Corran
2017-05-06, 05:56 AM
TL;DR
How would you do it?
How would you optimize a melee character for elven accuracy?
Below I make some general points one should have in mind, when trying to maximize the effect that elven accuracy will have on a character build. If you are in the mood of a long read, then go ahead. If not, you can take a quick look at the A,B,C,D,E points (just the titles, they are the essentials for optimizing elven accuracy) for getting the jist of it. Or you can skip everything and post your idea of an optimal use of elven accuracy when applied on a melee build. You can also check my own try at it, at the very end of this post.
As I mention later too, I am more interested approaching this subject from an end game perspective, without examining how viable the builds are during the various levels. This is because at this early stage, I am NOT interested in making optimal characters around elven accuracy, but in maximizing the effect that elven accuracy can have on characters, specifically on their final builds. This is the first step. Once I have enough insight about how to best optimize elven accuracy, I can approach the subject of optimizing characters with that feat with more ease, as this will just have to do with tweaking the resulting builds in such a way, so that it will make them more viable during their level progression. As I said, it is a two-step process, and optimizing elven accuracy is just the first step, which this thread is about.
That said, if you feel confident and want to talk about character optimization around elven accuracy, while still keeping elven accuracy as a central mechanic to the build's function, you can go ahead and do it. Though imo it is easier and yields to better results if we just take a step at a time.
Introduction
Intrigued by the discussion in this (http://www.giantitp.com/forums/showthread.php?523632-elven-accuracy) thread, and inspired by this (http://www.giantitp.com/forums/showthread.php?523966-Complete-Front-Line-Leader-Knight-Support-Build) thread, I started thinking on how to best optimize elven accuracy on a melee character.
Optimizing elven accuracy on a melee character, is the first step to optimizing a melee character around elven accuracy. They are not the same thing. That means the the optimal setup for elven accuracy might not be an actual optimal setup for our melee character, and if we are aiming for character optimization (with elven accuracy included) we might need to tone down the focus on elven accuracy a bit, in order to increase the build's overall efficiency.
But lets take the first step, and see how we can make the most out of elven accuracy on melee characters, for starters. Then it's a long way adjusting the final build to make it stronger overall, and ofc playable if possible. This thread mainly aims to explore how elven accuracy interacts with the game mechanics already in place, and tries to find silly ways to make the best of it.
The basics
What do we need in order to make elven accuracy matter? First and foremost we need a reliable way to have advantage on our melee attacks. Secondly, we need a high dpr to profit from the triple advantage. Otherwise what's the point of having triple advantage if our damage is low? Thirdly, we need high crit damage. Critical hits profit from rolled dice and not from flat damage bonuses, so that is something we have to take into account. Forthly, we need to increase if possible our crit range, as triple advantage has a very nice effect when combined with expanded crit range.
For example, while tripple advantage gives a 14.75% crit chance with a normal crit range, it gives a 27.1% crit chance when combined with 19-20 crit range; now, this is a very significant boost (it's an increase of our critical damage weighted down by 12.35%, and that's only how much it increases from just expanding our crit range; it's a significant boost, assuming we put some effort in having a good crit damage; and it gets even better if we throw GWM on top of it).
Last but not least, we need to be wary of effects that can impose disadvantage on us, as if our advantage is canceled, then we lose quite a bit of steam.
Lets make it into a list, for better reference:
A) Advantage
B) dpr
C) crit damage
D) crit range
E) Avoid disadvantage
The specifics
A) Ways to gain advantage
1) Reckless attack (Barbarian 2)
2) Vow of enmity (Paladin 3, vengeance)
3) Mounted Combatant (feat)
4) Darkness + devil's sight (warlock 2 if you get darkness from elsewhere, warlock 3 else)
5) Eyes in the dark (sorcerer 2, shadow)
6) Trip attack (Fighter 3, battlemaster)
7) Feinting attack (Fighter 3, battlemaster)
8) Faerie Fire (Bard 1, druid 1, drow, magic initiate feat)
9) Foresight (Bard 17, druid 17, warlock 17)
10) Greater invisibility
11) Shoving prone an enemy (extra attack, shield master)
12) Master of hexes (warlock 14, hexblade)
13) fighter samurai 3 = 3 times short rest advantage for two rounds (turn you cast it and the next one) + resistance to most common damages. Bonus action to activate. (copied and pasted from joaber's post)
14) revised beast master 3 with constrictor snake = at will, no action restrainned. (copied and pasted from joaber's post)
15) mystic = lots of stuffs, wu jen 6 can convert psi points to spellslots, great for smites and invocations. Aura of Bloodletting is reckless attack for everybody, but cost bonus action and you can lose concentration. (copied and pasted from joaber's post)
Now, there are a lot more ways to get advantage in melee combat (like blinding an enemy, or attacking from stealth, and many more), I just included whatever I could think of and thought it was a reliable enough source of advantage, to consider using it as the basis of our elven accuracy. Chances are that I may have underestimated some features and completelly forgot some others, so please take a moment to point out anything I missed and you consider it to be important in regard to elven accuracy, and I'll edit it to the list above (same goes as far as removing options from the list is concenred, if you make a good argument about it).
Edit: Ok, 14 levels of warlock is just as reliable, though with some bonus action opportunity cost. And some other good ways to gain advantage have been brought into my attention. So reckless attack might not necessarily be the best way to go about this, still, I am tempted to insist that it is, though I am far from 100% sure. Two drawbacks of reckless attack to mention: 1) it does not work with reaction attacks (for example ones you might get from sentinel), and 2) it makes you vulnerable through making incoming attacks have advantage, and when when rage is not on this is a huge problem. Even with rage on, it might be a problem still if we went with too few barbarian levels, so these are good points to take into account when thinking about actual play, though this thread is just messing around with elven accuracy from a white board optimization perspective, and without taking much into account overall character optimization. Thus I will allow my initial opinion, as shown just in the next paragraph, to stand.
Ok, I'll go ahead and say this. IMO, reckless attack is the only optimal choice here, and everything else should act as a potential substitute if picked (used under the right conditions). But without reckless attack, you are risking too much. Spells may fail or they may be countered. A familiar can be killed. A mount might not be suitable for certain places, and it can be killed (bonus points for find steed + armor of agathys, and for small characters riding such buffed medium mounts though). Reckless attack is the most reliable way to gain advantage, and having advantage is EVERYTHING, as far as exploiting elven accuracy is concerned.
B) dpr
There are two basic approaches regarding this:
1) GWM*, extra attack(s)
2) Sneak attack (rogue)
I didn't do the math. But I dont think I need to. I am guesstimating that with triple advantage in play, GWM will easily surpass sneak attack, especially if we consider the bonus action attack GWM will give us if we kill a creature or if we just crit with one of our attacks (for example, with a crit range of 19-20 and with just two attacks per turn, we have a chance of 46.8559% just to crit in any given round; and without the increased crit range, again, with just two attacks, the chance lies at 27.324375; not bad, but we see again that the increased crit range makes quite a difference here too.
Sneak attack does seem to work better on crits at first glance, but GWM will still come out ahead if we invest on smites/invocations and on spell slots. And it is not difficult to do so while also grabing one instance of extra attack, while sneak attack requires continuous and uninterupted investment in the rogue class in order to scale.
If we move in the GWM direction, we can add various resources on our repertoir to further increase our dpr. And while a second instance of extra attack might be very hard to get (considering we need to boost other aspects of our build as well, there are stuff like haste, sentinel and PAM that can help us in that direction (though personally I would probably avoid PAM, especially considering that we are very likely to be going for an improved crit range; cause in that case, half of the time the bonus action attack would be redundant, and I consider this to be too much overlap to spend a precious ASI on it).
C) Crit damage
1) Warlock smite-like invocations (warlock 3*)
2) Divine smite (paladin 2)
3) Sneak attack (rogue)
* The smite-like invocations can be taken as soon as you hit warlock 3, by taking pact of the blade and retraining one of your current invocations at the same time.
Note: I did something silly. Counted expanded crit range to be 19-20 due to hexblade's curse at all times. Thinking of bringing it down to 1/3 of the times, for a better approximation, and run the numbers again. When I do this, this note will be removed. Till then, read at your own risk and know that these numbers dont estimate the reality of things accurately.
I'll go ahead and say it. The warlock smites are incredibly silly as they are. And I dont like them one bit. I like the idea (or at the very least I get the idea behind it), but I dont like the execution of it. Anyway, back on track.
One could argue that because sneak attack does not expend resources, will add more damage than smites on crits, on a long adventuring day. Lets touch on this, via some rough calculations.
Assuming GWM, 2 attacks, and a crit range of 19-20, the chance to crit each round (not counting reaction attacks), is 59.5538489% (included the chance to proc the bonus action attack based only on crits (46.8559%) and crit on it (27.1%), for an increase of 12.6979489% on top of our initial 46.8559%, thus for a total of 59.5538489% crit chance per round).
Let's take a very long adventuring day. Say 8 encounters. And lets assume that each encounter has 4 rounds of combat, for a total of 32 rounds of combat for the whole day.
With our 60% chance to crit (rounded up for easier calculations), we'll crit about 19 (19.2) times this day with the above build.
Assuming the following spells slots: 4, 3+2, 3, 3, 2, 1, 1, 1 (doable, say barbarian2/hexblade3/valorbard15), and at least 1 short rest, the increased to to invocation-smites damage on crits will average at (272d8) 1224 damage (or 1360.5 damage including the increased weapon damage on crits, but NOT the extra bonus attack from GWM).
Now lets take a barbarian2/rogue18, to see if the maximum amount of sneak attack (while keeping the most reliable source of advantage in reckless attack, and thus barbarian 2) can keep up. Let's make the rogue and arcane trickster, and have him use BB. With triple advantage, there is little benefit to twf in regard to dpr (especially at high levels), and 2 attacks (from twf) wont increase significantly our chance to land a crit-affected sneak attack (plus, it will make easier my calculations), and the extra 3d8 (6d8 on crits) from BB is not an insignificant amount either. What's the chance that sneak attack will apply on a crit for that rogue, per any given round? It's 14.75%, so in 32 rounds of combat, that means 5 rounds (4.72) of crits, for a total increase of 5*99 = 495 damage (or 517.5 with weapon dice included/ just their increased damage on crits).
Now, that is not even comparable, but we were not fair with the rogue. Lets reduce the rogue levels and take 1 warlock level in hexblade, for the increased crit range (so barbarian2/warlock1/rogue17). The crit damage will remain the same ofc as before, but now we will crit more times per day. We have a crit chance of 27.1% per round, so in 32 rounds of combat, that means we will crit in exactly 8.672 rounds per day (lets say 9). That means a total of 901 damage (or 941.5 damage with weapon dice included). Again, it is lagging behind significantly. And consider this was the case as far as a very long day was concerned.
Hence we see it is vital to include smites (preferably the warlock ones), and spell slots, as this way, even in very long days we will still be ahead in crit damage (and ofc we will be ahead in dpr too).
ps: Perhaps there are a lot that can be said about a barbarian2/hexblade3/rogue(AT)15, who combines increased crit range, sneak attacks and warlock-smites, though I think that his overall dpr per adventuring day, even in long adventuring days (and when I say the whole dpr, I mean everything, crits included), thet will still lag behind a GWM, as GWM adds a lot to dpr with that bonus attack and the vastly increased crit chance (due to more attacks) per round, despite the AT having some slots to throw in warlock-smites. But that is just intuition (perhaps in long day the barb2/hex3/AT15 will be better, though probaby still worse in not so very long days.
Actually, no, as the warlock smites dont work with finesse weapons, so this is not eligible for a melee build. A barbarian 2/ paladin 2/ hexblade 1/ rogue 15 would come as close as it could, but I dont think it is close enough since our spell slots are very limited (4+1, 3, 3), and besides, we dont have the strong warlock smites...
D) Crit range
1) Fighter 3, champion (19-20, passive)
2) Warlock 1, hexblade (19-20, short rest recharge, against 1 enemy, may be moved if you deal the killing blow)
3) Fighter 15, champion (18-20, passive)
Obviously champion is far better than hexblade in that respect, but hexblade comes along with the smites, so it is a hard choice to make. I lean more towards hexblade, since their smites are far better than the paladin's (greater damage, also work with higher level spell slots). Though if one does not go with hexblade, but with paladin instead (for other benefits, most notable aura of protection -see point E about avoiding disadvantage-, or if one goes with a heavy rogue build to rely more on sneak attack than smites (suboptimal choice imo), then champion immediately becomes more desirable.
Oh, and about champion 15 for the 18-20 crit range, personally I dont think it is worth the investment, as the crit damage is not enough to capitilize on such a good crit range.
E) Avoiding disadvantage
To my knowledge, effects that can potentially give us disadvantage, target saving throws. Being poisoned, restrained or frightened are the most common effects that I can think of. So generally, it is things that target our wisdom, constitution and dexterity saves. Thus anything that adds to our saving throws is a much desired form of defense we should be aiming at. Thus:
1) Resilient (one of wis, dex, con, according to what our starting class was). (feat)
2) Aura of protection (paladin 6)
3) Slippery mind (rogue 15)
4) Mindless rage (barbarian 6, berserker)
5) Aura of protection (paladin 10)
6) Foresight (bard 17, druid 17, warlock 17)
There are other things here too, but I mostly confined myself to features that relevant (to the focus of the build) classes get.
If you think something should be included/excluded, please take the time to make a suggestion.
End specifics.
My takes on it
I'll be brief (did a sloppy job detailing this, but I will put more care detailing my second build, in the next spoiler).
GWM route. Halfelf.
STR 16, DEX 14, CON 14, INT 8, WIS 10, CHA 14
Reckless attack is the most reliable source of advantage when in melee. Thus barbarian 2. Rage (2 times per day) is an added benefit that will help with our survivability, and it will compare nicely with armor of agathys that we will get from being a warlock. Essentially, with 2 rages per day, it is fair enough to assume that our 8th and 7th level spell slots will go to casting AoA, while the rest slots are reserved for smiting (the crits will be many!).
There is more to that, since being bard 15 opens up a lot of options regarding spellcasting, so we can pick up spells that dont require concentration and that have a good duration for us to cast them before entering rage, or spells that will work nicely for us when we are out of rages. But I have not really looked that into it).
Hexblade 3 for the the curse bringer invocation (smites, but better), and for the increased crit range against the enemy of our hexblade's curse (if we kill this enemy we can move the curse onto another enemy, so that allows us to maintain our improved crit range. Armor of agathys is just something we grabbed along the way, but it is very useful against melee opponents because of how it combines with the resistance to physical damage due to our rage. Always use them together, and always upcast AoA with your highest slot (unless you are expecting a lot of non-melee enemies, but you know that already). The second warlock invocation would probably be devil's sight, to combine it with darkness when not raging, to gain advantage without using reckless attack and thus risking our survivability. A good alternative to reckless attack for getting advantage, particularly when we are not raging.
The bard levels are for two reasons. To increase our spell slots (for more smiting and higher upcasted AoA; resulting in lots of slots: 4, 3+2, 3, 3, 2, 1, 1, 1), but also because of extra attack. I guess I forgot to say that I am going with a greatsword and GWM (if that wan't obvious from the curse bringer invocation; made sure to mention it at the start of the build). So we need spell slots and extra attack, and the valor bard is the best way to go about it, and the only way to accomodate both at the same time (I guess the first version of FS could work too, but I think I prefer bard). I could grab haste and use it when I am not raging, though my survivability would be at risk in these cases (though lots of added damage due to the additional attack).
The third and final ASI (besides GWM and elven accuracy) could go in one of the following (sentinel, resilient, blade mastery, or +2 STR). I would probably go with sentinel.
(I could go with mounted combatabt with the last feat, assuming I picked find steed as one of my magical secrets, but I doubt it is good value for money).
The big drawback of the build is that it leaves itself a bit open to things that target its saves, and that means that we might be open to enemies imposing disadvantage on us, if they have ways of doing so.
A smaller drawback, is that due to the hexblade's curse being recharged on short rests, we might not enjoy as much as we would like the increased crit range (which does make a difference dpr and nova wise). But I dont think champion 3 was worth the investment, since we had already gone for 3 levels in hexblade for the warlock smites, and it just so happened to pick the curse (that deals with the expanded crit range) along the way.
And that's about it.
First of all, this was inspired by PeteNutButter's work, found in this thread (http://www.giantitp.com/forums/showthread.php?523966-Complete-Front-Line-Leader-Knight-Support-Build).
Secondly, although not as attack-oriented as the first build, I like it better, for 2 reasons:
1) I find the warlock smite invocations a bit unbalanced and dont like to use them (and frankly, I am a sucker for paladins), and the only non-official material this build uses is elven accuracy (everything else is official).
2) This build gives a boost to the saving throws unlike the previous one, and that is important against enemies that have the ability to try and force disadvantage on us.
Last but not least, this build is only built up to character level 14, allowing 6 more levels for which the potential options numerous (and since it allows for more freedom, perhaps it will be easier to optimize it as a viable character build).
So, without further ado, lets jump into it.
Race: Halfelf
Stats: STR 16, DEX 12, CON 14, INT 8, WIS 10, CHA 16
Feats: Elven accuracy, GWM
Edit: Didnt calculate the +1 DEX from elven accuracy. Thinking of upping DEX to 13 (so that it becomes a 14 with elven accurcy), and lowering WIS to 9, taking resilient wisdom as my third ASI (after GWM and elven accuracy) to increase wisdom to 10 and get wisdom save prof as well (assuming I dont start as a paladin - probably starting as barb, for the hp and the better saves in STR & CON).
Barbarian 2
For reckless attack.
Rage 2 times per day. Even if the final build will not figure a single level of warlock, we can still pick armor of agathys as one of our additional magical secrets from being a lore bard 6, and profit from rage + AoA. The more spell slots we end up with, the better this combo will work (due to upcasting AoA at a higher level). Rage + AoA + reckless attack is something you should do at least 2 times per day (as many as you can rage, that is).
Paladin (vengeance) 6
Divine smite: Not as good as curse bringer (from hexblade), but still, very very good. Going with paladin offers us other benefits to balance out the difference between divine smite and curse bringer.
Vow of enmity: Alternative way to reckless attack for getting advantage. If against one strong opponent, it is worth using it (even if it does take a bonus action that might deny you a bonus attack from crit + GWM), because it will spare you attacking recklesly and thus bestowing advantage to incoming attacks. Good to have it.
Aura of protection: +3 on our saves, so that means better chances at resisting effects that can potentially impose disadvantage on your attacks. This is the single best thing you can do, to ensure working at max efficiency (triple advantage, for which you are working for). That's why 16 cha and just 12 dex. Also affects our inspiration and spellcasting when not raging (but that last one is kind of secondary).
Find steed: With shared AoA, your steed is one tough m***********. If your campaign is mount-friendly, you can even drop he barbarian levels and grab the mounted combatant feat instead. Otherwise, better not bother, and keep the barbarian levels. It will be funny if some poor soul chooses to attack your steed when AoA is up. Enjoy the better movement ( I wonder if being mounted expands the radius of your aura of protection, due to the mount taking up 2 squares; I think not). Better mobility, and a few extra damage if your mount can make attacks.
Extra attack: Since GWM, from somewhere you have to get extra attack. And since it is just one level before aura of protection, and at the same level as find steed, what better way to get it? It also allows you to go with lore as your bardic college, so that is additional value.
Spell slots: 3 levels towards determining how many spell slots you have. It will just take 6 levels in some fullcaster (hence lore bard 6), to max out your smiting slots (of up to 4th level spell slots).
Spells: Couple of useful spells too. Some can be used along with raging (like aid) if precast, and some can be used out of combat or when you are not raging, although lore bard will play a bigger part regarding this last thing.
Bard (Lore) 6
Spell slots: As mentioned earlier, with just 6 levels in bard, you maximize your smiting slots. Smiting is important, as it is one of the few things you can do when also raging, and it will also boost your damage significantly, since with tripple advantage (and hopefully with increased crit range, if we plan for it), you will crit quite a lot (of up to once per 2 rounds, as an average). Taking more bard levels might indeed be a good choice, as you can still smite with any extra spell slots above 4th slot level, but more importantly, you will have some high level slots to use AoA with.
Inspiration/Cutting words: Both can be used even while raging, though make sure to use them after you have attacked for the round, because you dont want to miss on any bonus action attacks from GWM due to critting/killing someone. Cutting words better used when you are not recklessly attcking or raging, or having AoA up.
Spellcasting: There are sure many possibilities here (ones that can pair with raging, others that can cover you when you are not raging), though I have not looked into it analytically. I place more value in spell slots (having maxed spells slots up to 4th level, and having that many higher than 4th level spell slots as many uses of rage you have per day, so that you can combine rage with AoA always.
What is missing?
First and foremost, 6 levels of proression. There are many choices on how to fill them.
The most crucial thing missing from that build (so far), is having a way to increase our crit range. There are two ways to go about it. Either grab at least 3 levels in champion, or just one level in warlock hexblade. Improving our crit range is very important, so I assume you have to pick one of these options. Not convinced which way to swing about this (and having some doubts whether I could just let my crit range stay as it is without gimping this build so much).
The second thing missing, is more spell slots. We need to match the number of higher than 4th level spell slots, with the number of rages we have per day (till now we have 2 rages per day). This way, we will be able to upcast AoA before entering rage (that is before combat), without scrificing any of our smite-intended spell slots (of up to 4th level). So if we stay we 2 barbarian levels, I would expect at least one more bard level (or two more paladin levels), for 7 levels of bard as a minimum on the final build. If we decide to take 1 (or 2) more levels in barbarian, for 3 rages per day, then we would need at least 2 more bard levels (or 4 paladin levels), to have enough high level spell slots to use AoA with.
Taking 1 more barbarian level (for barbarian 3 in the final build) seems like a very good choice, for one extra use of rage, and for picking a barbarian path (several good choices there). Even a 4th barbarian level has value for the ASI.
Since I mentioned ASI's, this build already picked the two important ones (elven accuracy, GWM), but it could profit a lot from others too (any or any combination of the following: sentinel, blade mastery, +2 str, +2 cha, mounted combatant, and many more, could prove useful).
Regarding the improved crit range and how we will go about it (if we end up going about it), there were two options I mentioned: champion 3, and hexblade 1. If going with the champion route, we could always take a 4th level in fighter for the ASI. And if we are going we the warlock level, we could perhaps take a second level for the invocations (and the additional 1st level slot that will recharge on short rests).
How I am tempted to proceed
So the build was standing at barbarian2/paladin6/bard6, for a total of 14 levels.
Thinking of it, I would probably include at least one more level in barbarian, as the benefits (1 more rage and an archtype) seem to good not to be picked since they are just 1 level away. That would make me take 2 more bard levels (becoming bard 8), so that I balance the above-4th level slots (intended for AoA) with the number of rages I gain from being a barbarian 3), and also for the ASI (which I would use on sentinel if the group comp was favourable towards it -in which case I would probably pick wolf as my barb totem, as making good use of sentinel implies a strong melee presense from the rest of the group; else, I would probably use with either blade mastery or with just a +2 to str).
So, all this brings me to barbarian3/paladin6/bard8, leaving me 3 levels. This allows me to make a choice:
1) Take champion with my 3 remaining levels? Or...
2) Take 1 more levels in barbarian and 2 more levels in paly for two more ASI's?
3) Take 2 more levels in bard for even more spell slots and magical secrets, and 1 more level in barbarian for the ASI?
This is just the outline of the build, I have not given this much thought...
Now, this uses advantage from reckless attack, str-based BB rapier attack, divine smites from paladin, spell slots as a 6th level full caster plus one 1st level slot that recharges on short rests from being a warlock, increased crit range from the hexblade, and sneak attack (8d6) from the rogue. It gets slippery mind, and surprisingly enough 4 whole feats (which I dont know how to best spend, as I have not looked at this in detail).
Ideas from the community
I'd go Half-Elf Arcane Trickster/Bard. Bard is there just for Faerie Fire. Make 'em glow, then strike. A sneak attack crit is monstrous. With Magical Ambush later on, folks will save at disadvantage against your FF, so it's always good. But this is far from the most optimal build.
Half-Elf Barbarian(Berserker) 9/Fighter(Champion) 11
Str 20 (2 ASIs)
Dex 14
Con 18 (1 ASI)
Int 8
Wis 10
Cha 10
Feats: Elven Accuracy, Great Weapon Master
Weapon: Greataxe
Armor Half-Plate
Doesn't have the super burst Nova rounds(outside of Action Surge) potential of Smites or Sneak Attacks, but it has the longevity and consistency to more than make up for it.
3 Attacks/Action
19-20 Crit Range
Reckless Attack at-will advantage
GWM Shennanigans
BA Attack redundancy
Bonus Damage on a Crit from Brutal Critical
Save rerolls via Indomitable
All this character really needs is something like a Flametongue or Frostbrand Greataxe and they would be set.
I planned an samurai 3/hexblade 17, and thebidea looks good, AoA with resistsnce, and party can help with advanayge in other turns.
Never tested, but bonus action economy looks a problem. You need one to curse, other for samurai, if you ceit or kill you can't use BA for GWM, and have hex as other opportunity.
I don't have any interest in doing the math, but Moon Druid 9/Champion 3/Frenzy Barb 5/Paladin 2/War Cleric 1 could pull off some insane damage.
In Quetzalcoatlus form its damage would be 6d6+4 with triple advantage and an increased crit range. 3 attacks of that on offense (2 from Extra Attack and 1 from Frenzy/War Priest) combined with Smites on Crits could pull off some serious numbers. It could also get a 4th attack per round via Sentinel although that one would not come with Advantage.
That last point there brings up a challenge to your optimal choice of gaining advantage. Reckless Attack does not work on other people's turns like Darkness + Devil's sight does, and that makes the Warlock's ability superior when one considers reaction attacks and such.
Hexblade gets to use Charisma as its primary stat (which is Half Elf's only +2), critically strikes on a 19-20 on the target of your hex, and thus critically strikes 27% of the time (http://anydice.com/program/badd) with good damage from Sneak Attack. Alternative build could be a Strength-based Hexblade that uses Curse Bringer and goes single-class. Hide with Cunning Action lets you get advantage on your own at least some of the time, though actually getting it reliably is probably more party-dependent. Additionally, you can use the Hex spell to get a little bit of extra damage on a target.
At 20th level with no magical equipment, your DPR on hexed targets with two-weapon fighting would be .9571(1d8+10d6+11)+.9571(1d8+1d6+6)+.271(1d8+10d6) , which amounts to about 72.4 damage by my calculations, which is respectable enough for a melee Sneak Attack build that is archetype-independent on the Rogue side.
Barbarian build is a little more self-reliant. It gets advantage for free at-will, three attacks, and critically strikes almost 39% of the time thanks to its 18-20 threat range (http://anydice.com/program/badd). It lacks a source of bonus damage for those criticals, but has reliable output, and isn't punished too harshly that it can't use Great Weapon Master, because of how 3d20b1 throws the curve so heavily in favor of higher dice rolls (if you assume you need to hit on a 13 thanks to the -5, with Elven Accuracy, you still manage that over 78% of the time). If you go Champion to 17, you get two Action Surges, which means six attacks with almost 40% chance to crit on each one (amounting to a 95% chance of getting at least one crit in that round with no BA attack, as opposed to about a 77% chance without the Action Surge or BA attack). You are highly likely to get a 7th attack in, but if you don't, you can Frenzy instead for insurance.
At 20th level with no magic equipment, your nova DPR would be 0.784(7[2d6r1,2+17])+0.3859(7[2d6r1,2]), which amounts to about 161.5 DPR by my calculations twice per short rest, and your DPR otherwise (with a BA attack from somewhere) would be 0.784(4[2d6r1,2+17])+0.3859(4[2d6r1,2]), which amounts to about 92.3 DPR at-will if you can get a BA attack reliably.
Hexblade 14-15/Sorcerer 5-6 wouldn't be a bad grab either, if you have someone like a Wolf Totem Barbarian to give you ready advantage. Grab Curse Breaker, Thirsting Blade, Superior Pact Weapon (or Ultimate if you go to 15) and Lifedrinker as invocations, Hex as a Warlock spell, either Greenflame Blade or Booming Blade as Sorcerer cantrips (maybe both, if you want to commit to versatile attack options?), Quicken Spell (and maybe Twinned Spell if you take Booming Blade, interactions start to get wonky and beyond my immediate rules understanding from here if you do though) as metamagics, and Great Weapon Master as a feat. Max Strength and probably go Stone Sorcerer for the naked AC keyed to CON (build gets too MAD otherwise). You have a ridiculously stylish nova: two magical greatsword attacks which benefit from your Strength, Charisma, proficiency bonus, and 10d8 bonus from the expended spell slot in damage, plus a Quickened attack cantrip that can also apply all of the above. This is where RAW gets confusing: if you can Twin a Booming Blade, you can strike two people with your bonus action attack cantrip... And you can empower one... Both?... Of those attacks? I'm going to ignore twinned Booming Blade smites for this, as even if it's RAW, it's obviously not an intended interaction. Also, you get to use your reaction to teleport-attack people, which is just so damn cool, and probably worth losing the +3 weapon and extra invocation and 8th-level Mystic Arcanum for this build, in my opinion.
At 20th level with no magical equipment aside from your sword, your nova DPR would be .875(2[3d6+10d8+26])+.271(2[3d6+10d8])+.875(3d6+13d8+26)+.271(3d6+13d8), which amounts to about 274.5 damage in that one nova round, and can be repeated 3/day at short rest intervals. If you make the more sensible decision to withhold your spells for smiting until you roll a crit, your DPR changes to .875(2[3d6+26])+.271(2[3d6+20d8])+.875(3d6+3d8+26)+.271(3d6+23d8), or 193 damage per "round" until you critically strike three times or run out of sorcery points.
A blade singer can do this. 14 BS/ 4 champion with greater invisibility and TWF with short swords
How would you do it?
How would you optimize a melee character for elven accuracy?
Below I make some general points one should have in mind, when trying to maximize the effect that elven accuracy will have on a character build. If you are in the mood of a long read, then go ahead. If not, you can take a quick look at the A,B,C,D,E points (just the titles, they are the essentials for optimizing elven accuracy) for getting the jist of it. Or you can skip everything and post your idea of an optimal use of elven accuracy when applied on a melee build. You can also check my own try at it, at the very end of this post.
As I mention later too, I am more interested approaching this subject from an end game perspective, without examining how viable the builds are during the various levels. This is because at this early stage, I am NOT interested in making optimal characters around elven accuracy, but in maximizing the effect that elven accuracy can have on characters, specifically on their final builds. This is the first step. Once I have enough insight about how to best optimize elven accuracy, I can approach the subject of optimizing characters with that feat with more ease, as this will just have to do with tweaking the resulting builds in such a way, so that it will make them more viable during their level progression. As I said, it is a two-step process, and optimizing elven accuracy is just the first step, which this thread is about.
That said, if you feel confident and want to talk about character optimization around elven accuracy, while still keeping elven accuracy as a central mechanic to the build's function, you can go ahead and do it. Though imo it is easier and yields to better results if we just take a step at a time.
Introduction
Intrigued by the discussion in this (http://www.giantitp.com/forums/showthread.php?523632-elven-accuracy) thread, and inspired by this (http://www.giantitp.com/forums/showthread.php?523966-Complete-Front-Line-Leader-Knight-Support-Build) thread, I started thinking on how to best optimize elven accuracy on a melee character.
Optimizing elven accuracy on a melee character, is the first step to optimizing a melee character around elven accuracy. They are not the same thing. That means the the optimal setup for elven accuracy might not be an actual optimal setup for our melee character, and if we are aiming for character optimization (with elven accuracy included) we might need to tone down the focus on elven accuracy a bit, in order to increase the build's overall efficiency.
But lets take the first step, and see how we can make the most out of elven accuracy on melee characters, for starters. Then it's a long way adjusting the final build to make it stronger overall, and ofc playable if possible. This thread mainly aims to explore how elven accuracy interacts with the game mechanics already in place, and tries to find silly ways to make the best of it.
The basics
What do we need in order to make elven accuracy matter? First and foremost we need a reliable way to have advantage on our melee attacks. Secondly, we need a high dpr to profit from the triple advantage. Otherwise what's the point of having triple advantage if our damage is low? Thirdly, we need high crit damage. Critical hits profit from rolled dice and not from flat damage bonuses, so that is something we have to take into account. Forthly, we need to increase if possible our crit range, as triple advantage has a very nice effect when combined with expanded crit range.
For example, while tripple advantage gives a 14.75% crit chance with a normal crit range, it gives a 27.1% crit chance when combined with 19-20 crit range; now, this is a very significant boost (it's an increase of our critical damage weighted down by 12.35%, and that's only how much it increases from just expanding our crit range; it's a significant boost, assuming we put some effort in having a good crit damage; and it gets even better if we throw GWM on top of it).
Last but not least, we need to be wary of effects that can impose disadvantage on us, as if our advantage is canceled, then we lose quite a bit of steam.
Lets make it into a list, for better reference:
A) Advantage
B) dpr
C) crit damage
D) crit range
E) Avoid disadvantage
The specifics
A) Ways to gain advantage
1) Reckless attack (Barbarian 2)
2) Vow of enmity (Paladin 3, vengeance)
3) Mounted Combatant (feat)
4) Darkness + devil's sight (warlock 2 if you get darkness from elsewhere, warlock 3 else)
5) Eyes in the dark (sorcerer 2, shadow)
6) Trip attack (Fighter 3, battlemaster)
7) Feinting attack (Fighter 3, battlemaster)
8) Faerie Fire (Bard 1, druid 1, drow, magic initiate feat)
9) Foresight (Bard 17, druid 17, warlock 17)
10) Greater invisibility
11) Shoving prone an enemy (extra attack, shield master)
12) Master of hexes (warlock 14, hexblade)
13) fighter samurai 3 = 3 times short rest advantage for two rounds (turn you cast it and the next one) + resistance to most common damages. Bonus action to activate. (copied and pasted from joaber's post)
14) revised beast master 3 with constrictor snake = at will, no action restrainned. (copied and pasted from joaber's post)
15) mystic = lots of stuffs, wu jen 6 can convert psi points to spellslots, great for smites and invocations. Aura of Bloodletting is reckless attack for everybody, but cost bonus action and you can lose concentration. (copied and pasted from joaber's post)
Now, there are a lot more ways to get advantage in melee combat (like blinding an enemy, or attacking from stealth, and many more), I just included whatever I could think of and thought it was a reliable enough source of advantage, to consider using it as the basis of our elven accuracy. Chances are that I may have underestimated some features and completelly forgot some others, so please take a moment to point out anything I missed and you consider it to be important in regard to elven accuracy, and I'll edit it to the list above (same goes as far as removing options from the list is concenred, if you make a good argument about it).
Edit: Ok, 14 levels of warlock is just as reliable, though with some bonus action opportunity cost. And some other good ways to gain advantage have been brought into my attention. So reckless attack might not necessarily be the best way to go about this, still, I am tempted to insist that it is, though I am far from 100% sure. Two drawbacks of reckless attack to mention: 1) it does not work with reaction attacks (for example ones you might get from sentinel), and 2) it makes you vulnerable through making incoming attacks have advantage, and when when rage is not on this is a huge problem. Even with rage on, it might be a problem still if we went with too few barbarian levels, so these are good points to take into account when thinking about actual play, though this thread is just messing around with elven accuracy from a white board optimization perspective, and without taking much into account overall character optimization. Thus I will allow my initial opinion, as shown just in the next paragraph, to stand.
Ok, I'll go ahead and say this. IMO, reckless attack is the only optimal choice here, and everything else should act as a potential substitute if picked (used under the right conditions). But without reckless attack, you are risking too much. Spells may fail or they may be countered. A familiar can be killed. A mount might not be suitable for certain places, and it can be killed (bonus points for find steed + armor of agathys, and for small characters riding such buffed medium mounts though). Reckless attack is the most reliable way to gain advantage, and having advantage is EVERYTHING, as far as exploiting elven accuracy is concerned.
B) dpr
There are two basic approaches regarding this:
1) GWM*, extra attack(s)
2) Sneak attack (rogue)
I didn't do the math. But I dont think I need to. I am guesstimating that with triple advantage in play, GWM will easily surpass sneak attack, especially if we consider the bonus action attack GWM will give us if we kill a creature or if we just crit with one of our attacks (for example, with a crit range of 19-20 and with just two attacks per turn, we have a chance of 46.8559% just to crit in any given round; and without the increased crit range, again, with just two attacks, the chance lies at 27.324375; not bad, but we see again that the increased crit range makes quite a difference here too.
Sneak attack does seem to work better on crits at first glance, but GWM will still come out ahead if we invest on smites/invocations and on spell slots. And it is not difficult to do so while also grabing one instance of extra attack, while sneak attack requires continuous and uninterupted investment in the rogue class in order to scale.
If we move in the GWM direction, we can add various resources on our repertoir to further increase our dpr. And while a second instance of extra attack might be very hard to get (considering we need to boost other aspects of our build as well, there are stuff like haste, sentinel and PAM that can help us in that direction (though personally I would probably avoid PAM, especially considering that we are very likely to be going for an improved crit range; cause in that case, half of the time the bonus action attack would be redundant, and I consider this to be too much overlap to spend a precious ASI on it).
C) Crit damage
1) Warlock smite-like invocations (warlock 3*)
2) Divine smite (paladin 2)
3) Sneak attack (rogue)
* The smite-like invocations can be taken as soon as you hit warlock 3, by taking pact of the blade and retraining one of your current invocations at the same time.
Note: I did something silly. Counted expanded crit range to be 19-20 due to hexblade's curse at all times. Thinking of bringing it down to 1/3 of the times, for a better approximation, and run the numbers again. When I do this, this note will be removed. Till then, read at your own risk and know that these numbers dont estimate the reality of things accurately.
I'll go ahead and say it. The warlock smites are incredibly silly as they are. And I dont like them one bit. I like the idea (or at the very least I get the idea behind it), but I dont like the execution of it. Anyway, back on track.
One could argue that because sneak attack does not expend resources, will add more damage than smites on crits, on a long adventuring day. Lets touch on this, via some rough calculations.
Assuming GWM, 2 attacks, and a crit range of 19-20, the chance to crit each round (not counting reaction attacks), is 59.5538489% (included the chance to proc the bonus action attack based only on crits (46.8559%) and crit on it (27.1%), for an increase of 12.6979489% on top of our initial 46.8559%, thus for a total of 59.5538489% crit chance per round).
Let's take a very long adventuring day. Say 8 encounters. And lets assume that each encounter has 4 rounds of combat, for a total of 32 rounds of combat for the whole day.
With our 60% chance to crit (rounded up for easier calculations), we'll crit about 19 (19.2) times this day with the above build.
Assuming the following spells slots: 4, 3+2, 3, 3, 2, 1, 1, 1 (doable, say barbarian2/hexblade3/valorbard15), and at least 1 short rest, the increased to to invocation-smites damage on crits will average at (272d8) 1224 damage (or 1360.5 damage including the increased weapon damage on crits, but NOT the extra bonus attack from GWM).
Now lets take a barbarian2/rogue18, to see if the maximum amount of sneak attack (while keeping the most reliable source of advantage in reckless attack, and thus barbarian 2) can keep up. Let's make the rogue and arcane trickster, and have him use BB. With triple advantage, there is little benefit to twf in regard to dpr (especially at high levels), and 2 attacks (from twf) wont increase significantly our chance to land a crit-affected sneak attack (plus, it will make easier my calculations), and the extra 3d8 (6d8 on crits) from BB is not an insignificant amount either. What's the chance that sneak attack will apply on a crit for that rogue, per any given round? It's 14.75%, so in 32 rounds of combat, that means 5 rounds (4.72) of crits, for a total increase of 5*99 = 495 damage (or 517.5 with weapon dice included/ just their increased damage on crits).
Now, that is not even comparable, but we were not fair with the rogue. Lets reduce the rogue levels and take 1 warlock level in hexblade, for the increased crit range (so barbarian2/warlock1/rogue17). The crit damage will remain the same ofc as before, but now we will crit more times per day. We have a crit chance of 27.1% per round, so in 32 rounds of combat, that means we will crit in exactly 8.672 rounds per day (lets say 9). That means a total of 901 damage (or 941.5 damage with weapon dice included). Again, it is lagging behind significantly. And consider this was the case as far as a very long day was concerned.
Hence we see it is vital to include smites (preferably the warlock ones), and spell slots, as this way, even in very long days we will still be ahead in crit damage (and ofc we will be ahead in dpr too).
ps: Perhaps there are a lot that can be said about a barbarian2/hexblade3/rogue(AT)15, who combines increased crit range, sneak attacks and warlock-smites, though I think that his overall dpr per adventuring day, even in long adventuring days (and when I say the whole dpr, I mean everything, crits included), thet will still lag behind a GWM, as GWM adds a lot to dpr with that bonus attack and the vastly increased crit chance (due to more attacks) per round, despite the AT having some slots to throw in warlock-smites. But that is just intuition (perhaps in long day the barb2/hex3/AT15 will be better, though probaby still worse in not so very long days.
Actually, no, as the warlock smites dont work with finesse weapons, so this is not eligible for a melee build. A barbarian 2/ paladin 2/ hexblade 1/ rogue 15 would come as close as it could, but I dont think it is close enough since our spell slots are very limited (4+1, 3, 3), and besides, we dont have the strong warlock smites...
D) Crit range
1) Fighter 3, champion (19-20, passive)
2) Warlock 1, hexblade (19-20, short rest recharge, against 1 enemy, may be moved if you deal the killing blow)
3) Fighter 15, champion (18-20, passive)
Obviously champion is far better than hexblade in that respect, but hexblade comes along with the smites, so it is a hard choice to make. I lean more towards hexblade, since their smites are far better than the paladin's (greater damage, also work with higher level spell slots). Though if one does not go with hexblade, but with paladin instead (for other benefits, most notable aura of protection -see point E about avoiding disadvantage-, or if one goes with a heavy rogue build to rely more on sneak attack than smites (suboptimal choice imo), then champion immediately becomes more desirable.
Oh, and about champion 15 for the 18-20 crit range, personally I dont think it is worth the investment, as the crit damage is not enough to capitilize on such a good crit range.
E) Avoiding disadvantage
To my knowledge, effects that can potentially give us disadvantage, target saving throws. Being poisoned, restrained or frightened are the most common effects that I can think of. So generally, it is things that target our wisdom, constitution and dexterity saves. Thus anything that adds to our saving throws is a much desired form of defense we should be aiming at. Thus:
1) Resilient (one of wis, dex, con, according to what our starting class was). (feat)
2) Aura of protection (paladin 6)
3) Slippery mind (rogue 15)
4) Mindless rage (barbarian 6, berserker)
5) Aura of protection (paladin 10)
6) Foresight (bard 17, druid 17, warlock 17)
There are other things here too, but I mostly confined myself to features that relevant (to the focus of the build) classes get.
If you think something should be included/excluded, please take the time to make a suggestion.
End specifics.
My takes on it
I'll be brief (did a sloppy job detailing this, but I will put more care detailing my second build, in the next spoiler).
GWM route. Halfelf.
STR 16, DEX 14, CON 14, INT 8, WIS 10, CHA 14
Reckless attack is the most reliable source of advantage when in melee. Thus barbarian 2. Rage (2 times per day) is an added benefit that will help with our survivability, and it will compare nicely with armor of agathys that we will get from being a warlock. Essentially, with 2 rages per day, it is fair enough to assume that our 8th and 7th level spell slots will go to casting AoA, while the rest slots are reserved for smiting (the crits will be many!).
There is more to that, since being bard 15 opens up a lot of options regarding spellcasting, so we can pick up spells that dont require concentration and that have a good duration for us to cast them before entering rage, or spells that will work nicely for us when we are out of rages. But I have not really looked that into it).
Hexblade 3 for the the curse bringer invocation (smites, but better), and for the increased crit range against the enemy of our hexblade's curse (if we kill this enemy we can move the curse onto another enemy, so that allows us to maintain our improved crit range. Armor of agathys is just something we grabbed along the way, but it is very useful against melee opponents because of how it combines with the resistance to physical damage due to our rage. Always use them together, and always upcast AoA with your highest slot (unless you are expecting a lot of non-melee enemies, but you know that already). The second warlock invocation would probably be devil's sight, to combine it with darkness when not raging, to gain advantage without using reckless attack and thus risking our survivability. A good alternative to reckless attack for getting advantage, particularly when we are not raging.
The bard levels are for two reasons. To increase our spell slots (for more smiting and higher upcasted AoA; resulting in lots of slots: 4, 3+2, 3, 3, 2, 1, 1, 1), but also because of extra attack. I guess I forgot to say that I am going with a greatsword and GWM (if that wan't obvious from the curse bringer invocation; made sure to mention it at the start of the build). So we need spell slots and extra attack, and the valor bard is the best way to go about it, and the only way to accomodate both at the same time (I guess the first version of FS could work too, but I think I prefer bard). I could grab haste and use it when I am not raging, though my survivability would be at risk in these cases (though lots of added damage due to the additional attack).
The third and final ASI (besides GWM and elven accuracy) could go in one of the following (sentinel, resilient, blade mastery, or +2 STR). I would probably go with sentinel.
(I could go with mounted combatabt with the last feat, assuming I picked find steed as one of my magical secrets, but I doubt it is good value for money).
The big drawback of the build is that it leaves itself a bit open to things that target its saves, and that means that we might be open to enemies imposing disadvantage on us, if they have ways of doing so.
A smaller drawback, is that due to the hexblade's curse being recharged on short rests, we might not enjoy as much as we would like the increased crit range (which does make a difference dpr and nova wise). But I dont think champion 3 was worth the investment, since we had already gone for 3 levels in hexblade for the warlock smites, and it just so happened to pick the curse (that deals with the expanded crit range) along the way.
And that's about it.
First of all, this was inspired by PeteNutButter's work, found in this thread (http://www.giantitp.com/forums/showthread.php?523966-Complete-Front-Line-Leader-Knight-Support-Build).
Secondly, although not as attack-oriented as the first build, I like it better, for 2 reasons:
1) I find the warlock smite invocations a bit unbalanced and dont like to use them (and frankly, I am a sucker for paladins), and the only non-official material this build uses is elven accuracy (everything else is official).
2) This build gives a boost to the saving throws unlike the previous one, and that is important against enemies that have the ability to try and force disadvantage on us.
Last but not least, this build is only built up to character level 14, allowing 6 more levels for which the potential options numerous (and since it allows for more freedom, perhaps it will be easier to optimize it as a viable character build).
So, without further ado, lets jump into it.
Race: Halfelf
Stats: STR 16, DEX 12, CON 14, INT 8, WIS 10, CHA 16
Feats: Elven accuracy, GWM
Edit: Didnt calculate the +1 DEX from elven accuracy. Thinking of upping DEX to 13 (so that it becomes a 14 with elven accurcy), and lowering WIS to 9, taking resilient wisdom as my third ASI (after GWM and elven accuracy) to increase wisdom to 10 and get wisdom save prof as well (assuming I dont start as a paladin - probably starting as barb, for the hp and the better saves in STR & CON).
Barbarian 2
For reckless attack.
Rage 2 times per day. Even if the final build will not figure a single level of warlock, we can still pick armor of agathys as one of our additional magical secrets from being a lore bard 6, and profit from rage + AoA. The more spell slots we end up with, the better this combo will work (due to upcasting AoA at a higher level). Rage + AoA + reckless attack is something you should do at least 2 times per day (as many as you can rage, that is).
Paladin (vengeance) 6
Divine smite: Not as good as curse bringer (from hexblade), but still, very very good. Going with paladin offers us other benefits to balance out the difference between divine smite and curse bringer.
Vow of enmity: Alternative way to reckless attack for getting advantage. If against one strong opponent, it is worth using it (even if it does take a bonus action that might deny you a bonus attack from crit + GWM), because it will spare you attacking recklesly and thus bestowing advantage to incoming attacks. Good to have it.
Aura of protection: +3 on our saves, so that means better chances at resisting effects that can potentially impose disadvantage on your attacks. This is the single best thing you can do, to ensure working at max efficiency (triple advantage, for which you are working for). That's why 16 cha and just 12 dex. Also affects our inspiration and spellcasting when not raging (but that last one is kind of secondary).
Find steed: With shared AoA, your steed is one tough m***********. If your campaign is mount-friendly, you can even drop he barbarian levels and grab the mounted combatant feat instead. Otherwise, better not bother, and keep the barbarian levels. It will be funny if some poor soul chooses to attack your steed when AoA is up. Enjoy the better movement ( I wonder if being mounted expands the radius of your aura of protection, due to the mount taking up 2 squares; I think not). Better mobility, and a few extra damage if your mount can make attacks.
Extra attack: Since GWM, from somewhere you have to get extra attack. And since it is just one level before aura of protection, and at the same level as find steed, what better way to get it? It also allows you to go with lore as your bardic college, so that is additional value.
Spell slots: 3 levels towards determining how many spell slots you have. It will just take 6 levels in some fullcaster (hence lore bard 6), to max out your smiting slots (of up to 4th level spell slots).
Spells: Couple of useful spells too. Some can be used along with raging (like aid) if precast, and some can be used out of combat or when you are not raging, although lore bard will play a bigger part regarding this last thing.
Bard (Lore) 6
Spell slots: As mentioned earlier, with just 6 levels in bard, you maximize your smiting slots. Smiting is important, as it is one of the few things you can do when also raging, and it will also boost your damage significantly, since with tripple advantage (and hopefully with increased crit range, if we plan for it), you will crit quite a lot (of up to once per 2 rounds, as an average). Taking more bard levels might indeed be a good choice, as you can still smite with any extra spell slots above 4th slot level, but more importantly, you will have some high level slots to use AoA with.
Inspiration/Cutting words: Both can be used even while raging, though make sure to use them after you have attacked for the round, because you dont want to miss on any bonus action attacks from GWM due to critting/killing someone. Cutting words better used when you are not recklessly attcking or raging, or having AoA up.
Spellcasting: There are sure many possibilities here (ones that can pair with raging, others that can cover you when you are not raging), though I have not looked into it analytically. I place more value in spell slots (having maxed spells slots up to 4th level, and having that many higher than 4th level spell slots as many uses of rage you have per day, so that you can combine rage with AoA always.
What is missing?
First and foremost, 6 levels of proression. There are many choices on how to fill them.
The most crucial thing missing from that build (so far), is having a way to increase our crit range. There are two ways to go about it. Either grab at least 3 levels in champion, or just one level in warlock hexblade. Improving our crit range is very important, so I assume you have to pick one of these options. Not convinced which way to swing about this (and having some doubts whether I could just let my crit range stay as it is without gimping this build so much).
The second thing missing, is more spell slots. We need to match the number of higher than 4th level spell slots, with the number of rages we have per day (till now we have 2 rages per day). This way, we will be able to upcast AoA before entering rage (that is before combat), without scrificing any of our smite-intended spell slots (of up to 4th level). So if we stay we 2 barbarian levels, I would expect at least one more bard level (or two more paladin levels), for 7 levels of bard as a minimum on the final build. If we decide to take 1 (or 2) more levels in barbarian, for 3 rages per day, then we would need at least 2 more bard levels (or 4 paladin levels), to have enough high level spell slots to use AoA with.
Taking 1 more barbarian level (for barbarian 3 in the final build) seems like a very good choice, for one extra use of rage, and for picking a barbarian path (several good choices there). Even a 4th barbarian level has value for the ASI.
Since I mentioned ASI's, this build already picked the two important ones (elven accuracy, GWM), but it could profit a lot from others too (any or any combination of the following: sentinel, blade mastery, +2 str, +2 cha, mounted combatant, and many more, could prove useful).
Regarding the improved crit range and how we will go about it (if we end up going about it), there were two options I mentioned: champion 3, and hexblade 1. If going with the champion route, we could always take a 4th level in fighter for the ASI. And if we are going we the warlock level, we could perhaps take a second level for the invocations (and the additional 1st level slot that will recharge on short rests).
How I am tempted to proceed
So the build was standing at barbarian2/paladin6/bard6, for a total of 14 levels.
Thinking of it, I would probably include at least one more level in barbarian, as the benefits (1 more rage and an archtype) seem to good not to be picked since they are just 1 level away. That would make me take 2 more bard levels (becoming bard 8), so that I balance the above-4th level slots (intended for AoA) with the number of rages I gain from being a barbarian 3), and also for the ASI (which I would use on sentinel if the group comp was favourable towards it -in which case I would probably pick wolf as my barb totem, as making good use of sentinel implies a strong melee presense from the rest of the group; else, I would probably use with either blade mastery or with just a +2 to str).
So, all this brings me to barbarian3/paladin6/bard8, leaving me 3 levels. This allows me to make a choice:
1) Take champion with my 3 remaining levels? Or...
2) Take 1 more levels in barbarian and 2 more levels in paly for two more ASI's?
3) Take 2 more levels in bard for even more spell slots and magical secrets, and 1 more level in barbarian for the ASI?
This is just the outline of the build, I have not given this much thought...
Now, this uses advantage from reckless attack, str-based BB rapier attack, divine smites from paladin, spell slots as a 6th level full caster plus one 1st level slot that recharges on short rests from being a warlock, increased crit range from the hexblade, and sneak attack (8d6) from the rogue. It gets slippery mind, and surprisingly enough 4 whole feats (which I dont know how to best spend, as I have not looked at this in detail).
Ideas from the community
I'd go Half-Elf Arcane Trickster/Bard. Bard is there just for Faerie Fire. Make 'em glow, then strike. A sneak attack crit is monstrous. With Magical Ambush later on, folks will save at disadvantage against your FF, so it's always good. But this is far from the most optimal build.
Half-Elf Barbarian(Berserker) 9/Fighter(Champion) 11
Str 20 (2 ASIs)
Dex 14
Con 18 (1 ASI)
Int 8
Wis 10
Cha 10
Feats: Elven Accuracy, Great Weapon Master
Weapon: Greataxe
Armor Half-Plate
Doesn't have the super burst Nova rounds(outside of Action Surge) potential of Smites or Sneak Attacks, but it has the longevity and consistency to more than make up for it.
3 Attacks/Action
19-20 Crit Range
Reckless Attack at-will advantage
GWM Shennanigans
BA Attack redundancy
Bonus Damage on a Crit from Brutal Critical
Save rerolls via Indomitable
All this character really needs is something like a Flametongue or Frostbrand Greataxe and they would be set.
I planned an samurai 3/hexblade 17, and thebidea looks good, AoA with resistsnce, and party can help with advanayge in other turns.
Never tested, but bonus action economy looks a problem. You need one to curse, other for samurai, if you ceit or kill you can't use BA for GWM, and have hex as other opportunity.
I don't have any interest in doing the math, but Moon Druid 9/Champion 3/Frenzy Barb 5/Paladin 2/War Cleric 1 could pull off some insane damage.
In Quetzalcoatlus form its damage would be 6d6+4 with triple advantage and an increased crit range. 3 attacks of that on offense (2 from Extra Attack and 1 from Frenzy/War Priest) combined with Smites on Crits could pull off some serious numbers. It could also get a 4th attack per round via Sentinel although that one would not come with Advantage.
That last point there brings up a challenge to your optimal choice of gaining advantage. Reckless Attack does not work on other people's turns like Darkness + Devil's sight does, and that makes the Warlock's ability superior when one considers reaction attacks and such.
Hexblade gets to use Charisma as its primary stat (which is Half Elf's only +2), critically strikes on a 19-20 on the target of your hex, and thus critically strikes 27% of the time (http://anydice.com/program/badd) with good damage from Sneak Attack. Alternative build could be a Strength-based Hexblade that uses Curse Bringer and goes single-class. Hide with Cunning Action lets you get advantage on your own at least some of the time, though actually getting it reliably is probably more party-dependent. Additionally, you can use the Hex spell to get a little bit of extra damage on a target.
At 20th level with no magical equipment, your DPR on hexed targets with two-weapon fighting would be .9571(1d8+10d6+11)+.9571(1d8+1d6+6)+.271(1d8+10d6) , which amounts to about 72.4 damage by my calculations, which is respectable enough for a melee Sneak Attack build that is archetype-independent on the Rogue side.
Barbarian build is a little more self-reliant. It gets advantage for free at-will, three attacks, and critically strikes almost 39% of the time thanks to its 18-20 threat range (http://anydice.com/program/badd). It lacks a source of bonus damage for those criticals, but has reliable output, and isn't punished too harshly that it can't use Great Weapon Master, because of how 3d20b1 throws the curve so heavily in favor of higher dice rolls (if you assume you need to hit on a 13 thanks to the -5, with Elven Accuracy, you still manage that over 78% of the time). If you go Champion to 17, you get two Action Surges, which means six attacks with almost 40% chance to crit on each one (amounting to a 95% chance of getting at least one crit in that round with no BA attack, as opposed to about a 77% chance without the Action Surge or BA attack). You are highly likely to get a 7th attack in, but if you don't, you can Frenzy instead for insurance.
At 20th level with no magic equipment, your nova DPR would be 0.784(7[2d6r1,2+17])+0.3859(7[2d6r1,2]), which amounts to about 161.5 DPR by my calculations twice per short rest, and your DPR otherwise (with a BA attack from somewhere) would be 0.784(4[2d6r1,2+17])+0.3859(4[2d6r1,2]), which amounts to about 92.3 DPR at-will if you can get a BA attack reliably.
Hexblade 14-15/Sorcerer 5-6 wouldn't be a bad grab either, if you have someone like a Wolf Totem Barbarian to give you ready advantage. Grab Curse Breaker, Thirsting Blade, Superior Pact Weapon (or Ultimate if you go to 15) and Lifedrinker as invocations, Hex as a Warlock spell, either Greenflame Blade or Booming Blade as Sorcerer cantrips (maybe both, if you want to commit to versatile attack options?), Quicken Spell (and maybe Twinned Spell if you take Booming Blade, interactions start to get wonky and beyond my immediate rules understanding from here if you do though) as metamagics, and Great Weapon Master as a feat. Max Strength and probably go Stone Sorcerer for the naked AC keyed to CON (build gets too MAD otherwise). You have a ridiculously stylish nova: two magical greatsword attacks which benefit from your Strength, Charisma, proficiency bonus, and 10d8 bonus from the expended spell slot in damage, plus a Quickened attack cantrip that can also apply all of the above. This is where RAW gets confusing: if you can Twin a Booming Blade, you can strike two people with your bonus action attack cantrip... And you can empower one... Both?... Of those attacks? I'm going to ignore twinned Booming Blade smites for this, as even if it's RAW, it's obviously not an intended interaction. Also, you get to use your reaction to teleport-attack people, which is just so damn cool, and probably worth losing the +3 weapon and extra invocation and 8th-level Mystic Arcanum for this build, in my opinion.
At 20th level with no magical equipment aside from your sword, your nova DPR would be .875(2[3d6+10d8+26])+.271(2[3d6+10d8])+.875(3d6+13d8+26)+.271(3d6+13d8), which amounts to about 274.5 damage in that one nova round, and can be repeated 3/day at short rest intervals. If you make the more sensible decision to withhold your spells for smiting until you roll a crit, your DPR changes to .875(2[3d6+26])+.271(2[3d6+20d8])+.875(3d6+3d8+26)+.271(3d6+23d8), or 193 damage per "round" until you critically strike three times or run out of sorcery points.
A blade singer can do this. 14 BS/ 4 champion with greater invisibility and TWF with short swords