Maximum Average DPR by Class, No Feats or Multiclassing, Need Help Calculating

[Morphic tide]

This is partly to be a resource for anyone curious and partly asking to get an answer so that I can make a good, informed attempt at a well-balanced homebrew. Also as a resource for others to try to do the same.

As the thread title suggests, I want to have the basic DPR of the classes with the limited optimization of subclass picking and pointbuy. Feats and multiclassing disallowed, because they are both technically variant rules and they lead to massive combinatorial hell.

I'll be wanting the answers in the following table format:



Class(subclass)
Level
Casual DPR
Perpetual DPR
Nova DPR
All Crits DPR


Rogue(none)
1
13.5
23.5
N/A
41



Here's the code:
[#table='width: 500, class: grid']

Class(subclass)
Level
Casual DPR
Perpetual DPR
Nova DPR
All Crits DPR


Rogue(none)
1
13.5
23.5
N/A
41

[/table]

Remove the # to make the code work. Copy and paste the stuff inside for more rows. Copy and paste the to each row for another column if you want to add healing or some other type of DPR for some reason.

This is the basic form of a 1st level Rogue, which I chose due to simplicity of calculation. The reason for the Class column is for anyone who wants comparisons, like when classes get past particular milestones of damage, what class has the highest DPR at particular levels, who has more DPR than who at a specific level and so on.

Casual DPR is for the basic, expected DPR average from just playing normally. It does include expendables like Rage and lower-level spells, though stuff like Paladins blowing their highest level spell slot on Smites fall under Nova. Generally, if there's more than one use per short rest of something, or three or more per long rest, it can be put here.

What Perpetual DPR is for is the highest DPR the class can keep up forever, nothing that uses up limited resources. Will usually be lower for casters and higher for martials than casual DPR.

Nova DPR is going all out. Opportunity attacks, highest level spells, metamagic spam, everything you can pull in one round.

All Crits DPR is the Nova DPR with maxed rolls and crits on everything. Expect it to be huge for most of the classes. It's 77 for 2nd level Paladins.

Please give your answers in the form of tables, as they are easier to make sense of and compare with. This is wanting to have a set of tables to display the DPR of bare bones versions of the classes and subclasses.

As of March 3rd, 2017, I'm adding the mercy of having the tables only list the levels where at least one DPR entry changes. This is both for the sake of saving everyone's time and for the sake of making sure as much fits in the post as possible.



Class(subclass)
Level
Casual DPR
Perpetual DPR
Nova DPR
All Crits


Barbarian(none)
1
10
20
24
38


Barbarian(totem)
4
11
22
26
40


Barbarian(totem)
5
22
33
39
60


Barbarian(totem)
8
24
36
42
63


Barbarian(totem)
9
24
36
45
76.5


[tr]
Barbarian(totem)
13
24
36
45
87


Barbarian(totem)
16
24
36
48
90


Barbarian(totem)
17
24
36
48
100.5


[tr]
Barbarian(totem)
20
36
54
54
106.5

[retaliation08]

If the rogue is hitting with an attack action and a reaction, he should be getting sneak attack twice as SA is limited by turn usage, not round usage.

Also it is not clear whether the average dpr is assuming auto hits with attacks, auto hits with attacks and reactions, variable hit rate with attacks, or variable hit rate with attacks and reactions.

Regardless, I think your numbers are off here, though the idea is a very clever and useful one.

[Morphic tide]

If the rogue is hitting with an attack action and a reaction, he should be getting sneak attack twice as SA is limited by turn usage, not round usage.

Also it is not clear whether the average dpr is assuming auto hits with attacks, auto hits with attacks and reactions, variable hit rate with attacks, or variable hit rate with attacks and reactions.

Regardless, I think your numbers are off here, though the idea is a very clever and useful one.

Thanks for the correction on Sneak Attack. That adds 3.5 to the average for Perpetual and 6 to the Nova.

[Phoenix042]

I'm not sure how you're doing your DPR calculations, but they seem wrong to me. As far as I can tell, a level 1 rogue using a rapier to sneak attack should have a "casual" DPR calculation like this:

1d8 (4.5) + 1d6 (3.5) + 3 = 11

Note, however, that when dual-wielding shortswords, that DPR rises to

3d6 (10.5) + 3 = 13.5

This assumes that both attacks hit, but interestingly, the gap actually WIDENS when you factor in miss chance because dual-wielding has a higher chance to sneak attack successfully at least once.

Perpetual DPR for a rogue with a rapier should look like this:

(1d8 + 1d6 + 3)*2 = 22

While for a dual-wielder with short swords, it's:

(2d6 +3)*2 + 1d6, or 23.5


I'm not sure I like your definition of "nova" including critical hits, as they just inflate the numbers arbitrarily. It might be a good idea to include a FOURTH column that factors in crits (to demonstrate the damage potential of special features that multiply on a crit), but including it as is waters down features like the assassin's auto-crit feature. I'm thinking a "crit" column that assumes all crits

So I'd submit this as an alternative:



Class(subclass)
Level
Casual DPR
Perpetual DPR
Nova DPR
All Crits


Rogue(none)
1
13.5
23.5
N/A
41



For comparison, a barbarian actually DOES have a "nova" of sorts, if you don't count rage as a "casual" thing.



Class(subclass)
Level
Casual DPR
Perpetual DPR
Nova DPR
All Crits


Barbarian(none)
1
10
20
24
38



This assumes a greatsword, and that "raging" counts as going nova, which might be considered a stretch (especially at later levels). Note that the "crits," "nova," and "perpetual," columns assume that you hit with a reaction attack just like the rogue.

[Phoenix042]

I'm going to try to flesh it out a little.



Class(subclass)
Level
Casual DPR
Perpetual DPR
Nova DPR
All Crits


Barbarian(none)
1
10
20
24
38


Cleric(war)
1
10
20
30
51


Fighter(none)
1
11.3
22.6
N/A
39.3


Monk(none)
1
13
20.5
N/A
32


Paladin(none)
1
10
20
N/A
34


Ranger(none)
1
10
20
N/A
34


Rogue(none)
1
13.5
23.5
N/A
41




For the cleric, I assumed a greatsword (we're looking for MAX average DPR, so we're going to see the greatsword a lot). I gave the fighter a greatsword, even though two-weapon fighting has a 2-point higher casual DPR at this level, because the greatsword is better in every other column for now, and at level 5, will be better in EVERY column.

The monk is using either a spear or quarterstaff in two hands, and kicking or headbutting people with his bonus action. Note that this makes his perpetual and all-crits column slightly better than a TWF-style fighter, because his main weapon has a larger die (d8).

I think it's interesting to see that the cleric clearly has the best nova at level 1, although not that surprising since he has the only real "burn-and-done" damage boost (the barbarian's lasts longer though, and has other benefits).

Note that the paladin and ranger have no features that contribute to this table in ANY WAY at level one, unlike all the other martial classes. I've always felt that was a design flaw, but there it is. Also note that I'm giving the STR ranger a greatsword. That may be unfair to the classic ranger (STR isn't typically the primary ranger stat), in which case their DPR is going to suffer at level 1 even more.

[Phoenix042]

Let's take a look at some of the changes that happen at level 2. Some classes gain a nova at this level, while others come back onto the map because of their fighting style. Note the drawback in assuming that every attack hits; accuracy buffs like the barbarian's reckless attack and archery ranger's fighting style all come online at this level, but they don't change the numbers in this table at all, making archery continue to look inferior to greatsword-wielding rangers with the defense fighting style. Or the archery fighting style, but still wielding a greatsword...



Class(subclass)
Level
Casual DPR
Perpetual DPR
Nova DPR
All Crits


Fighter(none)
2
11.3
22.6
34
59


Monk(none)
2
13
20.5
26
46


Paladin(none)
2
11.3
22.6
41.5
77


Ranger(none)
2
10
20
27
48



So a lot happens at this level, and the math gets weird and full of assumptions.

For example, the paladin must burn both of his spell slots at once to nova like that (a bonus action for thunderous smite, followed by using the divine smite feature, which does not require an action), while the ranger has to use hunter's mark before he attacks with his greatsword, and must successfully concentrate until he gets his reaction attack. The barbarian, cleric, and rogue did not change at all at level 2 on this table, so I didn't include level 2 entries.

[Morphic tide]

Finally, a person who's actually making tables! Sadly, I want to see the 1-20 progression by-class, and the "peak" of each level in each category. After all, the primary point of this thread is to have an index of the DPR values of various types, for reference purposes. The reason I, personally, made it is that I want the numbers for the DPR of all the classes so that I can give a decent try at making well balanced homebrew classes. After all, feats and multiclassing are optional rules in 5e.

[Phoenix042]

I want to see the 1-20 progression by-class

I could take a stab at that... But for most classes, it's going to get into some kind of complicated comparison of several branches and set ups that get ever-more convoluted and unlikely to actually happen. The most useful such table should really include an analysis of hit chance and it's impact on DPR vs. monsters of various CR and AC. That's a really, really complicated analysis to make, although this one is still useful for setting boundaries on damage potential.

The simplest might be the totem barbarian (doesn't matter which totems, since none of them directly increase max DPR)

Lets find out!



Class(subclass)
Level
Casual DPR
Perpetual DPR
Nova DPR
All Crits


Barbarian(none)
1
10
20
24
38


Barbarian(none)
2
10
20
24
38


Barbarian(totem)
3
10
20
24
38


Barbarian(totem)
4
11
22
26
40


Barbarian(totem)
5
22
33
39
60


Barbarian(totem)
6
22
33
39
60


Barbarian(totem)
7
22
33
39
60


Barbarian(totem)
8
24
36
42
63


Barbarian(totem)
9
24
36
45
76.5


Barbarian(totem)
10
24
36
45
76.5


Barbarian(totem)
11
24
36
45
76.5


Barbarian(totem)
12
24
36
45
76.5


Barbarian(totem)
13
24
36
45
87


Barbarian(totem)
14
24
36
45
87


Barbarian(totem)
15
24
36
45
87


Barbarian(totem)
16
24
36
48
90


Barbarian(totem)
17
24
36
48
100.5


Barbarian(totem)
18
24
36
48
100.5


Barbarian(totem)
19
24
36
48
100.5


Barbarian(totem)
20
36
54
54
106.5



So a few notes: First, at level 9 brutal critical comes online, and the greatsword is no longer STRICTLY superior to the greataxe; a greataxe will do slightly more damage on a crit now. Overall, the higher average DPR of the greatsword is probably worth more, though, so I kept comparing the levels using that. Do note that the gap widens at higher levels (13 and 17) when brutal critical improves, and that the gap is more significant if you happen to be a half-orc. It might be worth looking at the table with a greataxe instead.

Also, level 20 is ridiculous, because at that level you still have nothing to use your bonus action for aside from rage, which now has no drawbacks and no limitations at all, which is why I included it in the "casual" and "perpetual" DPR columns (realistically, when are you ever going to make even ONE non-raging attack at that point?). Also, you get a 2-point damage boost to every attack in the form of Primal Champion, which the game designers were carefully balancing against the Ranger capstone, Foe Slayer.

I don't have time to do another one right now (and I suspect barbarian was by FAR the easiest to calculate, considering how many levels don't change at all), but I might contribute more later.

[Bahamut7]

As a Monk player I am a bit curious, as most DPR calculations I have seen seem to overlook this...static damage mod. Up until Level 5 on Martials, the Monk is the only class (martials) that can make 2 - 3 attacks...2 attacks consistently at least. Each one of those attacks have a higher minimum damage. All assuming a +3 in stat of course and not including crits.

1d8 + 3 (staff) + 1d4 + 3 (unarmed) = minimum 8 damage, max 18

or

1d8 + 3 (staff) + 1d4 + 3 (unarmed) + 1d4 + 3 (unarmed) = 12 - 25

When seeing the damage ranges, I can never figure out how the average is calculated as it always seems to not include this factor. I would love to know how everyone gets the 13.5. Thanks for the info.

[Phoenix042]

As a Monk player I am a bit curious, as most DPR calculations I have seen seem to overlook this...static damage mod. Up until Level 5 on Martials, the Monk is the only class (martials) that can make 2 - 3 attacks...2 attacks consistently at least. Each one of those attacks have a higher minimum damage. All assuming a +3 in stat of course and not including crits.

1d8 + 3 (staff) + 1d4 + 3 (unarmed) = minimum 8 damage, max 18

The average result of rolling 1d8 is 4.5, which I like to calculate by adding up each possible result (1 + 2 + 3 + 4 + 5 + 6 + 7 + 8) divided by the likelihood that you get that result. In this case, each roll is one out of eight equally likely rolls, so you just divide the whole thing by 8, for 36/8 = 4.5.

So in this formula, we have 1d8 (average 4.5) + 3 + 1d4 (average 2.5) + 3, which is 4.5 + 2.5 + 3 + 3 = 13. So a monk that hits with both attacks in a round should do about 13 damage at level 1 (and that obviously includes the static damage mod from dex or str).



or

1d8 + 3 (staff) + 1d4 + 3 (unarmed) + 1d4 + 3 (unarmed) = 12 - 25

Here we have 1d8 (average 4.5) + 3 + 1d4 (average 2.5) + 3 + 1d4 (average 2.5) + 3, which is 18.5, and that again includes the static modifiers.

I'm pretty sure that's all correct and includes the best possible options, but if I made any mistakes or something wasn't clear or written well, please say so.

[BurgerBeast]

In order to avoid the auto-hit problem, and thereby properly assess the damage in the context of hit chance, maybe you could assume the AC of the target is that given in the table on page 274 of the DMG.



Level
Monster AC


1-3
13


4
14


5-7
15


8-9
16


10-12
17


13-16
18


17-20
19

[Morphic tide]

In order to avoid the auto-hit problem, and thereby properly assess the damage in the context of hit chance, maybe you could assume the AC of the target is that given in the table on page 274 of the DMG.



Level
Monster AC


1-3
13


4
14


5-7
15


8-9
16


10-12
17


13-16
18


17-20
19



Well, the thread is labeled Maximum average DPR by class... Although this would be helpful for a second set of overall tables that does include hit chances in the calculations. It's quite simple math compared to Fighters using that fracking reroll of 1s and 2s on a weapon with two damage dice.

edit: More useful, and less complicated, would be attack roll bonuses and save DCs of what the class has, with two columns. One for what they have all the time and one for how big their in-class buffs are. Are there any things that let you boost DCs temporarily for spells?

[Tanarii]

Well, the thread is labeled Maximum average DPR by class... Although this would be helpful for a second set of overall tables that does include hit chances in the calculations. It's quite simple math compared to Fighters using that fracking reroll of 1s and 2s on a weapon with two damage dice.DPR without hit chance and crit chance is effectively useless information. Some things, like a Rogues Sneak Attack, are only useful once per turn, so multiple attacks increase the DPR of a Rogue relative to single attack in a non-linear ratio. Other things, such a paladin's smite, can be reserved to use on a crit only.

Edit: Just assume as .6 Hit chance for attacks, and a .5 hit chance for save spells, and you're good to go.

[Morphic tide]

DPR without hit chance and crit chance is effectively useless information. Some things, like a Rogues Sneak Attack, are only useful once per turn, so multiple attacks increase the DPR of a Rogue relative to single attack in a non-linear ratio. Other things, such a paladin's smite, can be reserved to use on a crit only.

Edit: Just assume as .6 Hit chance for attacks, and a .5 hit chance for save spells, and you're good to go.

I did edit in a note about considering listing the average attack roll and the save DC of spells in three extra columns, you know... And besides, I also mentioned using it for summery tables.

This is about maximum DPR, not the DPR counting all the ways things can go wrong. So, it won't be made part of the basic tables' calculations, it will be reserved for noting restrictions on the DPR, like the opportunity attacks in Perpetual DPR.

[Tanarii]

This is about maximum DPR, not the DPR counting all the ways things can go wrong.In other words, it's about a meaningless stat. Got it.

[Asmotherion]

may I suggest major curve points instead of level by level progression? Most people will be bored to do it level by level.

Also, DPR in my oppinion should only reflect at-will, or at least sustainable damage ability.

I'll contribute some information about Nova; The king of Nova Rounds as a single-class character would be the Sorcerer, as not only can he cast Meteor Swarm, the only thing that deals 40d6 damage with a single action, he can also Quicken and cast a fire cantrip as well, and if we're talking subclass, the Fire Dragon ancestry would give a +5 bonus per casting for a total of +10. A Dual Wielding Paladin probably comes second thanks to his Divine Smite, and the Two Weapon Fighter with his Action Surge 3rd.

That said, I do not find this method realistic, because, eventhough Feats and Multiclassing are indeed variant rules, they are rarelly if ever disabled, and are considered default unless the DM says otherwise (at least in my experiance), which is not much different from house-rulling and homebrewing.

[Morphic tide]

may I suggest major curve points instead of level by level progression? Most people will be bored to do it level by level.
It'd certainly make the tables smaller to only include when the DPR changes... I'll edit that into the first post, adjusting the Barbarian table to fit.


Also, DPR in my oppinion should only reflect at-will, or at least sustainable damage ability.
"Casual" is for typical play options and "perpetual" is the peak the class can reach with at-wills. "Nova" is just for trying to do as much damage as possible in one round.


That said, I do not find this method realistic, because, eventhough Feats and Multiclassing are indeed variant rules, they are rarelly if ever disabled, and are considered default unless the DM says otherwise (at least in my experiance), which is not much different from house-rulling and homebrewing.

It's a lot less number crunching and will keep the thread valid basically forever, because the basic classes' DPR numbers define where the classes stand here.

There's also the matter that allowing those two things turns this into a high-end optimization thread where people will argue about how to add another tenth of a point to the average Nova DPR of a character, caring nothing for how utterly insane the assumptions they are making are. Some feats allow theoretically infinite DPR, provided there's a large enough number of enemies.

The game is not balanced around multiclassing and probably not feat combinations, because trying to leads to utter insanity of number crunching for hours on every single feat and potentially days of number crunching on classes. 5e is made to be a non-broken game, and the lack of ability to control combinations makes it far, far harder to make new things without breaking the game. 5e has a lot of things built on assumptions of party capabilities and has a lot of controls for those capabilities available.

I'm wanting to make a subsystem that is properly balanced against the other classes in the game, primarily Warlock, Rogue and Paladin due to partial similarities to each. As shown by Paladin/Warlock and Paladin/Sorcerer, the game is not balanced for multiclassing, and feats aren't exactly all that great for game balance either. So I'm ignoring them to make this all go by faster so that I can get to making a proper subsystem with more than one class faster.

Seriously, I made this thread primarily to research the balance of the game for individual classes so that I have numbers to use for homebrew balance checking. So that I can throw numbers at people who made 5e classes to tell them their classes are OP or ****. So that I know what I'm getting into when I start actually playing 5e. And so that other people can do the same things.

[CaptainSarathai]

It's gonna get wonky for casters. A martial character usually only gets single-target damage, but compare something like Fireball. If there is 1 creature in the area, it's no big deal. If there are 10 creatures, suddenly you're doing a TON of damage on your turn, it's just spread out. Also, casters' slotted spells mean that they can only throw that massive damage a few times per day - you have a similar issue with the Paladin's Smite.

Single-Target vs Multi-Target even matters with Martials, as well. Again, take the Paladin, and also the Rogue. If they Smite or get SA, they do a LOT of damage to a single creature. Dishing 155 damage to a lone goblin is hilarious, but also totally wasted. Conversely, a Fighter might only deal about 40 damage per attack, but broken up over 4 attacks - they would kill 4 goblins, thereby removing 4x the HP from the enemy force.

I would also absolutely include Hit chances. This has a huge impact on the Barbarian's DPR, provided they always use Reckless Attack. I would also add in Critical chances, as some classes are clearly built around exploiting their crits.

I'm assuming you're also not factoring in any racial abilities (HOrc's better Crits) and are only using a Standard Array and ASI progression?

The reality is that there are a lot of variables that are going to effect an accurate picture of damage numbers.

[Morphic tide]

It's gonna get wonky for casters. A martial character usually only gets single-target damage, but compare something like Fireball. If there is 1 creature in the area, it's no big deal. If there are 10 creatures, suddenly you're doing a TON of damage on your turn, it's just spread out. Also, casters' slotted spells mean that they can only throw that massive damage a few times per day - you have a similar issue with the Paladin's Smite.

Single-Target vs Multi-Target even matters with Martials, as well. Again, take the Paladin, and also the Rogue. If they Smite or get SA, they do a LOT of damage to a single creature. Dishing 155 damage to a lone goblin is hilarious, but also totally wasted. Conversely, a Fighter might only deal about 40 damage per attack, but broken up over 4 attacks - they would kill 4 goblins, thereby removing 4x the HP from the enemy force.

I would also absolutely include Hit chances. This has a huge impact on the Barbarian's DPR, provided they always use Reckless Attack. I would also add in Critical chances, as some classes are clearly built around exploiting their crits.

I'm assuming you're also not factoring in any racial abilities (HOrc's better Crits) and are only using a Standard Array and ASI progression?

The reality is that there are a lot of variables that are going to effect an accurate picture of damage numbers.

Well, Perpetual DPR is specifically going to be Cantrips-only for most casters, given that it's the at-will section. Casual DPR is for anything that a character can do more than once per short rest or more than three times per long rest, so it applies to casters at a fairly early point. And it makes Barbarian rather boring, as they get Rage applying to it quite early on.

As for hit chances, I'd list that as how many attacks/spells are made/cast, how much of the damage is made on each attack/spell and the attack roll modifier/save DC of each. Possibly using the typical attack notation of (damage per attack)X(number of attacks) (attack roll modifier), (other things, listed the same way). Only altered to have it be the average damage. Maybe I'll do a bit of conversion work on the Barbarian table, to include this.

Now, can someone get me a full table of Rogue or Warlock, please? Subclass doesn't really matter to me. Those two are the main balance point of the subsystem I'm wanting to make, which prompted me to make this thread. The big point of this thread, for me, is to have a reference for homebrew balancing.

[Bahamut7]

The average result of rolling 1d8 is 4.5, which I like to calculate by adding up each possible result (1 + 2 + 3 + 4 + 5 + 6 + 7 + 8) divided by the likelihood that you get that result. In this case, each roll is one out of eight equally likely rolls, so you just divide the whole thing by 8, for 36/8 = 4.5.

So in this formula, we have 1d8 (average 4.5) + 3 + 1d4 (average 2.5) + 3, which is 4.5 + 2.5 + 3 + 3 = 13. So a monk that hits with both attacks in a round should do about 13 damage at level 1 (and that obviously includes the static damage mod from dex or str).



Here we have 1d8 (average 4.5) + 3 + 1d4 (average 2.5) + 3 + 1d4 (average 2.5) + 3, which is 18.5, and that again includes the static modifiers.

I'm pretty sure that's all correct and includes the best possible options, but if I made any mistakes or something wasn't clear or written well, please say so.

Ah! Now I get it. Thanks.

[BurgerBeast]

Well, the thread is labeled Maximum average DPR by class...

Yeah... maximum average damage per round.

If you read that as: maximum damage in one round, then that's fine with me. But that's not what it means. Not even close.

For example, you could calculate the average DPR of multiple different rogue builds, but the rogue build with the highest DPR (that's DPR, not maximum one-round damage) would be accurately called the maximum DPR build.

So, yeah... I'm not wrong.

[Morphic tide]

Yeah... maximum average damage per round.

If you read that as: maximum damage in one round, then that's fine with me. But that's not what it means. Not even close.

For example, you could calculate the average DPR of multiple different rogue builds, but the rogue build with the highest DPR (that's DPR, not maximum one-round damage) would be accurately called the maximum DPR build.

So, yeah... I'm not wrong.

Dude, I have four different types of DPR taken into account. At-wills and going all out with every option available are both covered, as is stuff you can do more than once per recommended short rest, namely the three to one ratio of short to long rest. So there's plenty of room for having more than one subclass given, and number crunching available subclasses to check the best in each category is still worth doing.

Also, no multiclassing or feats, because that leads to the thread dissolving into optimization arguments. Which gets in the way of compiling the stuff I made this thread I to get. Namely, tables for the DPR of the basic, no optimization effort beyond ASIs and subclass picks, classes. Because I want a resource for making balanced homebrew classes, which means cross referencing of the classes.

[LudicSavant]

It seems like a great deal of relevant variables simply aren't taken into account.

If you want to create a practical resource, you should be factoring in advantage/disadvantage, riders, crit chance, hit chance, effects like fighting styles or empowered spell, rounding effects, etc.

By contrast, things like "all hits, all crits" DPR has no practical value.

[Tanarii]

Dude, I have four different types of DPR taken into account.No. You don't. Because DPR includes hit chance in its calculation. That is his point. You are not calculating DPR.

[Morphic tide]

No. You don't. Because DPR includes hit chance in its calculation. That is his point. You are not calculating DPR.

And how, exactly, can one give a reliable hit chance? Different campaigns have different monsters with different ACs, and I already mentioned a way to note to-hit bonuses and number of attacks. I'll show the quote:


As for hit chances, I'd list that as how many attacks/spells are made/cast, how much of the damage is made on each attack/spell and the attack roll modifier/save DC of each. Possibly using the typical attack notation of (damage per attack)X(number of attacks) (attack roll modifier), (other things, listed the same way). Only altered to have it be the average damage. Maybe I'll do a bit of conversion work on the Barbarian table, to include this.

I'll change goal posts to satisfy you people, so long as the DPR maximum is still listed. Including hit rate in the calculations is very much not listing the maximum, and averages with hit rate are unhelpful because some DMs like to use swarms of lower level enemies, and some prefer enemies with ACs above what the party "should" be facing. Having a monster with AC 20 and a 20 in every save is a valid encounter well before the party can expect to reliably get through it, provided the monster is a pure meat wall with practically no offensive power. Similarly, a DM can have a monster with AC 0 and the literal minimum in all saves be a party level appropriate encounter by making it into the ultimate in glass cannons.

[LudicSavant]

Because I want a resource for making balanced homebrew classes

Then you should care about measuring damage per round accurately. Things like your "all hits all crits" are not going to give you meaningful figures for evaluating the balance of anything at all.


I'll change goal posts to satisfy you people, so long as the DPR maximum is still listed. Including hit rate in the calculations is very much not listing the maximum, and averages with hit rate are unhelpful because some DMs like to use swarms of lower level enemies, and some prefer enemies with ACs above what the party "should" be facing. Having a monster with AC 20 and a 20 in every save is a valid encounter well before the party can expect to reliably get through it, provided the monster is a pure meat wall with practically no offensive power. Similarly, a DM can have a monster with AC 0 and the literal minimum in all saves be a party level appropriate encounter by making it into the ultimate in glass cannons.

Your calculations are incorrect even for those values. A homebrewed creature with AC 0 still avoids an attack that rolls a 1, for example.

Moreover, you can simply print out the DPR result for every possible AC and saving throw value.

[Morphic tide]

Then you should care about measuring damage per round accurately. Things like your "all hits all crits" are not going to give you meaningful figures for evaluating the balance of anything at all.



Your calculations are incorrect even for those values. A homebrewed creature with AC 0 still avoids an attack that rolls a 1, for example.

Moreover, you can simply print out the DPR result for every possible AC and saving throw value.

And having one value that's basically meaningless for game balance is not that bad, when you have three more that are useful.

And I again point at the "maximum" part of the thread title. Hit rates aren't part of that, and the reason I'm not including it in the DPR is because it's a variable that the players have no control over.

And do you really want 20 versions of every single table? Because that's far, far crazier than just noting attack roll modifiers and number of attacks.

You won't get variables beyond those for damage itself accounted for, because the assumptions build up insanely high and it becomes basically worthless because it's turned into perfect circumstances, rather than just everything hitting.

[LudicSavant]

And having one value that's basically meaningless for game balance is not that bad, when you have three more that are useful. As others have pointed out, no. All four of your values do not account for basic game mechanics such as hit chance, and therefore are not accurate or useful DPR calculations. As such, they will not be useful for evaluating game balance, nor will they be useful in making optimization choices.

Contrast this:
http://i.imgur.com/HfAx49t.png

[Morphic tide]

As others have pointed out, no. Your values do not account for basic game mechanics such as hit chance, and therefore are not accurate or useful DPR calculations.

Contrast this:
http://i.imgur.com/HfAx49t.png

...What, exactly, is wrong with giving the attack roll modifiers? Seriously, this is about the maximum DPR, stop telling me to reduce it by including hit chance, one of the biggest campaign depended variables, in the DPR numbers.

Make a table with the attack roll modifiers and numbers of attacks included, because this is about the classes comparing to each other, not about how the classes deal with every AC value in the game.

Hit rate is relevant to this thread only in how it differs from class to class. Which varies rather little for most of them.

Oh, and of the tables in this thread, I calculated precisely none of them myself. You want hit chances included? Then make the tables yourself to include them. Or do as I say and make the tables note attack roll modifiers and number of attacks, giving both the maximum DPR as I've been defining it and the format to have them easily converted to account for hit rate, which is a variable itself and thus not conductive to tables including it.

[LudicSavant]

reduce it by including hit chance

Actually, including hit chance can sometimes *raise* the DPR over your formula, because of critical hits. For example, if you have a 1d12 attack that hits on a 2+, and 1d12 from a half-orc's critical hit rider racial feature, your DPR is higher if factoring in hit chance than if you assume all normal hits (6.825 > 6.5). So, even there, your math is wrong.


It's quite simple math compared to Fighters using that fracking reroll of 1s and 2s on a weapon with two damage dice.

The GWF reroll math is quite simple. It simply adds (Die size/2-1)/(Die size/2) to the normal average. So, for example, it would add 2/3s to a d6's average (raising it to 4.16 repeating).

Or, since you like tables:
d4 = 3 (2.5 + 1/2)
d6 = 4.16 repeating (3.5 + 2/3)
d8 = 5.25 (4.5 + 3/4)
d10 = 6.3 (5.5 + 4/5)
d12 = 7.3 repeating (6.5 + 5/6)

Also, you can calculate the average the same way you calculate the average of anything. For a 1d6, for example, you get 3.5 by adding (1+2+3+4+5+6)/6 =3.5. Since 1s and 2s are rerolled, they produce the average of a d6, so you just add (3.5+3.5+3+4+5+6)/6=4.16 repeating.

[Morphic tide]

Actually, including hit chance can sometimes *raise* the DPR over your formula, because of critical hits. For example, if you have a 1d12 attack that hits on a 2+, and 1d12 from a half-orc's critical hit rider racial feature, your DPR is higher if factoring in hit chance than if you assume all normal hits (6.825 > 6.5). So, even by the rather impractical way you have redefined DPR, your numbers are incorrect.

How, exactly, have I redefined DPR? Can you point me to a thread where DPR is accounting for every variable in the game?

And again, what is the issue with having the attack roll modifier rather than accounting for the "recommended" AC? If you want the crits included, take the additional damage of the crit and divide by 20, then add to the damage value of the attack. Hit chances are variable and having a fixed, assumed AC based on level is utterly redundant when you can just list the attack roll modifier.

There's no reason to include hit rate in the DPR calculations because either you are asking for a mind meltingly impractical use of graphs in a thread that's taking every level into account, or you are asking for an utterly redundant assumed AC that can be easily replaced with attack roll modifiers.

Again, give me tables listing the number of attacks and attack roll modifier. Stop trying to get me to have outright misses included, because the numbers compound in confusing ways when you take multiple attacks into the question. I have only ever seen optimization mention attack roll modifiers, rather than hit rates, and I'm doing the same here because it's more informative to give attack roll modifiers than to use assumed ACs, as you can then plug in the numbers to a very simple equation that gives you the type of graph that you showed me. Only with less of a headache because most of the variables are already taken care of.

Graphs may be a bit more informative, but they are much more space filling than tables which can give you more relevant numbers for generalized play than the graph that will need at least five iterations per class to describe the same information as a table that takes up less screen space than one graph.

Let me give you an example of the formatting:

If you have two attacks which average out to 13 damage each and a third that averages out to 16.5 damage per hit, and all these attacks have a +4 to the attack roll, then the resultant table cell is [13x2 +4, 16.5 +4]. How hard is that to make sense of?

This is what I've been asking you to do instead of continuing to yell at me to be doing stuff that tables can't do in a remotely reasonable way. Not everyone can make sense of the graph you showed as quickly as a table, and that graph only represents one level. Totem Barbarian would need nine of those graphs.

[LudicSavant]

This is what I've been asking you to do instead of continuing to yell at me

Nobody has yelled at you. We have patiently and politely offered you helpful suggestions, mathematical formulas, and corrections to your errors, apparently in return for hostile rants. :smallconfused:


If you want the crits included, take the additional damage of the crit and divide by 20, then add to the damage value of the attack. This formula will give incorrect results in a wide variety of cases. For example, the average Reckless Attacking Barbarian has a 9.75% chance to crit on any given attack, not 1/20.

[JNAProductions]

DPR is based on damage dealt and chance of hitting. If you really want to give good information, you'll dig through your Monster Manual, find the common ACs for the various levels, and include those to-hit chances in your DPR calculations.

[Morphic tide]

DPR is based on damage dealt and chance of hitting. If you really want to give good information, you'll dig through your Monster Manual, find the common ACs for the various levels, and include those to-hit chances in your DPR calculations.

...or just give the attack roll modifier and break up the DPR into the attacks made... Are you people completely ignoring my attempted compromise? I put "Maximum" in the title of the thread specifically because I don't want the DPR altered by a variable as swingy as hit chance.

[JNAProductions]

...or just give the attack roll modifier and break up the DPR into the attacks made... Are you people completely ignoring my attempted compromise? I put "Maximum" in the title of the thread specifically because I don't want the DPR altered by a variable as swingy as hit chance.

Well, you're kinda half-assing it, then. It's not useless info, to be sure, but it's incomplete.

Also, why is the Perpetual DPR higher than Casual?

[Morphic tide]

Well, you're kinda half-assing it, then. It's not useless info, to be sure, but it's incomplete.

Also, why is the Perpetual DPR higher than Casual?

Don't blame me, blame the guy who made the 1-20 version of that table which I condensed to the levels where the DPR changes. So I don't know where the guy went wrong. Check the Barbarian's level 1 DPR to check what went wrong there.

[BurgerBeast]

Okay, I wasn't trying to derail to the extent that this has, so I apologize for that. I was, however, trying to point out that "maximum average DPR" doesn't mean what you appear to think it means (and I admit I may be wrong about this - it's not the best form of communication).

I do think you've come up with a cool idea and I like it. Essentially, if I understand you, you just want a simple baseline or the classes. I think it would be a valuable tool.

What you've proposed is really not too far off of how the DMG sets up the determination of Monster CR. (I know it's different, but it avoids the math of in-depth considerations of hit chance, crits, advantage, etc.)

You might do something akin to what the MM does, and make a few explicit assumptions about abilities. I'm a bit of an over-analytical nerd, so I like the idea of considering the 6 encounters and two short rests per day that the game was balanced for, and then maybe assuming each battle is three rounds. Figure out the DPR on that basis (18 rounds of combat per day). For example, in this case the rogue would apply sneak attack damage every round (the game is designed on the assumption that the rogue will achieve it), but the barbarian would only get rage bonuses over 6 rounds (2 combats) at level 1. Likewise, you could figure out the wizard's damage by assuming he cast his best damage spells of each available level over the 6 battles (of course the wizard wouldn't always use his best spells, but assuming an intelligent player, he chooses other spells because they work out to be even better than his damage spells).

At the end of the day, the others on the forum (myself included) may disagree with your assumptions, but who cares? The assumptions make the degree of utility clear to everyone. As you've said form the start, you're looking for a baseline, not a math textbook.