Class Comparisons for Ranged Damage

[Darkmatter]

Xetheral pointed out an error in these calculations. I've not yet updated the new conclusions in this post. Some major conclusions change at high levels.

The common consensus is that when you're building a ranged attacker, the class to choose depends on your expected level of play. Since average damage is easy to calculate, I figured I'd put together a quick spreadsheet to share exactly when the switchovers are. I'm only looking at sustained damage output - considering spike damage is a much more complex task. The spoiler below has the criteria I used for the calculations for those who are interested.


This is looking only at single-target damage against a large pool of HP. There are too many variables for a straight calculation when we try looking at multiple targets. I'm also only going to consider single classed characters. Because of the way the classes leapfrog each other in being the top damage dealer, multiclassing invariably results in a slower progression with a bit better damage in the top end.

Average damage is calculated by considering the abilities listed below. Critical hits and misses are considered, as is advantage. An average damage per round was calculated for each AC between 12 and 20. These numbers were then averaged to get each character's average damage indicator at a given level.

In order to be considered in these calculations, the damage output has to be the same for a 10 round combat, and the character must be capable of doing this at least twice per day. 1 minute buffs count, single rounds of spike damage are not considered. This means the results show the suitability of each class for delivering consistent ranged damage - the standard archer niche.

The classes I looked at are: Bard, Fighter, Ranger, Rogue, and Warlock. The spoiler below shows the builds used. Each is fairly well optimized for these calculations - I tested a few iterations of each to find the best performers.



There are two cases - with feats and without. In both cases every build uses human and standard point buy array - under standard array it's not possible to get higher than a +3 in your primary stat, even if you're not human, and the only racial bonus which affects these calculations is proficiency in the longbow for Elf Bards and Rogues. If there are feats used, each build takes Sharpshooter at 1st. Sharpshooter -5/+10 almost always gives higher damage. At level 1 you're better off using Sharpshooter against any AC less than 18 (less than 20 if you're a fighter), and as proficiency and Dex increases, so does the threshold.

Each class increases Dex (or Cha for Warlock) every 4 levels, and gains proficiency bonus normally. Below are the relevant class features.

Bard (Valor)
1st: Shortbow
3rd: Longbow proficiency
6th: 2 attacks/round
10th: Takes Swift Quiver (4 attacks/round)
17th: Takes Foresight (attacks with advantage)

Fighter (Eldrich Knight)
1st: Longbow, Archery style
5th: 2 attacks/round
11th: 3 attacks/round
13th: Takes Haste (4 attacks/round)
20th: 5 attacks/round

Ranger (Hunter)
1st: Longbow
2nd: Takes Hunter's Mark (+1d6/hit)
3rd: Colossus Slayer (+1d8/round if hits)
5th: 2 attacks/round
17th: Takes Swift Quiver (no more Hunter's Mark, 4 attacks/round)

Rogue (Arcane Trickster)
1st: Shortbow, Sneak Attack (+1d6 per round if hit, scales)
2nd: Cunning Action (Modeled as giving the rogue advantage every round)
12th: Takes Skulker feat (to allow advantage if missing on first attack)
14th: Takes Haste (2 attacks/round)

Warlock (Any)
1st: Eldrich Blast, Hex
2nd: Agonizing Blast
5th: 2 attacks/round
11th: 3 attacks/round
17th: Takes Foresight for constant advantage, 4 attacks/round



Due to popular demand, I've added an analysis of Crossbow Expert as well.


This analysis assumes a liberal interpretation of the feat, so you can use the bonus action attack every round using a single hand crossbow. I don't necessarily agree with this interpretation of the feat, but this analysis will be helpful for those who do. In each case (except Warlock, obviously), it is beneficial to get the Crossbow Expert feat at 1st level, and Sharpshooter at 4th.

Bard: Crossbow Expert is recommended if you want to fully optimize. It's not as highly recommended as for the other classes due to the higher opportunity cost. Instead of Swift Quiver at level 10, you take Elemental Weapon. At level 15, you can cast it as a 7th level spell 2x/day. Note that this will compete with a magic weapon; getting a magic weapon will make everyone but the Bard (or Warlock) more effective. It also uses all of your 7th+ spell slots until 20th level.
Fighter: Crossbow Expert is highly recommended. You have no better use of your bonus action, and Sharpshooter gives you tons of bonus damage. Also, you have ability boosts to burn.
Ranger: Crossbow Expert is highly recommended. You have no better use of your bonus action for most of your career, and you have lots of bonus damage to pile on top of your extra attacks in addition to Sharpshooter. The only caveat is that you'll be killing enemies quickly enough that your bonus action will be eaten retargeting Hunter's Mark regularly.
Rogue: Crossbow Expert is not a must have, but it's not bad. It's slightly better than using your bonus action to gain advantage, assuming you have a melee ally around to ensure you can Sneak Attack.
Warlock: Crossbow Expert is not useful.



The results are in the charts at the bottom of the post. Here is a handy chart with the best classes at each level range. Any of the 5 classes get to be the best under at least one condition.

Level 1-5: Ranger or Rogue
Level 6-10: Ranger
Level 11-15: (With feats) Bard or Fighter (Without feats) Bard or Warlock
Level 16-20: (With feats) All but Warlock (Without feats) Warlock

I'll touch briefly on the results for each class.

Bard (Valor): The Bard is a strictly subpar ranged damage dealer until level 10 and Swift Quiver. Above level 10 the Bard is quite competent, and with Sharpshooter for bonus damage, is the best around at levels 10-12 and 17-19.

Fighter (EK): The Fighter is competent throughout his career due to a high attack bonus and number of attacks. The Fighter benefits a great deal from the Sharpshooter feat, but it takes a few levels to see this benefit since the -5 attack penalty is much stiffer at lower levels. With Sharpshooter, the fighter is the best around from 13-16 and at level 20.

Ranger (Hunter): The Ranger is the king of ranged attackers from levels 1-10, and does well after level 17 with Swift Quiver if feats are in play. The Ranger gets lots of bonus damage (especially Hunter's Mark) which is huge here. Volley is a pretty good at-will AOE which might make up for the Ranger's shortcomings from 11-16.

Rogue (AT): This one is a little suspect since I gave free advantage on every attack once Cunning Action shows up. Still, Sneak Attack bonus damage scales pretty well, especially if there are no feats to give most other classes bonus damage. There is a huge spike when the rouge gets Haste, since that doubles the chances of landing Sneak Attack that round. The Rogue is the top from level 1-4 (close to the ranger) as well as level 14-16 with no feats in play.

Warlock (Any): If there are no feats around, the Warlock will be the best ranged damage dealer once it takes the title from the Ranger at level 11. Hex and auto-scaling cantrips are quite effective. However, if other classes get bonus damage in the form of Sharpshooter, the Warlock falls behind. Warlock is never the best if feats are available, but with no feats it's the top from 11-13 and from 17 to 20.

http://i1368.photobucket.com/albums/ag187/altervir/NoSharpshooter_zps691710b8.png
http://i1368.photobucket.com/albums/ag187/altervir/WithSharpshooter_zps55c3dfd9.png
http://i1368.photobucket.com/albums/ag187/altervir/CrossbowExpert_zps352c3573.png

[Yorrin]

An interesting analysis! Did you account for stat bumps as they level eventually getting up to 20s?

I also notice Clerics aren't on here, despite Temptest and War Clerics both having martial weapon proficiency. War Clerics especially would be interesting, since Guided Strike offsets Sharpshooter to actually make it more accurate than a normal attack. Combined with Divine Strike and they'll probably look something like a ranger at levels 8+

[Grayson01]

How come you went with AT and not Assassin Rouge? If you were giving advantage everyshot I think the Assassin would net more damage?

[MeeposFire]

Another interesting idea is getting more damage via the hand crossbow (whether you require two or not I don't care so lets leave that debate out of this) for those classes that do not use their bonus actions in the above analysis. Lower damage die but lots of damage potential if you are not a class that can use swift quiver.

[MaxWilson]

Eldritch Knight can take Magic Initiate (warlock) to get the Hex spell from the Warlock list, which adds +1d6 to each of his attacks as well as imposing a (situationally-useful) disadvantage on one ability check. (Edit: oh, that doesn't stack with Haste I guess.)

On another note, I'm puzzled by the Warlock build. How does the Warlock get Foresight at 17th level? He only has up to 5th level spell slots according to the table on page 106 of my PHB, and there aren't any invocations that I see which grant Foresight.

[MeeposFire]

Are things like hex and hunters mark being considered for their basic classes in this diagram?

[Grayson01]

Are things like hex and hunters mark being considered for their basic classes in this diagram?

Yes the OP had them in the class build section.

Edit
Ranger (Hunter)
1st: Longbow
2nd: Takes Hunter's Mark (+1d6/hit)
3rd: Colossus Slayer (+1d8/round if hits)
5th: 2 attacks/round
17th: Takes Swift Quiver (no more Hunter's Mark, 4 attacks/round)

[MeeposFire]

Is EK the best for DPR comparisons for the fighter? I just wonder because I have not figured the increased crit range of champion and compare it to the couple times a day that an EK gets haste (which has an opportunity cost of not having somebody else casting it for him which granted being able to buff yourself is awesome).

[Darkmatter]

@Yorrin, Yes, each class starts with a 16 in the relevant score, and goes up by 2 at 4th and 8th (6th for Fighter). I didn't include Cleric because they do not have a way to get an extra attack for 20 rounds a day (which was my threshold). Without extra attacks or a level-scaling damage bonus like Sneak Attack, you can't really compete. Divine strike for +2d8 is nothing compared to 1 more attack at 1d8+15. Let me know if I've overlooked a key spell on the Cleric spell list which might offset this shortcoming.

@Grayson: This is an analysis for sustained ranged damage over a 10 round combat, not including spike damage. Assassin only gets their free crit on surprise round attacks, so for this analysis it's inferior - it basically comes out to a little bit less than an extra hit. I may do a 1 round spike damage analysis at some point, but that's a much more involved project since so many more classes can compete.

@MeeposFire: You're right about the hand crossbow (for Fighter, but not the Rogue, who really needs that bonus action to hide and gain advantage to remain competative). - extra attacks are the biggest driver of extra damage. I didn't include it in the discussion because I really think sustaining that bonus action attack with a "loaded hand crossbow" isn't going to fly in most groups.

The extra attack that Haste gives vastly outweighs the bonus crit from Champion. +1 crit range increases damage by ~2.5% with no feats and ~1% with feats, while +1 attack increases damage by 25-33%. If you got bonus damage dice from a magic weapon or something that would start to tip things toward the Champion, but it would need to be a whole lot of dice. You could weigh the opportunity cost given that Haste requires an action to cast and can only be done 2x/day until 16th level, but in this analysis, the Champion is clearly inferior after 14th level.

@MaxWilson: My rule about this applying to two fights per day made me decide not to apply Magic Initiate feat to this analysis, though technically you could have two 10 minute battles within the 1 hour duration. I did test out the effects, and it is a great way to boost fighter damage at lower levels. It still doesn't beat the ranger, though, who still has the +1d8 vs. a wounded enemy. Hex gets edged out by Haste once that comes on the scene, even with 4 attacks.

Warlocks have an ability that's separate from Pact Magic called Mystic Arcanum through which they can cast 6th-9th level Warlock spells, one of which is Foresight.

[MaxWilson]

Much obliged for pointing out Mystic Arcanum. I was relying too much on the chart and had completely overlooked the Mystic Arcanum. That makes warlocks significantly more okay than I thought.

Also I hadn't realized that you couldn't cast Magic Initiate-gained spells with regular spell slots. Ah well.

[Yuukale]

any chances to add multi-classing ? I mean, rogue 3 (assassin) might do something to help almost everyone's damage, no?

[Tenmujiin]

Isn't there a feat that duplicates sharpshooter for spell-casters?

Edit: nevermind, it doesn't give the -5/+10

[Darkmatter]

@MaxWilson: So there is a little debate on whether the intention was that Magic Initiate can be added to existing spell slots. My reading (and I think the most common reading) is that you just get that spell once/day.

@Yuukale: Multiclassing will give you the best damage if you're targeting a narrow range of levels, no question. However, it makes you fall behind otherwise since both sets of class features (and ability boosts) get delayed. There are way too many permutations of multiclassing for me to run all of them. I tested assassin 3/fighter 17, just for kicks. It performs well relative to everyone else at levels 2-3 (it's a Rogue) 8-9 (close to Ranger), at 14-15 (close to Fighter), and 16-20 (better than everyone else until Fighter edges it out at 20). However, this is still modeling Cunning Action as giving always on advantage, like Foresight. This is absolutely not relevant in the 4 attacks per round fighter case (you'd only get advantage until your first attack hits). Without that constant advantage, this build is a poor to average performer until levels 14+, at which point it becomes competitive with the Fighter, Bard, and Ranger. You do a bit better at lower levels by going Rng5->Rog3, but this build loses steam at higher levels and also as competing bonus actions (Cunning Action vs. Hunter's Mark retarget).

I've run a handful of multiclass tests with similar results - you get a few levels where you're the best in the business, but then someone else gets a cool feature which will overshadow you. So, basically just the same as single classes, but you can shuffle around the levels at which you're the best. I found only one corner case where this general rule that you have to suck now to be awesome later (or vice versa) didn't apply. In the absence of feats, you can build a Warlock 2 / Bard X which is as good as a single classed Warlock at almost every low level, jumps ahead with Swift Quiver + Eldrich Blast at level 12 (Brd10), then really steamrolls everyone at levels 16+ (Brd14) by pumping out 4 EB's plus one longbow shot a round boosted by Hex. Of course, you have to take Swift Quiver and EB as Bard spells for this to work, so it's a pretty specialized build. Also, it's not really that good compared to other classes if Sharpshooter is available - in that case you'd be better off going pure Fighter at almost every level.

Overall, this exercise has given me a lot of respect for the balance of multiclassing in 5e. There are very few dips without consequences.

[numerek]

Obviously it doesn't matter once the warlock gets foresight but if your giving the rogue advantage on all attacks, a pact of chain warlock could get advantage from their familiar every round and none of the rest of the analysis takes into account any action by the enemy.

Also arcane trickster gets a more consistent way to get advantage at 13th level no need for skulker feat.

Also in this analysis the sharpshooter feat builds are getting subsidies from the low ac creatures. I have done similar analysis before and for low ac creature the damage from sharpshooter and great weapon master go up fast. So yes against a mob with 900hp but 12ac a 20th level fighter with sharpshooter will probably win. I'm not so sure about a mob with 560hp and 20ac. when the fighter has +8 to hit or +13 to hit but -10 to damage. and the warlock has +11 to hit with advantage.

regarding the multiclass bard/warlock if you are going to only attack once with your weapon you could do heavy crossbow for 1 extra damage. hex on 5 attacks is 17.5*accuracy%, an attack with the crossbow with the sharpshooter feat would be 20.5*lower accuracy%. and two eldritch blasts is 21*accuracy%
It seems like this build would beat the single class bard as well with feats, if you use swift quiver instead of hex. the four eldritch blasts deal more than 2 sharpshooter arrows and with higher accuracy and the additional sharpshooter arrow deals more than 5 hex boosts. Sharpshooter from human variant, then 2 ASI to dex 2 ASI to cha. Which I believe puts it in 2nd place behind 20th level fighter. If you give up 2 damage and go with shortbow this multi class doesn't need extra attack or battle magic which means you could go lore bard and get d20+8+d12 with advantage for initiative giving it a good chance of being first to act and can do this for 2 combats(actually 5).

[Darkmatter]

@Numerek

The Warlock's familiar giving advantage is good for one attack - that's a bit annoying to simulate, and actually doesn't make a huge difference in the end. Warlock will still be pretty good if feats aren't available, and it doesn't make up the difference if competing against Sharpshooter.

Good catch on the AT Mage Hand advantage. I was thinking it was just on the first attack - that is a much better ability to combo with haste on rereading.

I'm not too worried about a skew toward low AC. First, I did check the AC's of creatures in the Basic Rules when choosing 12 to 20. The average AC of all those creatures with CR 1 and above is 13.3, so if anything the analysis is skewed toward high AC unless there are a high proportion of custom humanoid enemies. Also, as part of the analysis I checked when it is a good idea to use sharpshooter. For every class and level (including Rogue) it is always better to use Sharpshooter at AC's of 17 or below, at all levels. This goes up steadily with level as attack bonus increases. By level 9 the fighter's attack bonus is high enough that it's better to use Sharpshooter even vs. the 25 AC of the Tarrasque. All the classes get to that point by level 17.

The build you're describing (Warlock Eldrich Blast plus Swift Quiver from Bard) was actually the first multiclass I thought of testing when I started this. At 19th and 20th levels all its advantages come online and it's better than single classed Bard. At almost every other level (the exception being 5th) it's worse than single classed Bard. It's especially poor at levels 10 and 11, when the single classed Bard is already enjoying Swift Quiver. It's another example of multiclassing being the best in a narrow level range, but inferior otherwise. The build I described is unique among the multiclass builds I tested because for this given game (no feat use) it is always better or equal to single classed Warlock or single classed Bard.

Basically, the game is rarely played at level 20, but it seems like that's where all the optimization is focused. This analysis is meant to help folks figure out what's likely to be effective at the level range they plan on playing.

[Demonicattorney]

@Numerek


Basically, the game is rarely played at level 20, but it seems like that's where all the optimization is focused. This analysis is meant to help folks figure out what's likely to be effective at the level range they plan on playing.

Couldn't agree more. Many "optimized" multi-class builds are fairly bad while actually playing. Remember that your campaign ends at 20 it doesnt begin there.

[mister__joshua]

In the absence of feats, you can build a Warlock 2 / Bard X which is as good as a single classed Warlock at almost every low level, jumps ahead with Swift Quiver + Eldrich Blast at level 12 (Brd10), then really steamrolls everyone at levels 16+ (Brd14) by pumping out 4 EB's plus one longbow shot a round boosted by Hex. Of course, you have to take Swift Quiver and EB as Bard spells for this to work, so it's a pretty specialized build. Also, it's not really that good compared to other classes if Sharpshooter is available - in that case you'd be better off going pure Fighter at almost every level.

I don't really understand this analysis and I'm curious, as Archer Bard and Warlock Bard were the 2 builds I considered for my character.

You said it's 'pumping out 4 EB's plus one longbow shot a round boosted by Hex'. Is this switching back away from using Swift Quiver then? Why wouldn't it be 4EB's plus 2 longbow shots and no Hex?

I also don't get why you'd need to take EB as a Bard spell. Isn't that the point of the Warlock levels?

[Darkmatter]

I don't really understand this analysis and I'm curious, as Archer Bard and Warlock Bard were the 2 builds I considered for my character.

I wasn't too specific on the build, and I can see the confusion. The build uses different tactics at different times as they become more or less optimal.

Level 1: (Warlock 1) EB, Hex
Level 2: (Warlock 2) Agonizing Blast
Level 3-11: (Bard 1-9)
Level 12: (Bard 10) Take Swift Quiver
Level 14: (Bard 14) Battle Magic and take EB with Magical Secrets

Levels 1-11 you're just using Eldrich Blast like a Warlock. You're slightly behind in damage four of these levels due to your delayed Cha boost, but it's a pretty small amount since most of your damage is EB+Hex base damage. At level 12 you get Swift Quiver, and it becomes slightly more beneficial to spend concentration on Swift Quiver instead of Hex. At level 14 you get Battle Magic, and the best trick of the build surfaces, which is using EB + Battle Magic and Hex to deliver 4x 1d10+1d6+5 EB damage plus 1d10+1d6+5 heavy crossbow damage. You can't use Swift Quiver and Hex because both are concentration, and Hex on 5 attcks (5d6) is better than one extra attack at 1d8+5. In order to use Battle Magic with EB, you need to take it as a Bard spell, unfortunately. Starting at level 12, this build is superior to the single classed Warlock, and from level 16, it deals around 25% higher damage.

Note that this build is only this comparatively effective in an environment with no feats available, which is likely to be unusual. If Sharpshooter is available, a Dex based Swift Quiver build is equivalent or superior (except at level 5). One thing this build does have going for it is allowing your Bard to pump Cha at lower levels before turning to Dex at higher levels without falling behind too far in damage.

Finally, damage isn't everything, even though it's the only thing I'm concerned with in this analysis. The opportunity cost of always being one spell level behind in Bard spells is more significant than I think most optimizers realize.

[Easy_Lee]

In the warlock's defense, I think it's important to note the consistency with which eldritch blast brings the damage. It's force and based purely on charisma, infinite times per day, and long range (extremely long if you take spell sniper or the invocation).

Also, you can get a huge selection of aoe spells for even more consistent dps by going the popular warlock 2 / sorcerer x build. Since spell slots scale with total caster level this edition, the build doesn't suffer as much from delayed progression. Talk your DM into force dragon sorcerer ancestry for even more damage.

That said, I don't know if OP's analysis takes magic items into consideration. If not, a +3 bow could shift things pretty heavily in the ranger's favor for DPR under optimal circumstances.

[Easy_Lee]

Finally, damage isn't everything, even though it's the only thing I'm concerned with in this analysis. The opportunity cost of always being one spell level behind in Bard spells is more significant than I think most optimizers realize.

It also doesn't take into consideration the other benefits of each build.

A dex build has great initiative and survivability. The rogue variant will often get the surprise round and win initiative, yielding early burst.

A warlock with CHA and other spells has enormous benefits for playing party face, casting spells for all kinds of useful effects, etc. And for DPR, the biggest boon here is that it's force damage that never runs out. That means minimum chance of damage reduction and no reliance on gear or ammo. Agonizing EB brings the damage, no questions asked.

The swiftquiver bard may end up with the best of everything, since it has skills out the whazo and great utility spell selection. But it also has MAD, and is a late bloomer.

As person_man said in another thread, I don't think damage is a huge consideration. It appears the damage-focused character choices are all reasonably close. It's encouraging that playing well will have a much larger impact in battle than character selection.

[MaxWilson]

Also, you can get a huge selection of aoe spells for even more consistent dps by going the popular warlock 2 / sorcerer x build. Since spell slots scale with total caster level this edition, the build doesn't suffer as much from delayed progression. Talk your DM into force dragon sorcerer ancestry for even more damage.

Per the multiclassing rules, warlock levels don't count as caster levels for spell progression--warlock slots get added on separately.

[Easy_Lee]

Per the multiclassing rules, warlock levels don't count as caster levels for spell progression--warlock slots get added on separately.
Ah, that's unfortunate. Puts a bit of a damper on the build, then.

[Rummy]

This is a fantastic analysis. One question... Would going battle master and using precise attack to turn misses into hits up the damage from EK? It obviously would before haste is avaliable. Also, having buddies cast haste and bless is the best course of action.

[Shining Wrath]

I notice that both with Sharpshooter and without, best and worst are within 50%; that is, best DPR < 1.5*worst DPR.

I won't argue that 50% doesn't matter, but I will argue that you might want to consider other factors of the build even if your primary focus is ranged damage. Bards, rogues, and Warlocks may have more out-of-combat utility.

[obryn]

I would certainly include the dual hand-crossbow numbers, here. It's totally functional for any kind of RAW analysis as you're doing here.

I would also try the Battlemaster Fighter, since those added damage dice will add up. (And some of them increase accuracy.)

[MeeposFire]

I would certainly include the dual hand-crossbow numbers, here. It's totally functional for any kind of RAW analysis as you're doing here.

I would also try the Battlemaster Fighter, since those added damage dice will add up. (And some of them increase accuracy.)

Yes regardless of whether you think you need one or two crossbows or even if you can do it after the first round we can put up the numbers for the first round which there is no doubt it would work (you can have two loaded hand crossbows on round one the argument is really whether you can do it after round one when trying to reload). Since it is possible in that first round you can have those numbers up for that one round and that way if you do allow it work after round one you can use the analysis (individual games can decide that on your own and your tables would be more complete).

[Darkmatter]

I don't know if OP's analysis takes magic items into consideration.

It does not, and this point is something I meant to have mentioned. This will definitely skew things a bit further away from the Warlock, unless the DMG provides for enchanted implements.

We're in complete agreement that damage is not the only selling point of a build - it's just the only thing that can be arithmetically compared (well, apart from AC, HP and healing, I guess). Is it better to have telepathy or a pet wolf? The answer is "It depends." A lot of what it depends on is simple player preference.


I would certainly include the dual hand-crossbow numbers, here.

Since there have been so many requests, I'll append an analysis with Crossbow Expert as well.


I notice that both with Sharpshooter and without, best and worst are within 50%; that is, best DPR < 1.5*worst DPR.

I won't argue that 50% doesn't matter, but I will argue that you might want to consider other factors of the build even if your primary focus is ranged damage. Bards, rogues, and Warlocks may have more out-of-combat utility.

Absolutely. This is just to compare builds whose primary purpose is laying down the hurt at range. Secondary purposes are really important - the Bard will have a good deal more utility than the Fighter, for example, even if delaying Cha increases.

Regarding Battlemaster, there are too many unknowns to properly analyze the maneuvers. My feeling is that after getting Haste, the EK will still be a good bit better; I can tell you for sure that an extra 1d12 damage 6 times per combat does not compete with an extra attack at 1d8+15 each round from haste. Using them for precision would help much more at lower levels, when you're missing a more significant amount of the time.

@MeeposFire: The one round test you're describing is a spike damage analysis, which requires much more time that I'm willing to devote at the moment. Limiting myself to abilities which last 10 rounds or longer cuts down drastically on the builds I need to simulate to make a reasonable analysis.

[Easy_Lee]

In regards to the updated post with crossbow expert...I did not expect crossbow expert fighters to perform so well. For both rangers and fighters, the feat gives consistent damage at any range, less MAD, and more consistent damage with bonus-action usage. Throw a magic crossbow on top of that and you've got one hell of a machine gun.

Since it's DEX, no reason those characters couldn't pick up some impressive stealth skills too, though nothing to rival the rogue of course.

[MeeposFire]

It does not, and this point is something I meant to have mentioned. This will definitely skew things a bit further away from the Warlock, unless the DMG provides for enchanted implements.

We're in complete agreement that damage is not the only selling point of a build - it's just the only thing that can be arithmetically compared (well, apart from AC, HP and healing, I guess). Is it better to have telepathy or a pet wolf? The answer is "It depends." A lot of what it depends on is simple player preference.



Since there have been so many requests, I'll append an analysis with Crossbow Expert as well.



Absolutely. This is just to compare builds whose primary purpose is laying down the hurt at range. Secondary purposes are really important - the Bard will have a good deal more utility than the Fighter, for example, even if delaying Cha increases.

Regarding Battlemaster, there are too many unknowns to properly analyze the maneuvers. My feeling is that after getting Haste, the EK will still be a good bit better; I can tell you for sure that an extra 1d12 damage 6 times per combat does not compete with an extra attack at 1d8+15 each round from haste. Using them for precision would help much more at lower levels, when you're missing a more significant amount of the time.

@MeeposFire: The one round test you're describing is a spike damage analysis, which requires much more time that I'm willing to devote at the moment. Limiting myself to abilities which last 10 rounds or longer cuts down drastically on the builds I need to simulate to make a reasonable analysis.

To be honest the only reason I mention that particular type of spike damage is that it just so happens to be the same damage that is constant assuming you allow either dual hand crossbows or the single hand crossbow working with crossbow expert (both types have the same damage results assuming they are the same type of crossbow). So in reality I am jsut asking for the crossbow expert results and just dressing it up so that way if you don't like either dual crossbows every round or the single hand crossbow idea it won't drive them crazy since you can list it as "spike damage" for a corner case (with a wink wink nudge nudge to those of us that actually allow for dual crossbows or single hand crossbow expert users).

EDIT: Never mind I just saw your update. As expected it makes the fighter a bit more damaging and gives the rogue a little bit more of a boost. Not too bad. It looks like it also makes an alternative for a bard that does not want to step on the rangers toes by stealing swift quiver. For the ranger it seems to make an alternative where you spend a feat so you don't have to rely on your higher level slots for similar damage (hunters mark instead of swift quiver). Very interesting.

[numerek]

@Numerek

The Warlock's familiar giving advantage is good for one attack - that's a bit annoying to simulate, and actually doesn't make a huge difference in the end. Warlock will still be pretty good if feats aren't available, and it doesn't make up the difference if competing against Sharpshooter.

I'm not too worried about a skew toward low AC. First, I did check the AC's of creatures in the Basic Rules when choosing 12 to 20. The average AC of all those creatures with CR 1 and above is 13.3, so if anything the analysis is skewed toward high AC unless there are a high proportion of custom humanoid enemies. Also, as part of the analysis I checked when it is a good idea to use sharpshooter. For every class and level (including Rogue) it is always better to use Sharpshooter at AC's of 17 or below, at all levels. This goes up steadily with level as attack bonus increases. By level 9 the fighter's attack bonus is high enough that it's better to use Sharpshooter even vs. the 25 AC of the Tarrasque. All the classes get to that point by level 17.



If you are referring to one attack per turn then yes, but at 3rd and 4th level that is every attack. and the warlock in your chart at 3rd level is doing far less than 1d10+1d6+3=12 so accuracy is affecting the result.

I'm really just wanting it to be clear that the warlock's damage is less effected by creatures ac. Another thing to think about is the fighter has to carry around at least 75 arrows/90 bolts (this is with retrieving 25/30 after the first fight) to pull this off 50/60 of which need to be easily accessible to allow drawing and firing 5 in 6 seconds. I didn't find much online but the largest quiver I found held 2 dozen and I believe that was custom.

Another option for bard build is elemental weapon from the paladin spell list. With it a longbow can do without feats d8+3d4+8=20*higher accuracy%*2 vs d8+5=9.5*accuracy%*4.

crossbow expert hand crossbow 3 attacks would deal 1d6+3d4+18=29*higher accuracy%*3 vs d8+15=19.5*accuracy%*4 and the accuracy better than archery fighting style. does it at level 17(the damage would actually be there at 15th using 7th level and 8th level slot to cast elemental weapon twice as a 7th level spell, the next 2 levels would be increasing accuracy).

Of course this build competes with any magical weapons you may find but you can always trade out the spell at the next level up or let someone else in the party have the magical weapon.

I came up with a single class addition that I think does well on the without feats chart. A draconic bloodline sorcerer can cast 2 firebolts a turn around 20 times a day starting at level 9 via quicken spell and 40 sorcery points, then at 17th level you can wish for hex or hunter's mark, then at 19th level you can cast scorching ray and firebolt without wish via 97 sorcery points, then at 20th level you can cast wish again. 4d10+10d6+20=77*accuracy% vs warlock 4d10+4d6+20=56*higher accuracy%

As for multiclassing messing up damage at different level it is easy to apply a multiclass dip to classes at the end, bard maxes out at 17th level so if he took 3 levels of fighter or ranger or rogue after 17th level rogue would be 2d6 or would get +2 to hit with the elemental weapon build his extra to hit would match the feats negative to hit and either improved critical which is something like 9.25% higher critical chance with advantage or the 1d8 from ranger's hunter. Ranger could do the same thing with fighter levels and grab improved critical. rogue maxes out at 19th level so it could take one fighter level for the archery fighting style and longbow proficiency.

The only thing warlock gets after 4th level is more ASI and foresight, if you switch to sorcerer after 4th by 13th character level(slightly less if you allow short rest) you will be able to double eldritch blast every turn. Which would end up being 8d10+8d6+40=112*accuracy%

[Darkmatter]

A short treatise

So, you've got a couple of nice high level builds there, and I hope this analysis helps you find more. Multiclassing at the end of the progression can definitely pump some numbers, but that is going to be a minority of games in actual play.

Elemental weapon is inferior than SQ on a Bard with 2 attacks, but it's a better than SQ with 3, especially with a level 7 slot. It also lasts longer, so if your DM goes for it, that's the way to roll.

That single class sorcerer build is going to be pretty poor at low levels, and even at high levels it's burning through daily abilities pretty prodigally. It's a nice feat of high level optimization tuned specifically to perform well at this analysis, though.

The Warlock / Sorcerer burns through daily abilities as well, but not as quickly. It's a pretty solid SAD build, competent at most levels.d

I did look into your point about the warlock's familiar granting advantage at low levels. The familiar comes along at 3rd, at which you'd get the biggest boost. That boost is significant: from ~7.5 to ~10.4, which is enough to be competitive with a no-feat Ranger from level 3 on. If you want to ballpark the calcs, it will be a little less than +40% damage at 3rd and 4th, around +20% damage from 5th to 10th, and something like +10% from 11th to 16th as attack bonus increases.

[Suichimo]

Gotta say, I really hate how much Paladins ABSOLUTELY SUCK at ranged combat. They are so disincentivized to it that it isn't funny. Heavy armor and CHA to saves downplays DEX, no ranged combat style, and they can't use their Smite or Improved Smite with ranged attacks.

If you want range you have to go with a ****ty thrown weapon or use a feat/multiclass to get Eldritch Blast.

[numerek]

Gotta say, I really hate how much Paladins ABSOLUTELY SUCK at ranged combat. They are so disincentivized to it that it isn't funny. Heavy armor and CHA to saves downplays DEX, no ranged combat style, and they can't use their Smite or Improved Smite with ranged attacks.

If you want range you have to go with a ****ty thrown weapon or use a feat/multiclass to get Eldritch Blast.

Yeah same thing goes for barbarian, and monk. I have tried to figure out how to make a thrown weapon ranged build but as far as I can tell you can't RAW draw more than 2 thrown weapons a turn.

[Suichimo]

Yeah same thing goes for barbarian, and monk. I have tried to figure out how to make a thrown weapon ranged build but as far as I can tell you can't RAW draw more than 2 thrown weapons a turn.

On top of that, I forgot to mention, their maximum effective range is 30' for a javelin. Anything after that and you're at disadvantage. You may as well just walk your happy ass up to the enemy and hit it with a real weapon.

[numerek]

Elemental weapon is inferior than SQ on a Bard with 2 attacks, but it's a better than SQ with 3, especially with a level 7 slot. It also lasts longer, so if your DM goes for it, that's the way to roll.


I'm guessing here you are meaning with feat cause without feat unless I'm missing something I'm pretty sure elemental weapon beats SQ 2 attacks vs 4. At fifth level slots it is a little lower base damage but higher accuracy so I think even if sq wins here it is not by much and as I mentioned the bard could swap out spells at 15th level when he can use 2 level 7 slots.

I appreciate all the work you have done, this has been a fun analysis to play around with.

[Darkmatter]

I'm guessing here you are meaning with feat cause without feat unless I'm missing something I'm pretty sure elemental weapon beats SQ 2 attacks vs 4. At fifth level slots it is a little lower base damage but higher accuracy so I think even if sq wins here it is not by much and as I mentioned the bard could swap out spells at 15th level when he can use 2 level 7 slots.

I appreciate all the work you have done, this has been a fun analysis to play around with.

Yeah, I looked into the math on elemental weapon a bit more. It makes Crossbow Expert much better for the Bard assuming you take the interpretation that you use a single crossbow. With no feats, SQ beats Elemental Weapon (12% swing) until the 7th level boost becomes available (level 15-16), in which case Elemental Weapon is a little better (8% swing). Once Foresight hits at 17, the accuracy boost isn't as important, and you're only 3% better with Elemental Weapon. According to my rules, we can't apply the 7th level boost until you get two of those per day (so 15th level with a 8th level slot). With just Sharpshooter, SQ is always better, as you might imagine. I'd say the competition with a magic weapon makes EW not worth it in the no feat case. I'll update the OP for the Crossbow Expert feat at least.

It's worth noting that trading out Magical Secrets is also shaky rules territory. RAW you can trade them, but in return you can only choose Bard spells. There's nothing stopping you from just taking both Elemental Weapon and Swift Quiver, it just eats up more resources.

[numerek]

I came up with a single class addition that I think does well on the without feats chart. A draconic bloodline sorcerer can cast 2 firebolts a turn around 20 times a day starting at level 9 via quicken spell and 40 sorcery points, then at 17th level you can wish for hex or hunter's mark, then at 19th level you can cast scorching ray and firebolt without wish via 97 sorcery points, then at 20th level you can cast wish again. 4d10+10d6+20=77*accuracy% vs warlock 4d10+4d6+20=56*higher accuracy%

I realized that this build could get a small boost from feats, magic initiate and elemental adept. The elemental adept damage adjustment is minimal(1-1.5 damage depending on whether it applies to hex dice for fire spells) though could still be useful against creatures resistant to fire. Magic initiate gives access to eldritch blast and depending on interpretation either 1/day hex or you could use spell slots to cast it. You don't get charisma modifier added to damage but you do get to apply hex which would beat the damage that firebolt would do. if nothing else it changes the high end calculation to 4d10+13d6+15=82.5(83.5/84 w/ elemental adept), if "You learn that spell" means you can cast it with your spell slots casting it as a 3rd level spell has it last 8 hours and as a 5th level spell has it last 24 hours. Then it would improve make eldritch blast beat firebolt at 5th level then at 6th level firebolt damage wins til 11th level when the sorcerer can switch to eldritch blast as the additional hex damage would beat the charisma modifier. If you go by this it definately helps the build do better in the lowers levels, still will be the worst til you can cast 2 cantrips a turn but not by as much. Also if you would like to use your daily abilities for other things this gives better damage when only using 40 points to double cantrip of 8d10+8d6=72 vs 8d10+2d6+10=61(61.8/62.13 w/ elemental adept)

83.5 compares pretty well with the non-crossbow expert feat chart(which I'm pretty sure it wasn't the designers intent to put in a way to attack faster with crossbows than with bows), and has higher accuracy than all but the warlock whose damage can't compete with the feats.

[numerek]

Another high level optimization, This may seem crazy but I'm going to use 8 fifth level spell slots or above. And that's not all I'm going to give up 3 rounds of combat and use up 2 more magical secrets, So first round contagion blinding sickness, then second round bestow curse targeting constitution, third round contagion flesh rot. So first round applies the blinded condition and gives disadvantage on wisdom saves thus setting up round two which adds d8 to all your damage rolls and gives disadvantage on constitution saves thus setting up round three and backwards sets up round one which makes them vulnerable to all damage from anybody. After that is business as usual swift quiver it up. Oh and thanks to battle magic we actually get to make 3 attacks during those turns.

Spell casting turns: 3d8+10+2*(2d8+5)=51.5 total for the three turns
No Feat Damage: 2*4*(2d8+5)=112 per turn
Damage average over 10 turns: 83.55 + all the extra damage everybody else in the party is doing.

Spell casting turns: 3d8+30+2*(2d8+15)=91.5 total for the three turns
Feat Damage: 2*4*(2d8+15)=192 per turn
Damage averaged over 10 turns: 143.55 + all the extra damage everybody else in the party is doing.

But wait a minute this is a ranged thread and contagion and bestow curse are touch spells. That is where the second magical secret that I mentioned comes into play, find familiar can deliver touch attacks for you. Touch attack with a dislocated fly speed is ranged in my book. That blinded condition will help keep that familiar alive. If the familiar dies before the 3rd spell is cast, you start using swift quiver as normal and saved yourself some high level spell slots and the action economy is alittle balanced since the enemy had to use some of its attacks to kill the familiar, first turn got all your allies advantage and target disadvantage also more susceptible to will saves, second turn disadvantage on constitution saves and must make a wisdom save at disadvantage or lose their action and an extra d8 on all your attacks, this close to makes up for the 2 rounds you spent casting spells (it does for no feat version).

But wait another minute bestow curse is a concentration spell, not if cast with a level 5 slot or above.

Unfortunately this can only be accomplished at level 19. Though at 13th level you can still use your familiar to deliver bestow curse, the target will have a better chance to save against it but if it hits it will make up for the damage lost in the first round plus other benefits using up 2 fifth level slots 1 sixth level slot and 1 seventh level slot for twice a day, at level 14 when you can get your 3rd and 4th magical secret, you could do any of the contagions slimy doom seems the most likely since it makes contagion more likely to last for the duration also the stunned condition gets you advantage much sooner than the bard was previously getting it. At 17th level you can cast 2 contagions most likely slimy doom and flesh rot.

One other note instead of battle magic you could swift quiver plus extra attack first round then on rounds 2-4 cast the spells while getting 2 attacks with your bow, but you will be doing less total damage if all 3 spells stick, if however flesh rot doesn't stick then you are better off going this route.

And finally, obviously this build needs to spend 2 of its ASI on charisma for harder save dcs.

I think this damage is nothing short of amazing, especially since this is a single classed build with only one feat.

[Xetheral]

Also, as part of the analysis I checked when it is a good idea to use sharpshooter. For every class and level (including Rogue) it is always better to use Sharpshooter at AC's of 17 or below, at all levels. This goes up steadily with level as attack bonus increases. By level 9 the fighter's attack bonus is high enough that it's better to use Sharpshooter even vs. the 25 AC of the Tarrasque.

I believe you have made a mistake somewhere in your Sharpshooter analysis, or maybe I have. Consider:

A third level hunter ranger with a longbow, archery style, and colossus slayer has +7 to hit and deals 2d8+1d6+3 damage. That's 15.5 average damage. Against AC 17 the Ranger has a 55% chance to hit, for average applied damage of 8.525. With sharpshooter the Ranger has a 30% chance to hit and deal 25.5 points of average damage, for an average applied damage of 7.65. Sharpshooter only becomes better at AC 15 and below for this character, which contradicts your assertion that the breakeven point is AC 17 for all classes.

A third level arcane trickster rogue with a shortbow has +5 to hit (with advantage) and deals 3d6+3 damage. That's a 69.75% chance to hit AC 17 to deal an average of 13.5 points of damage. Average applied damage is therefore 9.416. With sharpshooter that becomes a 36% chance to deal 23.5 points of damage. Average applied damage with sharpshooter is therefore 8.46. Sharpshooter only becomes worthwhile for this character at AC 16 and below.

The discrepancy isn't only at low levels, either:

A thirteenth level arcane trickster rogue with a shortbow has +9 to hit (with advantage) and deals 8d6+5 damage. That's a 87.75% chance of dealing 33 damage, for an average applied damage of 28.958 points. With sharpshooter that becomes a 64% chance of dealing 43 damage, for an average applied damage of 27.52 points of damage. Sharpshooter only becomes worthwhile for this character at AC 15 and below. (And in rounds where the Rogue doesn't get Advantage, Sharpshooter is only worth using at ACs of 10 and below.)

Being off by a point or two of AC for some builds isn't a huge problem, and isn't likely to change your numbers much. However, it does suggest a flaw in the methodology. Another possibility is that maybe I missed some source of attack bonus you were relying on when calculating my examples?

If you find it helpful, I've posted a table of breakeven points (http://www.giantitp.com/forums/showthread.php?373572-GWM-Reference-Table&highlight=gwm) for GWM that also applies for Sharpshooter.

For spreadsheet purposes, the formula for breakeven points that I used to make the table was: 10/([%HitNormal]/[%HitSharpshooter]-1)

Edit: Fixed Typo

[Easy_Lee]

On top of that, I forgot to mention, their maximum effective range is 30' for a javelin. Anything after that and you're at disadvantage. You may as well just walk your happy ass up to the enemy and hit it with a real weapon.

Find steed helps with that, at least. Riding a mount gives you a free dash on your turn, so you can move double the mount's speed in one round. Take mounted combatant, and you get advantage on attacks against anything smaller than your mount, in addition to your mount gaining evasion (which effectively means you have it). A glaive-wielding mounted paladin is a force to be reckoned with.

[Darkmatter]

I believe you have made a mistake somewhere in your Sharpshooter analysis, or maybe I have.

Wow, thanks for the check. This was the sort of thing I did regularly when I was starting the spreadsheet, but it looks like I made an error when making the thing prettier. I'm pretty upset with myself for not double checking afterwards. I wound up double counted prof bonus. As you might imagine, this changes the spread very significantly. Within a single class it doesn't do much, but the class to class comparisons are completely thrown around. Foresight becomes much more significant. I'm going to put a placeholder on the first post since I don't have time at the moment to regenerate my charts and analysis.

[Xetheral]

Wow, thanks for the check.

Glad I could help! I look forward to seeing the revised results... you've done a very interesting analysis here.

[imaginary]

No update on this thread or the calculations I imagine then?

Anyway, even with the error, it still is pretty useful. Thanks!

[tchad78]

Edited - I somehow missed a key wording!

[Foxhound438]

Is EK the best for DPR comparisons for the fighter? I just wonder because I have not figured the increased crit range of champion and compare it to the couple times a day that an EK gets haste (which has an opportunity cost of not having somebody else casting it for him which granted being able to buff yourself is awesome).

not to mention precision attack's synergy with sharpshooter

[PoeticDwarf]

The bard comes really high. How do you use swift quiver. He can't use it always...
Everyone can get hex with a feat next to that...

[Foxhound438]

Lvl 1: Fighter with Crossbow Expert Feat and Dueling style for +5 damage on each shot with your 1handed xbow (per sage advice 2015/01/12 crossbow-expert-bonus-action)


look at the actual rules. dueling fs only applies to melee weapons. says so in plain english.

[tchad78]

look at the actual rules. dueling fs only applies to melee weapons. says so in plain english.

Oh cool, not sure how I missed that - that makes it def not a worthwhile multi-class option. Nice.

*EDIT* Ahh, the character builder App doesn't specify melee weapon... That's how I missed it.

[Finger6842]

I love theory crafting as the parameters often drive the conclusion. As I understand this thread it's ranged damage of any kind against a single target. It does not take into account anything the target will do to nullify or reduce attacks so we are talking about a target here. A mobs ability to close to 5' and make non feat attacks at disadvantage is also not accounted for. What kind of change to the numbers would you get if you put each class in the same support party?

I was recently part of a group that was pretty optimized for survival (all level 20). A wizard, A warrior, A Barbarian, A Priest and me, A Bard18/Wizard2. For magical secrets I wound up with Counterspell, Haste (target’s speed is doubled, it gains a +2 bonus to AC, it has advantage on Dexterity saving throws, and it gains an additional action on each of its turns), Finger of Death (Zombie army and 100 damage), Simulacrum (Double 1 individual), and Meteor Swarm (40d6). My 9th bard spell was Foresight (can’t be surprised and has advantage on attack rolls, ability checks, and saving throws. Additionally, other creatures have disadvantage on attack rolls against the target). Other spells of relevance are Find Familiar (advantage on all attacks for the rest of us), Bless (+1d4 to all attacks for everyone), Shield and Ray of Frost (10' movement reduction cantrip).

The Champion fighter is optimized for ranged damage with Crits on 18,19 and 20, Sharpshooter (no disadvantage on range) +Lucky (no 1's) + Martial Adept(precision (+1d6 attack) / pushing attack (15' push away) feats and the Archery (+2 Attack) / Defense (+1 AC) fighting styles.

The wizard and I both have a Simulacrum of the warrior. I toss Foresight on the warrior and then haste. The wizard puts haste on the Great Weapon Master Bear Totem Barbarian OR slows the Target (if single) OR Meteor Swarm/Web (if AOE). The Priest casts Bless on the entire party (+1d4 to all attacks).

SO now to the numbers:
As a party:
Warrior 5x1d8+15 (Advantage on all attacks, +1d4+2 to hit (-5 sharpshooter/+5 dex wash out)
Wizard 40d6, then adhoc primary battlefield control and dps
Simulacrum 1 4x1d8+15 (Advantage on all attacks, +1d4 to hit)
Simulacrum 2 4x1d8+15 (Advantage on all attacks, +1d4 to hit)
Barbarian 3x 1d12+17 Greataxe, Advantage on all attacks, +1d4+2 to hit (-5 sharpshooter/+5 of str wash out) +3d12 Brutal Critical
Priest doesn't really dps other than his animated hammer. He can, he just doesn't, focusing instead on healing and control
Me, Finger of Death = up to 100, shortbow 2x d6+5 (Advantage on all attacks, +1d4 to hit), +Bardic Inspiration +32 zombies so 32x1d8/turn

Solo:
Me, Finger of Death = up to 100, shortbow 2x d6+5 (Advantage on all attacks through familiar), +Bardic Inspiration(cutting words) +32 zombies (current number) so 32x1d8/turn
Simulacrum 1 5x1d8+15 (Haste + Advantage on all attacks through foresight)

I'm probably off on some of the numbers since I don't min/max much. Had the warrior taken Crossbow expert then ranged could be 1d10 for him and the simulacrums. Had the wizard chosen a school other than transmuter his DPS would be higher, but then he wouldn't have been able to cast raise dead on me when I needed it. The warrior has an oathbow but I didn't count that since any class theoretically could, still with more attacks than any other class it would make that much more difference every round. Everyone will hit more due to inspiration and bless and in practice the 18, 19, 20 critical hits are brutal. The Barbarian might be the MVP since he keeps anything attacking us squishies at disadvantage and lays quite the smackdown while doing it. While solo I've developed a habit of hiding behind the larger undead for partial cover and to prevent anything from closing to melee distance. The simulacrum can keep anything at a distance with the speed increases from haste. Casters the wizard and I can control for a short period with counterspell, dispel magic, web, plant growth or silence. Even a short period is long enough for 3 archers to drop most things, often before they get into casting/attack range. More importantly the whole party helps each other and we spend a lot more time out of combat than in. Downtime on the trail is usually worry free with a ritual Leomund's hut and a song of rest.

As the bard you'd think i'd be the face but since I don't enjoy that function our Priest most often does it. It doesn't hurt that I give him illusion pretty man as needed. I tend to stay out of town anyway since many people react to zombies as if they are somehow evil, I mean really, they can play with children for hours without getting tired, what could be better in a babysitter? Oh, I forgot, my mount is a centaur, so there's another longbow with 2x1d8+x... I'd kill for a rogue in the party, these people think I can sneak... We earn gold pretty well, I bless crops with plant growth and perform amazing routines involving complicated illusions at the guild hall events. The wizard mostly makes nobles look younger and fills his spellbook. I don't know what the warrior and priest do but they always have money. The barbarian is a champion arm wrestler, loves his mead and a good barroom brawl so he's always poor but I don't mind because in a fight, I know who to trust. Though really if I'm adventuring it's the Barbarians fault, he needs money again. Now if I can just find a fabled Manual of Golems...

[Citan]

Xetheral pointed out an error in these calculations. I've not yet updated the new conclusions in this post. Some major conclusions change at high levels.

The common consensus is that when you're building a ranged attacker, the class to choose depends on your expected level of play. Since average damage is easy to calculate, I figured I'd put together a quick spreadsheet to share exactly when the switchovers are. I'm only looking at sustained damage output - considering spike damage is a much more complex task. The spoiler below has the criteria I used for the calculations for those who are interested.


This is looking only at single-target damage against a large pool of HP. There are too many variables for a straight calculation when we try looking at multiple targets. I'm also only going to consider single classed characters. Because of the way the classes leapfrog each other in being the top damage dealer, multiclassing invariably results in a slower progression with a bit better damage in the top end.

Average damage is calculated by considering the abilities listed below. Critical hits and misses are considered, as is advantage. An average damage per round was calculated for each AC between 12 and 20. These numbers were then averaged to get each character's average damage indicator at a given level.

In order to be considered in these calculations, the damage output has to be the same for a 10 round combat, and the character must be capable of doing this at least twice per day. 1 minute buffs count, single rounds of spike damage are not considered. This means the results show the suitability of each class for delivering consistent ranged damage - the standard archer niche.

The classes I looked at are: Bard, Fighter, Ranger, Rogue, and Warlock. The spoiler below shows the builds used. Each is fairly well optimized for these calculations - I tested a few iterations of each to find the best performers.



There are two cases - with feats and without. In both cases every build uses human and standard point buy array - under standard array it's not possible to get higher than a +3 in your primary stat, even if you're not human, and the only racial bonus which affects these calculations is proficiency in the longbow for Elf Bards and Rogues. If there are feats used, each build takes Sharpshooter at 1st. Sharpshooter -5/+10 almost always gives higher damage. At level 1 you're better off using Sharpshooter against any AC less than 18 (less than 20 if you're a fighter), and as proficiency and Dex increases, so does the threshold.

Each class increases Dex (or Cha for Warlock) every 4 levels, and gains proficiency bonus normally. Below are the relevant class features.

Bard (Valor)
1st: Shortbow
3rd: Longbow proficiency
6th: 2 attacks/round
10th: Takes Swift Quiver (4 attacks/round)
17th: Takes Foresight (attacks with advantage)

Fighter (Eldrich Knight)
1st: Longbow, Archery style
5th: 2 attacks/round
11th: 3 attacks/round
13th: Takes Haste (4 attacks/round)
20th: 5 attacks/round

Ranger (Hunter)
1st: Longbow
2nd: Takes Hunter's Mark (+1d6/hit)
3rd: Colossus Slayer (+1d8/round if hits)
5th: 2 attacks/round
17th: Takes Swift Quiver (no more Hunter's Mark, 4 attacks/round)

Rogue (Arcane Trickster)
1st: Shortbow, Sneak Attack (+1d6 per round if hit, scales)
2nd: Cunning Action (Modeled as giving the rogue advantage every round)
12th: Takes Skulker feat (to allow advantage if missing on first attack)
14th: Takes Haste (2 attacks/round)

Warlock (Any)
1st: Eldrich Blast, Hex
2nd: Agonizing Blast
5th: 2 attacks/round
11th: 3 attacks/round
17th: Takes Foresight for constant advantage, 4 attacks/round



Due to popular demand, I've added an analysis of Crossbow Expert as well.


This analysis assumes a liberal interpretation of the feat, so you can use the bonus action attack every round using a single hand crossbow. I don't necessarily agree with this interpretation of the feat, but this analysis will be helpful for those who do. In each case (except Warlock, obviously), it is beneficial to get the Crossbow Expert feat at 1st level, and Sharpshooter at 4th.

Bard: Crossbow Expert is recommended if you want to fully optimize. It's not as highly recommended as for the other classes due to the higher opportunity cost. Instead of Swift Quiver at level 10, you take Elemental Weapon. At level 15, you can cast it as a 7th level spell 2x/day. Note that this will compete with a magic weapon; getting a magic weapon will make everyone but the Bard (or Warlock) more effective. It also uses all of your 7th+ spell slots until 20th level.
Fighter: Crossbow Expert is highly recommended. You have no better use of your bonus action, and Sharpshooter gives you tons of bonus damage. Also, you have ability boosts to burn.
Ranger: Crossbow Expert is highly recommended. You have no better use of your bonus action for most of your career, and you have lots of bonus damage to pile on top of your extra attacks in addition to Sharpshooter. The only caveat is that you'll be killing enemies quickly enough that your bonus action will be eaten retargeting Hunter's Mark regularly.
Rogue: Crossbow Expert is not a must have, but it's not bad. It's slightly better than using your bonus action to gain advantage, assuming you have a melee ally around to ensure you can Sneak Attack.
Warlock: Crossbow Expert is not useful.



The results are in the charts at the bottom of the post. Here is a handy chart with the best classes at each level range. Any of the 5 classes get to be the best under at least one condition.

Level 1-5: Ranger or Rogue
Level 6-10: Ranger
Level 11-15: (With feats) Bard or Fighter (Without feats) Bard or Warlock
Level 16-20: (With feats) All but Warlock (Without feats) Warlock

I'll touch briefly on the results for each class.

Bard (Valor): The Bard is a strictly subpar ranged damage dealer until level 10 and Swift Quiver. Above level 10 the Bard is quite competent, and with Sharpshooter for bonus damage, is the best around at levels 10-12 and 17-19.

Fighter (EK): The Fighter is competent throughout his career due to a high attack bonus and number of attacks. The Fighter benefits a great deal from the Sharpshooter feat, but it takes a few levels to see this benefit since the -5 attack penalty is much stiffer at lower levels. With Sharpshooter, the fighter is the best around from 13-16 and at level 20.

Ranger (Hunter): The Ranger is the king of ranged attackers from levels 1-10, and does well after level 17 with Swift Quiver if feats are in play. The Ranger gets lots of bonus damage (especially Hunter's Mark) which is huge here. Volley is a pretty good at-will AOE which might make up for the Ranger's shortcomings from 11-16.

Rogue (AT): This one is a little suspect since I gave free advantage on every attack once Cunning Action shows up. Still, Sneak Attack bonus damage scales pretty well, especially if there are no feats to give most other classes bonus damage. There is a huge spike when the rouge gets Haste, since that doubles the chances of landing Sneak Attack that round. The Rogue is the top from level 1-4 (close to the ranger) as well as level 14-16 with no feats in play.

Warlock (Any): If there are no feats around, the Warlock will be the best ranged damage dealer once it takes the title from the Ranger at level 11. Hex and auto-scaling cantrips are quite effective. However, if other classes get bonus damage in the form of Sharpshooter, the Warlock falls behind. Warlock is never the best if feats are available, but with no feats it's the top from 11-13 and from 17 to 20.

http://i1368.photobucket.com/albums/ag187/altervir/NoSharpshooter_zps691710b8.png
http://i1368.photobucket.com/albums/ag187/altervir/WithSharpshooter_zps55c3dfd9.png
http://i1368.photobucket.com/albums/ag187/altervir/CrossbowExpert_zps352c3573.png
Hi!
Thanks for this effort in comparing classes and their potential output.
However, I find it slightly reductive especially on the aspect of chances to hit and how they affect the damage potential (unless I'm mistaken).

For example, why consider that Rogue has always advantage on attacks (which is heavily reliant on DM and situation unless using Fog Cloud / Darkness and change hiding position regularly), but totally disregard all ways a Bard or a Warlock has to gain advantage on their attacks?
Bard: Faerie Fire, Blindness, Greater Invisibility to quote the most obvious ("direct" effect).
Warlock: same quoted above depending on patron choice, Darkness + Devil's Sight combo, Command (only with Crossbow Expert to make them lie prone).

By the way, why make a cumulative average for all grades of AC for all levels? I don't really understand the logic there...


Yeah same thing goes for barbarian, and monk. I have tried to figure out how to make a thrown weapon ranged build but as far as I can tell you can't RAW draw more than 2 thrown weapons a turn.
Could you please give the source on that? I believe you but I'm quite surprised WoTC wouldn't have taken this hypothesis into account (Fighter has easily more than 2 attacks per round to start)


@MaxWilson: So there is a little debate on whether the intention was that Magic Initiate can be added to existing spell slots. My reading (and I think the most common reading) is that you just get that spell once/day.

@Yuukale: Multiclassing will give you the best damage if you're targeting a narrow range of levels, no question. However, it makes you fall behind otherwise since both sets of class features (and ability boosts) get delayed. There are way too many permutations of multiclassing for me to run all of them.

I've run a handful of multiclass tests with similar results - you get a few levels where you're the best in the business, but then someone else gets a cool feature which will overshadow you. So, basically just the same as single classes, but you can shuffle around the levels at which you're the best.
Overall, this exercise has given me a lot of respect for the balance of multiclassing in 5e. There are very few dips without consequences.
What you say has some very serious reality to it, with that said there are quite a few multiclass which easily give better results than pure classes for a good part of career.

Simple example: single-level dip in War Cleric for any class having a minimum spellcasting (otherwise might as well just take Magic Initiate): Bless adding +1d4 to-hit makes a very big difference, especially at lower levels.
So on an Eldricht Knight, while you will probably regret it when (if!) you reach lvl 20, from level 2-3-6 onwards (depending as how early you dip), it just makes you MUCH better overall (very specially true on a ranged Fighter: low threat + Concentration proficiency + 1d4 save).
On a Bard it's even better: people usually don't care about the capstone and it allows you to take Crossbow Expert earlier, whether you plan on getting Sharpshooter or not in the end.
On a Ranger it's somewhat a matter of opportunity: although it does compete with Hunter's Mark, it also makes easier to use Sharpshooter benefit so it does end as a positive change for any mid-level threat.
Also, chose War because it gives you some additional bonus action attack: even if it's only a long rest WIS mod resource, until you get Crossbow Expert it's a decent boost. And you get Divine Favor for when you prefer doing more damage than increasing your chances to hit.

Another simple example: single level of Fighter on any other class except Ranger: immediate and permanent +2 to hit on all attacks is largely worth waiting one more level for an additional attack, although it may sound counter-intuitive at first, as long as you take it early.

[Specter]

You should consider Greater Invisibility for Valor Bards; that's the same attack effect of Foresight except at level 7.