D&D 5E (2014) Tidbit for monster design

[dave2008]

Here's a brief tidbit. While the sample size I've been working with that suggests this is still relatively small, it stands out.

When a 2014 monster doesn't work with the formula--the 2024 version almost always does. The opposite is not true. If this finding continues as I go through more monsters, it suggests that 2024 uses the same formula as 2014 (and adjusted the stats of monsters that didn't calculate correctly with it so that they do), and that I have a formula that is either correct or extremely close to what was intended for both 2014 and 2024.

As I start including additional factors (right now I've been looking at basic monsters, virtual damage for conditions, and secondary damage (damage or virtual damage value locked behind a save after an attack roll), hopefully that pattern will continue, but I'll see.

OK, so this thread is long, can you summarize your formula (or have you been updating the OP)?

 

[Sword of Spirit]

OK, so this thread is long, can you summarize your formula (or have you been updating the OP)?

I'm away from my computer, but the equation is in post #59, and it's intended to be used with the data entry form in post #56 (which Mike Mearls used).

Ongoing work is determining if the data elements actually use the same values as the DMG says or something different, so the actual math can be replicated, ad perhaps a spreadsheet can be created that does exactly the same thing theirs does.

The 2024 monsters are producing evidence that there were a few inadvertent errors in the original MM, where the wrong data got entered into the form and CR is off by one. Which means that I'm not sure I can ever really know if I manage to correctly reverse engineer, since it means no formula or spreadsheet will produce results that fully conform to the MM.

But if I can get it really close and all (or almost all) of the wrong CR monsters produce the correct CR with the 2024 changes, I suppose I'll have to accept that as the best reasonable outcome.

 

[dave2008]

I'm away from my computer, but the equation is in post #59, and it's intended to be used with the data entry form in post #56 (which Mike Mearls used).

Ongoing work is determining if the data elements actually use the same values as the DMG says or something different, so the actual math can be replicated, ad perhaps a spreadsheet can be created that does exactly the same thing theirs does.

The 2024 monsters are producing evidence that there were a few inadvertent errors in the original MM, where the wrong data got entered into the form and CR is off by one. Which means that I'm not sure I can ever really know if I manage to correctly reverse engineer, since it means no formula or spreadsheet will produce results that fully conform to the MM.

But if I can get it really close and all (or almost all) of the wrong CR monsters produce the correct CR with the 2024 changes, I suppose I'll have to accept that as the best reasonable outcome.

Thank you for the direction (to the post). Personally I don't really care if it works for the 2014 MM. If it works for the 2024 MM that is good for me!

 

[NotAYakk]

That said, these assumptions have some problems:

Firstly, many classes do not scale attacks per turn past 2, which trashcans their DPR at higher levels. 5e was not designed for high level play for many classes. Ranger and Monk should have attacks scale as Fighter. Barbarian should get a third attack at 11, but their level 20 feature makes up for a fourth attack. Paladins do not need a third attack at 11, but do need a third attack at level 17/20. I don't consider a Warlock a weapon user, but the pact of the blade probably needs some scaling as well.

Feats are technically optional rules, but are hard assumed in this math.

Magic item guidance by level is sparse at best. Magic items are not an optional rule, but the advice is pretty much "do whatever you want". In 2024, they printed a magic item by level chart for the party that does not scale with party size.

Spell outliers like Fireball and Meteor Swarm make this relationship in actual play tepid at best. 70 damage at 20? Meteor Swarm deals 140.

Why should a monster last 4+ turns against a PC? if you halve monster hp, they last just over 2 turns, meaning you can simply slot one monster per PC and have a working combat. Much simpler. 4e doubled monster hp, did 5e do that too?

You have to assume that the Paladin etc are using their non-weapon abilities to deal damage.

At level 20, assume the Paladin has Holy Weapon up. Their +3 longsword hits for 4d8+8 (26) times 2, or 52.

The "similar" Fighter with a +3 longsword deals (1d8+10)*4 or 58 damage.

Same ballpark.

Throw in smites and action surges etc and you still remain in the same ballpark.

---

My back of napkin math for the XP calculator was

1/6 (DPR) * (HP) * (AC+ATK)/10

At level 10, using the DMG numbers, we get 220 HP 7 ATK 17 AC 68 DPR = 2.4 * 68 * 220/6 = 5984 (instead of 5900)
At level 5, 38*145*2.1/6 = 1929 (instead of 1800)
At level 15, 12528 instead of 13000
At level 20, 27000 instead of 25000
At level 2, 533 instead of 450

At CR 1 and below this doesn't work, but CR 1's 85 HP always seemed a bit off.

It is also super-easy to remember.

 

[Steady Vane]

Paladin smites are equivalent-ish to Fighter action surge, and Barbarian's added rage damage + other rage benefits.

No, a paladin with 2 attacks at level 20 is not equivalent to a Fighter with 4. For every "cast a spell" there's an equivalent action surge of 8-9 attacks in a single round nova from the Fighter.

Afaik, the DMG numbers are wrong in spirit and in practice, which is why we use averages per level, such as the ones by Teos. If you can find something in the DMG numbers however, I'd be fascinated to see them!

Hope that helps

 

[NotAYakk]

Paladin smites are equivalent-ish to Fighter action surge, and Barbarian's added rage damage + other rage benefits.

No, a paladin with 2 attacks at level 20 is not equivalent to a Fighter with 4. For every "cast a spell" there's an equivalent action surge of 8-9 attacks in a single round nova from the Fighter.

Afaik, the DMG numbers are wrong in spirit and in practice, which is why we use averages per level, such as the ones by Teos. If you can find something in the DMG numbers however, I'd be fascinated to see them!

Hope that helps

Very few monsters have features that make them worse than their raw DPR and toughness.

Many monsters have features that make them stronger than their raw DPR and toughness.

Using monster averages gives you deflated DPR and Toughness values for a given level, because the DMG numbers effectively apply to "dumb brutes".

Gorilla: 44 DPR, +9 ATK, 12 AC, 157 HP
44*157*2.1/6 = 2418 XP (vs 2900)

Averaging HP/DPR/etc of creatures with non-dumb-brute features gives you bad input data. Even the t-rex has restrained and a 2nd attack that can't target the same target, both of which change how dangerous they are in some way that isn't fully clear.

Hill giant 36 DPR, +8 ATK, 13 AC, 105 HP. 2.1 * 105 * 36/6 = 1323 XP (vs 1800)

Gladiator 33 DPR, +7 ATK, 16 AC (+3 parry), 112 HP - 1602 XP (vs 1800) (parry is almost as good as +3 AC - better than +2 - as to be worse you have to be hit by 3 or less twice in the same round. At about 10 incoming swings on the same turn it becomes as bad as +2 AC; before that it is better)

Dragon Turtle 58 DPR, 104 dragon breath (1.67 uses/3 rounds), +13 ATK +10 save DC, 341 HP.

83.7 DPR, +11 ATK, 20 AC, 341 HP
3.1 * 83.7 * 341 / 6 = 14746.545 XP (vs 18000).

I get, for brutes, CR values within 1 of the described CR using a simple formula.

---

A baseline sword and board fighter who has 1 fight per day. They have +3 weapon, dueling fighting style, 20 strength. Paladin has defensive fighting style (for 1 more AC). Subclasses are ignored for now.

Fighter attacks are 1d8+10 x 4 per action, or 58. They get 2 action surges so over 3 rounds do 290 (modulated for accuracy). Assuming 75% accuracy and 5% crit rate, they deal 217.5 + 4.5 from crits (20 attacks, 5% chance per attack landing a crit), for 222 damage.

Paladin casts holy weapon before the fight, as it lasts an hour, and is also sword and board. They drop level 1 and 2 smites no normal attacks, and level 3 and 4 smites on crits. Over 3 rounds they get 6 attacks.

Each attack deals 4d8+8 damage, has a 75% chance of hitting and 5% chance of critting. They get 0.3 crits over the 3 round fight, and 4.5 hits. Each hit deals +2d8 to 3d8 damage from smite. Each crit deals +12d8 to +14d8 damage. To be conservative we'll use the lower level smite.

So 4.2 hits for 6d8+8, 0.3 crits for 16d8+8, or 30d8 + 36 = 143 damage over a 3 round fight. This paladin probably only used 1st level slots.

Going all out, it gets 4.5 hits and uses 3 4th and 1.5 3rd level slots in its 4.5 hits, an average of +4.67d8 per hit. Add in 4d8 from improved smite and holy weapon and we get 47 damage per hit, times 4.5 is 211.6 damage, basically matching the above naive fighter.

Over a day with 1 short rest, the fighter gets 4 action surges worth 44.4 damage each (total 177.6), plus at-will damage of 44.4 per round.

The paladin with a holy sword deals 40.8 nearly at-will damage (4d8+8 times 2, 75% accuracy, 5% crit rate). Over the day they can dump out 15+12+9+8 = 44d8 of smite damage. If they smite without saving big smites for crits, 1/15 are crits, so this comes to 198 smite damage per day.

So on a 1 short rest day, this naive paladin and a naive fighter are doing ... almost the same damage. Within 10% at-will damage (where the paladin self-buffs with an hour+buff), and action surge vs smite ends up being pretty close.

Add in a 2nd short rest and the fighter pulls ahead, not the least because the paladin runs out of holy weapon slots.

TL;DR; The paladin's damage boost from 11 to 20 is from their spell slots and spells. If you give them a substantial boost at 17, naive paladins will outclass naive fighters.

---

Now, the Fighter does scale nicely with many magic items, better than the paladin does. A +3 weapon that adds +2d6 elemental damage massively boosts the fighter's damage output and doesn't nudge the paladin's nearly as much. And GWF ups the fighter's damage output as well.

But I think optimization and itemization causing a problem should be treated differently than the baseline class causing a problem.

 

[Steady Vane]

An easy way to prove this is as such: how much added damage does a single weapon Feat grant in 5e? They all perform remarkably similarly, because the Fighting Styles make them all perform the same.

If that level of numerical care was taken with the Fighting Style+Associated Feat combo, then it would be strange if class features around weapons were not equivalent as well.

How do the added damage from class features therefore compare in 5e? Let's check.
DnD 5e DPR Calculator back at it again, praise be to RPGbot


A level 1 character with a +0 greatsword and no weapon feat does 7.71 damage

A Fighter can Action Surge, doing an extra round of damage in a three-turn combat, so four rounds of damage over three turns.
7.71x4/3 =10.28
A Barbarian has two parts to Rage: extra damage per hit, and freedom to Reckless Attack. Let's assume they Reckless Attack one round per combat. They get more damage because the attack is more accurate/misses less, and is also more likely to crit that round.
(11.21+8.15+8.15)/3 =9.17
Wow, really close, only held back by lacking a Fighting Style. What if the barbarian is Reckless two rounds instead?
(11.21+11.21+8.15) =10.19

Okay, looks like the Barbarian is expected to Reckless Attack about two rounds a combat, I can buy that, it seems plausible enough.

Let's try at level 5, with a bare minimum +1 weapon, to see if this still holds up
Fighter 19.36x4/3 =25.81
Barbarian (26.84+26.84+20.3)/3 =24.66
Ranger with Hunter's Mark
24.61x3/3 =24.61
Paladin who smites once, highest slot, 3d8=13.5
(19.36x3+13.5)/3 =23.86
Paladin who smites once on a crit, highest slot, 6d8=27
(19.36x3+27)/3 =28.36
Paladin crit and non-crit smite average
(28.36+23.86)/2 =26.11
*interestingly, if you do a spell list Smite that uses d6s instead of a Divine Smite that uses d8s, the average of crit/non crit Smite is 24.61
Barb 24.66
Ranger 24.61
Paladin 24.61
Fighter 25.81

Conclusion:
The class features are very equivalent to each other.

This comparison breaks at level 11, because the Barbarian and Ranger do not get a feature at that level. It is exceptionally clear that class features are, in fact, pretty balanced with each other, but only when those class features exist. A pretty great example is a first-level Paladin, who has no smites, and so a first-level Fighter will always deal more damage than in comparison. Action Surge beats no Action Surge; who knew. Similarly, an 11th-level Paladin or Fighter will always deal more damage than a same level Barbarian, because the Barbarian does not have a level 11 feature.

 

[Steady Vane]

Once again, this relies on some assumptions: the Barbarian Reckless attacking two rounds every combat on average for example. The Paladin saving their Smite for a critical hit half the time is another. And while published adventures tend to be generous with loot, it is ultimately DM fiat over whether the party by level 5 has any +1 weapons at all.

But like, are these unreasonable assumptions about the classes here? Theoretically you could play any class and not use its class features in combat, but why would you balance around that. It would be folly to balance around a Wizard who never casts spells or a Paladin who never Smites, or a Fighter who never uses Action Surge and Second Wind. It only makes sense to balance around people using the features a class has.

And in a very real sense, the class features are all balanced around each other. Pretty much everything in 5e does the same damage. 5e simply takes great pains so it does not appear or feel that way. Because 4e had everyone do the same damage while looking and feeling like everyone did the same damage and people hated it.

Once again, I am not joking. Everything in 5e does the same damage, on average.

 

[Steady Vane]

TL;DR; The paladin's damage boost from 11 to 20 is from their spell slots and spells. If you give them a substantial boost at 17, naive paladins will outclass naive fighters.

Let's try this then
A Fighter, a Paladin, and a Barbarian walk into a bar. They all have GWM and PAM as the optimized 5e build, with +3 weapons even though both DMGs tell DMs to give their players +4.

The three agree to a contest of arms attacking a training dummy to see who has the highest damage.

We will do both baseline DPR per round as a comparison, and then compare when using resources from their class. Barbarian will attack Recklessly twice per three round combat as before, Paladin will Smite with their highest spell slot, Fighter at this level can Action Surge twice.

Baseline:
Fighter 67.7
Paladin 47
Barbarian 58.3

Class Resource Used:
Fighter (67.7x5/3) =112.8
Paladin (47x3+(5d10aka 27.5x3/2))/3 =60.8
Barbarian (58.3+80.3+80.3)/3 =73


Please tell me how these numbers compare at all in any way. The Fighter does literally double the damage of the Paladin.

 

[Steady Vane]

For fun, here's the math if Paladin and Barbarian actually get 3 attacks instead of being stuck with only 2, and use +4 weapons as I suggest.
Baseline:
Fighter 75.2
Paladin 71.4
Barbarian 75.9

Wow look they actually compete with each other again, who would have thought