D&D 5E Level = Challenge Rating

[EzekielRaiden]

I did 70 pages of errata is not good game design. That's the proof right there.

It absolutely is not, and unless you're even remotely willing to consider any other perspective, it's not possible for us to have a productive conversation on this topic.

Being aware enough to see your mistakes and fix them is not a flaw. It is a virtue. Treating it as though it is some damning criticism literally tells people not to bother fixing their mistakes. It directly contributes to worse game design.

You also have provided 0 proof either in your claims BTW.

I absolutely have. Repeatedly. I gave you multiple examples of games that are widely held, by both fans and critics, as well-balanced games. You have responded with shifted goalposts ("it wasn't popular, so it doesn't count") and other irrelevancies, rather than actually engaging with literally anything I've said.

We're done here.

 

[Zardnaar]

It absolutely is not, and unless you're even remotely willing to consider any other perspective, it's not possible for us to have a productive conversation on this topic.

Being aware enough to see your mistakes and fix them is not a flaw. It is a virtue. Treating it as though it is some damning criticism literally tells people not to bother fixing their mistakes. It directly contributes to worse game design.


I absolutely have. Repeatedly. I gave you multiple examples of games that are widely held, by both fans and critics, as well-balanced games. You have responded with shifted goalposts ("it wasn't popular, so it doesn't count") and other irrelevancies, rather than actually engaging with literally anything I've said.

We're done here.

I would argue B/X is the best balanced D&D at levels that matter combined with speed of play. Most of the borked spells don't exist and by the time fireball is an issue the games either over already or close enough to it.


70 paged of errata for a slow grinding ggame is a key reason why people don't wanna play it. That's a bigger issue than CR not working. Its so bad people voted with their feet.

What I did say was square peg round hole. 70 pages of errata is exactly that. It didn't work keep banging on it still doesn't work make more errata. Whose gonna read 70 pages of errata?

Any half decent DM can figure out how to use monsters well completely ignoring CR. 4E had stinkers for CR as well.

 

[Umbran]

70 paged of errata

. 70 pages of errata... 70 pages of errata?

Mod note:
Dude.
First off, repetition doesn't make the point stronger.

Second, and more importantly, ER has withdrawn from the conversation - continuing to pursue with this drum beat rhetoric is not cool. It is badgering.

How about you allow folks to disengage with a little grace, hm? Thanks.

 

[NotAYakk]

I mean, writing up a 4e style 5e monster building system isn't hard?

Around level 1 it is tricky. But beyond that

Defensive Level X base stats (X from 0 to 20)
(X+1)*5 HP
12+X/3 AC

Offensive Level X base stats (X from 0 to 20)
(X+2)*3 damage if everything hits
14+X/2 attack attribute (14 to 24)

Overall level is average of Offensive and Defensive level.

AOE abilities count for (1+X/2), where X the estimated number of extra targets beyond 1.

Save-for-half abilities count as +50% damage.

Proficiency derived from Overall level.

Each +1/-1 attack modifier above target multiplies damage by +/-10%.
Each +1/-1 AC modifier above target multiples HP by +/- 10%.

Doing saves/resistance/invisibility/etc ends up futzing with your offensive/defensive levels.

Elites

An Elite(N) monster has N times as much HP and (N-1) legendary actions. These legendary actions do 1/2 standard Offensive damage (or equivalent). They also have (N-1) "get out of jail free" cards (legendary resists or whatever).

They count as N monsters of a given level.

XP Value
Level X Elite(N) monsters are worth N times as much XP as a level X monster.

Rough first draft XP table:

1: 100 XP
2: 150 XP
3: 200 XP
4: 250 XP
5: 300 XP
6: 400 XP
7: 500 XP
8: 600 XP
9: 700 XP
10: 800 XP
11: 900 XP
12: 1000 XP
13: 1200 XP
14: 1400 XP
15: 1600 XP
16: 1800 XP
17: 2000 XP
18: 2200 XP
19: 2400 XP
20: 2600 XP
21: 2800 XP
22: 3000 XP
23: 3250 XP
24: 3500 XP
25: 3750 XP
26: 4000 XP
27: 4250 XP
28: 4500 XP
29: 4750 XP
30: 5000 XP
31: 5250 XP
32: 5500 XP
33: 5750 XP
34: 6000 XP
35: 6250 XP

Base encounter budget is (sum of PC XP values) above, and is a challenging encounter. To work out encounter budget, just add up XP values of monsters (no count multiplier).

If the monsters add up to x1.3 it is in the deadly range.

If the monsters add up to x0.7 it is in the easy range.

Calibrate based on past expectations of party capabilities, and how rested they are.

---

Please note: this is a first pass. There is some mathematics underlying it, but I didn't crunch all of the numbers and eyeballed a bunch. I quite possibly made the XP curve too shallow or steep (I used (L+1)*1.5-1 produces roughly 2x XP value).

Before using it, I'd both check existing monsters against it, and I'd prototype creating monsters, and compare it to a baseline D&D party.

The results are not intended to be totally accurate, but just a way to quickly make encounters in the right ballpark of difficulty level.

The underlying math (which I failed to follow perfectly) is that a monster's threat is equal to its (DPR) times its (HP), with a factor based on accuracy/AC. Two monsters with A damage and B HP are very roughly equivalent to one monster with A*1.5 damage and 2B HP, because half way through the fight with 2 monsters the first monster dies (and 4 monsters with A DPR and B HP are roughly equivalent to one monster with 2.5A DPR and 4B HP).

To make comparing "more monsters" to "bigger monster" easy, we want to balance N times as many monsters with a monster with N^Q times as much HP and N^Q as much damage; Q values around 0.7 work reasonably well.

We want to add up XP on multiple monsters linearly (if one is worth 300 XP, two should be worth 600 XP), and not use multipliers. So we start off with that - a level 1 monster is worth 100 XP, so N level 1 monsters is worth N00 XP.

What level of monster is roughly as tough as 2 level 1 monsters? That is worth 200 XP. (In the above chart I said level 3)
What level of monster is roughly as tough as 3 level 1 monsters? That is worth 300 XP. (In the above chart I said level 5)

We then repeat the same logic with level 3 and level 5 monsters - 3 level 3 monsters is worth 600 XP. What single monster has about as much threat at 3 level 3 monsters? That is worth 600 XP. It should line up with the threat of 2 level 5 monsters if we are doing our math right!

This gives us our XP table.

 

[Yaarel]

I mean, writing up a 4e style 5e monster building system isn't hard?

Organizing character features into "feats" with one feat of design space per level, is doable enough.


The tricky part is demonstrating transparency between player Level and monster Challenge.

What makes it tricky is, monsters tend to have more hit points than characters. "Solo" boss monsters are more like three to five different monsters functioning as a single monster. The same is true for "Elite" monsters but less extreme.

Thus the equivalent monster Level is lower than the Hit Points suggest. The Hit Dice cannot be taken at face value with regard to Level, and merely quantify the amount of Hit Points.

The actual tell for what Level a monster is, is its Proficiency bonus. Many (most?) monsters in Mordenkainen Presents have a Challenge and equivalent Level that are the same number. When determining the Level from the Proficiency, it normally is the same number Level, but with Hit Points equal to "d12 Hit Dice + 5 Constitution", rather than "d8 Hit Dice + 0 Constitution".

In other words, when determining Level, the monster Challenge is more like Barbarian Levels, with the corresponding Proficiency and Hit Points, and combat competence.

Note again, a Solo is more like several Barbarians put together. Maybe a way to think about Solos and Elites, is they are like a player character that has a pet, sidekick, or summoning with it, whence more Hit Points and more attacks. But it is still at the lower level that tends to correlate with the Proficiency.


The reason for the high Hit Points of monsters is to stabilize the combat game engine. By making the monster lower Level (with lower Proficiency), the monster is less likely to one-shot insta-kill a player character. Likewise, by having more Hit Points, the group of player characters are less likely to one-shot insta-kill a monster that they are ganging up on.

This abundance of Hit Points makes combat less swingy. There is attrition, and players and monsters get a chance to perceive their Hit Points dwindle before deciding what to do about it. The tactical game is more solid. But the "bag of Hit Points" can sometimes make combat feel like a slog, especially at the high tiers. Sometimes players want to face more dangerous opponents (higher Level) but with speedier resolution (fewer Hit Points). The DM can tinker with Level and Hit Points to suit to taste between "rocket tag" and "bag of Hit Points".

 

[Yaarel]

Around level 1 it is tricky. But beyond that ...

When organizing features into a design space of a feat, Level 1 is worth several feats.

Plus, the 2024 background is worth some feats, including the ability improvements, background feat, and proficiencies.

Plus the species traits are worth some feats.

Level 1 characters are hugely front-loaded.


There is a "Level 0" background, that is worth a number of feats before even gaining a Level in a class.

It is possible to divine this Level 0, into several Levels, each with feats. It is tempting to divide Level 0 into:
• Level 0 − species traits
• Level ⅛ − background proficiencies with tools, skills, and languages
• Level ¼ − rudimentary combat training (perhaps borrowing from class weapon, armor, or cantrip)
• Level ½ − background feat

Defensive Level X base stats (X from 0 to 20)
(X+1)*5 HP
12+X/3 AC

"(Level + 1)*5" Hit Points are reasonable for "typical" Humanoid player characters, utilizing a d8 Hit Die.

But note, monsters are more often: (Level + 1)*(7 + 5 Con).

Humanoids, including player characters, should typically have "5 Hit Points" (d8) during their background at Level 0. The rudimentary combat at Level ¼ might most this to a total of 6 Hit Points (d10) or 7 (d12), depending on choices at this Level.

Generally speaking, one can build a monster while deciding d8+0 Con (typical Humanoid), d10+2 Con (soldier), or d12+5 Con, for its Hit Points, while handling Solo and Elite separately.

Offensive Level X base stats (X from 0 to 20)
(X+2)*3 damage if everything hits
14+X/2 attack attribute (14 to 24)
AOE abilities count for (1+X/2), where X the estimated number of extra targets beyond 1.
Save-for-half abilities count as +50% damage.

For damage, I am looking at the following, comparable to player characters with reasonable optimization but nothing convoluted.

Per-Round single target (depends on design standard): 6 + Level*1.5 + Ability

Per-Spell single target: 4.5 + Level*4.5 (namely d8)
Per-Spell multi target: 3.5 + Level*3.5 (namely d6)

So for example:
A Level 9 monster casting a (slot 5) Fireball deals about: 35 damage (10d6).
A Level 17 monster casting a (slot 9) death spell at single target deals about: 81 damage (18d8).

Overall level is average of Offensive and Defensive level. ... Proficiency derived from Overall level.

I look at the Proficiency bonus, then build a player character at one of these four Levels, that can match the monster combat stats.

XP Value
Level X Elite(N) monsters are worth N times as much XP as a level X monster.

Rough first draft XP table:

1: 100 XP
2: 150 XP
3: 200 XP
4: 250 XP
5: 300 XP
6: 400 XP
7: 500 XP
8: 600 XP
9: 700 XP
10: 800 XP
11: 900 XP
12: 1000 XP
13: 1200 XP
14: 1400 XP
15: 1600 XP
16: 1800 XP
17: 2000 XP
18: 2200 XP
19: 2400 XP
20: 2600 XP
21: 2800 XP
22: 3000 XP
23: 3250 XP
24: 3500 XP
25: 3750 XP
26: 4000 XP
27: 4250 XP
28: 4500 XP
29: 4750 XP
30: 5000 XP
31: 5250 XP
32: 5500 XP
33: 5750 XP
34: 6000 XP
35: 6250 XP

Heh. I stopped using XP a while ago, so the XP numbers are starting to look like gibberish to me.

I do get the XP is part of encounter building, but it is better to refer to Levels for encounter building, rather than either XP or Challenge.

To do this, there needs to be a clear Level/Challenge equation.

 

[NotAYakk]

When organizing features into a design space of a feat, Level 1 is worth several feats.

Plus, the 2024 background is worth some feats, including the ability improvements, background feat, and proficiencies.

Plus the species traits are worth some feats.

Level 1 characters are hugely front-loaded.

I mean, I don't care?

My goal is to make a monster such that 4 of them are a reasonable challenge for 4 level 1 PCs. They don't have to look like level 1 PCs particularly. The challenge at level 1 can even be easier than at later levels, so long as the difference isn't crazy.

"(Level + 1)*5" Hit Points are reasonable for "typical" Humanoid player characters, utilizing a d8 Hit Die.

The HP of a level X foe is based on how long we want a level X fight to last and the damage output of level X PCs, not HD.

Also, I chose 5 because it is easier to do the math with. How exactly fights are calibrated can be done in a post-pass; if it turns out the ideal monster HP due to some other work should be 5.72 *(Level+1.5) then I'd rather fix that somewhere else than in the base monster building math (like maybe in the encounter building math, or the assumption of how hard the default encounter is) than have annoying math here.

(Level+1)*5 means it is easy and fast to take an existing monster and map it over to defensive level, or take a defensive level and map it over to HP, without a huge amount of mathematical skill. And no need for a spreadsheet.

I mean, ideally I'd like monster math to be easy enough that I could improvise a level X monster on the fly at a table with 0 prep time.

But note, monsters are more often: (Level + 1)*(7 + 5 Con).

Levels aren't HD. My goal is not, and I don't think it should be, to build a "monster level system" where monsters gain levels in some chart like a PC class. Nor do I want to make a system where a PC can pick a level X monster and play it in a party with level X players.

Generally speaking, one can build a monster while deciding d8+0 Con (typical Humanoid), d10+2 Con (soldier), or d12+5 Con, for its Hit Points, while handling Solo and Elite separately.

So, 5e uses the assumption that monster HD size is based on monster size. This has basically no effect on the actual game, but it sort of feels fun.

In 5e, tiny has d4, HD, small has d6 HD, medium d8, large d10, and larger has d12 HD.

The challenge of a monster doesn't map to number of HD in 5e, and I have no intention of making that happen.

HD have almost no use in 5e other than recovery during a short rest, which few monsters experience in practice.

I have spitballed the idea of a HD-based monster balancing system, but that should be more of an OSR-system rebuild than something I'd retrofit onto 5e. (Ie, monsters have HD and stars, where stars refer to how elite they are: each star counts as an extra monster of that HD (or maybe has half an extra monster?); PCs have HD, and probably start with 2. You balance encounters by adding up HD.) I don't recommend doing this in 5e, as the game wasn't built around it.

For damage, I am looking at the following, comparable to player characters with reasonable optimization but nothing convoluted.

Per-Round single target (depends on design standard): 6 + Level*1.5 + Ability

Per-Spell single target: 4.5 + Level*4.5 (namely d8)
Per-Spell multi target: 3.5 + Level*3.5 (namely d6)

So for example:
A Level 9 monster casting a (slot 5) Fireball deals about: 35 damage (10d6).
A Level 17 monster casting a (slot 9) death spell at single target deals about: 81 damage (18d8).

So, monster damage should scale with PC HP and Healing.
Monster HP and other defences should scale with PC damage.

You are scaling monster damage with PC damage, and monster HP with PC HP, which is sort of backwards and only semi works.

I look at the Proficiency bonus, then build a player character at one of these four Levels, that can match the monster combat stats.

Heh. I stopped using XP a while ago, so the XP numbers are starting to look like gibberish to me.

I do get the XP is part of encounter building, but it is better to refer to Levels for encounter building, rather than either XP or Challenge.

To do this, there needs to be a clear Level/Challenge equation.

The point of having a challenge value that isn't "the level" is that how does a level 2X creature compare to 2 level X creatures?

Making a level 2X creature be equivalent to 2 level X creatures in challenge places very strict design requirements on monster and PC damage curves. Those restrictions are not obeyed by 5e. In essence, you can't do it in 5e.

You can map level X monsters to a certain Threat value, and level 2 X to another (larger) Threat value, and have encounter building that adds up Threat values. This is sort of like how vanilla 5e XP based encounter building works, but they left in another fiddly bit (encounter size multiplier) which can be eliminated by doing the XP math better.

4e has an XP system that is similar to this.

In 4e, the power curve of monsters is dominated by +ATK/+DEF modifiers, and it doubles every 4 levels (a level X+4 monster is as dangerous as 2 level X monsters). As 5e's ATK/DEF doesn't scale nearly as fast, you really can't use a similar exponential encounter building system like you do in 4e.

That in turn means that 4e's "Build an encounter with a L+2 solider, and 6 L-2 artillery" doesn't work, as the +2/-2 are (in 4e) exponential multipliers on monster threat (x1.4 and x0.7 off baseline). The power curve in 5e isn't exponentially shaped.

(Character Power in 5e is closer to linear than it is to exponential.)

In 5e, the rough equivalent to the 4e encounter patterns is closer to "Build an encounter with a L*1.5 soldier, and 6 L*0.5 artillery". Most of 5e's power budget comes from more/less HP and more/less damage, not more/less accuracy and more/less AC like 4e.

 

[Yaarel]

Levels aren't HD. My goal is not, and I don't think it should be, to build a "monster level system" where monsters gain levels in some chart like a PC class. Nor do I want to make a system where a PC can pick a level X monster and play it in a party with level X players.

On this point, the goal is exactly, "A PC can pick a Level X monster and player it in a part with Level X players."

The reason is, Level (of monster) must be the same as Level (of player character). There needs to be a single of unit of measurement that the 5e game engine uses consistently and meaningfully. This useful use of "Level" as the measure of power of any creature helps the DM understand clearly and precisely how much power they are introducing when building a combat encounter against the player characters.

Level = Level = Level

 

[Yaarel]

The point of having a challenge value that isn't "the level" is that how does a level 2X creature compare to 2 level X creatures?

There probably needs to be a reference table, saying something like "Level 5 + Level 5 ≈ Level 8", or whatever the numbers actually are.

This is important to help the DM understand exactly how much power is present when putting together various kinds of combat scenarios.

 

[NotAYakk]

On this point, the goal is exactly, "A PC can pick a Level X monster and player it in a part with Level X players."

The reason is, Level (of monster) must be the same as Level (of player character). There needs to be a single of unit of measurement that the 5e game engine uses consistently and meaningfully. This useful use of "Level" as the measure of power of any creature helps the DM understand clearly and precisely how much power they are introducing when building a combat encounter against the player characters.

Level = Level = Level

No, that is a bad idea.

A level X PC is designed for a single human being to spend a large percentage of their attention managing and controlling.

A level X monster should not be designed for this.

Attempts to make PCs simple enough that a DM would want to control 5 of them make PCs too simple; players want more levers and bells and whistles.

Similarly, DMs shouldn't be pulling the levers and bells and whistles of 5 complex tokens; it is a waste of the DM's mental effort, and it doesn't make encounters more fun. DMs have better things to do.

PCs in 5e are not designed to fight against other PCs - the defence vs offence balance isn't tuned. Pacing is just whackey. I mean, you can try to do it, but it makes things worse. And quite possibly making it happen would make other desirable properties of the game not work, like how PCs interact with higher and lower level foes and the like.

If you succeeded at the described task - made a monster template that could be swapped for a PC template - the result demonstratively sucks, as within lived and tested experience we can see that fights against 5e PC-build monsters aren't really that fun.

So you are first talking about having to rework the entire 5e PC side of things, then you have to solve for pacing issues, work out the power curves that give the experiences you desire with relative foe levels, etc.

...

I'll give a concrete example.

If PC damage starts out at 10 DPR and goes up by 3 per level, while PC HP starts out at 10 and goes up 6 per level.

PC accuracy goes up by 1 every 2 levels, while PC AC goes up 1 every 4 levels.

If we want fights to last ~3 rounds, drain PCs of 50% of their HP on average, and PC accuracy against a level 1 foe is 50%, and level 1 monster accuracy against level 1 PCs is 30%, then level 1 foes should have ~15 HP and do 5 damage on a hit.

At level 10, we might expect 60% accuracy and monsters hit 50% of the time; PC have gained +4 to hit and +2 AC, while monsters have gained +6 to hit and +2 AC. PCs have 37 DPR before accuracy, and have 64 HP. The same 3 round pacing puts even level monsters at 21 DPR and 67 HP.

At level 20 we might expect 70% accuracy and monsters hit 60% of the time. PCs have gained another +5 to hit and +3 AC, while monsters have gained +5 more to hit and +3 more AC. Players have 67 DPR before accuracy and 124 HP. The same 3 round pacing puts level 20 monsters at 141 HP and 35 DPR.

L1 PC: 16 AC, +5 to hit, 10 HP, 10 DPR
L10 PC: 18 AC, +9 to hit, 64 HP, 37 DPR
L20 PC: 21 AC, +14 to hit, 124 HP, 67 DPR

L1 monster: 16 AC, +1 to hit, 15 HP, 5 DPR
L10 monster: 18 AC, +7 to hit, 67 HP, 21 DPR
L20 monster: 21 AC, +12 to hit, 141 HP, 35 DPR

This model makes monsters that are designed to fight level X PCs, not monsters designed to emulate level X PCs. We can compare threat volumes of monsters of various levels and check if we get good experiences when (say) a level 20 PC fights a bunch of level 1-3 monsters.

If we are taking PC development as a given (say, we are making our own monsters for 5e, but leaving PCs alone), then we calibrate the PC model and derive a monster model from it. We have a number of degrees of freedom in the monster model - we can tweak how much of its offensive power comes from +ATK and how much from damage, and similar for AC/Saves vs HP - as those have a large impact on different level PCs interact with a given monster.

If we have the freedom to change the PC model, we get more degrees of freedom.

For example, we might want monsters to gain HP faster than they do damage output, and PCs go gain damage output faster than they do HP, in order to (a) make low level monsters feel threatening but weak, easy to clear out but dangerous if ignored and (b) make high level monsters feel oppressive but not instant-death for lower level PCs. We can do this while maintaining the same combat pace against even level monsters.

That might create tension you might like: if 50 low level soldiers are aiming at you in a PC-defence-scales slowly model, you are threatened even at higher level. Meanwhile, if you face up against a dragon too early, you won't be able to take it out because monster-defence-scales-fast. If you are anywhere approaching the right level you won't instant-die from the first attack, however. (It will hurt a pile!)

In any case, my point is that making monsters designed to fight against PCs in fun ways is a different problem and opposed to making monsters that scale just like PCs.