On Feats (D&D 3.x)
By Mike Bourke
Recently it became necessary for me to review several collections of Feats for inclusion in my D&D 3rd Edition campaign, and in the process, I was struck by a number of similarities between the Feats being offered from different sources. Reflecting on these common patterns gave rise to some more general thoughts on the nature of Feats in the d20 systems in general. Realising that others might also find these observations, deductions, and opinions to be of value, I have compiled them into the following analysis.
Part I - A general classification system for Feats
Feats come in a number of basic flavours. One way that can be used is "Official" vs "Unofficial" content, but that's too smplistic to be all that useful - the game designers don't have a monopoly on good ideas (or bad ones for that matter). A better system is as follows:
- Enhancement Feats - these take an ability that a character has, and make him better at it. The original ability may be restricted in terms of who has access to it.
- Avoidance Feats - these let a character stop an opponant from doing something they would normally be capable of.
- Extension Feats - these let a character do something they are normally forbidden to do.
- Customisation Feats - these let a character do something extra when they succeed in a given task, but often do nothing to make it easier to carry out the task. If the "something extra" is particularly spectacular, they may even make it harder.
- Clarification Feats - these take a situation that is a little murky in the rules and define, as part of the feat, both a normal way of handling the situation, in addition to containing one of the above types of Feat.
Any of these types of feat can be game unbalancing, giving too great an advantage for what they cost. Some may be uninentionally unbalancing; others are intended to let the characters who have taken them get away with murder. I have seen one feat on the internet, for example, that lets a character subtract hit points from his own total and add them to the damage done by any attack, regardless of whether or not the attack hit. So there is another type of classification system for Feats: "Good" ones and "Bad" ones (from the GMs perspective, of course)!
But which ones are good and which ones bad? How can a GM tell the difference? What standards should be applied, what criteria?
Part II: Good Feats Vs Bad Feats - some definitions
In seeking to define the differences between a good feat and a bad feat, let's start by listing all the possible good things that a feat can achieve. The perfect feat:
- benefits the character that takes it
- is unique
- clarifies rules gray areas
- extends or enhances campaign background
- does not assume campaign background elements that may not be in place
- individualises characters
- applies in a variety of situations
- is specific in the benefits that it brings
- is simple to apply
- is consistant in level of benefit in comparison with existing feats
- has some other redeeming value
While it's fair to say that a feat that does not achieve ANY of these is a Bad feat, there are obviously going to be gray areas in between - feats that have some, but not all, of these characteristics. I have yet to see the perfect feat! That said, a feat does not have to be perfect to be perfectly acceptable. It follows that any proposed feat has to be considered in terms of each of these criteria, and that failure in any specific respect may or may not constitute a downcheck for the feat.
Part III: Assessing Feat Criteria: 11 Questions
Does the feat benefit the character that took it?
If a feat holds no value to the character, they won't take it, that much is self evident. But there is an equally-important distinction to be made here - there is a HUGE difference between benefiting the character who takes the feat and penalising everyone who does not, and that difference is not as obvious without a little deep thought. Ultimately, a feat of the latter type encourages everyone to take it, making characters less distinctive. Even a feat that everyone of a certain class or type has to take goes too far in this respect. No matter how good a feat might be in other areas, a failure with respect to this criteria is an automatic rejection, in my book. The rejected feat, if it has enough other virtues, might provide the inspiration or even the foundation for an original feat, but in its original form, the feat is as dead as a dodo.
Is the Feat unique?
If a new feat is exactly the same as a pre-existing one, why do you need it? That's the obvious question. Once again, though, there is a deeper issue to be considered in this section, and it derives from the obvious answer to that question - that you can only take most feats once. Game systems break when pushed too far (like most things in life); that's the reason that restrictions on the number of times a feat can be taken exist in the first case. Having a feat that gives exactly the same benefits as another feat pushes the system, potentially beyond its breaking point. Once again, this is grounds for an immediate rejection.
But, there are cases where this is not quite so clear-cut. A feat that gives a benefit under specific circumstances is not the same as one that gives the same benefit under different circumstances, or is it? This is one case where the devil is in the details, particularly if a third feat, or class ability, or whatever, lets the character change the circumstances so that they overlap. If the referee is completely convinced that the two sets of circumstances can never overlap, then despite the same benefit, the two feats are unique and distinct. If the referee is uncertain, he has a harder assessment to make. His choices are to permit one feat but not the other; to permit both and risk an overlap; or to actually amend one or both feats benefit descriptions to specifically prohibit stacking of the benefits.
These are not easy choices. If the arguements given above are held as valid, the "take the risk" option is obviously unsatisfactory. If only one of the two is to be excluded, which one should get the green light and which should be erased? That answer can be a relatively easy one if the circumstances of one feat completely contain those of the other - the choice is then between a broadly-appliccable feat or the more specific version - but this is not often the case. In general, it's better - and far easier - to adjust both feats to specifically state that the benefits do not stack with each other, and then reasess them both. In addition to the usual questions, there is one very specific question that should be asked in this circumstance:- "is there any clear benefit to taking one feat over another in the general case"? A yes answer again gives a reasonable basis for rejecting one feat over the other. A no answer permits each feat to be considered on its merits.
Does the feat clarify the rules?
Some readers may ridicule the entire concept of a rules clarification existing within a feat. The rest of us know better. A good example is a feat from Fantasy Flight Games "Monster's Handbook", which spells out a procedure for Playing Dead. When a feat offers a rules clarification as part of it's description, the first question to be addressed under this criteria of judgement is "do you agree with the rules clarification, or do you have a better way of handling this subject?". Work very hard at answering this question objectively; it's very easy to fall in love with a house rule because you wrote it! If you don't like the feat's suggested rules change/clarification, then you have only two choices: Change the feat to suit the way you're handling that aspect of the rules, or reject it out of hand. If, on the other hand, the suggestion embodied by the feat is an improvement on the current situation, or is something you hadn't even thought of before, then the feat can be assessed on it's own merits - but the fact that you already like the way the author's mind works is certainly an encouraging sign.
Too encouraging! It can be easy to overlook flaws in the feat if you've been prejudiced by the rules clarification. The correct procedure in this case is to completely reassess the feat, under the assumption that the rules clarification is already in force. (Heck, if the clarification is that good, the change SHOULD be in force immediatly!) It might be that the clarification was written on a good day, suggested by some other GM's handling of the situation, while the feat itself was written on a bad day. By taking the rules clarification out of the picture, you can be more objective about the rest of it.
However, where a feat contains a clarification that makes sense, it's a good indication that there are grounds for SOME sort of rules dealing with the subject. So even if you don't use the original, you should make the effort to have SOMETHING, even if you have to write it yourself.
Does the feat extend or enhance the campaign background?
Once again, this is not something that many feats embody. But there are a few that have this attribute, and when it exists, it should be considered extremely carefully. One example is a feat that makes translating spells from one spell book to another when the mages are of different races. Right away, this feat extends not only the campaign background but also the rules that interpret that background into game mechanics, - stating that by default, mages can't translate spells from one race to another, which implies that each race has it's own unique style of magic. Why, it's even possible that each school of magic actually originates with a different races' mages! The implication is that there should be some sort of racial modifier to spells of the given type when cast by the race in question - again, something that can easily be expressed through the introduction of a feat available only to members of that race, or a feat that lets others gain the same advantage as the race in question. What if a race's pre-existing modifiers are merely the most overt and obvious manifestations of these potentials? What if these facts have been lost, but are awaiting revelation in some long-lost tome? This one hypothetical feat has sparked enough concepts to completely reinvent the campaign world!
In fact, I have seen several "new character classes" and even "new races" that could be better described as existing types with one or two additional feats, something that really shows the power of the Feat as a conceptual tool. A combat-oriented character class with lots of abilities pertaining to the riding of Dragons, which is how this class travels, and which is it's sole point of uniqueness? Why not just describe the pertinant class abilities as a series of Feats, and list membership in the social organisation "Riders Of The Wyrm" as one of the prerequisites? Suddenly, Wyrmrider Mages and Wyrmrider Clerics and all sorts of other combinations become possible, extending the background far beyond the original fighter variant). Several of the feats in the various Faerun supplements fall into this category.
Assessing this kind of feat can be extraordinarily difficult, because they connect to the campaign and the rules in so many ways and in so many places - so much so that maybe its a subject for a whole seperate article sometime! But, in general, the principles used to assess rules clarifications can be brought to bear here. The first step is to extricate the feat from the campaign elements, and to assess each on their own merits - if that's possible. You are under no obligation to accept the whole package!
Either way, lets start by looking at the campaign elements. Do they contradict things already established within your campaign world? Do they contradict plans you have for the future? A "Yes" to either question is an almost-certain repudiation of the background elements at least, or of the whole package if you were unable to extricate the rules elements.
Rules issues should be considered for all feats, whether or not they fall into this particular characterisation. Broadly stated, the rules elements are the proposed amendments to the art of the possible within the rules, and the consideration that comes under this criterion's heading are how those rules integrate with the campaign and its history. Does permitting the feat enable or a contradiction in campaign history, for example by opening the door to a more sensible solution to past problems? For example, a feat that permitted mass healing of disease by clerics would fail this test if a plague were an integral part of the campaign history. Does the feat create an easy escape clause from a problem that is currently endemic to your campaign world, or that is central to current or future scenario plans? Feats that can cause these sorts of problems vary from the obvious to the very subtle. A feat that prevents misunderstanding of magical communications is trivial, so narrow in scope that it will certainly fail later tests - but if you intend to have a war start because of a misunderstood communication, the existance of this feat will certainly complicate the GMs situation. In a nutshell: is the campaign world better off if this feat is excluded, not because it's inherantly bad in general, but because it's specifically bad for this campaign? If yes, regardless of the acceptance or otherwise of campaign elements contained within, or inspired by, the feat, it's GONE.
But it's always worth noting these things - in writing - for future referance and for future campaigns, and because thinking about them helps inspire new ideas - for campaigns, for scenarios, for character motivations.
Does the feat mandate the inclusion of new campaign elements?
In some ways, this has already been addressed in the previous discussion, but the forced inclusion of campaign background elements is one of my pet peeves . To a very large extent, this is my biggest criticism of the D&D rules in their current form; each racial description contains sociology and personality traits that are almost inextricably entwined with the rules. Ditto each class description, even more so. These are things that should be in a worldbook or campaign setting, NOT the core rules. But, setting that to one side, if a feat makes certain behaviours easier or makes sociological assumptions, for example by assuming (or stating outright) that elves - or dwarves, or whatever - have certain abilities or attributes, it deserves to get a more rigourous scrutiny. If I don't want mages to be able to do create scrolls, I do away with the scribe scroll feat - which means that any feat that lists scribe scroll as a prerequisite, or that has anything to do with scribing spells on or into anything, is immediatly suspect.
The questions raised in this section are more about the underlying assumptions of the feat and the applicabilty of those assumptions to the campaign world, than they are the more overt ingredients discussed in question 4. The time to consider those issues is when the review from question 4 is still fresh. The hard part about this area is identifying those underlying assumptions. These essentially come down to two areas - the requirements to be met in order to qualify for the feat, and the circumstances under which the feat will be beneficial (which is not the same thing).
So let's look at the requirements. What do you have to do to meet them? Is a feat required that is not normally taken by the character class? Or a skill? Will the feat be useless or unneccessary or redundant by the time the qualifications are met? And, ultimately, what assumptions about the campaign world do the requirements make, and how viable and appropriate are those assumptions?
And then there are the circumstances under which the feat will be beneficial. Is there an alternative that could be called on that would be equally valid on such occasions? Is the feat only useful for a given race, or a given class - or a given class from a given race? Alternatively, does any character meeting the requirements for traking the feat automatically gain benefits ALL THE TIME from the feat? And finally, one of the most significant questions of the section: does the feat change or eliminate one of the defining characteristics of the race or class?
The answer to any of these questions COULD be enough to reject a proposed feat out of hand; it could be reason enough to accept the feat as written; but, it mostly likely points out flaws in the design of the feat. These flaws could kill the notion, but if they don't, they will define what if anything needs to be amended or clarified at this point of the review process in order to make the feat acceptable.
Does the feat individualise characters?
Which comes first, the chicken or the egg? In so many games, characters are more distinctive at low levels, when they have differences in what they are good at and what they aren't; as the characters grow in experience, they gradually grow towards archetypical examples of their character class, good at everything the class represents and at least capable in others.
To some extent, adding more feats to the list of choices available solves this problem, or seems to, and that's one reason why every GM worth his salt is always going to at least look at any new feat that gets put under his nose. In reality, this is only deferring the problem - possibly to beyond the planned retirement point of the characters, but possibly not. It's always easier, in some ways, to craft adventures for established characters, whose interactions can be predicted, whose capabilities are known, whose responses can be anticipated.
Some feats encourage specialisation, and hence individuality. In particular, feats whose value to a character is based on that character's ability in one very specific area; and to some extent, given a limited pool of characters in play at any given time, these work to create individuals. After a while, though, you begin to notice that these are simply archetypical subtypes of a given character class with different personalities; the capabilities are the same. Combating these tendancies and trends are the concept of Prestige Classes - classes that, like feats, have significant requirements in both a gameplay and a characteristics sense. But the introduction of a new Prestige Class is a Very big deal. EVERY feat must be rescrutinised in the context of a character of the Prestige Class type, for example, because ANY prestige class brings with it the risk of game imbalance. EVERY Prestige Class carries inherant extensions to the Campaign world, introducing new character dynamics, organisations, reactions to the same, and so on - all the baggage given such tight scrutiny under criteria 5 and 6.
As a side note, a lot of players assume that if it's in the DMG or some other official publication, it is automatically available, and that if they meet the game mechanics requirements, all they have to do is take a level in the Prestige Class. One even once suggested that doing so then made it the GMs responsibility to put in place the implied social requirements, ie the gameplay requirements. Disabuse players quickly of any such notions! Any prestige class must be explicitly permitted or refused in advance, and the GM should ensure that players have to actually work towards achieving one, in more than the game mechanics sense. The players should plan their advancement, and make sure that the referee knows their plans, so that if a particular requirement has to be met, he can put the potential for it in place; and if a prestige class is not permitted, the referee can ensure that it is quite impossible for ANYONE to meet the requirements, so that there can be no accusation of bias in letting one character have something and not another. But, I digress....
Some feats go further than simply encouraging specialisation, they encourage individuality. Feats that can have multiple different effects, amongst which the character has to choose at the time of taking the feat, for example - if it can only be taken once. Feats that explicitly prohibit certain combinations of classes and feats, or that specifically prohibit the taking of alternative feats. Feats that use their requirements to mandate the taking of other feats. Make no mistake about it, restriction may be irritating - but it's a good thing in the long run. If you have 7 character classes, each of which has 3 or 4 archetypes, each of which has 3 different character development paths in terms of prestige class opportunities and 8 different paths in terms of a linked series of feats over many levels, you have 252 or more distinctively different character types even before races and personalities are taken into account.
While that much variety is almost certainly a pipe dream, GMs can work actively to create it. Whenever you look at a new class, prestige or otherwise, consider making its class abilities a linked series of feats. You can use prerequisites to exclude certain classes or races; or to mandate that the feats individualise a certain race or class, to control the pace of development within the feat series, to effectively create a new archetype within some or all of the character classes. And, of course, there are some feats that already do this.
As usual, there's a downside, some danger to watch out for. In this case, beware of feats that let one character class do something that is the unique province of another character class. Don't let it, in other words, act to reduce the individuality of characters. (The same thing applies to a Prestige class!)
Such feats are always worth close examination, and considerable effort on the part of the GM to make them compatable with the rest of the campaign. So far, a lot of these criteria have been focussed on the negative, examining reasons for rejection. Feats of this type, in contrast, demand that you look for ways to keep them. That's not to say that they can't be dangerous in terms of making the character TOO good at something - or at too many things. Which brings us to the next criteria:
Does the benefit apply in a variety of stuations?
The more narrowly focussed a feat is, the more dangerous it can be, by pushing part of the system, or the roleplay, beyond breaking point. If a character makes a substantial investment in being good at something, they will try and use it as much as possible. They will try to change circumstances into situations that permit the use of the ability, because it gives them an edge. The personality of the character will slowly change, to justify all this - and the new personality will then be used to justify further enhancing the character's capabilities in the given area.
As a rule of thumb, the more narrowly defined a benefit, the greater that benefit becomes. I have yet to see any feat take this trend to a blatantly absurd extreme, but I have seen some come close. Where a skill can be used for several things, but the character only cares about one of them, any feat that enhances that skill's use specifically and explicitly in that way risks issues of game balance.
A freind of mine often makes the point that if you ask a referee 10 times for something that gives them +10, they will probably say no, but if you ask for ten different things, each of which gives +1, they will probably get at least 3 of them, and possibly more. Throw in rephrasings and alternate justifications, and over time, the character will get that +10. Or more. That's the danger of narrowly-defined benefits.
Whenever I examine a feat that is so narrowly defined that it will be useless most of the time, I always ask myself what I'm missing. Is there some class ability or racial ability, or whatever, that will, or can, become truly overwhelming in this situation? I was once presented with a proposed feat that enabled a cleric character to use his ability to turn undead to "turn" (ie make flee in fear) the living worshippers of deities to which his deity was opposed, provided that he was standing on ground hallowed to his deity, by making "the wrath of God" aparrant to the foes. But, the mage of the party was carrying items that enhanced any fear attack that the character made, and had a feat that let him synergise with the cleric to add his INT bonus to any other character's skill checks and rolls. And another member of the party had a class ability that automatically sanctified the ground for a given radius around him. (NB: I was not the GM). Add all of that together, and you can soon reach the point where on any roll other than a 1, no sentient enemy could come within 60' of the party if the party weren't ready and waiting for them. On the surface, the feat presented was not all that unreasonable, but the compound of effects was such that the party would quickly become near-invulnerable. They could loot dungeons and ruins with near-total impunity to gain the treasures within, treasures that would only add to their overwhelming might. Even taking one of the three elements of the coctail produced an ability that was just too strong to be permitted.
Particularly dangerous are feats that introduce a new type of bonus, or a bonus of an unspecified type. Be especially wary of bonuses that imply that they are of a given type or for a given reason; as a rule of thumb, assume that these will be interpreted by at least one player in the most favourable way possible. This is what makes the Bard class so dangerous, as several GMs have found to their utter despair - they generate Morale Bonuses, a new category of bonus that therefore stacks with everything else. Is a "blessed weapon" bonus the same as a "sacred" bonus? Assume the worst and act accordingly - Beware the over-specific!
Is the feat specific?
Having waxed reasonably eloquent (I hope) on the dangers of feats that are too specific, we have the other side of the coin: feats that are too general. One website I have visited contained a feat that gave +1 to everything per level, provided that the character remain honourable - with scant definition of what that meant. The requirements mandated that only characters of 10th level or more could take the feat. A strict interpretation of the feat as written meant that on taking it, the character would get +10 to all rolls and checks - saves, skill checks, attack rolls, damage rolls..... the list goes on. Looking more closely at the context of the feat, it became clear that this +1 was to apply only to levels of a specific 5-level prestige class, suggesting that things were not so bad as first thought - but this is still a HUGE benefit to ANY character. Too big, in fact!
There are two good reasons to reject feats that are too broad (of which the above is an extreme example). The first is that the more broadly-defined a benefit is, the more likely it is to stack with other benefits from other feats - and that is something that you have to keep a very close eye on, as explained previously. The second is that it acts to reduce the uniqueness conferred by more specific feats. Instead of a +1 on all attack rolls, why not half-a-dozen feats each giving +2 to a specific attack manouver, or to a specific weapon type? The more specific a feat, the more it encourages the creation of other feats. Reject the generic!
Does it overcomplicate things?
This is my weak point, I have to admit. Things that look completely functional on the page turn into a dog's breakfast when used in play. For example, I introduced a skill called Piety that was, to a priest, the combined equivalent of a spell points system and a measure of how faithful the cleric had been to the tenets of his faith. On paper it looked fine, and I was preparing a list of feats that would enhance various aspects of its use in play; but in play, it bogged things down terribly, the players hated it, and it has been a miserable failure. As a result, the Piety skill - and the feats - are about to join the scrap heap.
As a means of self-control in this respect, I have come up with a rating scale for how much additional complication a rules change incurs. Modifying that scale for feats gives the following:
||add a fixed number to a reasonably static number, every time.
||add a fixed number to a die roll, or a die roll to a reasonably fixed number, every time.
||if a die roll required is one that has to be made anyway
||for each circumstance or condition that must be met for the action to take place that will not AUTOMATICALLY be satisfied by everyone who qualifies for the feat
||for each circumstance or condition that must noy apply in order for the action to take place
||every additional die roll required
||each thing that the feat or ability lets you do that you normally can't do
||each thing that the feat or ability stops the opponants from doing that they could normally do
||every thing that you can do because you have the feat that needs additional explanation
||every additional variable that has to be tracked and modified, eg Piety; -2 if it's one that has to be tracked and modified anyway, eg Hit Points.
When you rate a feat for complication, the lower the score, the better in terms of simplicity, but the more prone to being overgeneralised the feat is. Middle scores then to be a little more complicated, but are more likely to have substantitive issues within the text, or to be excessively narrow in definition. High scores tend to have real risks in terms of the substance, real issues in terms of the assumed facts, real dangers and rewards in terms of additional campaign elements, and extreme concern over the complications that will be introduced. Any score over 10 gets my alarm bells ringing.
But the level of complexity is not enough, in and of itself, to reject or approve a feat. These results must be interpreted in the context of what contribution the feat makes. No matter how good feat might be, at anything over twelve points I would look at simplifying the proposal.
Is the feat Excessive?
A number of the issues raised have had elements of game balance concern. This section is the bottom line on how good the feat is, based on the benefits given by preexisting feats. For example, the Players Handbook lists a feat that gives a +4 modifier to one specific skill, and a feat that gives +2 to two different but specific skills. It has feats that give +1 to specific kinds of rolls, like attack or damage rolls, where there is only the one type of roll to be made under the given circumstance, and others that give +2 to rolls where there are multiple different rolls that could apply, like saving rolls. If a feat lets you do one thing that's normally forbidden, under one specific circumstance, it has reasonably low requirements; if it lets you do multiple things or has multiple circumstances under which it applies, it usually has at least one and more often 2 other feats as prerequisites. If a feat is specifically more useful to a given character class, its power level reflects the rate at which characters of that class gain levels, and hence, gain feats - fighter feats tend to be less spectacular, individually, than mage feats - unless they have requirements that are extremely difficult to meet. Metamagics will typically allocate a low level spell to the sort of power level where you might expect to find a spell of the given power level once the metamagic is taken into account - a doubling of one simple numeric value is +1 spell slot, eliminating a requirement is 2 slots, replacing a variable with a fixed number equal to its maximum is 3 slots, and completely overcoming a major restriction like casting time is 4 slots.
Using these pre-approved feats as guidelines, it can soon be determined whether or not the feat is excessive. The +1 per level feat mentioned in section 9 clearly violates these guidelines - a sure sign that it needs amendment or outright rejection.
Specific mention should be made at this point of Feats that enhance a single aspect of character activity all the time, especially those which offer an advantage in determining initiative. Since Initiative is only rolled once per combat, it can be suggested that feats which add +1 in this area are less powerful than a feat that adds +1 to an attack or damage roll. Don't believe whoever is making the suggestion! The tactical advantage that can result from always or even frequently going first in combat is worth AT LEAST +1 on each attack - the correct method of analyzing the benefits of such feats is 'how large a combat bonus can result from permitting the character to act first?' Going first may permit the character to reach cover, conferring up to 90% automatic miss chance when attacked by missile weapons. Consider how much additional AC the character would need in order to reduce the chance of a hit to 4/-, and a different value can be perceived. Then consider the reduced effectiveness of mages and the like due to damage that may be inflicted before they can act even once. Then consider the additional options that the character has in combat by virtue of acting before the opposition can do anything to stop them. +1 to initiative is substantially more massive than a mere +1 to hit, which can easily be countered by any of innumerable adds to AC. Worse, feats that add to initiative always seem to stack, and never seem to have ANY conditions that have to be met - they are 'always on'. Another way to look at +1 to initiative is to count the number of levels that a character needs to achieve in order to get +2 stat (and hence +1 to the stat bonus) - usually 8. How many +1s to the base combat value does a character get in that many levels? Again, usually +8, and an extra attack in some rounds to boot!
It's also possible that a feat is underpowered, not giving enough to justify its being taken by a character until they are running short on ideas and should be thinking about retirement!
Does it have some other redeeming value?
I won't pretend to have covered everything in this discussion, and certainly not to have covered the issues exhaustively. There is always the potential of a feat having some other virtue that I havn't thought of. For example, it might make the GM's life easier in some fashion. It might explicitly ban something the GM doesn't want anyone to be able to do, or make possible something that the GM wants people to be able to explicitly do. One example is a feat I created that lets characters creating undead give those undead additional character levels - at the expense of their own. It might encourage something the GM wants to encourage, or discourage something that he wants to discourage. All these are virtues that might not necessarily be covered in the analysis thus far.
Part IV: The final assessment: Approve, Reject, or Modify?
So, having looked at just what value a feat has, and what the penalties for inclusion of that feat are, you are now in a position to make a rational choice about that feat's inclusion. Some of the sections raised the potential for positive aspects of a feat - new character development options, new campaign directions and ideas, and so on. Others discussed grounds for objecting to a feat.
In making a final determination, I will go over a feat for a second time. If the feat has something positive to contribute under a given heading - be it a rules clarification, a new background element that I like, a character development path, or simply scope for character individualisation, or whatever - I give it a + sign, one for each. If a given section dwells on the negative, grounds for rejection or modification, and the feat falls fowl of one, I give it a minus sign, one for each. If a feat has no plusses, its gone. If a feat has two or three negative signs more than it has positives (or worse), it's probably not worth the effort to salvage - extract any worthwhile bits and use them in some other way. But, if it has some virtue, and only a couple of negative factors, I look at the complication score. If that is high, then again, steal the good bits and forget the rest; if it's not, then perhaps it's time to start tinkering.
Let's be frank - almost any feat sourced from anywhere beyond the PHB will need tweaking. There are exceptions, but they are just that. Each minus sign is a weakness in the feat's design that you should try and eliminate. If it's too general, make it more restricted; if too strong, reduce the benefit; if too complicated, try and reduce the work involved. Look at other ways of handling a feat - for example, using combat bonuses in the same way that metamagics raise the spell slot, to confer some advantage in battle, instead of making something possible outright. Do the same thing with skill DCs - raise the DC required but let the characters do something special if they succeed. Throw incongrous aspects of the character together - a feat that is triggered when your attack bonus is higher than a character's INT bonus. Look at building choices into feats - and consequences of those choices. Try to base a feat on the things that are happening anyway under those circumstances, instead of adding an extra task.
And leave yourself open to ideas. For every 3 feats you approve, even if you have had to modify them to make them acceptable, you will probably think of another one all on your own. While feats are normally player-centric, think about specific feats and how they might apply to NPCs, or to creatures of various kinds. That having been said, never permit an NPC to have anything that you would be unwilling to let a PC have!
Part V: Closing Thoughts: some unusual feat applications
Feats can be taken by just about anything. Most GMs think of imbuing a feat into a magic item sooner or later; but how many think of imbuing a magic item with a feat that empowers the item, not the character weilding it? Feats that are only useful when lots of characters have them, for armies? How about a Druid's Grove? Or a mountain range? Castle Walls?
Any feat can be treated as a magical effect - and that gives rise to a mechanism for using feats in unusual places and unusual ways. Creating specific feats for these different applications - assuming that everything is sentient on some level - can open up new worlds of thoughts and ideas. Caradras, The Cruel - with "Summon Snowstorm" as a feat? Why not?
Why not, indeed....