D&D 5E Project Monsters by Level (not CR)

[dave2008]

Monsters by Level (not CR)​

I loved that 4e (and now PF2e) had monsters by level and not CR. It made things conceptually clear: 1 standard monster = 1 PC at the same level. I have often thought about doing something like this for 5e, so today is the day I give it a go!

Project Goals​

• Redefine monsters based on PC level, no more CR.
• Bring back monster breadth via 4e style monster roles: Solo, Elite, Standard, Grunt, Minion
• Make encounter building easier

Project Assumptions​

• A monster of equal level to a PC is a challenge for that single PC. Roughly 1 Monster = 1 PC.
• Use the same Monster math as the DMG, MM, etc.

What is a Challenge​

For this concept to work, I need an easy / simple definition of what constitutes a challenge. To me, a challenge is encounter where there is reasonable chance of failure. That is not the case in any of the levels of difficulty described in the DMG. Therefore, I am developing a new metric to use for this exercise.

I have decided to base a monster's challenge on the Monster's XP verses a single PCs Adventuring Day XP (AD XP) in the DMG. My first thought was to base the monster's XP on the full AD XP. However, I don't trully want a challenge to be 50/50. It should be difficult, IMO, but generaly winnable. The DM can always have more or higher level monster's to make things tougher. So I have currently decided a monster of equal level to a PC has an XP value of 50% of the PC's AD XP.

​

Spoiler: Adventuring Day XP

Level	Adjusted XP per Day per Character
1st	300
2nd	600
3rd	1,200
4th	1,700
5th	3,500
6th	4,000
7th	5,000
8th	6,000
9th	7,500
10th	9,000
11th	10,500
12th	11,500
13th	13,500
14th	15,000
15th	18,000
16th	20,000
17th	25,000
18th	27,000
19th	30,000
20th	40,000

​

​

Based on this definition of a challenge, a monster should fall within the following exp values per level

PC Level	Standard Monster Expected XP
1	150
2	300
3	600
4	850
5	1,750
6	2,000
7	2,500
8	3,000
9	3,750
10	4,500
11	5,250
12	5,750
13	6,750
14	7,500
15	9,000
16	10,000
17	12,500
18	13,500
19	15,000
20	20,000

​

What's this about Monster Roles?​

More to come.

Examples​

Ok, let's look at a few examples using the table above. Remember this is a monster equivalent to 1 PC at the given level:

Level	Monster	Expected XP	Actual XP
1	Orc (CR 1)	150	100
5	Ettin (CR 4)	1,750	1,100
10	Frost Giant (CR 8)	4,500	3,900
15	Arcanaloth (CR 12)	9,000	8,400
20	Death Knight (CR 17)	20,000	18,000

Issues​

Immediately I am seeing some issues with my idea. Clearly the 5e design monster design was not intended for this type of balance. I am also seeing that monsters are very dangerous at lower level and less so at higher levels. I will have to think about this more, but it may not be possible to convert existing monsters. I may need to make them from scratch. Also, think the metrics of a "challenge" may need to be revised.

Anyway, I would love to hear your thoughts.

 

[FarBeyondC]

I suppose the first question would be, how will you determine what an average PC of a given level is?

PCs being as unequal as they are at various levels depending on their composition.

 

[GuyBoy]

I wish you luck with this.
I find this type of monster-math to be fascinating to watch and incredibly hard to accomplish due to so many special powers, often situational in nature eg flying, swim speed, resistances etc.
I fondly remember Don Turnbull doing his Monstermark in the early White Dwarfs. Brilliant articles.
I’ll follow with interest, @dave2008

 

[dave2008]

I suppose the first question would be, how will you determine what an average PC of a given level is?

PCs being as unequal as they are at various levels depending on their composition.

While I agree, I provided my assumption for this exercise: 50% the daily XP budget. That means I am assuming all PCs of the same level are roughly equal. Will that is not strictly true, I think it is good enough for this type of exercise. This also doesn't account for party and monster synergies. Those would have to be adjusted for separately.

 

[Quickleaf]

I'm also fascinated by these sorts of explorative maths breakdowns, even if I find they usually break down when put to the actual play test.

For example, trying to go backwards with this kind of system doesn't work. "If we say an ettin is level 5 challenge suitable for one 5th level PC, then five 1st-level PCs should be able to..." – and then you end up with some chunky salsa PCs.

Not that the CR system works well in this regard anyhow. Plus we have recent confirmation from the D&D Creators Summit that the WotC internal monster design tools do not reflect what's in the DMG (we can see this in monsters like the Quickling, for example, which the DMG scores as being CR 2 to 3, compared to VGtM which lists it as CR 1).

Recognizing that all those guidelines are pretty borked...

I usually pay more attention to monster damage output compared to PC hit points (I use a d8 Hit Die and 12 Constitution as a "typical adventurer"). Basically, with my "at a glance" evaluation of a prepared battle, I'm wondering: Is there a decent probability that this monster would knock out the entire fresh party in one turn or outright kill a fresh PC in one turn? (i.e. without allowing player agency to save themselves barring reaction powers) If I answer "yes", that's a sign that the fight is probably too much and if I decide to go forward with it, I need to flex my GM artistry. That rule-of-thumb has, in my 5e games, been a pretty good guiding light.

To look over your chart with those values in mind...

Est. Level	Monster	Monster Damage (Average)	PC Hit Points (Average)	Difference
1	Orc	9	9	= 0
5	Ettin	28	33	= 5
10	Frost Giant	50	63	= 13
15	Arcanaloth (finger of death)	62	93	= 31
20	Death Knight (staggering smite)	95	123	= 28

 

[FallenRX]

Hasnt Giffyglyph done this already with his monster maker here?

 

[Quickleaf]

Hasnt Giffyglyph done this already with his monster maker here?

Giffyglyph's "Monster Levels" seem a little different.

First, I don't see on the website that there's any correlation between Monster Levels and PC levels.

EDIT: Instead they have a rough assumption that the GM uses monsters whose Monster Level is +/-3 of the PCs' level, and if the GM exceeds that then double soemthing called Monster Points for encounter building.

Second, they never describe how they created the CR to Level conversion table. It appears to be done "by feel" or personal experience. For example, they list...

Orc (CR 1/2) = Monster Level 2
Ettin (CR 4) = Monster Level 8-10
Frost Giant (CR 8) = Monster Level 16-17
Arcanaloth (CR 12) = Monster Level 21
Death Knight (CR 17) = Monster Level 26

EDIT EDIT: I'm sure they have some system behind all this, but the way they're presenting it on their website is extremely opaque. For instance, according to their guidelines, a GM would usually only use an Arcanaloth with a party whose level was 18th or higher... which is ridiculous.

 

[dave2008]

Hasnt Giffyglyph done this already with his monster maker here?

Not really, as @Quickleaf noted, other using levels and 4e type monster roles, it is quite a different approach.

 

[dave2008]

I'm also fascinated by these sorts of explorative maths breakdowns, even if I find they usually break down when put to the actual play test.

For example, trying to go backwards with this kind of system doesn't work. "If we say an ettin is level 5 challenge suitable for one 5th level PC, then five 1st-level PCs should be able to..." – and then you end up with some chunky salsa PCs.

Not that the CR system works well in this regard anyhow. Plus we have recent confirmation from the D&D Creators Summit that the WotC internal monster design tools do not reflect what's in the DMG (we can see this in monsters like the Quickling, for example, which the DMG scores as being CR 2 to 3, compared to VGtM which lists it as CR 1).

Recognizing that all those guidelines are pretty borked...

Thank you for the input and feedback!

I usually pay more attention to monster damage output compared to PC hit points (I use a d8 Hit Die and 12 Constitution as a "typical adventurer").

Is there any particular reason you picked this as your "typical."

Basically, with my "at a glance" evaluation of a prepared battle, I'm wondering: Is there a decent probability that this monster would knock out the entire fresh party in one turn or outright kill a fresh PC in one turn? (i.e. without allowing player agency to save themselves barring reaction powers) If I answer "yes", that's a sign that the fight is probably too much and if I decide to go forward with it, I need to flex my GM artistry. That rule-of-thumb has, in my 5e games, been a pretty good guiding light.

To look over your chart with those values in mind...

Est. Level	Monster	Monster Damage (Average)	PC Hit Points (Average)	Difference
1	Orc	9	9	= 0
5	Ettin	28	33	= 5
10	Frost Giant	50	63	= 13
15	Arcanaloth (finger of death)	62	93	= 31
20	Death Knight (staggering smite)	95	123	= 28

What I am realizing after my first dive is that it may be impossible, or very difficult, to convert existing monsters to equivalent level without a deeper dive into the math with per CR/level/Tier adjustments or something similar. However, I may be able to come up with a method to make custom monsters by level work the existing CR math.

 

[Quickleaf]

Thank you for the input and feedback!

Is there any particular reason you picked this as your "typical."

What I am realizing after my first dive is that it may be impossible, or very difficult, to convert existing monsters to equivalent level without a deeper dive into the math with per CR/level/Tier adjustments or something similar. However, I may be able to come up with a method to make custom monsters by level work the existing CR math.

Sure! It's a rabbit hole, that's for sure. It's always interesting to see what you come up with, especially with all the monster design you've been engaged with.

I went with d8 Hit Dice as "typical" because that's almost in the middle of all the D&D classes assuming roughly equal representation of classes across all tables (see below). And I choose 12 Constitution because most of the PCs I've seen at my table seem not to dump stat Con, but also not to sink their highest scores in Con. But there's no maths involved, just eyeballing.

d6 = Sorcerer, Wizard
d8 = Artificer, Bard, Cleric, Druid, Monk, Rogue, Warlock
d10 = Fighter, Paladin, Ranger
d12 = Barbarian

 

[dave2008]

Sure! It's a rabbit hole, that's for sure. It's always interesting to see what you come up with, especially with all the monster design you've been engaged with.

I went with d8 Hit Dice as "typical" because that's almost in the middle of all the D&D classes assuming roughly equal representation of classes across all tables (see below). And I choose 12 Constitution because most of the PCs I've seen at my table seem not to dump stat Con, but also not to sink their highest scores in Con. But there's no maths involved, just eyeballing.

d6 = Sorcerer, Wizard
d8 = Artificer, Bard, Cleric, Druid, Monk, Rogue, Warlock
d10 = Fighter, Paladin, Ranger
d12 = Barbarian

I'm think that something like your approach makes some sense. Maybe I will create a PC by level table, like the Monster by CR table, with expected HP, AC, DPR, and attack bonus. That way it is very clear what is the comparison. The use that as a basis for monster level.

 

[Quickleaf]

I'm think that something like your approach makes some sense. Maybe I will create a PC by level table, like the Monster by CR table, with expected HP, AC, DPR, and attack bonus. That way it is very clear what is the comparison. The use that as a basis for monster level.

Yeah, that's a really useful tool that I wish the designers provided because it says a lot about the assumptions of the game. There's an analysis – can't find the link – a fan did of how many goblins a 1st level fighter kills throughout the editions before the fighter is taken down. That sort of thing is invaluable to establish a baseline.

I'll take a stab at a chart for the Rogue (Thief), since it's an easier starting point.

Assumptions:

• Overall, I'm assuming a non-optimizing player who is experienced enough to run the class effectively in combat.
• Rogue with Thief subclass
• 16 in prime ability score (Dex) at 1st, up to 18 at 4th (ASI), and 20 at 8th (ASI).
• 12 Constitution.
• Max HP at 1st, average HP after 1st level.
• Wearing studded leather (AC = 12+Dex) & wielding a rapier one-handed.
• At 10th level, 12th, 16th, and 19th levels picks up feats or ASIs to flesh out the character that don't necessarily affect the numbers on this table.
• Reliably adds Sneak Attack damage once per round, thanks to subclass features, Hiding, allies working together, or opportunity attacks.
• Cunning Action (2nd level) is roughly worth a +1 AC bonus for getting distance from melee, avoiding opportunity attacks, and being hidden. This is a low value assumption because PCs have more uses for bonus actions typically, and sometimes you just don't have a great way to Hide. I think you could safely increase this instead to +2 AC, assuming that the rogue Hides about half the time with Cunning Action.
• Starting at 5th level, Uncanny Dodge allows the rogue to halve damage from one attack per round. First, I'm assuming that a rogue gets to use Uncanny Dodge about once per day/rest (due to a combo of avoiding being targeted in the first place, some monsters forcing saves, or the rogue using their reaction for other things). This is a low-ball assumption...Uncanny Dodge might contribute up to 3x that in a particularly intense melee. Second, we need a sense of what average monster damage actually looks like by CR (tl;dr the SRD monsters' avg damage caps close to 20 at CR 11+). This number might look different factorign VGtM, MToF, MMotM, and newer books. Third, we need to make some kind of assumption about "Typical Monster CR" encountered at each level – I'm going with CR = Level - 2 (i.e. CR 3 at 5th, etc). So I've created an Avg Monster Atk Dmg column, and added half of that to the rogue's HP (rounded down) starting at 5th level when they get Uncanny Dodge. Though this is a BIG assumption at lower levels, if that ~20 damage cap holds true, then it stops being a significant assumption at higher levels.
• Thief's Reflexes (17th level) allow for an extra turn at the start of combat. This equates to approximately a +33% increase in damage, assuming typical 3-round combats.
• Elusive (18th level) is roughly worth another +1 AC bonus. Just eyeballing it.
• Stroke of Luck (20th level) is roughly worth a +1 bonus to attack (i.e. comparable to Advantage once per combat, which is +4 divided by 3... rounded down to +1).
• I wasn't sure about how to factor critical hit damage, so for simplicity I left it out for now.
• For comparison purposes, I did not include magic weapons/items.
• While I did not include Advantage in the Attack Bonus, my GM hunch is that a rogue has Advantage on maybe half of its attack rolls. The median bonus from Advantage over all target rolls is +4. So, I think a safe assumption may be increasing the rogue's attack bonus universally by +2.

Rogue Level	Avg Monster Atk Dmg (of a monster whose CR = level -2)	HP	AC	DPR	Attack Bonus
1	–	9	15	11 (1d8+3+1d6)	+5
2	–	15	16	11	+5
3	–	21	16	14.5 (1d8+3+2d6)	+5
4	–	27	17	14.5	+6
5	10	38 (33+5)	17	19 (1d8+4+3d6)	+7
6	12	45 (39+6)	17	19	+7
7	15	52 (45+7)	17	22.5 (1d8+4+4d6)	+7
8	15	58 (51+7)	18	22.5	+8
9	17	65 (57+8)	18	27 (1d8+5+5d6)	+9
10	17	71 (63+8)	18	27	+9
11	20	79 (69+10)	18	30.5 (1d8+5+6d6)	+9
12	20	85 (75+10)	18	30.5	+9
13	20	91 (81+10)	18	34 (1d8+5+7d6)	+10
14	20	97 (87+10)	18	34	+10
15	20	103 (93+10)	18	37.5 (1d8+5+8d6)	+10
16	20	109 (99+10)	18	37.5	+10
17	20	115 (105+10)	18	54.5 [1.33* (1d8+5+9d6)]	+11
18	20	121 (111+10)	19	54.5	+11
19	20	127 (117+10)	19	59 [1.33*(1d8+5+10d6)]	+11
20	20	133 (123+10)	19	59	+12

Now we can take some of those values for the Rogue (Thief) and cross-reference with the example monsters you gave – Orc, Ettin, Frost Giant, Arcanaloth, Death Knight – and see what shakes out...

Est. Level	Monster	Monster DPR (Average)	Monster Attack Bonus	Rogue's AC	Factored Damage (%hit * DPR)	Rogue's HP	% Reduction to HP
1	Orc	9	+5	15	5 (.55 * 9)	9	55%
5	Ettin	28	+7	17	15 (.55 * 28)	38	39%
10	Frost Giant	50	+9	18	30 (.6 * 50)	71	42%
15	Arcanaloth (finger of death)	62	save DC 17	CON save +1	46 (.75 * 62)	103	45%
20	Death Knight (staggering smite)	95	+11	19	62 (.65 * 95)	133	47%

 

[NotAYakk]

Multiply CR by 4. Add up monster levels.

80% of PC levels is easy
100%of PC levels is medium
133% of PC levels is hard
160% of PC levels is deadly

Divide solos by 4 and elites by 2 to taste.

 

[dave2008]

Yeah, that's a really useful tool that I wish the designers provided because it says a lot about the assumptions of the game. There's an analysis – can't find the link – a fan did of how many goblins a 1st level fighter kills throughout the editions before the fighter is taken down. That sort of thing is invaluable to establish a baseline.

I'll take a stab at a chart for the Rogue (Thief), since it's an easier starting point.

Assumptions:

• Overall, I'm assuming a non-optimizing player who is experienced enough to run the class effectively in combat.
• Rogue with Thief subclass
• 16 in prime ability score (Dex) at 1st, up to 18 at 4th (ASI), and 20 at 8th (ASI).
• 12 Constitution.
• Max HP at 1st, average HP after 1st level.
• Wearing studded leather (AC = 12+Dex) & wielding a rapier one-handed.
• At 10th level, 12th, 16th, and 19th levels picks up feats or ASIs to flesh out the character that don't necessarily affect the numbers on this table.
• Reliably adds Sneak Attack damage once per round, thanks to subclass features, Hiding, allies working together, or opportunity attacks.
• Cunning Action (2nd level) is roughly worth a +1 AC bonus for getting distance from melee, avoiding opportunity attacks, and being hidden. This is a low value assumption because PCs have more uses for bonus actions typically, and sometimes you just don't have a great way to Hide. I think you could safely increase this instead to +2 AC, assuming that the rogue Hides about half the time with Cunning Action.
• Starting at 5th level, Uncanny Dodge allows the rogue to halve damage from one attack per round. First, I'm assuming that a rogue gets to use Uncanny Dodge about once per day/rest (due to a combo of avoiding being targeted in the first place, some monsters forcing saves, or the rogue using their reaction for other things). This is a low-ball assumption...Uncanny Dodge might contribute up to 3x that in a particularly intense melee. Second, we need a sense of what average monster damage actually looks like by CR (tl;dr the SRD monsters' avg damage caps close to 20 at CR 11+). This number might look different factorign VGtM, MToF, MMotM, and newer books. Third, we need to make some kind of assumption about "Typical Monster CR" encountered at each level – I'm going with CR = Level - 2 (i.e. CR 3 at 5th, etc). So I've created an Avg Monster Atk Dmg column, and added half of that to the rogue's HP (rounded down) starting at 5th level when they get Uncanny Dodge. Though this is a BIG assumption at lower levels, if that ~20 damage cap holds true, then it stops being a significant assumption at higher levels.
• Thief's Reflexes (17th level) allow for an extra turn at the start of combat. This equates to approximately a +33% increase in damage, assuming typical 3-round combats.
• Elusive (18th level) is roughly worth another +1 AC bonus. Just eyeballing it.
• Stroke of Luck (20th level) is roughly worth a +1 bonus to attack (i.e. comparable to Advantage once per combat, which is +4 divided by 3... rounded down to +1).
• I wasn't sure about how to factor critical hit damage, so for simplicity I left it out for now.
• For comparison purposes, I did not include magic weapons/items.
• While I did not include Advantage in the Attack Bonus, my GM hunch is that a rogue has Advantage on maybe half of its attack rolls. The median bonus from Advantage over all target rolls is +4. So, I think a safe assumption may be increasing the rogue's attack bonus universally by +2.

Rogue Level	Avg Monster Atk Dmg (of a monster whose CR = level -2)	HP	AC	DPR	Attack Bonus
1	–	9	15	11 (1d8+3+1d6)	+5
2	–	15	16	11	+5
3	–	21	16	14.5 (1d8+3+2d6)	+5
4	–	27	17	14.5	+6
5	10	38 (33+5)	17	19 (1d8+4+3d6)	+7
6	12	45 (39+6)	17	19	+7
7	15	52 (45+7)	17	22.5 (1d8+4+4d6)	+7
8	15	58 (51+7)	18	22.5	+8
9	17	65 (57+8)	18	27 (1d8+5+5d6)	+9
10	17	71 (63+8)	18	27	+9
11	20	79 (69+10)	18	30.5 (1d8+5+6d6)	+9
12	20	85 (75+10)	18	30.5	+9
13	20	91 (81+10)	18	34 (1d8+5+7d6)	+10
14	20	97 (87+10)	18	34	+10
15	20	103 (93+10)	18	37.5 (1d8+5+8d6)	+10
16	20	109 (99+10)	18	37.5	+10
17	20	115 (105+10)	18	54.5 [1.33* (1d8+5+9d6)]	+11
18	20	121 (111+10)	19	54.5	+11
19	20	127 (117+10)	19	59 [1.33*(1d8+5+10d6)]	+11
20	20	133 (123+10)	19	59	+12

Now we can take some of those values for the Rogue (Thief) and cross-reference with the example monsters you gave – Orc, Ettin, Frost Giant, Arcanaloth, Death Knight – and see what shakes out...

Est. Level	Monster	Monster DPR (Average)	Monster Attack Bonus	Rogue's AC	Factored Damage (%hit * DPR)	Rogue's HP	% Reduction to HP
1	Orc	9	+5	15	5 (.55 * 9)	9	55%
5	Ettin	28	+7	17	15 (.55 * 28)	38	39%
10	Frost Giant	50	+9	18	30 (.6 * 50)	71	42%
15	Arcanaloth (finger of death)	62	save DC 17	CON save +1	46 (.75 * 62)	103	45%
20	Death Knight (staggering smite)	95	+11	19	62 (.65 * 95)	133	47%

Awesome! That is more in-depth than what I was thinking, but great info. Thank you!

 

[dave2008]

Multiply CR by 4. Add up monster levels.

80% of PC levels is easy
100%of PC levels is medium
133% of PC levels is hard
160% of PC levels is deadly

Divide solos by 4 and elites by 2 to taste.

Did you read the OP? That is not at all what I am trying to do.

 

[Quickleaf]

@dave2008 Yeah, it's white room stuff, so take with a grain of salt, but at least it's informed by my experiences running and playing 5e. Probably the most valuable part of my analysis is being up front about the sorts of assumptions I had to make – because any effort to decode this stuff involves getting into underlying design assumptions. The important thing is to distinguish those assumptions that REALLY move the dial vs. those assumptions that if we change only change the numbers a little bit.

I think the rogue is a pretty good baseline to compare from – middle of the road hit points and damage that "feel" like a solid typical adventurer.

EDIT: One thing from the rogue analysis that I noticed is that – without help from their friends, a clever ploy, a magic item, or lucky rolls – the rogue would be felled in about 3 rounds by any of these foes consistently hitting him. That might be a good starting design guideline to try out – i.e. a monster whose level equals a PCs' level deals 1/3rd of the PCs' hit points in damage on its turn.

 

[NotAYakk]

But, a level 7 monster in 4e was not equal in power to a level 7 PC. In fact, 4 of them where a medium difficulty challenge for 4 level 7 PCs.

The monster level was advice about what band of PC levels the mosnter was used for.

This made encounter building easy (together with how it measured XP: but you can pretty much ignore XP in 4e and use relative levels instead. 4e XP was basically 100* 2^((level-1)4) with rounding.

Which means a Level X+4 monster is worth 2 Level X monsters, a X+2 was worth about 1 and a half, X-4 is half, etc, regardless of X.

Easy encounter building needs easy math. 5e alread has it, mostly hidden.

But 5e PCs are balanced around adventuring days, not encounters like 4e PCs are.

 

[Oofta]

I'm just not sure this isn't just CR by another name.

While I applaud the effort to fix things, I'm not sure it can be fixed. Let's take 2 groups. Group 1 is made up of either dwarves or PCs with high con modifiers. Group 2 has a mix, no dwarves and average-for-adventurers constitutions.

Now suppose they run into a green dragon. In group 1, the dwarves will have advantage on their poison saves and take half damage from the breath weapon. Group 2 will be at much higher risk. But wait ... what if group 1 significantly lacks ranged attacks while group 2 is full of archers and PCs with ranged spells? Suddenly the picture could look far different. My own experience is that you have to balance the encounters on a group by group basis.

Any system we come up with is never going to be fool proof. I think there are better calculations for balancing encounters based on CR than what we get in the DMG, I'm just not sure how this helps much.

 

[Quickleaf]

I'm just not sure this isn't just CR by another name.

While I applaud the effort to fix things, I'm not sure it can be fixed. Let's take 2 groups. Group 1 is made up of either dwarves or PCs with high con modifiers. Group 2 has a mix, no dwarves and average-for-adventurers constitutions.

Now suppose they run into a green dragon. In group 1, the dwarves will have advantage on their poison saves and take half damage from the breath weapon. Group 2 will be at much higher risk. But wait ... what if group 1 significantly lacks ranged attacks while group 2 is full of archers and PCs with ranged spells? Suddenly the picture could look far different. My own experience is that you have to balance the encounters on a group by group basis.

Any system we come up with is never going to be fool proof. I think there are better calculations for balancing encounters based on CR than what we get in the DMG, I'm just not sure how this helps much.

I think the goal is just a better lingua franca (so many GMs get CR confused and use the "level" language) that is more consistent than what we have already.

Definitely going to have the issues you point out about situational benefits/drawbacks.

The DMG lightly touches on this, but in the Green Dragon vs. High Con/Dwarf Party, I would mark down those PCs having a situational benefit. If I wanted to preserve the threat for that party the same as if they were not High Con/Dwarf, then I'd add a situational drawback – e.g. have the Green Dragon ambush the party, restrict the PCs' ability to see the dragon, present some ongoing damage in the environment, or catch them in the midst of another demanding activity like defending an emissary or fixing a bridge.

 

[FitzTheRuke]

I'm very interested to see what you come up with here. I'd even be interested in helping to build/convert some monsters once you have a working framework.