How to Design Feats

[Vadskye]

There has been some discussion recently of how to fix feats. I've tried to write up four guiding principles for feat design. If you disagree, I'm very interested to hear how you design feats!

Feats should be relevant at every level
Good: Most feats. Blind-Fight, Improved Initiative, Skill Focus, Weapon Focus
Bad: Endurance, Diehard, Run, Two-Weapon Fighting, Toughness (3.5), Vital Strike (PF)

Feats which become obsolete with level are bad. I think we all agree on that. However, many people misapply this principle. "Scaling" isn't about arbitrarily increasing numbers. It's about making sure that a feat is always useful. Toughness should scale because HP is a real number that you add and subtract with. Vital Strike and Two-Weapon Fighting should scale because you get more attacks with level. The single attack granted by those feats becomes increasingly weak with level.

Bonuses to d20 rolls (attacks, checks, and so on) are completely different. A +2 bonus at 1st level and a +2 bonus at 20th level mean the same thing: you increase your chance of success by 10%. That's it. Feats like Weapon Focus, Skill Focus, and Iron Will are relevant at every level for the same reason that Improved Initiative and Blind-Fight are. The bonuses granted by those feats should not increase with level. (This is different from saying that those feats are good enough. Weapon Focus may be too weak. However, if it is, it's too weak at every level; scaling the bonus is not a solution.)

Weapon damage is a special case. Hit points scale with level, which suggests that damage should scale. However, # of attacks per round also scales, which means that a +1 bonus to damage at 20th level adds more damage than a +1 bonus to damage at 1st level. (EDIT: I found a problem with my previous calculations. I would estimate that damage should scale, but only very slightly. Not one per two levels; maybe one per five, roughly speaking.)

Finally, feats like Endurance and Run violate this principle in a different way. They grant bonuses which are technically relevant at every level. However, at high levels, it is extremely rare for characters to make rolls that Endurance helps with or take a run action. The feat mechanics are fine, but they help irrelevant things.
Every feat is unique
Good: Blind-Fight, Manyshot, Mounted Combat, (Improved) Precise Shot
Bad: Many. Weapon Focus and Weapon Specialization, Two-Weapon Fighting and improvements, Vital Strike and improvements, any tree where the advanced feat is "the same thing, but better".

Every feat should contribute something unique to a character. Improved Two-Weapon Fighting and Improved Vital Strike add nothing. They are merely a feat tax which limits the use of abilities the character has already invested in. This encourages overspecialization and limits character flexibility. Since the core problem with warriors is a lack of flexibility, not a lack of raw power, this is very bad. If nothing else, it's boring.

The problem with Weapon Specialization is less obvious. Mechanically, Weapon Focus and Weapon Specialization do different things. Doesn't that mean they are unique? No. They are the exact same thing: "Get generically better with a specific weapon". We don't need four different feats which simply contribute numerical bonuses. This encourages overspecialization and makes characters more generic. There should only ever be one feat which means "get bonuses with X".

There are two ways to solve this problem. One is to simply add the effects of the other feats to the original feat. That is a good approach for feat trees which granted scaling effects, like Two-Weapon Fighting and Vital Strike. However, that is a poor solution for Weapon Focus: a single feat should not grant a +2 bonus to attack rolls and a +4 bonus to damage. Instead, Weapon Specialization might grant a new ability while using the chosen weapon which is not a numerical bonus, or it might grant a bonus while imposing a penalty when using any non-specialized weapon. There are several possible approaches here, and that will be dealt with later.
Feats are balanced against other feats
Good: Depends on what balance point you set for feats. Probably Endurance, at least.
Bad: Melee Weapon Mastery (PHB2), most "fighter fix" homebrew feats, Power Attack (3.5), Endurance, Run

"Balance" is a tricky term. Higher level feats can be more powerful than lower level feats. Base attack bonus requirements are a good way of doing this. Similarly, feats with ability score requirements can be more powerful than feats without such requirements, since gaining the ability score limits the character in other ways. However, feat and class prerequisites should not affect a feat's power level.

Suppose we make an overpowered feat and try to balance it by assigning four feats as a prerequisite. Alice wants to take this overpowered feat. If the prerequisite fleats work well together, this will not decrease Alice's power, and he will still be overpowered after completing the feat chain. If the feats are unrelated and don't help Alice much, she will be weaker than normal, so Alice is balanced overall. However, that is poor design. Alice will be unusually weak until the final feat is acquired. Additionally, it doesn't make sense for Alice to take unrelated feats which don't help her fulfill her character concept. It's just bad all around.

This principle is commonly violated with class-based requirements in homebrew. Suppose we think a class are weak. One common fix is to create a "Fighters Are Awesome" feat - or set of feats - which dramatically increases the power of the chosen class. This is a terrible approach. Because the feat is so much more powerful than other feats, every member of the class must take it or be weak. This is particularly bad when the feat has several prerequisites (most commonly, Weapon Specialization), since it forces everyone in the class down a particular path. The class as a whole becomes less flexible and dynamic.

Feat and class prerequisites are not a balancing mechanism. They are simply a way to enforce game logic. EDIT: Snatch Arrows doesn't require Deflect Arrows because Snatch Arrows would be overpowered otherwise. It's because it just makes sense.

Finally, there is the well known problem of feats which are simply too weak relative to other feats. Those can simply be improved to match other feats.
Feats should not have irrelevant prerequisites
Good: Feats without prerequisites, Cleave, Snatch Arrows, Two-Weapon Defense
Bad: Improved Overrun, Whirlwind Attack, Defensive Throw (CW), all ranged feats relying on Point Blank Shot, Moonlight Stalker (PF)

Why does the ability to shoot farther (Far Shot) require the ability to shoot very close targets more effectively (Point Blank Shot)? Why does the ability to attack many foes at once (Whirlwind Attack) require the ability to focus on a single opponent and avoid their attacks (Dodge)? Why does the ability to fight better while concealed (Moonlight Stalker) require... any of its bizarre prerequisites (Blind-Fight, Combat Expertise, Bluff, darkvision or low-light vision)? This doesn't make sense.

Feat prerequisites should be closely related to the feat. Removing irrelevant prerequisites like Point Blank Shot allows characters to be more flexible, makes the game less complicated, and just makes sense.

[ddude987]

Good tips, I like it. These are definitely things I keep in mind when working on rewriting the 3.5 feat system.

[Mighty_Chicken]

Thank you for such a well written OP. It's clarity will certainly influence the quality of any discussion, not the mention its merits as a guide itself.

My sincere acknowledgement apart, I disagree with you in several parts.



Feats which become obsolete with level are bad. I think we all agree on that. However, many people misapply this principle. "Scaling" isn't about arbitrarily increasing numbers. It's about making sure that a feat is always useful. Toughness should scale because HP is a real number that you add and subtract with. Vital Strike and Two-Weapon Fighting should scale because you get more attacks with level. The single attack granted by those feats becomes increasingly weak with level.

Bonuses to d20 rolls (attacks, checks, and so on) are completely different


You are forgetting that in the context of a 1st level character, +5% to hit means something and in the context of a 20th level character +5% to hit means something different. At 20th level, it's a bonus for 4 attacks instead of one; but the third and fourth attacks are 50% and 75% behind already.

And this doesn't happen in a vaccum; it's in the context of, among other things, "Quadrantic wizard, linear fighter". Wizard abilities escalate, almost all of them, at the same time they get new abilities (spells); but not all fighter abilities escalate (BAB and HD do, many feats don't). Yes, the point of escalating Weapon Focus (just an example) is to make it more relevant in 20th level than it was at 1st level. You can say it's bland design; I'd agree. But Metamagic feats are escalating too, because first you were empowering 10d6, now you're empowering 15d6. There are plenty of feats that become more powerful with time. This isn't bad design, this is just keeping them quiet at early levels when they'd be overpowered, and let them shine later.



Weapon damage is a special case. ... I would estimate that damage should scale, but only very slightly. Not one per two levels; maybe one per five, roughly speaking.

You're talking about keeping the relevance of the feat stable. Other people may think it's important to make a feat more and more relevant with time.

Why should feats keep the same relevance instead of growing more relevant? "Personal taste" is a fair answer - "good design" isn't. Both designs can be good, but they have different methods and different objectives.



Every feat should contribute something unique to a character.

Just a comment: in AD&D Weapon Specialization gave +1 attack, +2 to damage and +1/2 attack/round. What made it unique was that only the fighter could have it; other than that, it was an almost purely numerical bonus.

Then again, I'd agree pure numerical bonuses are bland. But saying it's bad design isn't necessarily true. Pure numerical bonuses might be part of what makes a character unique. It surely worked with Weapon Specialization in AD&D. And what is Rage in 3.5 but a numerical bonus with non-numerical penalties?



Feats are balanced against other feats

Yes, that's why when criticizing a homebrew feat, one must consider if the feat is supposed to be used with RAW feats or not.

There's also another thing to think about: are feats supposed to be equally good to any class, or are some classes and characters supposed to profit more on feats than others? Then again, it's more a matter of taste and preference than of good or bad design.


However, feat and class prerequisites should not affect a feat's power level.

In this specifically I think you're simply wrong. Feat prerequisites are level prerequisites, because almost every character get the exact same number of feats; so if the feat I want has 3 other feats as a requirement, I'm not getting it before 9th level in 3.5 or 7th level in PF; the exceptions are humans and fighters. In fighter's case, cleaving through feat requirements faster is a hidden class feature; actually, this is something many homebrewers tend to enhance giving fighters class features that reduce ability requirements.

(A side note: this reminds me of how monks have some feat and ability requirements reduced for them)

Feat requirements are a limitation. Even considering a feat tree that is well designed as does not tax you, it limits the number of possible combos.

About class requirements: why should they allow only uniqueness, and not power? Feats shouldn't only be balanced against other feats; they should be also balanced in combination with class features and spells. A certain combination of spells and class features can make a feat broken. A class requirement limits the number of class features and spells that can be used to exploit it; therefore, its power level has space to increase.

We go back to the question of whether feats should work similarly for all classes or not. A class-specific feat is an optional class feature. If you dislike optional class features, it's "bad" design; if you like it, it's "good" design. That's a matter of preference and not of quality.

The problem with a powerful Weapon Focus tree that's exclusive for fighters is that it's the class' only exclusive feat tree. If the class had more "optional class features", it wouldn't be a false option, but instead would allow for modular class design.


Feats should not have irrelevant prerequisites

I agree, of course, but here I think you contradict yourself a little. Or maybe you don't; for example, are you ok with Cleave requiring Power Attack for fluff reasons only? Because if Cleave would be balanced as a prerequisite-less feat, why not have it as it? I could flavor Cleave in a dozen of ways that do not involve Power Attacking.

[ddude987]

In this specifically I think you're simply wrong. Feat prerequisites are level prerequisites, because almost every character get the exact same number of feats; so if the feat I want has 3 other feats as a requirement, I'm not getting it before 9th level in 3.5 or 7th level in PF; the exceptions are humans and fighters. In fighter's case, cleaving through feat requirements faster is a hidden class feature; actually, this is something many homebrewers tend to enhance giving fighters class features that reduce ability requirements.

(A side note: this reminds me of how monks have some feat and ability requirements reduced for them)

Feat requirements are a limitation. Even considering a feat tree that is well designed as does not tax you, it limits the number of possible combos.

I disagree. Feat prerequisites are not a sufficient level barrier for feats. There are numerous ways to get extra feats, from human to dark chaos shuffling. If a feat is designed with a specific minimum level in mind, it should require a skill requirement, which is something harder to overcome than feat prereqs, or require a specific character level.



I agree, of course, but here I think you contradict yourself a little. Or maybe you don't; for example, are you ok with Cleave requiring Power Attack for fluff reasons only? Because if Cleave would be balanced as a prerequisite-less feat, why not have it as it? I could flavor Cleave in a dozen of ways that do not involve Power Attacking.

But I would say, regardless of fluff reasons, power attack is not a irreverent feat as compared to say, toughness. Of course, if you take into consideration the re balancing of 3.5e's feat system, following these rules, then there will be no irrelevant prerequisites as "prereq" feats like point blank shot, endurance, et cetra, either will not exist or will not be irrelevant.

[Vadskye]

Thank you for such a well written OP. It's clarity will certainly influence the quality of any discussion, not the mention its merits as a guide itself.
You are the nicest person!

You are forgetting that in the context of a 1st level character, +5% to hit means something and in the context of a 20th level character +5% to hit means something different. At 20th level, it's a bonus for 4 attacks instead of one; but the third and fourth attacks are 50% and 75% behind already.
I've run some calculations across level for various characters against average AC. As far as I can tell, a +1 attack bonus translates into a 5-10% increase to damage per round against average AC, regardless of level. There are complexities (I ignored critical hits), but none of those seem to the change the fundamental logic that +1 to attack is always relevant.

And this doesn't happen in a vaccum; it's in the context of, among other things, "Quadrantic wizard, linear fighter". Wizard abilities escalate, almost all of them, at the same time they get new abilities (spells); but not all fighter abilities escalate (BAB and HD do, many feats don't). Yes, the point of escalating Weapon Focus (just an example; I never did that) is to make it more relevant in 20th level than it was at 1st level. You can say it's bland design; I'd agree. But Metamagic feats are escalating because first you were empowering 10d6, now you're empowering 20d6.
Feat design is not the place to fix "quadratic wizard, linear fighter". Class and magic imbalances should be addressed with class and magic fixes. Feats can address math imbalances to a very limited degree; if you rely on feats to fix the math, you force everyone to take the "math-fixing" feats.

You're talking about keeping the relevance of the feat stable. Other people may think it's important to make a feat more and more relevant with time.

Why should feats keep the same relevance instead of growing more relevant? "Personal taste" is a fair answer - "good design" isn't. Both designs can be good, but they have different methods and different objectives.
If every single feat became equally more relevant with time, it would be possible to call that good design. But I've never seen a system that did that. I'm not sure it's possible. Instead, you get a mishmash where people decide that certain feats (Weapon Focus) should actively become more relevant with level and eventually become "must-take" feats, while the vast majority of feats continue be equally relevant across level. That's because it's really, really hard to make everything become equally more relevant with level.


Just a comment: in AD&D Weapon Specialization gave +1 attack, +2 to damage and +1/2 attack/round. What made it unique was that only the fighter could have it; other than that, it was an almost purely numerical bonus.

Then again, I'd agree pure numerical bonuses are bland. But saying it's bad design isn't necessarily true. Pure numerical bonuses might be part of what makes a character unique.
There's nothing wrong with numerical bonuses! I'm sorry if I gave that impression. I think you should have the ability to take a feat to improve virtually any numerical aspect of your character. Numbers are essentially how we differentiate characters, and you should definitely have control over that. But you don't need four feats that increase the same number.

Imagine if every feat had "this feat may be taken multiple times. Its effects stack". That would be an incredibly boring and broken system! Weapon Focus, Weapon Specialization, Greater Weapon Focus, and Greater Weapon Specialization are bad for the exact same reason.

Yes, that's why when criticizing a homebrew feat, one must consider if the feat is supposed to be used with RAW feats or not.

There's also another thing to think about: are feats supposed to be equally good to any class, or are some classes and characters supposed to profit more on feats than others? Then again, it's more a matter of taste and preference than of good or bad design.
A system where every class gained the same value from every feat would be bizarre. I would go so far as to say that good design dictates that feats should be have varying utility for different characters.

In this specifically I think you're simply wrong. Feat prerequisites are level prerequisites, because almost every character get the exact same number of feats; so if I need a feat has 3 other feats as a requirement, this means I'm not getting it before 12th level in 3.5 or 9th level in PF; the exceptions are humans and fighters. In fighter's case, cleaving through feat requirements faster is a hidden class feature; actually, this is something many homebrewers tend to enhance giving fighters class features that reduce ability requirements.
There are problems with use feats as quasi-level prerequisites, as ddude pointed out. If you want level prerequisites, use level prerequisites. Are you saying that feats with multiple feat prequisites should be more powerful than other feats?


(A side note: what about the way monks reduce feat and BAB requirements for some feats?)
I like it in principle. In practice, it's implemented poorly; when you have a choice between two feats, they should both have their prerequisites reduced similarly. Improved Grapple vs. Stunning Fist is a poorly designed choice.

The Ranger mechanic for ignoring ability prerequisites is poorly designed, however. Rangers are thematically dexterous, and ignoring a Dex requirement is only useful if the ranger is low Dex. Which is weird.


Feat requirements are a limitation. Even considering a feat tree that is well designed as does not tax you, it limits the number of possible combos.
Why is that good?


About class requirements: why should they allow only uniqueness, and not power? Feats shouldn't only be balanced against other feats; they should be also balanced in combination with class features and spells. A certain combination of spells and class features and make a feat broken. A class requirement limits the number of class features and spells that can be used to exploit it; therefore, its power level can be increased.
If class-specific feats are more powerful than comparable feats, they impose a feat tax, limiting flexibility for characters of that class. Feats can expand options for members of a class; Extend Rage, Extra Turning, and so on are all great. However, it's a mistake to make those class features more powerful than normal feats.

Imagine if there was "Dodge" for most characters, but a Rogue-specific "Rogue Dodge" that was the exact same thing, but with a higher bonus. What's the point? It's just a confusingly implemented class feature - and if the bonus is too high, it's a feat tax. Class features belong in class fixes; feats are different.


We go back to the question of whether feats should work similarly for all classes or not. A class-specific feat is an optional class feature. If you're against optional class features, it's "bad" design; if you like it, it's good design.
Class-specific feats as optional class features is cool. However, they don't need to be more powerful than other feats. They can just be different.


The problem with a powerful Weapon Focus tree that's exclusive for fighters is that it's the class' only exclusive feat tree. If the class had more "optional class features", it wouldn't be a false option, but allow for modular class design.
Generally, I like to implement modular class design by making the class design modular in its own right, akin to the Pathfinder rogue talent / rage power / etc., and let feats be for things that aren't class features.


I agree, of course, but here you contradict yourself a little. Or maybe you don't; for example, are you ok with Cleave requiring Power Attack for fluff reasons only? Because if Cleave would be balanced as a prerequisite-less feat, why not have it as it? I could flavor Cleave in a dozen of ways that do not involve Power Attacking.
I, uh, am not sure why I used Cleave as an example. You're totally right - in my own feat system, I removed Power Attack as a prerequisite for Cleave a long time ago. My other examples (Snatch Arrows, Two-Weapon Defense) are much better, I think.

[Grod_The_Giant]

I've run some calculations across level for various characters against average AC. As far as I can tell, a +1 attack bonus translates into a 5-10% increase to damage per round against average AC, regardless of level. There are complexities (I ignored critical hits), but none of those seem to the change the fundamental logic that +1 to attack is always relevant.
While that may be mathematically true, it's not necessarily true perception-wise. After all, we don't see the feat as providing "+5% hit chance," we see it as "+1." At first level, that +1 might be 20% of your total attack bonus. At 20th, it might be 2%. That's a huge drop in perceived value. The feat looks like it's getting less and less useful. And in a game, particularly one where most people aren't looking at the metagame numbers, that's an important thing to keep in mind.

[Vadskye]

While that may be mathematically true, it's not necessarily true perception-wise. After all, we don't see the feat as providing "+5% hit chance," we see it as "+1." At first level, that +1 might be 20% of your total attack bonus. At 20th, it might be 2%. That's a huge drop in perceived value. The feat looks like it's getting less and less useful. And in a game, particularly one where most people aren't looking at the metagame numbers, that's an important thing to keep in mind.
Yes and no. You're absolutely right that perception is important. Players should never have to choose between the interesting and the good; things that are good should also look interesting (and vice versa!). Weapon Focus doesn't have a lot of perceived value after the early levels, regardless of its actual value.

However, I would argue that that's exactly what should happen. I implied this in one of the principles, but maybe it deserves to be its own principle: Higher level feats should be better than lower level feats. The goal isn't to make low-level feats be equal in power to high level feats. That means there's no character advancement! If taking a 1st level feat at 18th level is a good idea, something has probably gone wrong. The goal is to make low-level feats relevant at every level. And a +1 bonus accomplishes that goal.

Basically, Weapon Focus isn't supposed to be a 20th level feat. It's a 1st or 3rd level feat. Most low-level feats will - and should - have a low perceived value at high levels. That's not a problem as long as the feat isn't actually useless (like Endurance).

[Mighty_Chicken]

I disagree. Feat prerequisites are not a sufficient level barrier for feats. There are numerous ways to get extra feats, from human to dark chaos shuffling.

They're still a barrier. Fighters and humans have this "hidden feature" of having an easier time breaking these barriers, just like monks have the "open feature" of breaking the barriers for some specific feats.

And even in fighter's case, a feat prerequisite means that if you have to invest a limited resource. The practical effect is limiting the number of possible combos.



Feat requirements are a limitation. Even considering a feat tree that is well designed as does not tax you, it limits the number of possible combos.

Why is that good?

I don't know if it's good. But it's surely a balancing mechanic.

I personally think it would be interesting to go radical and throw away all feat requirements. It would open up a lot of interesting possibilities, and balancing would still be possible with other kinds of requirements.

This, however, do not mean feat prerequisites are uneffective to balance feats. A feat requirement is similar to level requirement, but you're accepting that certain characters will be able to break the level requirement. It's ok, some classes get some spells earlier, and some classes get iterative attacks earlier. Some characters may get an specific feat earlier than others, too, that's a feature, not a bug.



Feat design is not the place to fix "quadratic wizard, linear fighter". Class and magic imbalances should be addressed with class and magic fixes.

But that is just, like, your opinion, man :) Then again it's a matter of preference and not of design quality.

Why isn't it the place to fix it? Wizards have spells, fighters have feats. Grossly speaking, nerf spells, buff feats is a way to go. "But everyone can have feats" - well, I think everyone should have limited access to spells, too. Everyone has access to hit points and BAB, too, but some characters use each mechanic better than others.

There are many possible solutions. Arguing "this will not work" is one thing, but arguing "this mechanic isn't supposed to work this way" is more about personal preference than effective design.


I've run some calculations across level for various characters against average AC. As far as I can tell, a +1 attack bonus translates into a 5-10% increase to damage per round against average AC, regardless of level.

Erm, sorry if I'm being lazy, but could you calculate how other feats escalate? For example, do metamagic feats translates like this, too? Or do they become more and more significant?


You're talking about keeping the relevance of the feat stable. Other people may think it's important to make a feat more and more relevant with time.



Why should feats keep the same relevance instead of growing more relevant?...

If every single feat became equally more relevant with time, it would be possible to call that good design. But I've never seen a system that did that. I'm not sure it's possible.

Well, perfect equality is hard. Aproximate equality, for most feats? Can be done. Aproximate equality, but just for certain categories of feats? That's a design decision. In a vacuum I'd agree it's bad design, but in the context of 3.5 D&D, it means feats would mean one thing for one class and a different thing for others, just like magic means one thing for sorcerers and another thing for clerics.



If class-specific feats are more powerful than comparable feats, they impose a feat tax, limiting flexibility for characters of that class.

Fair enough, but only if the power difference is too big. If, for example, rogues had a number of good feats waiting for them, "Rogue Dodge" wouldn't always be the best option, so it wouldn't be a tax or a false choice.

In such case I have to agree with you feats must be balanced against each other, otherwise there's no choice. That's, ironically, where feat trees as a balancing method makes the most sense. If I'm building a tripping fighter and Ultimate Damage Weapon Focus Feat is easily gainable, I'll get this feat, the grappling fighter will get this feat, the two handed guy will get it too, so it's not a real choice- it's a feat tax. But if its feat requirements are too much, trying to get it will drain away the resources I have to focus with I want my Ultimate Tripping Feat.

But I agree that this "feat arms race" is complicated. If there are, for example, 4 "Feat Tree Ultimate Capstones", then there are just 4 possible fighter builds, with minor alterations depending on how much space does the feat trees leave. The versatility really suffers a bit.


A system where every class gained the same value from every feat would be bizarre. I would go so far as to say that good design dictates that feats should be have varying utility for different characters.

Of course, but what I also meant was: the feat mechanic itself could mean something different to each character, just like BAB means something different to each class. To some classes feats might be additional side features, to other classes they might be their backbone.


Generally, I like to implement modular class design by making the class design modular in its own right, akin to the Pathfinder rogue talent / rage power / etc., and let feats be for things that aren't class features.

I was reading your homebrew today and this philosophy is effective within it.


Higher level feats should be better than lower level feats. The goal isn't to make low-level feats be equal in power to high level feats. That means there's no character advancement!

It's a resonable design philosophy, but it's also something restricting. After all, it means that once I get my 18th level feat, it would be a waste of resources to choose a low-level feat. But sometimes, there is no high-level feat which fits the concept or kind of ability I'm looking for. I'm left with fewer reasonable options.

So, should a 20th level character have 2 awesome feats, 3 useful feats and 2 meh feats, or 7 awesome feats?

EDIT:
Just remembered something.


(A side note: what about the way monks reduce feat and BAB requirements for some feats?)
I like it in principle. In practice, it's implemented poorly; when you have a choice between two feats, they should both have their prerequisites reduced similarly. Improved Grapple vs. Stunning Fist is a poorly designed choice.

Stunning Fist had a quite radical improvement for monks, as you know... insanely reduced prerequisites, and multiplied daily uses. The option was, "you can be about as good grappling as other characters, or the master of Stunning Fist". Yet, a lot of players chose Grapple! Class specific bonuses don't always mean feat taxing.

[3WhiteFox3]

Yes and no. You're absolutely right that perception is important. Players should never have to choose between the interesting and the good; things that are good should also look interesting (and vice versa!). Weapon Focus doesn't have a lot of perceived value after the early levels, regardless of its actual value.

However, I would argue that that's exactly what should happen. I implied this in one of the principles, but maybe it deserves to be its own principle: Higher level feats should be better than lower level feats. The goal isn't to make low-level feats be equal in power to high level feats. That means there's no character advancement! If taking a 1st level feat at 18th level is a good idea, something has probably gone wrong. The goal is to make low-level feats relevant at every level. And a +1 bonus accomplishes that goal.

Basically, Weapon Focus isn't supposed to be a 20th level feat. It's a 1st or 3rd level feat. Most low-level feats will - and should - have a low perceived value at high levels. That's not a problem as long as the feat isn't actually useless (like Endurance).

Weapon Focus doesn't have to be a 20th level feat, it just shouldn't be a 1st level feat through out it's entire existence or it ceases to have mechanical impact on the character build soon after it's inclusion.

Lets take two different feats at level 20.

Weapon Focus
Improved Initiative

Both give flat bonuses to numbers, but Improved Initiative is not only more useful, it feels more useful. Both are level one feats, improved initiative feels equally strong at 1st through 20th, at no point would a 20th level character feel he wasted a feat slot. Weapon Focus, past say level 5, is likely to feel like a wasted choice that could have been filled in with something else.

That's not good design, and it's not simply a problem of Improved Initiative being 'better' initiative is only calculated once per fight, while you'll be making numerous attack rolls, so the higher number is logical. The problem is, Weapon Focus should not exist as a feat, it's boring, perceived as weak (in 3.5; in PF it's often considered much stronger thanks to there being far fewer mundane feats worth taking) and only really useful for fighters with too many open feats.

Wrap Weapon Focus/Weapon Specialization into a class feature that fighters get.

EDIT:
Erm, sorry if I'm being lazy, but could you calculate how other feats escalate? For example, do metamagic feats translates like this, too? Or do they become more and more significant?

I don't know about calculations, but Metamagic is going to be scale far better thanks to two factors.

1. It's tied to spell level, with each new spell level a caster gains allowing him to apply the same meta-magic feat to newer spells.

2. This isn't merely a caster/martial imbalance problem either. Unless you make spells not scale at all and make it so that new spell levels don't increase power, Quicken Spell will still work better as you get more and more powerful spells. The very mechanic Metamagic uses necessitates becoming more powerful with each level

To do otherwise would require scrapping Metamagic in it's current form and/or a massive change to spells (far out of scope than anything on these forums; it would need to be a complete rehaul).

[Vadskye]

I don't know if it's good. But it's surely a balancing mechanic.

I personally think it would be interesting to go radical and throw away all feat requirements. It would open up a lot of interesting possibilities, and balancing would still be possible with other kinds of requirements.
That's essentially what I'm doing with Rise. However, I agree with your earlier point - this falls under the category of personal preference/experimentation, not "good design".


This, however, do not mean feat prerequisites are uneffective to balance feats. A feat requirement is similar to level requirement, but you're accepting that certain characters will be able to break the level requirement. It's ok, some classes get some spells earlier, and some classes get iterative attacks earlier. Some characters may get an specific feat earlier than others, too, that's a feature, not a bug.
Feat requirements are only a level requirement if a character must spend every feat meeting those requirements. That's not a sustainable model.

Suppose we have Ultra Mega Feat which is very powerful, so we balance it by assigning four feat prerequisites. A human fighter can get this feat at 3rd level, a normal fighter gets it at 4th, a human nonfighter can get it at 9th level, and a generic character can get it at 12th. Great! Fighters have their "hidden feature" of breaking feat barriers, and everything makes some degree of sense. (Though I'm not sure humans need a 3-level advantage.) Now, Ultra Mega Feat has to be balanced so it doesn't break anything at 3rd level, but is still powerful and worth the four feat investment at 9th level. That's challenging, but doable.

But what happens after you get Ultra Mega Feat? Every other feat will be worse - forever! There's also Super Deathkill Feat, which is also balanced by assigning four feat prerequisites. The human fighter could also pick that up by 9th level. But at that point, the fighter is in the exact same position as any other human picking up the exact same feat. See, the fighter "hidden feature" isn't actually getting early access to feats locked behind feat requirements. That advantage happens exactly once, at low levels - when the fighter was already doing pretty well anyway. The real feature is having more feat combinations than anyone else. That's a much, much weaker advantage; see my discussion near the bottom about food.

But that is just, like, your opinion, man :) Then again it's a matter of preference and not of design quality.

Why isn't it the place to fix it? Wizards have spells, fighters have feats. Grossly speaking, nerf spells, buff feats is a way to go. "But everyone can have feats" - well, I think everyone should have limited access to spells, too. Everyone has access to hit points and BAB, too, but some characters use each mechanic better than others.

There are many possible solutions. Arguing "this will not work" is one thing, but arguing "this mechanic isn't supposed to work this way" is more about personal preference than effective design.
It won't work because you're fixing a class-based problem with a class-independent solution. If you rely on Ultra Mega Feat to fix caster/noncaster imbalance, the most powerful character will be the (human) CoDzilla who picks up Ultra Mega Feat. You bring up BAB/HP, which raises an interesting point: the more impact feats have on your combat abilities, the less important BAB and HP are. At that point, you're actually making the class imbalances worse by nerfing the one advantage noncasters not named Fighter had.

Erm, sorry if I'm being lazy, but could you calculate how other feats escalate? For example, do metamagic feats translates like this, too? Or do they become more and more significant?

You're talking about keeping the relevance of the feat stable. Other people may think it's important to make a feat more and more relevant with time.
It's tough to evaluate because spells inherently have a nonlinear progression. Technically, any feat which improves spellcasting also has a nonlinear progression. Let's ignore that problem for the moment. Generally speaking, metamagic feats that directly increase power (Empower, Maximize, and Quicken) have a nonlinear progression, while metamagic feats that change aspects of a spell (Still, Silent, Energy Substitution) have a more linear progression.

Well, perfect equality is hard. Aproximate equality, for most feats? Can be done. Aproximate equality, but just for certain categories of feats? That's a design decision. In a vacuum I'd agree it's bad design, but in the context of 3.5 D&D, it means feats would mean one thing for one class and a different thing for others, just like magic means one thing for sorcerers and another thing for clerics.
Trying to figure out what that means. So all skill feats are balanced against all other skill feats, and all combat feats are balanced against all combat feats, but we don't worry about balance between skill and combat feats? That makes sense to me. If you're talking about subdivding combat feats into different categories that each have separate balance points, I'm not so sure.

Fair enough, but only if the power difference is too big. If, for example, rogues had a number of good feats waiting for them, "Rogue Dodge" wouldn't always be the best option, so it wouldn't be a tax or a false choice.

In such case I have to agree with you feats must be balanced against each other, otherwise there's no choice. That's, ironically, where feat trees as a balancing method makes the most sense. If I'm building a tripping fighter and Ultimate Damage Weapon Focus Feat is easily gainable, I'll get this feat, the grappling fighter will get this feat, the two handed guy will get it too, so it's not a real choice- it's a feat tax. But if its feat requirements are too much, trying to get it will drain away the resources I have to focus with I want my Ultimate Tripping Feat.
But then there will only be...

But I agree that this "feat arms race" is complicated. If there are, for example, 4 "Feat Tree Ultimate Capstones", then there are just 4 possible fighter builds, with minor alterations depending on how much space does the feat trees leave. The versatility really suffers a bit.
Oh, yeah, that's what I was going to say. I'm suspicious of capstones for this reason. More accurately, I'm suspicious of inflexible capstones. If you group feats into coherent categories ("Mobility", "Precision", "Power"...) then you might be able to have "capstone" feats without invoking the 4-build problem. Rise has the categories now, but not the capstones.

Of course, but what I also meant was: the feat mechanic itself could mean something different to each character, just like BAB means something different to each class. To some classes feats might be additional side features, to other classes they might be their backbone.
Plausible. I'm suspicious of this in part because feats need to fulfill class-independent roles, like skill and defense improvements. But I certainly wouldn't call that "bad design".

I was reading your homebrew today and this philosophy is effective within it.
Huzzah!

It's a resonable design philosophy, but it's also something restricting. After all, it means that once I get my 18th level feat, it would be a waste of resources to choose a low-level feat. But sometimes, there is no high-level feat which fits the concept or kind of ability I'm looking for. I'm left with fewer reasonable options.

So, should a 20th level character have 2 awesome feats, 3 useful feats and 2 meh feats, or 7 awesome feats?
Suppose we're trying to decide what to eat today. We have delicious steak, pretty okay salad, this fruitcake that Aunt Betty gave us for Christmas, and a dead rat. Obviously we eat the steak first. Then we have to choose another food. I guess we'll choose the salad. If we have to choose another food, we'll try to stomach the fruitcake.

If you choose repeatedly from the same pool of options, every choice gets progressively worse. Feats work the same way. Every feat chosen from the same pool is worse than the previous feat. If your pool of feats is stagnant, each new feat you get is worse than your previous feats. Feat prerequisites don't change that fundamental logic - they just delay it. Suppose that, after eating the cake and the meatloaf, we unlocked ice cream! Yay! But we still have to choose the fruitcake next time we get hungry.

You also can't solve this by making all feats balanced. Even if the steak, the salad, fruitcake, and rat were all prepared by a professional chef, diffent eaters will prefer different foods. Some people are vegetarian, so they'll start with the salad. Some people are eating with Aunt Betty in the room, so they'll start with the fruitcake. Everyone has preferences, and there will always be an order of "best" to "worst".

The only way (that I have found) to prevent this problem is to provide more options with time. That means level-based prerequisites. Now, that doesn't technically mean that high-level feats are more powerful than low-level feats - just different. But you're never going to get 7 equally awesome feats; there's always an order. Since the entire rest of the system gets more powerful with level, I think it makes the most sense to make the high-level feats more powerful.


EDIT:
Just remembered something.


Stunning Fist had a quite radical improvement for monks, as you know... insanely reduced prerequisites, and multiplied daily uses. The option was, "you can be about as good grappling as other characters, or the master of Stunning Fist". Yet, a lot of players chose Grapple! Class specific bonuses don't always mean feat taxing.
I think Stunning Fist vs. Improved Grapple is fairly balanced. But the fact that they are balanced just means Stunning Fist's prerequisites were way, way too high to start with. It was a problem because sometimes people who wanted Grapple would take Stunning Fist simply because it was virtually impossible to get without ignoring the feat prerequisite.


Weapon Focus doesn't have to be a 20th level feat, it just shouldn't be a 1st level feat through out it's entire existence or it ceases to have mechanical impact on the character build soon after it's inclusion.

Lets take two different feats at level 20.

Weapon Focus
Improved Initiative

Both give flat bonuses to numbers, but Improved Initiative is not only more useful, it feels more useful. Both are level one feats, improved initiative feels equally strong at 1st through 20th, at no point would a 20th level character feel he wasted a feat slot. Weapon Focus, past say level 5, is likely to feel like a wasted choice that could have been filled in with something else.

That's not good design, and it's not simply a problem of Improved Initiative being 'better' initiative is only calculated once per fight, while you'll be making numerous attack rolls, so the higher number is logical. The problem is, Weapon Focus should not exist as a feat, it's boring, perceived as weak (in 3.5; in PF it's often considered much stronger thanks to there being far fewer mundane feats worth taking) and only really useful for fighters with too many open feats.
This is part of the perception vs reality problem. Suppose initiative increased by 1 per character level. Improved Initiative would be just as powerful - but people would think it got weaker with level, just like Weapon Focus. It would be unfairly good if you scaled Improved Initiative with level. I'm still not sure how to fix it - but scaling Weapon Focus isn't a good solution.

Wrap Weapon Focus/Weapon Specialization into a class feature that fighters get.
As long as being a generalist fighter is still a viable option, I agree.


To do otherwise would require scrapping Metamagic in it's current form and/or a massive change to spells (far out of scope than anything on these forums; it would need to be a complete rehaul).
You mean... like a Spell Reformation (http://www.giantitp.com/forums/showthread.php?t=327102)?

[Ziegander]

There are shadows of interesting and potentially awesome design floating around in here that I've never considered before. Which is somewhat sad if you'll consider just how many feats I've written for the d20 system in the past.

Speaking specifically about the Fighter, his feats, and other classes, consider for a moment that the Fighter gets nothing but feats as his class features. He gets 11 of them and 9 dead levels in place of the class features other classes get, which normally fill up all 20 levels of those other classes (caster classes, who get spells every level, notwithstanding). Non-Fighters normally get 7 or 8 feats (if Human or something), and maybe a couple pre-selected bonus feats if they're lucky. The Fighter gets those + 11 more. So, here's the thing: the benefits that other classes get from feats don't actually have to be level-appropriate for those classes. They get level-appropriate power and versatility from their class features. Fighters don't get that. So, really, once we're talking about strictly Combat Feats, they only need to be level-appropriate for the Fighter that's taking them.

What this means is that, if feats are designed in such a way so that the Fighter can gain Whirlwind Attack at 3rd level or 4th level, but a Ranger can't until 12th level, that's okay. The Ranger will do just fine without it because, at least in theory, he has level-appropriate class features that will make sure he's effective without Whirlwind Attack. For other classes, such feats are just fun little things they can do on the side. It isn't, or shouldn't be, the job of feats to give the Ranger level-appropriate abilities. But for the Fighter, feats are the only thing he can do. If he doesn't get level-appropriate benefits from them, then he isn't playable.

So, if what would amount to "capstone" feats for non-Fighters are available to Fighters pre-10th level, then we just need to re-imagine post-10th level play for Fighters to ensure that they get feats that only they could possibly take once they've reached the pinnacle of other classes' feat achievements.


If you choose repeatedly from the same pool of options, every choice gets progressively worse. Feats work the same way. Every feat chosen from the same pool is worse than the previous feat. If your pool of feats is stagnant, each new feat you get is worse than your previous feats. Feat prerequisites don't change that fundamental logic - they just delay it. Suppose that, after eating the cake and the meatloaf, we unlocked ice cream! Yay! But we still have to choose the fruitcake next time we get hungry.

Let's just start with, I think this is a really bad foundation to begin designing feats from. There is just no way this is true unless there is a feat famine or something. Each new feat is always worse than your previous feats? What the hell gave you that idea? What evidence do you have that remotely supports that? Using that logic to inform your design of feats is not only fundamentally flawed, but actively, profoundly damaging to the fun and stability of the game as levels increase.


The only way (that I have found) to prevent this problem is to provide more options with time. That means level-based prerequisites. Now, that doesn't technically mean that high-level feats are more powerful than low-level feats - just different. But you're never going to get 7 equally awesome feats; there's always an order. Since the entire rest of the system gets more powerful with level, I think it makes the most sense to make the high-level feats more powerful.

Now you're contradicting yourself again. You say that feats will always be worse than any previous feat you've taken, that technically high-level feats aren't more powerful than low-level feats, but also that it makes the most sense to make the high-level feats more powerful. Of course the high-level feats should be more powerful than the low-level feats. That's how D&D works. If high-level feats aren't more powerful than low-level feats, then why do they require a higher level? You shouldn't have to worry about getting 7 equally awesome feats. You should be getting 7 feats that not only are totally awesome for the level that you're getting them, but each new feat should, in full-on, total opposition to what you've been saying up to this point, get progressively more awesome than any other feat you've taken previously. Or at least you should have the capability of choosing feats in that way (the option to take a lower-level feat at high-level should always be there if the player chooses to take it).

[Vadskye]

There are shadows of interesting and potentially awesome design floating around in here that I've never considered before. Which is somewhat sad if you'll consider just how many feats I've written for the d20 system in the past.
Thanks! I just hope we're thinking of the same things. :smalltongue:


What this means is that, if feats are designed in such a way so that the Fighter can gain Whirlwind Attack at 3rd level or 4th level, but a Ranger can't until 12th level, that's okay. The Ranger will do just fine without it because, at least in theory, he has level-appropriate class features that will make sure he's effective without Whirlwind Attack. For other classes, such feats are just fun little things they can do on the side. It isn't, or shouldn't be, the job of feats to give the Ranger level-appropriate abilities. But for the Fighter, feats are the only thing he can do. If he doesn't get level-appropriate benefits from them, then he isn't playable.
So instead of building combat feats so they are level-appropriate for every class, you want to build combat feats so they are only level-appropriate for fighters, and bollocks on all the rest of the martial classes? If you do that, you'll have to make substantial rewrites to every non-Fighter martial class just so they stay competitive - and even after doing that, they just lost access to a great deal of customization and flexibility. Instead of fixing every non-Fighter martial class, it would be better to give Fighters actual class features and throw away the horrible "every other level is a dead level" design.


Let's just start with, I think this is a really bad foundation to begin designing feats from. There is just no way this is true unless there is a feat famine or something. Each new feat is always worse than your previous feats? What the hell gave you that idea? What evidence do you have that remotely supports that? Using that logic to inform your design of feats is not only fundamentally flawed, but actively, profoundly damaging to the fun and stability of the game as levels increase.
Um, I think you need to look at the context a bit more carefully. Try it with this bolding:

Every feat chosen from the same pool is worse than the previous feat
That's what life is like if you don't use level-based prerequistes. Of course it's terrible design - that was the point. It was an argument for using level-based prerequisites so your feat pool expands with level, preventing the "I guess I have to eat fruitcake now" problem.


Now you're contradicting yourself again. You say that feats will always be worse than any previous feat you've taken, that technically high-level feats aren't more powerful than low-level feats, but also that it makes the most sense to make the high-level feats more powerful. Of course the high-level feats should be more powerful than the low-level feats. That's how D&D works. If high-level feats aren't more powerful than low-level feats, then why do they require a higher level? You shouldn't have to worry about getting 7 equally awesome feats. You should be getting 7 feats that not only are totally awesome for the level that you're getting them, but each new feat should, in full-on, total opposition to what you've been saying up to this point, get progressively more awesome than any other feat you've taken previously. Or at least you should have the capability of choosing feats in that way (the option to take a lower-level feat at high-level should always be there if the player chooses to take it).
We completely, 100% agree. Read what I said more carefully? I was disagreeing with Mighty_Chicken, who said that feats getting more powerful with level was not necessarily a very good design philosophy.

[Ziegander]

So instead of building combat feats so they are level-appropriate for every class, you want to build combat feats so they are only level-appropriate for fighters, and bollocks on all the rest of the martial classes? If you do that, you'll have to make substantial rewrites to every non-Fighter martial class just so they stay competitive - and even after doing that, they just lost access to a great deal of customization and flexibility. Instead of fixing every non-Fighter martial class, it would be better to give Fighters actual class features and throw away the horrible "every other level is a dead level" design.

I'm talking about current class design, and the theory behind it. If the Fighter gets nothing but feats as class features, then, yes, combat should only be level-appropriate for Fighters, because other classes, in theory, should have class features that make them level-appropriate characters.

I'm arguing that this is a fully acceptable avenue of design that, until now, hasn't actually been respectfully designed for. No, it doesn't work with the game as written, but the game could have been written this way, and it might work. It might even work very well.


We completely, 100% agree. Read what I said more carefully? I was disagreeing with Mighty_Chicken, who said that feats getting more powerful with level was not necessarily a very good design philosophy.

My bad. I didn't get that from Mighty_Chicken's posts. Instead, I got the opposite; that you thought that feats getting more powerful with level was not good game design. Sorry. Indeed, we do seem to agree 100%. That's how D&D is designed, and if feats are part and parcel of level-based advancement, and they do not follow suit, then they are poorly designed. We as informed, intelligent designers can fix that.

And we should be getting paid for it.

But that's a personal problem for each of us to deal with, I suppose. ;)

[Vadskye]

I'm talking about current class design, and the theory behind it. If the Fighter gets nothing but feats as class features, then, yes, combat should only be level-appropriate for Fighters, because other classes, in theory, should have class features that make them level-appropriate characters.

I'm arguing that this is a fully acceptable avenue of design that, until now, hasn't actually been respectfully designed for. No, it doesn't work with the game as written, but the game could have been written this way, and it might work. It might even work very well.
Ah, I see what you're saying. Yes, I think that this was the (poorly executed) way the game was originally designed. You see hints of that in the incredibly powerful tactical feats in Complete Warrior, and in the high-level feats in the PHB2. However, I still don't think it works, because it relies on a fundamentally incorrect assumption. I'll just quote myself:

See, the fighter "hidden feature" isn't actually getting early access to feats locked behind feat requirements. That advantage happens exactly once, at low levels - when the fighter was already doing pretty well anyway. The real feature is having more feat combinations than anyone else. That's a much, much weaker advantage.
Another way to put it is that any single thing a fighter can do, any other class can also do. They just have to invest more of their resources into it. There's no "firewall" that separates the fighter from the other classes, and that's a critical gap. You could solve that gap by creating stupendously deep feat trees that are literally impossible for non-fighters to take, but that just creates more problems.

Actually, if you really want to take that approach, you could use "feat prerequisites" that don't require specific feats, but instead require X combat feats. That would solve the "specific build" problem, and as long as you created very powerful feats that required some ludicrous (10+) number of combat feats as prerequisites, you could give fighters a unique late game while still offering some interesting options to other martial characters, since any martial character could go down the feat "trees" to some degree.

Alternately, instead of having special feats that required huge prerequisites, you could each feat a special scaling mechanic that says "If you have X or more combat feats, you also get this ability." It reminds me of a feat fix somewhere on these boards that simply scaled every combat feat more or less linearly with the number of combat feats you have, thereby granting quadratic scaling overall.

Note that I don't think that the above is the best solution. It encourages one-dimensional, combat-only characters (as if skill feats weren't already weak enough...). Additionally, it's likely to lead to number inflation, since increasing numbers is the laziest way to grant that sort of scaling. A better solution is to just fix fighters with class features properly. But it's functional, and could be interesting to see implemented.


My bad. I didn't get that from Mighty_Chicken's posts. Instead, I got the opposite; that you thought that feats getting more powerful with level was not good game design. Sorry. Indeed, we do seem to agree 100%. That's how D&D is designed, and if feats are part and parcel of level-based advancement, and they do not follow suit, then they are poorly designed. We as informed, intelligent designers can fix that.

And we should be getting paid for it.

But that's a personal problem for each of us, I suppose. ;)
Hahaha. Hey, one of these days I'm going to publish Rise! I just need a proper Monster Manual and DMG. Just you wait... I'll make at least negative however much it costs to print copies dollars.

[Ziegander]

Just you wait... I'll make at least negative however much it costs to print copies dollars.

:)

Good luck, my friend.

A word to the wise: over-cautiousness has never served me well in this sort of endeavor.