Feats Not Created Equal

[Solirion]

Y'know, I only took one probability course, so I'm a bit hazy on the technicalities - maybe hong could fill us in here, I don't know. But if I recall correctly, if you've got a bunch of things weighted towards a central value (in this case, feats with an average power value), you're pretty much guaranteed of a few things. Some will be below average, some will be above average, most will be around the median. The more sample points you have (more feats from more supplements), the more likely it is that some will be more significantly above or below the central point.

I think what you are going for here is an application of the central limit theorem.
Unfortunately, the CLT relys on a number of assumptions on the random variables in question. One would be, that all the random variables (e.g. feat power levels in this case) are independent and identically distributed. In order to be sure that the requirements are met, the first thing you would need is a proper definition of the concept of 'feat power level' and an argument for the independence assumption. Without that any conclusions based on probability theory are dubious at best.

 

[Psion]

As tends to be the case with classes, I find feats very context dependant when it comes to power. For example, I didn't see one feat a character took -- a dragon blood feat that let him hold his breath for a long time -- to have much impact.

I totally forgot about the feat until the feat made a combat I expected to be challenging (a blood wierd) rather trivial.

Not that I have any problem with that. I find it lame when you take a feat and it never does anything for you because the GM figures that it's not worth throwing an encounter at you because you are prepared for it. I like players to feel good about their feat investments.

 

[Fast Learner]

Given that only extraordinarily rigorous mathematical analysis of feat power levels could tighten the distribution (which isn't likely to happen), and having more options is, in general, a good thing, I find it pretty hard to complain about the current situation with the number of feats available and the power levels they encompass.

No, such analysis couldn't tighten the distribution, at least not meaningfully. Tell me, which feat is better, Power Attack or Alertness? Figure out some kind of mathematical model for attack/damage and skill success or something, then put the characters in a hallway where a trap is about to go off and drop a block on the group. Now which feat is better, again?

I totally agree. I never liked it that, unless you had Spring Attack, you couldn't move after you attacked.

I take it you mean "if you moved before the attack," since obviously you can move after an attack just fine, if you haven't already moved.

 

[MerricB]

Feats allow variation of characters and monsters. You don't need to know about what they do until a PC or monster has one, and that is likely to be a very limited number in the course of the campaign.

As with others, I dislike feats that represent what should be standard combat manuevers, preferring feats that remove penalties (Improved Grapple and similar feats are good examples of what I prefer).

Mostly, however, feats are fine.

Cheers!

 

[Impeesa]

No, such analysis couldn't tighten the distribution, at least not meaningfully. Tell me, which feat is better, Power Attack or Alertness? Figure out some kind of mathematical model for attack/damage and skill success or something, then put the characters in a hallway where a trap is about to go off and drop a block on the group. Now which feat is better, again?

I was actually more inclined to say it's impossible, as you assert, but theoretically one could take into account all those variables (relative frequency of use, and the difference it makes in the outcome of an encounter, etc). It's just highly unlikely, which I did say. I think you're basically agreeing with me and don't realize it.

--Impeesa--

 

[Duncan and Dragons]

Maneuver Feats

Personally, I just wish there weren't any maneuver feats. Everyone should be able to cleave and whirlwind attack. The feat should just mitigate penalties associated with using the maneuver. Maneuvers could still have prerequisites, but anyone meeting the base prerequisites should be able to try the maneuver. Whirlwind attack is a -5 penalty an attacks made against all targets within reach. Cannot be attempted until BAB 4+.

I just bought the Book of Iron Might from Malhavoc Press. Although I have not finished it yet, I bought a lot of D&D stuff, I believe this system allows charcters to do most maneuvers 'unrestricted'. There are just penalties such as AoO or penalty to hit (like -20). Feats help you over come the penalties.

 

[seankreynolds]

I'd much rather see the system cater to weaker characters.

Actually, it does ... the CR system assumes a party of four characters, specifically these four characters: Jozan, Mialee, Lidda, and Tordek. Those four characters are designed sub-optimally (some of them have really crappy skill and feat selections that don't do them any good in combat) because the core designers didn't want to assume that everyone who played would be a master minmaxer. Instead, they assume characters that are weak-to-average for their character level, so your average PC can face CR-appropriate challenges without fear of being squashed just because they're not expert character-tweakers.

 

[Chaos Drake]

Yes all feats are not created equal. Feats only correspond to one 'power level' and it's almost impossible to balance them out. Look at Power Critical and Improved Critical, for example. Does a +4 to score a Crit correspond to a doubling in threat range? Hardly.

PS: I did a PDF called The Book of Half-Feats to tackle this, and some folks have liked it. Not trying to toot my own horn, but I think there definitely needs to be something in between, especially if some feats are taken for character flavor.

 

[hong]

I think what you are going for here is an application of the central limit theorem.
Unfortunately, the CLT relys on a number of assumptions on the random variables in question. One would be, that all the random variables (e.g. feat power levels in this case) are independent and identically distributed.

There are versions of the CLT that don't require either independence or identically distributed RVs. IIRC the most important requirement is that the 2nd moment is finite. If you have that, then with some additional, very broadly applicable assumptions, the result follows.

In any case, what Impeesa said has nothing to do with the CLT. Conceptually speaking, you can think of feats as potentially having a very wide range of power. The more feats you observe, the more likely you are to see some that are extremely low or high-powered, just due to random chance. This has nothing to do with mathematicalese, and everything to do with common sense.

 

[Greylock]

In Reply To Tessarael...

While I agree wholeheartedly that a DEX bonus shouldn't require a feat to substitute bonus', several of the feats you mention really aren't weak.

The char I'm playing right now has, from your above list, Armor Proficiency Heavy (as yet unused), Dodge, Mobility, Exotic Weapon Proficiency, and Mounted Combat, among others. Ahhh, the Fighter class.

Mounted Combat has proved suprisingly successful as a stand-alone feat, much more so than I imagined when I took it as a pre-req. for other feats. Haven't even followed the feat chain yet (can't decide where to go now: weapon specific feats, Iron Will, or RBA and the MC tree).