's chassis is better than Rogue's combat wise, but it's still a much weaker class. Unless your chassis actively makes some class features redundant (like giving a poor BAB to Barbarian), the class features are much more important.

As for dead levels, they're ok if it's a caster. Their spells are their class features. I'd still give every caster a scaling bonus, so they lose something for PRCing out. If it's a mundane class, it'd better get something.



Pools versus Spell Slots

That's no different from normal "n times per day" class features, unless your class features are somehow hard to manage otherwise.



Options versus Too Many Roles

I think every class should be able to participate in social encounters. Otherwise, you end up with bored players trying not to fall asleep as their bard friend solos the two hours long social session. However, you can have roles in social encounters, too. For example, a character with lots of Knowledge skills is useful socially, but doesn't steal the bard's thunder.

T.G. Oskar
2014-01-28, 05:31 AM
I've been having three recurring issues for a while with the various classes I've designed for 3.5/PF and I'm looking for some playgrounder input on each subject.

Dead Levels versus Power Creep :smallannoyed:
The first is my aversion to dead levels. In an effort to fill every level I often end up creating features which, whilst minor in effect/power on their own, add up to a more powerful class than I originally envisioned. Sometimes this power is overarching and dangerously/game-breakingly wide in scope - but more often it's so narrow it handicaps the class in other respects (think 'this is the sole purpose of the class, to use it for anything else is suboptimal'). Are dead levels sometimes useful to space out class features and provide optimal 'stepping off' points for prestige classes?. Should I try follow a similar class feature as the 3.5 core rulebook classes with copious dead levels - or as the core PF rulebook classes or WotC dead (https://www.wizards.com/default.asp?x=dnd/cwc/20061013a) level articles (https://www.wizards.com/default.asp?x=dnd/cwc/20070227x) with flavorful but ultimately nigh-powerless features?

Most people have said this probably better than I do, but here's my point about it: dead levels are bad, and fluffy abilities that don't impact the class are equally bad. However, I believe it's important to mention a concept I call "level weight".

Essentially, each level has a certain "load". This "load" is mostly subjective, and not easy to notice, but if you know the terms "front-loaded" and "too little, too late", you know what I mean; these are the two extremes of the load a specific class level can have. For example: the Paladin's first four levels have a wealth of class features, but the remaining 16 are really poor. The load of the class is placed squarely on those first four levels. Of those, the first has a moderately useful skill (Detect Evil), an increasing ability (Smite Evil) and a class feature that exists to justify something (Aura of Good). All three make the first level of the class, as strange as it may seem, be well-loaded: Detect Evil is a very useful ability, but its effect is limited and doesn't really scale up. Smite Evil scales in two ways: damage is based by level, and you get extra uses every five class levels. Finally, Aura of Good only exists for purposes of determining the effect of Detect Good, so it's not something that boosts the character that much. Level 2, on the other hand, has both Lay on Hands (a nice boost healing move, even if usable only once) and Divine Grace, which is the ability just about everyone who wants to use Charisma properly seeks. Divine Grace on its own would have made the level reasonable enough, but when you stack Lay on Hands, it turns out to be a very heavy class; in fact, Divine Grace at such early level is so "heavy", it immediately becomes a dropping point for any multiclass build. It's so heavy, the only reason you'd like to stay for more than 2 levels is because you want Turn Undead without becoming a Cleric (or getting into Sacred Exorcist) and the good BAB progression. Level 3 has two immunities at once, which is also pretty heavy: you could move Lay on Hands to 3rd level and keep Divine Health and it'd be balanced, but you get two immunities there which is pretty "heavy". As you gain levels, you see that an extra use of Smite Evil AND an increase in damage won't have the same load as before, and can exist in conjunction with a "heavier" ability without making the level too good. Likewise, you could move Divine Grace to level 6 by itself and it'd still be a great class feature; level 11th would be pushing too far, but it's still reasonable enough.

What this means, regarding your question, is as follows: each level has its "load", and this "load" grows with each passing level (or rather, level range) because of Prestige Classes and the synergy provided by dipping in other classes. When working each level, you need to have the right balance between class features: the first few levels should have the starting versions of scaling abilities (Rage is an excellent example of a good scaling class feature and the right way to make a per-day class feature for mundanes) or meaningful class features left alone, while latter levels can have two or three abilities at once without affecting their worth. Again, this is something you get with practice and time; it's mostly instinctive rather than learned.


Options versus Too Many Roles :smallcool:
The second problem is that I always try to design classes with a multitude of paths/options for progression and with at least a slight capability for social capabilities or skillmonkey functions too. Sometimes this works out and I end up with a class that *could* fill any role needed (but not all at once) but it often ends up with a one-man party of a class. Should I endeavour to keep classes more traditionally focused (e.g. 'this is a melee class it hits things' and 'this is a skillmonkey it handles social/exploration roles) or is it preferable to allow for a multitude of playstyles at the risk of creating a one-man band?

This is a question of role and class balance. Ideally, you want to have a class that's at least good in one thing, and moderately good in others; this is the hallmark of Tier 3, or the "Rogue/High" balance point. Again, this is more instinctive than anything else, but each class should be capable of doing three things: be very good at one thing, decent in at least two or three, and capable of doing a few more when pooling resources. Remember that D&D is a cooperative game, so you don't want one-man armies, but you don't want people unable to share the spotlight just because they can't step up to the challenge. When you're good (or great, or even excellent) at one thing, that's your defined role ("this is a melee class, it hits things"/"this is the face, it handles social roles"); the two or three things you're decent at allow you to support the specialist or be the switch hitter ("the Paladin has decent Charisma, so he can aid the Bard on his Diplomacy check"/"the Ranger has Search and Spot to detect traps when the Rogue is out of commission, even though the Ranger can't disarm them"), and the rest requires pooling resources to work out, so that helps you support the switch hitter on hard challenges ("the Rogue can use a Wand of Bull's Strength to help the Cleric using Righteous Might to lift the Boulder, because the Barbarian is exhausted and the Cleric has no Restoration spells prepared"). The first is defined by the exact role your class excels at, and defines its class features; the second is defined by the supplementary class features of the class and the third is provided by any odd synergy or unusual option you can provide it. After a while, it'll feel natural to build Tier 3/High balance classes.


Saving Throws :smallfrown:
The third and least problem is to do with saving throws. I usually try to follow the standard chassis options presented by classes (High BAB, High Fort, 2+Int sp as a warrior class for example) but often I find myself allowing a class two high saving throws because it feels more appropriate to the theme than a single one. Is preserving the 'feel' of the class or keeping the saving throws in line with a traditional class more important? How common exactly is it for a class to have two, or even three, high saving throws? Do prestige classes tend toward multiple high saving throws more often than base classes or am I imagining that?

In this case, go with gut feeling. The "feel" of the class can be harmful sometimes. A Fighter, for example, might not need high Reflexes or Will to survive, but it definitely needs high Fortitude. A Paladin, on the other hand, might need higher Will if you intend to reduce its MAD because its Wisdom will most likely be low, so you may seek to grant it a better Will save. Likewise, if you were to go with feel other than with utility, the Bard wouldn't have a good Reflex save, if only because it makes no sense when dealing with a class that's meant for support and being the face/scholar of the group.

Going with base classes, it's pretty common to see one if the class has loads of other benefits: the Paladin has one good save, but it has d10, full BAB, heavy armor and shield proficiency and Divine Grace. On the other hand, the Ranger has full BAB, a d8, Evasion and prefers light armor, so it gets Reflex. The Cleric will naturally have high Will but it gets a d8 Hit Die, medium BAB and heavy armor/shield proficiency, so it makes for a decent front-liner and hence gets good Fort; thus, it's not as common as you may think. You can say that about half the classes have one good save and the other half has two; only Monk and Favored Soul has three good saves, and the most common choices are between Fort and Will for single good saves, and Fortitude/Will for two good saves.

But, again: gut feeling.


Chassis versus Class Features
Firstly should I scale the class's chassis on it's power level? Using my previously mentioned 'conjurer general' homebrew class, I originally give it medium bab, a decent hd, skills and two good saving throws because it didn't have anything that let it operate alone - everything it achieved or could achieve was done through proxy of minions. Later I ended up adding a feature which allowed it to temporarily take on aspects of it's minions to accomplish things they could do - a feature which drastically increased it's solo survivability. Should I reduce it's chassis to compensate for the increased power, scale back other class features or just go back to the older design? At the moment I'm thinking of converting it to a prestige class with strict requirements or labeling it a 'DM's should allow it for the big bad only' class since it would quite nicely fill the role of a world-conquering caster.

The class' chassis is important because it's shorthand for its role. If you see d10 HD, full BAB, good Fort and heavy armor/shield proficiency, you KNOW it's a front-liner. On the other hand, if you see poor BAB, good Will saves, d4 or d6 Hit Dice...before I say "and simple weapon proficiency" you KNOW I speak about a caster. Likewise, medium BAB with d6/d8 HD, good Reflex saves and proficiency with simple weapons + light armor will probably end up with 6+Int skill points or even 8+Int, because you know that's the skillmonkey.

One thing I always like to do when working with chassis is to default to 4+Int skill points if the class won't have much use for Intelligence AND isn't meant to be a skillmonkey. If the class is meant to use Intelligence, then I can drop it two skill points. A skillmonkey that uses Intelligence is the only one allowed to have 8+Int skill points, and I might even consider 10, because of its role. Afterwards, I let the role define its chassis. You should rarely, if ever, bind the chassis to its power level, because half of the class' role is based on its chassis; you define the class' power level by its class features, even if weapon and armor proficiencies are somewhat part of the chassis (however, they are easier to alter than, say, BAB or saves). You can tweak one or two things (maybe grant two saves to a full BAB class, most likely Fortitude and Will because the class is meant to be a tank), but they have to make sense (the same chassis with only Reflex and Will as good saves is pretty odd, because you'll expect a frontliner to have good Fort, period). Tweaking the class' chassis is an instinctive process, but this is the easiest one to attune with; watching how other classes show their intended role through their chassis gives you a good way to handle working that on your own 'brews.


Pools versus Spell Slots
Secondly, if I utilise a 'pool' (e.g. you have x amounts of points in this pool to power your class features, they return when you rest), is it messy to also use spell slots? Should I pick one and drop the other? Are Vancian spell slots easier to play with or is a pool easier to work with?

It depends on how you use the slots, honestly.

I'll use an example of a 'brew I haven't released yet: the Shugenja. All retooled classes with an "Oriental" feel WILL have the "Ki" pool stacked upon them. The Shugenja is a full spellcaster, which combines spontaneous specialization (they know all spells of their chosen element, but none of their opposed element) with Sorcerer-based casting (they get no Advanced Learning, but they get to choose spells from the other elements for their personal use, so they get a lot of leeway on that regard). The way they use the "Ki" Pool works as a hybrid of Turn Undead and Inspiration Points: they get a set amount of uses per day, but they can use them on a variety of class features. This involves easier use of metamagic, boosting caster level or spell save DC, or using Order abilities (think Domains, but you use the Order's spells as SLAs).

Just by mentioning this I used both a pool, spell slots, AND SLAs, but I can assure you the Shugenja is by no means a Tier 1 class, because it has the trappings of the Sorcerer and its spell list is far more limited. On the other hand, between the choices of Orders and the flexibility in choice of spells, it allows for a huge amount of builds, which is something I aim for. It has both the balance I seek (even if it is somewhat more powerful than Tier 3/High balance point and closer to Tier 2/borderline Very High balance point) and the flexibility I seek, and it has both a pool AND spell slots.

Just by mentioning the 'brew, I mentioned one class that does this well: the Cleric. The Cleric is Tier 1, but it's not because of Turn Undead; the spells do that. However, with Divine feats, they can use Turn Undead for a variety of purposes, fulfilling what you seek. If you were to restrict the spell list of the Cleric to its choice of domains, you could get a well-balanced class that uses both pools.

P.S.: No problem about essay-length replies. They come to me naturally, so here's an essay-length response :smalltongue:

PairO'Dice Lost
2014-01-28, 03:10 PM
Pools versus Spell Slots
Secondly, if I utilise a 'pool' (e.g. you have x amounts of points in this pool to power your class features, they return when you rest), is it messy to also use spell slots?

"Messiness" and complexity isn't based solely on the number and kind of resource, but also on (A) how complex each individual resource is and (B) how much they overlap.

For instance, the core paladin actually has a lot of resources to manage:
Smites per day
Turn Undead per day
Lay on Hands pool
Spells per day
Special Mount
SLAs per week

And yet very few people have difficulty playing a paladin, first of all because each of those individual resources is straightforward in how they're used (Smite this round, yes or no? Undead to turn, yes or no? Enough lay on hands to heal my ally fully, yes or no?) unless you choose to make them more complex (take some [Divine] feats to add more uses to Turn Undead, take an ACF to add resources in place of Detect Evil or Special Mount, micromanage your LoH HP, etc.).

Second of all, they don't overlap much: smites work with attacks, Turn Undead is a special attack, LoH heals HP, the weekly SLAs heal conditions, the spells are mostly buffs or utility. Not all of them are usable in combat, and on a turn-by-turn basis you're only looking at one or two of those resources (again, unless you deliberately opt for more complexity by adding [Divine] feats, Battle Blessing, etc.).


Contrast this with a druid. The druid only has two resources: spells and wild shapes. However, it has a lot of spells slots by mid level, drawn from the entire spell list, and many of them are situational (like entangle) so you have to pay attention to the scenario to use them well (or at all). Wild shape can turn you into any animal from every monster book with animals in it, later adding the ability to turn into elementals as well. While its companion isn't really a "resource" to manage because it doesn't change on a daily basis, summoned creatures do: you have to pick one (again, from among many choices in many books), control it in combat, and so forth...not to mention that a given druid could have a half-dozen going at once by mid levels.

Just looking at the resource management schemes, the druid looks simple and the paladin incredibly complicated, when in actual play it's the reverse because how those resources are used and how they interact. It's possible to make a complex class with few resources, a complex class with many resources, a simple class with few resources, a simple class with many resources, or a class that splits the difference somehow like T.G.'s shugenja. You really need to look at the whole package and decide what level of complexity you're comfortable with.

Newwby
2014-01-30, 01:39 AM
The chassis is overrated too often in D&D design IMO. Sure, having a poor BAB is worse than having a good BAB, but simply having a good chassis means very little. For example, Monk's chassis is better than Rogue's combat wise, but it's still a much weaker class. Unless your chassis actively makes some class features redundant (like giving a poor BAB to Barbarian), the class features are much more important.
Alright, that's good advice for considering what BAB progression to apply actually. Often I use BAB as the starting point to determine what the rest of the classes chassis should look like.


As for dead levels, they're ok if it's a caster. Their spells are their class features. I'd still give every caster a scaling bonus, so they lose something for PRCing out. If it's a mundane class, it'd better get something.
Since I still consider myself a novice homebrewer I've been trying to stray away from spells for the most part, instead focusing on either limited casters with very restricted spell lists or mundanes with a number of spell-like abilities. It's usually the mundanes/non-spell slot classes that end up with bulkier/more plentiful class features. I'll see about freeing up some levels (especially for classes whose main features require class progression) where I've stuck pointless 'flavour' abilities.


I think every class should be able to participate in social encounters. Otherwise, you end up with bored players trying not to fall asleep as their bard friend solos the two hours long social session. However, you can have roles in social encounters, too. For example, a character with lots of Knowledge skills is useful socially, but doesn't steal the bard's thunder.
Agreed - when attempting to make a class 'well rounded' I often initially design combat features and end up retroactively working social or environmental features in later (I'm getting better at remembering to include them in the first run through nowadays). I don't necessarily include just conversational features or skill sets under this umbrella though - the first 'full' class design I finished ended up with some serious crafting features. I figured acting as the party blacksmith/armoursmith was a role in itself.


Most people have said this probably better than I do, but here's my point about it: dead levels are bad, and fluffy abilities that don't impact the class are equally bad. However, I believe it's important to mention a concept I call "level weight".

-snip-
This is really good advice, something I hadn't actually considered in length at all. The Paladin is a fantastic example of this, 2 or 3 level dips have become common in a lot of cha-focused warrior builds I've made and I'm adverse to that. Not so much the idea of few level dips as a whole, but the idea that one class can be so distilled in to it's first few levels. I've recently cleaned up one class I've been working on and removed the two immunities it gained because I worried I was overloading the mid-levels, I think I might add them back in and push them a bit further up the progression. I can see several points where it might be favourable to prestige out and players should have to weigh pros and cons, not have such a clear cut choice.


When you're good (or great, or even excellent) at one thing, that's your defined role ("this is a melee class, it hits things"/"this is the face, it handles social roles"); the two or three things you're decent at allow you to support the specialist or be the switch hitter ("the Paladin has decent Charisma, so he can aid the Bard on his Diplomacy check"/"the Ranger has Search and Spot to detect traps when the Rogue is out of commission, even though the Ranger can't disarm them"), and the rest requires pooling resources to work out, so that helps you support the switch hitter on hard challenges.
Speaking specifically about classes that act as 'jack of all trades' or as you put it, 'switch hitters', if a class is capable of performing multiple roles it should by definition not outshine specialists in the same field, that's a given. What about a class with a limited capability of fulfilling many roles, and a limitation on what roles they can fill? To put it in perspective; my homebrew class Antecedent gains access to spheres of ability as they progress, one at first level and one at every odd level afterward. Each of these spheres allow them limited access to a small number of utility spells, a single skill and a special sphere-specific unique capability. The Antecedent cannot manifest more than a few spheres at once, and cannot indefinitely manifest any sphere. Obviously each Antecedent will not be capable of every role, as the choice of spheres they have access to determines that, but they will be able to tailor themselves around any number of roles (the spotter, the trap disarmer, the tank, the summoner, the healer, the buffer, the debuffer etc) for a limited time per day. To what degree should a class like that be capable of performing a role?


In this case, go with gut feeling. The "feel" of the class can be harmful sometimes. A Fighter, for example, might not need high Reflexes or Will to survive, but it definitely needs high Fortitude. A Paladin, on the other hand, might need higher Will if you intend to reduce its MAD because its Wisdom will most likely be low, so you may seek to grant it a better Will save. Likewise, if you were to go with feel other than with utility, the Bard wouldn't have a good Reflex save, if only because it makes no sense when dealing with a class that's meant for support and being the face/scholar of the group.
I try to consider attribute dependency when assigning saving throws but it grates with me so much if the class I envisioned as a fluid agile warrior doesn't have saving throws that reflect my image of the class. In cases where two saving throws don't fit (but I want them to) for whatever reason, I think I'm going to approach it the same way that was suggested earlier in the thread, providing bonuses so that it seems as if it's 1.5x high saving throws.


-snip-

Tweaking the class' chassis is an instinctive process, but this is the easiest one to attune with; watching how other classes show their intended role through their chassis gives you a good way to handle working that on your own 'brews.
I like the idea of seeing the chassis as a shorthand for the class, it helps mentally pigeonhole classes in to distinct roles/archetypes for easier associations and comparisons. I'm notoriously absent minded so sometimes, especially when happening significantly apart from each other, my various tinkerings will upset a chassis so much I have to throw it out and reset it. I like for the chassis to be both the first and the last thing I look at every rewrite.



-snip-

Just by mentioning this I used both a pool, spell slots, AND SLAs, but I can assure you the Shugenja is by no means a Tier 1 class, because it has the trappings of the Sorcerer and its spell list is far more limited. On the other hand, between the choices of Orders and the flexibility in choice of spells, it allows for a huge amount of builds, which is something I aim for. It has both the balance I seek (even if it is somewhat more powerful than Tier 3/High balance point and closer to Tier 2/borderline Very High balance point) and the flexibility I seek, and it has both a pool AND spell slots.
-snip-
I think I will make an attempt to use both slots and a pool then, I've still got both written versions for slots-only and pool-only. The class doesn't have a full spell list, so it should be totally manageable, and most class features don't require pool point expenditure - they only rely on keeping your pool of points filled (since I added a 'refresh one point' ability it's akin to a psionic focus I guess).

PS: Your Project Heretica designs are what caused me to get more serious about homebrewing. The image you used (of the winged Paladin) was something I'd come across before I came across Project Heretica and I built a four-path 'Hallowed' Paladin class based on it. Shortly after I finished it I came across your Project Heretica and saw that you'd done what I'd done, and then some (I cannot stress how much worse mine was, even though it was following a similar idea). I've been since trying to up my homebrewing game!


That's no different from normal "n times per day" class features, unless your class features are somehow hard to manage otherwise.


"Messiness" and complexity isn't based solely on the number and kind of resource, but also on (A) how complex each individual resource is and (B) how much they overlap.

-Snip-
This was great advice on resource management, I'm definitely going to take this on board when I go back through the design. Since the class is a partial caster (as mentioned above) at best, the slots and pool used simultaneously shouldn't be too confusing. If they are I can always change it.

T.G. Oskar
2014-01-30, 06:07 AM
This is really good advice, something I hadn't actually considered in length at all. The Paladin is a fantastic example of this, 2 or 3 level dips have become common in a lot of cha-focused warrior builds I've made and I'm adverse to that. Not so much the idea of few level dips as a whole, but the idea that one class can be so distilled in to it's first few levels. I've recently cleaned up one class I've been working on and removed the two immunities it gained because I worried I was overloading the mid-levels, I think I might add them back in and push them a bit further up the progression. I can see several points where it might be favourable to prestige out and players should have to weigh pros and cons, not have such a clear cut choice.

Chalk that up to early-installment weirdness combined with legacy. The classes in the Player's Handbook suffer from the construction of AD&D 1st and 2nd Editions, where you gained most of the class' abilities in the first few levels. For example: the Fighter had everything pretty much since 1st level, aside from its choice of specialization in weapons; the Cleric had all class features at 1st level aside from spells, and reaching 7th level spellcasting (then the highest level of spells a divine spellcaster could use) took so long that only the most dedicated players ever got to use them. When transitioning to 3e, classes had the same effect, so most of them got all of their abilities by 4th or 5th level, where the character was most likely to end its campaign. It's one of the reasons E6 is popular; most games rarely reach beyond 10th level or higher, for one reason or another (I, personally, have reached about 14th level with one character and have DMed up to 13th level, though fortunately my campaign is far from over). Thus, there's that necessity of having useful class features at early levels, and...well, they jumped the gun a bit.

Later on, with the advent of MMOs, class design went from "get everything as early as possible and then advance what's left" to "make each level interesting". You'll notice only the Fighter has some sort of "capstone" (the Cleric and the Wizard get 9th level spells at 17th level and the Rogue gets its last class feature at 19th; the Fighter at least gets one final bonus feat, though it won't really help that much), and the first class to have a true capstone is the Monk with Perfect Self. Time later, and pretty much all classes and prestige classes HAVE to end with a capstone, or something similar, not to mention how classes have fewer dead levels than before.

The idea behind level weight is that, if all levels are balanced, you'll be motivated to remain in the class. However, sometimes, you need to establish some sort of imbalance that could be fixed by multiclassing or PrC'ing out, generally when you work with a concept that can't be covered by the class itself. Personally, I find that's the problem with PF classes; they have their imbalances, but they focus too much on remaining on one class and make multiclassing harmful. Ideally, a good game will reward system mastery while providing multiple options to achieve the same concept at a similar degree; level weight helps with the latter, but you need a certain degree of imbalance between the classes to allow considering other options.

In short, and to sound less philosophical: level weight is a good class-building measure, but sometimes you have to make wild choices and allow people to consider PrCing out. If a player has difficulty thinking "should I PrC out or should I remain one more class?", then you're going right. Not usually the case, but when considering options it's usually better to stay in the class, it may have the best degree of balance between levels but it'll be pretty poor.


Speaking specifically about classes that act as 'jack of all trades' or as you put it, 'switch hitters', if a class is capable of performing multiple roles it should by definition not outshine specialists in the same field, that's a given. What about a class with a limited capability of fulfilling many roles, and a limitation on what roles they can fill? To put it in perspective; my homebrew class Antecedent gains access to spheres of ability as they progress, one at first level and one at every odd level afterward. Each of these spheres allow them limited access to a small number of utility spells, a single skill and a special sphere-specific unique capability. The Antecedent cannot manifest more than a few spheres at once, and cannot indefinitely manifest any sphere. Obviously each Antecedent will not be capable of every role, as the choice of spheres they have access to determines that, but they will be able to tailor themselves around any number of roles (the spotter, the trap disarmer, the tank, the summoner, the healer, the buffer, the debuffer etc) for a limited time per day. To what degree should a class like that be capable of performing a role?

Sure: a jack of all trades can't be better than a Specialist, but it should have enough skill and knowledge to assume the role successfully if the specialist is out. The idea is as follows: assume that, for a specific role, you aim for 100% success. A specialist should never struggle to reach 100% or easily reach it; the idea is that the specialist should exceed the chances of success but not so much that the task becomes irrelevant. A specialist achieving 120% chance of success by using all resources, 100% success with moderate resources and around 70-80% success without even trying is fair enough. A switch-hitter should aim for 20-25% lower: without even trying, it should succeed about half the time, about three-quarters of a time with moderate resources, and succeed if it focuses on the task. Once that's established, you can see why the specialist MUST be capable of reaching 120% or higher chance of success; anything that makes success more difficult will be easily surpassed by the specialist, but not that much by the jack of all trades. A dabbler shouldn't really succeed 100% with all of its effort (it makes the effort of the specialist irrelevant and the innate traits of the jack of all trades even worse), but if it pulls all resources, it should have a fair amount of success (about 70-80%, what a jack of all trades could pull off if it made a reasonable effort but not fully devote into). The exchange, of course, is that the specialist will be almost incapable of doing what another specialist (whether it dabbles on the first specialist's role or not) or the jack can pull off, and the jack can do one or two roles with roughly equal chances of success.

While the numbers are mostly subjective, it gives you an idea of how to work role balance out. From what I see, the Antecedent is sort of a Savant or Factotum in that it can do every role, but not that well. Since it dabbles on so many roles and essentially shifts its usefulness given time to prepare, the weight each of its components of the build contribute to the overall success of each role incline towards the class features (about a 50% of the overall success). If the Antecedent chooses to specialize in one role over others, it should reach pretty close to 100% with moderate effort (just as a specialist would), so the other 50% should come from feats, skills and magic items, naturally. In terms of skills, since jacks generally are good skillmonkeys, that means a good 60-70% of the skill's success depends greatly on its skill ranks, so your class features shouldn't focus that much on providing a good skill bonus. In terms of combat, for example, where feats are pretty scarce and they contribute roughly to 20-30% of your fighting potential, class features have to cover up a lot more. In terms of, say, healing or buffing, where class features cover almost the entirety of the action, they have to cover even more than with combat. Therefore, you'll notice that one role will be stronger than the other, but when you add up all factors, it'll end up with your class features providing around half of the required strength or chance of success, with your build providing the necessary rest. From there, the actual build takes care of the rest.

The weight of class features in determining the success of each role isn't really set in stone, or should be standardized. A trapbuster won't need as many resources in terms of class features as a healer because the latter only requires three good skills (Disable Device, Search and Spot), a class feature (trapfinding) and an item (Thieves' Tools) to do its work, while the latter needs to have all its utility drawn from class features (the actual spells, class features to boost healing, class features to compliment healing and make it viable in combat, etc.), so you'll notice the trapbuster will be pretty barebones and probably grant Trapfinding and a few bonuses to two skills, while the healer will be more extensive. The trapbuster might not need actual spells, so if you add them, you'll eventually surpass the specialist on its job (the Rogue, or the dedicated trapbuster), but if you compile it to a much larger role (a saboteur, which requires a much larger tool kit than just a pair of tools and skills), you allow the role to dabble only as much as you need, compact two or three roles that might be too specific to pad out, and also justify why such a role has access to spells or abilities that might be outside its influence (for example, a saboteur could have Shatter, Rusting Grasp and even Disintegrate because all three affect objects, but you'll never reach the true extent of someone who devotes its life to sabotage and demolish).

Again, in short: never better than the specialist, nor worse than one who simply dabbles. And, if the role causes you too many problems, sometimes it's best to consider if the "role" is merely a specialization in a much larger, more generalized role that you could tap up rather than simply delve in.


I try to consider attribute dependency when assigning saving throws but it grates with me so much if the class I envisioned as a fluid agile warrior doesn't have saving throws that reflect my image of the class. In cases where two saving throws don't fit (but I want them to) for whatever reason, I think I'm going to approach it the same way that was suggested earlier in the thread, providing bonuses so that it seems as if it's 1.5x high saving throws.

d20 Modern tapped the idea of creating "medium" saving throw progressions, which are pretty odd (they're roughly like poor saving throw progressions, except you start at +1 instead of 0 and you gain 2/3rds of a point at 1st level, 5th level and every 5 levels afterwards, IIRC. if you're not afraid of breaking with tradition, you could use that progression, which allows you to create a class that has one good save, but its other saves aren't necessarily poor. That way, the fluid and agile warrior could have good Reflexes and great Diplomacy, but if you don't want to stack good Fort on top, you could add medium Fort progression and make it only slightly healthier than similar classes. That's an option, though.

The other is going the way of the Swashbuckler, whose Grace ability effectively allowed you to have a Reflex save better than a Fighter, but not as good as a Rogue, directly as a class feature instead of a modification to saving throws. A similar thing I do is give a PrC all bad saves, but give them a 1st level ability that defines its role and thus grants them one good save, which replaces its bad save. This is a more direct approach, but one that works for classes with wildly differing methods of approach.


I like the idea of seeing the chassis as a shorthand for the class, it helps mentally pigeonhole classes in to distinct roles/archetypes for easier associations and comparisons. I'm notoriously absent minded so sometimes, especially when happening significantly apart from each other, my various tinkerings will upset a chassis so much I have to throw it out and reset it. I like for the chassis to be both the first and the last thing I look at every rewrite.

In my case, being such easy references, it's usually the first thing I work and rarely tinker with. I set them as templates to create classes more often than not, so once I place them, I rarely alter them any further. If it's a class with added stuff (such as a spellcasting class), I address it based on the closest chassis that works; about the only classes that ended up having a chassis of its own was the Retooled Monk, the Blademaster (because it uses the typical Martial Adept chassis but I gave it a pretty important alteration that made it bizarre to say the very least), and perhaps one or two more. PrCs have other stranger alterations, but usually remain the same. You'll notice, if you read the bit about chassis in the Project Heretica thread, that the Divine Champion chassis is pretty strict, but I've used it as a basis in MANY of my retools (Bez-Kismet, Zealot, the other Paladin variants, the Dreadlord and even the Ranger to an extent follow the same chassis, only differing on the content and in the case of the Ranger the actual class features being pretty different, what with no smite equivalent or turning pool or even an excess of supernatural abilities). Likewise, I adapted the Spontaneous Specialist chassis to the Healer, which shares some class features with the Warmage; the Summoner and the Shugenja are a hybrid between this and a Spontaneous Caster proper. It's best to rarely, if ever, tweak a chassis once you use it, because sometimes the template itself reins you in and prevents you from going overboard with class features and whatnot.

I think I will make an attempt to use both slots and a pool then, I've still got both written versions for slots-only and pool-only. The class doesn't have a full spell list, so it should be totally manageable, and most class features don't require pool point expenditure - they only rely on keeping your pool of points filled (since I added a 'refresh one point' ability it's akin to a psionic focus I guess).

The "Ki" pool works in similar ways, where keeping the pool filled grants a minor bonus. It seems in this case that many class features are tied into it, which is a problem by itself; if most, instead of few, class features don't require you to spend points from the pool and instead require that the pool remains filled, you're removing utility from the features that do require expending points from the pool as you'll want to keep the passive features rather than risk going with the active ones, and makes going nova a really bad idea. Going nova with the Factotum's Cunning Surge feature poses no problem because you'll replenish all points later on; if you can only replenish one point, Cunning Surge would be used less because the other class features are just as useful (esp. the spells), and you'd try to hoard them (much like what happens with the Ninja, where even if you had 14-15 points in your "Ki" pool, you wouldn't spend them all on going nova with Ghost Strike because you'd be out of uses real quick). Each time you tie a passive feature into that pool, it only makes that worse; I mean, losing a bonus to Will saves isn't that bad, but when your main offensive or defensive measure is tied to having that pool filled, then going nova is simply NOT an option, and every use you spend from the pool is one less feature you might have to deny yourself in the future because it may cause your pool to dry up.

Pool-based features have to be meaningful so that you can spend uses reliably. It's a reason why I consider Rage the defining example of a pool-based resource done right: it provides bonuses useful in a variety of ways, it has a decent duration and can be used to power up other abilities that are just as good and last just as long, not to mention that the ability itself keeps adapting to your adventuring needs. Using Rage at 1st level is just as good as using it at 20th level, and since it lasts for as long as you need it, eventually you end up with more uses than you actually need, so it motivates you to use that feature in just about every battle and then some (I mean, why not use Rage to break an obstacle, for example?)


PS: Your Project Heretica designs are what caused me to get more serious about homebrewing. The image you used (of the winged Paladin) was something I'd come across before I came across Project Heretica and I built a four-path 'Hallowed' Paladin class based on it. Shortly after I finished it I came across your Project Heretica and saw that you'd done what I'd done, and then some (I cannot stress how much worse mine was, even though it was following a similar idea). I've been since trying to up my homebrewing game!

Uh...thanks for the compliment, I guess. It's pretty inspiring to hear that, though (much like that Simpsons' episode where Homer becomes an inventor), every muse has its own muse as well. I might not have much of a muse in terms of homebrewing, but I know that in terms of credentials I'm not at the same level as others. I work well as a certain kind of 'brewer (one who does great fixes), but I can't claim to be THE best fixer, particularly since my magnum opus isn't usually considered THE end-all-be-all choice for Paladins. Surprisingly enough, I'm the go-to guy for spontaneous spellcasters, something that usually baffles me (I like to play gishes, so it's weird that I'm a bit more skilled with casters...) That said, I can't hold a candle to people like The Demented One, for example, or Fax, or Krimm, or Tribble, or those who make up systems of their own (whether system-wide fixes or new resources for the game), so compared to them, my homebrewing "game" is nowhere near. But, I guess that comes with your choice of specialization, that is.

Seerow
2014-01-30, 01:30 PM
On the subject of dead levels, I want to point out: Gaining a new spell known, new invocation, new maneuver, new psionic power, or something along those lines should ALWAYS count as a feature when considering dead levels.

A lot of people get hung up on the 'special abilities' table when designing a class, and seeing nothing there makes them feel like something is wrong. But if at that level you gained a new spell known, chances are you just gained a new ability at that level that is more powerful/interesting than whatever is going to be written to fill in that blank space.

I admit, I fall into the trap as well. Especially in a class that already has a number of features, seeing 1 or 2 levels without a new ability makes those levels stand out and makes you want to fill it in just to have something there. But it's really not needed.

Newwby
2014-01-31, 07:54 AM
Chalk that up to early-installment weirdness combined with legacy.
-snip
I think I need to go over the other editions in more detail. My knowledge of pre-2nd edition is purely theoretical (things I've gleaned from forum chat and posts on various blogs), and even though I've learned a lot about 2nd edition recently, I still can't say I have reasonable knowledge of any DnD system than 3.5.

It's funny you mention the Monk whilst discussing capstones. I'm not a big MMO player and the Monk was the first class I played up to 20th level and since I've been of the opinion that all classes (prestige included) should have capstones to reward 'sticking it out'. More recently I've become laxer on the idea (although yesterday I did design an Arcane Archer rebuild with a monstrous capstone so obviously not entirely), instead aiming to spatter endgame abilities from the 16th level to the 19th.


The idea behind level weight is that, if all levels are balanced, you'll be motivated to remain in the class. However, sometimes, you need to establish some sort of imbalance that could be fixed by multiclassing or PrC'ing out, generally when you work with a concept that can't be covered by the class itself. Personally, I find that's the problem with PF classes; they have their imbalances, but they focus too much on remaining on one class and make multiclassing harmful. Ideally, a good game will reward system mastery while providing multiple options to achieve the same concept at a similar degree; level weight helps with the latter, but you need a certain degree of imbalance between the classes to allow considering other options.

In short, and to sound less philosophical: level weight is a good class-building measure, but sometimes you have to make wild choices and allow people to consider PrCing out. If a player has difficulty thinking "should I PrC out or should I remain one more class?", then you're going right. Not usually the case, but when considering options it's usually better to stay in the class, it may have the best degree of balance between levels but it'll be pretty poor.
I'd actually not noticed the multiclassing disparity between 3.5 and PF (I'm going to put that down to very limited experience with PF) - I suppose that's one of the big differences between (particularly for homebrewing) the two? (Mentioning my Antecedent again, I allowed specific options for multiclassing with Paladin, Fighter or Monk classes but the class is so self-contained I worry it doesn't lend itself to multiclassing well).


Sure: a jack of all trades can't be better than a Specialist, but it should have enough skill and knowledge to assume the role successfully if the specialist is out. The idea is as follows: assume that, for a specific role, you aim for 100% success.
-snip-
It's a damn good way of visualising it, and it answered my question about how to design the Antecedent's spheres (the last piece of the class I'm working on) well.


While the numbers are mostly subjective, it gives you an idea of how to work role balance out. From what I see, the Antecedent is sort of a Savant or Factotum in that it can do every role, but not that well.
-snip-
Yeah, the factotum was the original inspiration for the class (that and the aforementioned fact the Monk was my first DnD class), although it's been so long since I've gone over the class I feel like going back over it (after this reply) again to revisit the origins of the 'brew. Your notes on not adding skill bonuses to complement skill ranks for a non-specialist got me thinking about the choices of spells the spheres grant (not sure if I mentioned before but the spheres also determine the Antecedents spell list), I think some of them need serious revision. I'm pretty sure somewhere on one of the earliest designed spheres it grants use of the open lock skill, and then grants the 'knock' spell. I really hope I'm remembering that wrong.


The weight of class features in determining the success of each role isn't really set in stone, or should be standardized. -snip-
I was struggling with the idea of spells for the rogue-like spheres and I love the idea of the saboteur! If you don't mind I think I will use that, the flavour is fantastic - much better than 'sneaky rogue sphere'. I'll try to reconcile the various rogue sphere/aspects in to smarter roles to justify spell lists now. I'm definitely going to stray from including 'knock' and 'find traps' anywhere though, lest I completely disqualify a skill from contributing.


d20 Modern tapped the idea of creating "medium" saving throw progressions, which are pretty odd (they're roughly like poor saving throw progressions, except you start at +1 instead of 0 and you gain 2/3rds of a point at 1st level, 5th level and every 5 levels afterwards, IIRC. if you're not afraid of breaking with tradition, you could use that progression,
-snip-
I actually always thought it strange 3.5 didn't include medium saving throw progressions. I'm loathe to add such a thing though, I get the feeling the first post commenting on the class would be 'that saving throw is wrong', or I'd have to include a disclaimer explaining it. I might experiment with it in future though - I'm already trying to envision a class which might have three medium-progression saving throws.


The other is going the way of the Swashbuckler, whose Grace ability effectively allowed you to have a Reflex save better than a Fighter, but not as good as a Rogue, directly as a class feature instead of a modification to saving throws. A similar thing I do is give a PrC all bad saves, but give them a 1st level ability that defines its role and thus grants them one good save, which replaces its bad save. This is a more direct approach, but one that works for classes with wildly differing methods of approach.
My favoured approach! With the Antecedent I honestly could see it using all three saving throws (as mentioned it's rather monk meets factotum inspired), but didn't want to grant three saving throws. I knew I couldn't get rid of Will (class mechanic relies on a high progression) so I decided just to give the player choice of Fortitude or Reflex.
Directly from my Antecedent.odt;
Hardened Resolve (Ex)
An Antecedent quickly learns that a strong will alone is a poor companion when paired with an untempered body. At 2nd level an Antecedent may choose either their Fortitude or Reflex save to improve to a 'high progression', following the same progression each level as their Will save. This change is immediate and retroactive.


In my case, being such easy references, it's usually the first thing I work and rarely tinker with. I set them as templates to create classes more often than not, so once I place them, I rarely alter them any further.
-snip-
Think I'll go back and read on your work on chassis's (chassi? no idea of the root) next time I'm working on a new class (which given my proclivity for short lived ideas will probably be in about two hours).


The "Ki" pool works in similar ways, where keeping the pool filled grants a minor bonus.
-snip-
In the case of the Antecedent the class features that expend the pool were designed as 'limited per day' abilities, solely things to be saved for emergencies or special situations (in particular, one granted you concealment as an immediate action against an attack in exchange for expended Ki). The reasoning I had behind the 'keep one point in the pool for features x, x and x' was something (which I may have removed entirely in the purely spell slot version, I don't remember) to do with the spheres and maintaining them in extreme situations (such as incapacitation, anti-magic fields and against failed dispel checks). To wit, the Antecedent expends a Ki point and ignores the effect. The idea being that even at their most exhausted (perhaps I may replace it with something to do with fatigue/exhaustion) and expended, the Antecedent can still muster enough energy to do it once before refocusing, at the cost of their class features until they do so.


Uh...thanks for the compliment, I guess. It's pretty inspiring to hear that, though (much like that Simpsons' episode where Homer becomes an inventor), every muse has its own muse as well. I might not have much of a muse in terms of homebrewing, but I know that in terms of credentials I'm not at the same level as others. I work well as a certain kind of 'brewer (one who does great fixes), but I can't claim to be THE best fixer, particularly since my magnum opus isn't usually considered THE end-all-be-all choice for Paladins. Surprisingly enough, I'm the go-to guy for spontaneous spellcasters, something that usually baffles me (I like to play gishes, so it's weird that I'm a bit more skilled with casters...) That said, I can't hold a candle to people like The Demented One, for example, or Fax, or Krimm, or Tribble, or those who make up systems of their own (whether system-wide fixes or new resources for the game), so compared to them, my homebrewing "game" is nowhere near. But, I guess that comes with your choice of specialization, that is.
I didn't mean to sound quite so intense when I wrote that - but it was your work that got me beyond scribbling down random ideas and up to writing full class designs. It was pretty weird when I first came across Project Heretica (and probably why your work more than anyone elses got me to this point) to see what I'd been trying to do for weeks done in a far superior fashion. I think the main thing it got me doing (specifically what I mean by 'upping my homebrewing game') is asking questions of others. I'm terrible at asking for help even when I need it, but this thread alone has given me a lot to think about.

Re: Your mention of systems,
I'm actually DnD Homebrewing as a way of refining the way I think about contemporary tabletop gaming. Most of my previous experience, especially design experience was with various White Wolf games. I've been working on building my own Sci-Fi tabletop system based on the computer game 'FTL' on-and-off for the past nine months (five rewrites and 500k words s'far!), and I've most recently taken a break to understand other systems. Homebrewing DnD3.5 classes is where I'm at with that at the minute.


On the subject of dead levels, I want to point out: Gaining a new spell known, new invocation, new maneuver, new psionic power, or something along those lines should ALWAYS count as a feature when considering dead levels.
Agreed, although I tend to wait the higher (well 4th+) levels of spells/maneuvers/powers as more deserving of a level alone. Whilst most of my designs have some form of limited spellcasting, most stick around the 4th spell level (Paladin/Ranger style progression) or at most, 6th spell level (Bard style progression). I think if I were designing a class with spells up to 9th level I'd be a lot more comfortable leaving dead levels (in terms of special column features and other abilities) spread out around the higher levels.


A lot of people get hung up on the 'special abilities' table when designing a class, and seeing nothing there makes them feel like something is wrong. But if at that level you gained a new spell known, chances are you just gained a new ability at that level that is more powerful/interesting than whatever is going to be written to fill in that blank space.
The higher up the spell (or equivalent) levels the truer that becomes. I've found designing purely special abilities can actually handicap a class versus it's spell-gaining compatriots. Shy of some ridiculous god-like class I designed for a laugh, I don't think I've ever created a semi-magical class whose features could match 8th/9th spells (nor do I expect I will).


I admit, I fall into the trap as well. Especially in a class that already has a number of features, seeing 1 or 2 levels without a new ability makes those levels stand out and makes you want to fill it in just to have something there.
Nine times out of ten, when I add a feature to a class that I didn't originally envision, it's because the gap in the special abilities section of the class table grates at me too much. It's not a habit I'm going to get over easily either I imagine.

UserShadow7989
2014-02-03, 10:25 PM
For the longest time, I was extremely bad at overloading classes with more abilities than they needed. Not only did I feel compelled to fill blank spots on the Special Abilities table, I didn't count improvements on scaling abilities (though I didn't hold that standard for casting classes, at least).

Filler abilities for the sake of just filling spots are a bad idea. Besides adding more to keep track of, they give the player nothing of value. The level is still empty and pointless in all but the most literal of senses. Look at the core Monk for an example of the end result of that: a bundle of abilities with a very loose theme, no synergy, and while interesting on paper are useless in practice.

The best way to ensure a player gets something out of each level is having existing abilities scale, especially those that act as the 'center' of the class. It's a benefit that actually matters mechanically, and it keeps things concise.

Newwby
2014-02-06, 11:51 AM
I think I'm back to overloading with actual class abilities - the most recent class I built was entirely based around scaling class features but kept gaining new ones anytime there was a gap in progression. I always end up overpowering - going to shift it to spell slots and a limited spell list with maybe custom spells rather than covering it's bases with spell-like abilities (or similar), tear a hole or two in the class table.