@Stormonu I wanted to answer your question a bit out of order because below I'll dive into some detail, but first and foremost I think the question you raise about
"what's the meaning of dividing Average Damage by the CR number?" is absolutely the right question. While I disagree that it's meaningless, I 100% agree that it is not a complete picture. I think the meaning is twofold...
First, it's a way to get the pulse on big trends. For example, the big jump from a very high value for fractional CR monsters (0, 1/8, 1/4, 1/2) to a much lower value for CR 1+ monsters. Or another example was the descending value as monsters increase in CR more generally.
Second, there's meaning in getting a feel for to what extent the threat a monster poses is encapsulated within these simple sort of maths VS. "other things" not accounted for in the maths, AND ALSO whether that ratio of how much "other things" are influencing this changes going from lower CR monsters to higher CR monsters. "Other things" might include PCs with damage mitigating powers like a barbarian's Rage or
blade ward, damage avoidance powers like
shield or Parry, monster damage being broken up into Multiattacks, etc.
I appreciated
@Fanaelialae and
@ezo sharing some more adjusted numbers and bringing PC hit points into the question... I still have to review those... but my point about "other things" is still going to apply to any deeper maths analysis we do.
Anyhow, I wanted to start with that – that you've got me asking "is there meaning? if so, what is that meaning?" Hopefully I can back it up with some better numbers and examples, but it may be in the end that
there's not ENOUGH meaning to be a valuable design tool.
OK, now I'm going to get into some specifics where it may seem like I'm being critical or cherry-picking examples. So apologies in advance if it comes across that way - not my intent, but I know how intent can get lost in these sorts of conversations.
"Easily shake off" is probably subjective, but this is a great example where those "other things" (e.g. damage mitigation, avoidance, and multiattack) make a big impact. The barbarian and rogue come to mind, so I must thinking of Conan...
If we use your example against a
20th level barbarian (who gets 24 Con as their capstone)... that barbarian has 285 hit points (12 +7 * 19 + 7 * 20). So the
132 damage is less than half (
about 46%) of their hit points. And if the damage is bludgeoning, piercing, or slashing, then the damage the barbarian takes is reduced to 66 damage, which is less than a quarter (
with mitigation about 23%) of their hit points. So that's one example which I think contradicts your assertion.
If we think of a
20th level rogue with 16 Con... that rogue has 163 hit points (8 + 5 * 19 + 3 * 20). At first blush, that
132 damage looks like an utterly devastating
80% of the rogue's HP, buuut the rogue has tricks. --Let's say this monster relies on a Multiattack (2x attacks for 66 dmg each) which is an arbitrary assumption because I need to make an assumption to evaluate the rogue's trickiness-- So what are the rogue's tricks? Uncanny Dodge halves damage from one attack, so that's -33 damage, and 133 - 32 = 99. I am guesstimating that it's also fair to say that the rogue's features like Cunning Action, Elusive, and Evasion
probably merit avoiding one attack throughout a hypothetical three-round combat... so that's 1/3 * 66 which is -22 damage, and 99 - 22 = 77 damage. The rogue has managed to reduce the damage to just under
50% (with mitigation) of their HP. Nothing to sneeze at, but our 20th level rogue could survive two rounds against this monster.
Right, so there's less mitigation/avoidance options for very low level PCs. Of course, there are still plenty of options (for example, I'll go on about how poorly designed the
Shield spell is), it's just that they are less prevalent at 1st-level both in the sense of PCs having fewer resources and more easily stretched thin & in the sense of some PCs not having those mitigation/avoidance options available yet.
You gave a high level example earlier, so I'm going to flip the script to look at a CR 1/2 monster (Hobgoblin) facing 1st level barbarian and then a 1st level rogue – this time it's Fafhrd and Gray Mouser, I guess. The Hobgoblin's
average damage is roughly 12, and its "damage divided by CR" is 24.
The
1st level barbarian has 15 hit points. So 12 damage would be a whopping
80% of their total - devastating. But they say "Hah you fool! I rage!" They mitigate it to 6 hit points, which is
40% (mitigated) of their hit points. Compare these % values to the 20th level barbarian (46% and 23% respectively). Because the higher level barbarian has that many more hit points and is mitigating more in relation to their hit points, this Hobgoblin is actually more dangerous to them than the whopping 132-damage dealing theoretical monster.
The
1st level rogue has it worse with 7 hit points. They don't have Uncanny Dodge, Elusive, or Evasion yet. The rogue takes that 12 damage – which is about
170% of their hit points – and they're on the sidewalk bleeding out.
No mitigation. Compared to the 20th level rogue (who without mitigation was taking 80% of their HP, or with mitigation just shy of 50%), this is night and day, and the 1st level rogue better stay hidden, stay at range with cover, get a lucky drop, get someone else to fight for them, etc.
To circle back to "what's the meaning of dividing CR by damage?", these examples comparing to PC hit points echo what I initially was noticing – that there's a big difference between "
21 damage for fractional CR monsters (CR 0, 1/8, 1/4, 1/2)" and monsters of CR 1+ (where it ranges from
8 to 6 damage).
So, for my process where I didn't want to dive into deeper maths/examples until I saw there was something there, seeing that really big jump from
21 damage to 8 damage was the "wow" moment – it was what got me to sit up and pay attention. And with these crude initial examples of a barbarian & rogue, hopefully I'm illustrating that yes, there is a meaningful difference in the threat presented that this big jump signals.
). As I pointed out, it is a line from CR 1 up to CR 19, not 20. At CR 20 you get the jump of 18 points, not 6. The increases from the previous CR to the next is 6 points from CR 1 to 19. From CR 20 to 30 it is 18, so, also a line in this segment, just with a steeper slope. The CRs below 1 have increases typically of 2, with the exception of CR 1/2, which is 3 points. Since these aren't discrete numbers for the CR, it gives us exaggerated slopes while working on a scale less than 1 unit for our indepentent variable (the CRs).
