|
(This was originally posted in the Racial Feats Category -thread)
I was taking a look at those 4E Racial Feats, and that got me thinking: Feats you pick at 1st level should matter at higher levels, too. Too many times that Iron Will or Dodge has felt pretty insignificant by the time my PCs have hit Level 10 or 15.
Here's the thing: what if Feats in PF could be organized into "Feat Trees" so all of them could "upgraded" via the following simple and elegant mechanics:
BASIC level: +1 to two rolls/+2 to one roll.
IMPROVED level: Once per day you may reroll any roll related to the abilities given in the basic level of the feat.
GREATER level: Your bonus from the basic level scales with your level, so that you get a bonus of +LVL/4 to any rolls related to the abilities given in the basic level of the feat. This bonus stacks with any bonuses from the basic level of the feat (to the maximum of +6 at 20th level).
For example:
* * *
*DODGE (Combat)
You have mastered a defensive stance that allows you to easily react to your opponents.
Prerequisites: DEX 13
Benefit: You receive +1 Dodge Bonus to your Armor Class until your next turn.
**IMPROVED DODGE (Combat)
Your mastery of defense allows you occasionally to evade blows in ways that border on miracles.
Prerequisites: DEX 15, Dodge
Benefit: Once per day you may call for a re-roll on any attack roll that would have hit you. You must accept the result of this second roll, even if it is higher than the first one.
***GREATER DODGE (Combat)
You have perfected your defensive stance to the degree that you easily evade almost any blows from your opponents.
Prerequisites: DEX 17, Improved Dodge
Benefit: You receive an additional bonus of +1 bonus/4 levels to your Armor Class until your next turn. This additional bonus stacks with the bonus from Dodge.
* * *
This way your Feats would truly matter at higher levels, too. If three feats per "Feat Tree" feels like too much, another possibility would be to just make *all* Feats that give bonuses to function like the "Greater Level" in the suggestion above -- i.e. the bonuses would scale up with your character level. That way when you pick Dodge, it grants you +5 to your AC at 20th level.
Any thoughts? Which way would be better -- "automatically" scaling Feats (the paragraph above) or these "Feat Trees"?
|
|
Are these automatic improvements or do characters still need to spend feat choices to acquire them? If it's the former, good. If it's the latter, it's bad. Players would rather choose feats that actually matter at high level - the once per day re-roll is too weak (once per encounter might be better but I see it failing at high levels when pitched against full attacks) and the bonus to armor class continues the current state of score wars (I think the option of spending an immediate action to make a move when attacked is a far better ability that would matter at high levels).
Either it's automatic (which I provided as one of the options) *or* you have to pick the first two "lesser" Feats in the Tree to get the bonus that "scales up" with your level. I'm not sure if the former is better -- if you automatically get +1/4 levels to your AC just by picking a single Feat (i.e. Dodge), it might be too good (especially if *all* such Feats scale up automatically) and completely throw the game balance out of the window. Therefore, I would suggest using the "Feat Tree"-system to get all the benefits -- every PC gets a lot more Feats in PF than in 3E, so this would allow you to "specialize" more than before. And similar Feats (Improved Two-Weapon Defense, for example) were far less powerful in 3E.
As for the re-roll being only once/day -- that's how the other "similar" Combat Feats work, too (e.g. Devastating Blow). Note that re-roll is actually pretty good -- you could announce that the DM must re-roll a Threat or confirmation roll, which makes it pretty handy. I would not suggest implementing any "per encounter" stuff into PF, because that's how it works in 4E (and frankly, I don't like it all).
|
|
Maybe three "tiers" or "levels" for each Feat would be too much? Maybe it would work best if you have the "lesser" version of the Feat (which provides a static bonus, such as +1 Dodge Bonus to AC) and the "greater" version (which provides a dynamic bonus, e.g. +1 Dodge Bonus to AC per 4 levels). Naturally, the effects would stack.
My suggestion would be that any +1 "base" bonus would scale up by +1 per 4 levels, and +2 by +1 per 5 levels. As for feats that apply bonuses to two different abilities, I'm not sure if there even should be a "scaling" bonus (i.e. "greater version")? For example, if a Feat grants you +2/+2 on two different skills, it is undoubtly more lucrative to pick such Feats instead of simple 'Skill Focus'. Or if you get +1 Dodge Bonus to AC *and* +1 to REF saves, it would definitely be broken if *both* of them would scale up with levels.
All in all, I don't think this would make characters too powerful, because the effect would scale with levels. For example, if you took 'Iron Will' (+2 to Will saves) and then 'Greater Iron Will' (+1 to Will Saves/5 levels), you would still end up with +6 at 20th levels -- exactly the same bonus as taking 'Iron Will' and 'Epic Will' in 3E.
I was also thinking of Feats (mentioned in my original post as 'Improved' versions of the Feats)that would allow you re-roll or Take 20 on any roll once/day -- would they work better as Racial Feats (such as 'Fortune's Favorite' for halflings, which would allow a halfling to re-roll any saving throw once/day)?
Any thoughts?
Here's the thing: what if Feats in PF could be organized into "Feat Trees" so all of them could "upgraded" via the following simple and elegant mechanics:
There is little question that the 2 most broken aspects of the 3.5 system were skills and feats.
I think your approach has merit, though it will change a number of feats. I'd want to playtest this, but offhand my sense is that the limited number of feats characters get might make it OK to do automatic progression. BUT, that automated progression would have to be by level. I'm inclined to make it 1st, 7th, 14th, which fills some traditional dead level slots nicely and seems about right in terms of power.
Advantages:
* When you take a feat at 14th level, it's guaranteed to be immediately useful, not just a bleah pre-requisite.
* It follows the same meta-logic as the school and domain powers, for system consistency.
* It may go a long way, all by itself, toward fixing the fighter. Right now, their power is built of "feat trees" that only they have the feat numbers to take. With automatic progression, their advantage really shines because now they have all kinds of cool abilities to draw on. Casters have lots of cool spells, Rogues have lots of special class abilities, Fighters have lots of feats to draw on and each does something cool at the fighter's level.
* This approach also boosts humans, whose additional abilities are otherwise a bit weak even in the new Pathfinder system because one more feat isn't a huge deal. Scale the feats, however, and it gets a lot better. More parity for humans is good.
Basically, adopting this approach means switching feats over to level-dependent feat trees. It would still be possible to create feat-tree add-on feats, however, because of the 3 set limit and progression type.
Let's look at the Dodge/ Mobility/ Spring Attack/ (Combat Expertise)/ Whirlwind Attack progression under this system.
Existing Feat Tree Example
"Dodge" feat tree. +1 bonus. Add a miss chance to any attack at 7th level. Then up to +4 bonus at level 14, rising to +6 bonus at level 20. Nice. When you're facing EL 20 monsters, you need that kind of bonus to matter at all.
"Battlefield Mobility" feat tree. Begins at +2 when moving out of a threatened square (down from +4). Adds a miss chance to any attack of opportunity at 7th level, so taking both feats gives the PC up to 2 miss chances per day to use. OK, that's decent compensation. At 14th level, bonus rises to +5, up to +7 at level 20. This is pretty good - and I would now sever this from Dodge so it has no prerequisites.
But if you do take both feats, you could pick a creature eligible for an AOO, declare a dodge bonus, stack it with the battlefield mobility bonus, and take +9 to your AC at 14th level against that opponent. Not broken at that level, and against other opponents you're still only +5. Now we have a feat tree that makes sense for certain character choices, but we don't have to enforce it with prerequisites. Instead, we offer a natural synergy advantage and many players will follow. But those who don't follow didn't waste a feat with something lame like the v3.5 Dodge.
Combat Expertise feat tree. We run into a problem here. Like Power Attack, this is designed to be a variable, player-set bonus up to +5. But your system doesn't really accommodate that. We could say that since a character can always fight defensively (-4 penalty on attacks, +2 dodge bonus to AC), we stack with that to say you can now take -4 penalty and get +4 AC. But that's a stiff penalty at lower levels. We could keep the variability by setting -4/+4 as the initial limit, since it's inherently a variable range ability, and letting players pick from -1/+1 on up to -4/+4. That's a bit weaker, but probably close enough for Combat Expertise to work. Plus, there are additional compensations. At 7th level, you get the force an attack reroll (so someone with Dodge + CE gets 2 per day, and maybe another AOO related re-roll if they have Mobility). At 14th level, you kick the chosen CE range up to -7 attacks/+7 AC, rising to -9/+9 range at 20th level. OK. That probably isn't broken.
Prestige Feats
Spring Attack is our first example of a... let's call it a "Prestige Feat" that doesn't follow these progression rules. Prestige Feats will be defined by needing prerequisites. Spring Attack would have the same prerequisites as now, and it gives the listed ability. Taking Dodge + Mobility gives big advantages, but Spring Attack gives the same concept as a guaranteed ability with additional movement, targeting one opponent.
The question with this revised approach may become "Why take Spring Attack, as opposed to another auto-scaling feat tree"? If Dodge and Mobility are progressing nicely, they may be enough to do many of the same things (this should be a playtest focus), and another feat tree may offer bigger advantages. My fear is that we turn "Spring Attack" into another 3.5 Dodge with this approach, useful mostly as a prerequisite feat.
One way to balance this is to give "Prestige feats" a 14th level kicker, too. Let's say that this kicker can be flexible to allow cool effects instead of straight numerical progression, but the usual approach is that at the 14th level kicker, you can use a Prestige Feat twice per round instead of once. So you can make 2 Spring Attacks now, or 2 Whirlwind attacks, or 2 Spirited Charge attacks. That prospect may keep stuff like Spring Attack balanced with another feat tree as a long term choice.
Whirlwind Attack also maintains the same prerequisites and abilities under this Prestige Feat system. Its effectiveness is obvious. So let's look closer.
It appears that a very focused human fighter could have this as early as 6th level, and an equally focused human barbarian would have to be 9th. That hasn't changed from the previous system. At 14th level, applying the standard "2 uses at once" rule for Prestige Feat abilities means that our esteemed pirate queen fighter (swashbuckler) does indeed continue to progress in line with her increased abilities, and that flashing rapier now strikes at every opponent around her twice. This is good - compare to a 14th level barbarian with no Whirlwind Attack but a strength focus for ability increases instead of dexterity, a Greataxe, and Great Cleave. Our Pirate Queen is still well short of his ability to mow through weaker opponents - but at least it's a slightly more balanced scale thanks to the Prestige Feat kicker.
Alas, Power Attack will suck if it's just -2/+2, as this system suggests. A re-roll may not really matter much, either, if the bonuses remain weak.
You'll want to shift the core progression mechanic to say that the 7th level mid-tier for Basic Feats EITHER forces a related re-roll, OR immediately shifts the bonus maximum to a stacking +1 per 4 levels (which is what it will do later, anyway). So a 7h level Ranger sees their Power Attack range rise to -3 attack / +3 damage since we're rounding down, but it rises again to -4/+4 range at 8th level, -5/+5 at 12th, -6/+6 at 16th, and -7/+7 at 20th. Since it starts smaller and rises only with increased levels, the fact that we've duplicated the "Greater Power attack" bonus at lower levels doesn't matter.
Now we have an ability that's weaker at lower levels (where, let's face it, not a lot of "I take -5" decisions anyway), and a bit stronger at higher levels. Keep it as a prerequisite for the "Cleave" Prestige feat (which becomes Great Cleave as a 14th level kicker), and we're good.
Yes, I think this could work. We'd want to test it on some other trees, but you may have discovered a solid core mechanic principle.
The investigation into Power Attack as an auto-scaling Feat Tree appears to have revealed an additional rule for this kind of core system:
Feats that grant combat attack capabilities NEVER get the mid-tier re-roll; at 7th level, they ALWAYS jump directly to an additional stacking bonus of +1 per 4 levels, rounded down."
So Basic Feats like Power Attack, Mounted Archery, Trample, Two-weapon fighting, et. al. would all fall under this rule. It would probably be a good idea for feats like Two-weapon defense to do so as well.
|
|
Wow, I have to take a closer look at your analysis (I'm in a hurry right now, so I'll just post some quick comments).
Prestige Feats -- I'm not sure if I would call them that, but I agree that there are Feats that should exist outside the system. Power Attack is a good example -- it would completely suck if you could only have +1/-1 or +2/-2 at, say, 7th level.
First of all, I would definitely change it as a shift between AC and damage -- not BAB and damage. I would probably have the 'Basic' level Power Attack function as -1/+1 per 2 levels (max. -5/+5 at 10th level) since for some classes taking a penalty on AC might not be as significant as a penalty to attack. Yet this modification would make *much* more useful for pretty much every melee-oriented character (at the moment I never use it 3.5 with any high-level characters). Of course, you would get double (or 1.5 times?) the penalty as bonus on damage with 2-H weapons.
The Improved Power Attack would allow you to apply half your total character level as a penalty to AC and bonus to damage (-10/+10 at 20th level).
I would probably have the 'Basic' level Power Attack function as -1/+1 per 2 levels (max. -5/+5 at 10th level).... The Improved Power Attack would allow you to apply half your total character level as a penalty to AC and bonus to damage (-10/+10 at 20th level).
You need a 3rd tier to make the system work as a consistent mechanic. Take a second look at the option suggested.
I like auto-scaling.
Asgetrion, thought about this last night. I worry that this discussion might be losing the plot, but really, it has 2 of them. One is in the title, feats that scale to remain useful at high levels. The other is your proposed core mechanic as an organizing principle for feats.
My sense is that the core mechanic is more important, because without that sort of simplification/ improvement, we're very unlikely to get the wholesale changes in the feat system that you're asking for.
So, let's revisit it and see if it works, and if it doesn't what change need to be made. At the end of the day, I think we're either going to have a new core mechanic for feats that shows it can handle all of the situations required, or we're going to show that the proposed core mechnic does not work, and we'll end up with pretty much the feat system Pathfinder has now.
FEATS, RELOADED:
There are 3 kinds of feats:
1. Basic Feats, which auto-scale at 3 different levels and have no feat prerequisites.
2. Prestige feats, which have feat prerequisites and scale only at 1 relatively high level.
3. Magic Feats (metamagic and item creation), which already auto-scale by becoming usable with more powerful magics and remain exactly as they are.
BASIC FEATS
* Basic Feats begin with a +1 to +2 bonus at Tier 1 (option: unless you're trading one capability for another, in which case the trade range doubles to -1/+1 to -4/+4).
* Some basic feats that do not grant attack bonuses or change saving throws auto-scale at Tier 2, by granting the PC one re-roll per day within its normal usage conditions.
* At Tier 2, Basic Feats that grant attack bonuses or change saving throws either remove a previous restriction on the ability in question, or go straight to the Tier 3 mechanic of increasing their bonus by a stacking bonus of +1/4 levels, rounded down.
* At Tier 3, all basic feats either add or continue a stacking improvement to their initial bonus of +1/4 levels, rounded down. At 20th level, therefore, Basic Feats confer a bonus of +5 to +6 ((1-2)+(20/4)).
* If we use the above option for "trade-off feats", adopting the stacking bonus would raise the level 20 figure to -9/+9.
PRESTIGE FEATS
* Prestige Feats require pre-requisites, and build on earlier feats. In this respect, they mirror the Prestige Class system. The pre-requisites can involve any aspect of a character, but always require a feat prerequisite of an earlier Basic or Prestige feat or feats. If a PC receives 2 or more feat slots at a certain level, it is possible to fill the last feat prerequisite and take a Prestige Feat at the same time.
* Prestige Feats aren't about straight plus bonuses, they're about multiples and doing cool things. Whirlwind Attack, Cleave, Spirited Charge, and Snatch Arrow are all good examples.
* Prestige Feats have an auto-scaling "kicker" of their own at a commonly designated high level (proposed: 14th), which "kicks them up a notch" to help these feats keep pace at the very highest levels of play.
* The default Prestige Feat kicker is the ability to use it 2x per round as a standard action. The kicker can be anything, however, as long as it remains balanced within the game.
Let's take that proposed core mechanic set, and see if we can break it by producing feats it either does not fit, or feats where its application breaks the game. Whether or not we would have designed the given feats exactly that way would not be the point, at least for now...
There are 3 kinds of feats:
Basic Feats, which auto-scale at 3 different levels and have no feat prerequisites.
Prestige feats, which have feat prerequisites and scale only at 1 relatively high level.
Magic Feats, which already auto-scale by becoming usable with more powerful magics and remain exactly as they are.
The question here is "what levels do they scale at?" That deserved a separate post from the foundational "here's the proposed strawman system, let's see if we can break it" post, because it's a separate issue.
* My original proposal for the 3 Basic Feat tiers was 1st, 7th, 14th.
I chose these levels because they are often "dead levels" in terms of exciting new abilities.
- 4th level is too low for the mid-tier in most campaigns.
- 5th level is big for wizards, clerics, and druids because they get 3rd level spells. For whatever reason, 3rd and 5th level spells offer a big step-change in he game. This is less true for Druids (although... call lightning?), but they also get Wild Shape.
- 6th is big for sorcerers (3rd level spells) and all fighter classes (2 attacks with full attack). The Ranger also sees their weapon path abilities kicked up a notch.
- 9th is big for Wizards and Clerics because it confers major abilities (Teleport, Raise Dead)
- 10th is big for Sorcerers, for the same reason.
Hence the 1/7/14 proposal. Although, the truth is that many campaigns never reach 14th. If you wanted these 3 to be accessible to more players, we might say that since fighters are the big beneficiaries, we can make the progression 1/5/10 or 1/6/12.
I'm actually beginning to lean toward 1st/ 6th/ 12th character levels as the best set points for Basic Feat auto-scaling. I also have to acknowledge that multiclassing will often make it an impossible challenge to guarantee hitting a "dead level".
5th level strikes me as too early. The good news: gives non-spellcasters a corresponding power-up at 5th level. Bad news: early in campaigns relative to challenges; and lost in too many other goodies for spellcasters, many of whom will not multiclass or Prestige Class yet.
7th level... Pros: Usually a less exciting level, and if you look at published modules, it tends to denote the threshold for the next tier of play. Cons: May be too late to wait for stuff like your Dodge feat to start getting really useful, and magnifies the effect by pushing the final stage to level 14.
Thoughts?
* The Prestige Feat kicker should be a pretty high level, somewhere in the upper 3rd of possible play. 2 Whirlwind Attacks, for instance, is something that should only belong to legends. Great Cleave, which can very conceivably kill 50 goblins in one round, also belongs here. 14th level is reachable, and also denotes the upper 1/3 of non-epic play.
Thoughts?
So, let's revisit it and see if it works, and if it doesn't what change need to be made. At the end of the day, I think we're either going to have a new core mechanic for feats that shows it can handle all of the situations required, or we're going to show that the proposed core mechnic does not work,
I'm going to look at saving throw-related feats under this system....
Iron Will. Old version: "You get a +2 bonus on all Will saving throws."
Iron Will, New version: Can't get a re-roll on saving throws at mid-tier, go straight to the additional bonus track.
Iron Will, under 6th level: You get a +1 bonus on all Will saving throws.
Greater Iron Will, 6th+ level: Add +1/4 levels, rounded down, as a stackable bonus with Iron Will. A 6th level character has a +2 bonus, and 8th level character +3, all the way to +6 at 20th level.
3rd Tier progression: No change, per level increases as before, up to +6 at 20th.
[u]Pros and Cons:[/u] There could be an argument that +1 is too weak to really help at low levels, where it's most needed. On the other hand, there may also be an argument that +6 to will saves is pushing it at 20th level, and +7 would be even worse. Let's see.
Example 1: A Level 20 Barbarian with Iron Will suddenly has 2 good saves, +12 for Fortitude and Will before magic and abilities kick in. Along the way, Will is a mid-way save between Fortitude and Reflex from 1st-7th level, then begins oscillating between mid-range and parity with Fortitude during levels 8-20.
That's a pretty substantial benefit for a single feat. On the Pro side, it certainly lives up to its name, and builds an iron-willed Conan type. On the minus side, this means that 1 feat can turn other classes into "3 good saves" classes, and the fighter with feats to burn can cover that gap using 2 feats. Then again, those 2 feats could have been going into other cool powers like Power Attack and Cleave instead, done as feat tree feats. Would your fighter really make that trade? I think some would, some would not. Which, to me, says "success!" within the system itself. I want more hard thought about which goodies to take.
But can we break it using another class? Can we make an argument that this approach pushes the whole set of feat benefits too far?
Example 2: A Cleric really doesn't want to be bothered by mental attacks, and takes Iron Will. At level 20, Will saves are +18 base, and assuming 22 Wisdom (probably a bit low), that's a +24 Will save. Without a single magic item, this is a PC who slips out of DC 34 Will saves 50% of the time. And I don't know who even has DC 34 will saves at their disposal, other than near-deity class figures. A 9th level spell still needs another +15 to get it to that point. Mind Flayers see this guy coming and immediately run up the white flag.
Then again, he is a level 20 cleric of Asmodeus - if anyone is going to do the dominating around here, it'll he him thank you very much. Now pass him his whips, "Mr. Mentalist the Great, Sorcerer of Reknown," and bite down on the nice rubber ball. If you entertain him, he may let you live.
Too uber? Or good game flavor (the cleric, not the rubber ball which tastes terrible)?
Example 3: Even 12th level is a +12 will save for a Cleric with Iron Will. Now add Wisdom 20 and a Cloak of Resistance +3. That's +20 Will save at 12th level. How many EL 12 monsters can even dent that enough to force a save? An enemy 12th level sorcerer with a 20 charisma, firing a 5th level spell (Dominate, DC 20), is an automatic save barring extraneous factors. Heck, the enemy sorcerer could have fired a Mind Fog last round, and our iron-willed cleric of Sarenrae still has a 50% chance to walk through it all like it isn't even there.
Hmm. Now we compare to her "old feat" counterpart with same Wisdom and item, who has a Will save of +8+5+2, or +15. The 12th level enemy Sorcerer above would succeed 20% of the time (because only 1-4 fail), not high but not zero either. With a Mind Fog the round before, odds of success jump to 2/3 instead of 1/2.
Example 4: Or take a 20th level Paladin with the analogous Great Fortitude, a 16 Charisma, a 16 constitution, and a +3 Cloak of Resistance. Fortitude Save? +12 base +3 Con + 3 Divine Grace + 3 Cloak +6 GF = +27 Fort. DC 26 Dragon Bile was the highest DC poison in the 3.5 rules, and the Paladin is immune.
But then, with the 3.5 rules and version of Great Fortitude, our Paladin's save is a 3, and only Purple Worm poison and Dragon Bile have any hope of working at all.
Query: is this difference enough of a difference for us to say that Iron Will is broken? Or is it OK to reward PCs who take the feat, even if it makes them uber-tough in this respect? Are the feats given up enough cost? Does it leach too much challenge from the game? That would be an interesting playtest set.
My instincts say it's powerful, but it may pass. PCs with save-boosting feats will be rewarded well, but they're also giving up good stuff from other feats they could have taken (see Dodge up above, for instance). And this lets me throw tough enemy Orc Barbarians at the party, without giving up magic items to them because I needed to keep the wizard from just charming him.
But we'd have to test using more scenarios to know.
|
|
Katman,
you could also get +6 on Will Saves in 3E by 21st level -- and it, too, required two feats: Iron Will (+2) and Epic Will (+4). In that sense, yes, maybe this new system would be more "powerful", but at least it would not just "leap up" by +4 as you hit level 21.
And fighters and barbarians *should* be able to boost their Will Saves, if the players really want to burn Feats to do that. If this system proves to be too advantageous, then it's easy to add that "Mid Tier" which would then be obligatory to achieve the "Greater Tier".
And I used to have a 15th level Paladin with all saves between +21-32 (Ref being the weakest), so I'm not really worried about PCs who might have a single saving throw modifier at 25+.
|
|
Oh, I seem to have missed your idea of Level Requirements for Tiers... sorry about that! Did I understand correctly that you meant that you could only get Tier 2 at LVL 7, and Tier 3 at LVL 14? If so, I think you're actually correct -- the system would probably work a lot better that way... :)
My original idea was that you could pick up all three Tiers at 1st level (which wouldn't really help you that much, because the bonuses wouldn't yet scale a lot).
Is the "gap" between Tiers 1 and 2 too long? I mean, should you be able to get Tier 2 at LVL 6 and Tier 3 at LVL 12? Or did you intend them for "dead" levels on purpose?
Oh, I seem to have missed your idea of Level Requirements for Tiers... sorry about that! Did I understand correctly that you meant that you could only get Tier 2 at LVL 7, and Tier 3 at LVL 14? If so, I think you're actually correct -- the system would probably work a lot better that way... :)
Yes, though I'm thinking more along the lines of 6th level for Tier 2 and 12th level for tier 3 now.
For feats without a different middle tier, of course, it would just be that steady level-based bonus progression. For a feat in this category that begins at +1 the bonus ratchet is:
1st: +1
6th: +2
8th: +3
12th: +4
16th: +5
20th: +6
...and as you point out, by 21st level an Epic feat can put that character in exactly the same situation as our 20th level model, at +6. So OK, not broken. Good!
Really, Epic Feats should pretty much all be Prestige Feats anyway....