D&D 5E Official Monsters and 3rd party monsters - power level

[Quickleaf]

Invisibility is worth 2 AC. Blurred movement is worse.

Really? I would disagree. Let's compare Superior Invisibility (from faerie dragon, defensively worth +2 AC according to DMG, leaving aside the offensive bonus which is separate) with Blurred Movement (from quickling).

Superior Invisibility. As a bonus action, the dragon can magically turn invisible until its concentration ends (as if concentrating on a spell). Any equipment the dragon wears or carries is invisible with it.

• Being invisible means you can't be targeted by spells/effects which stipulate "a creature you can see..."
• Requires a bonus action to activate.
• Being incapacitated (i.e. paralyzed, stunned, or otherwise unable to take an action) prevents activation.
• Requires concentration, which can be disrupted by damage.
• Is countered with faerie fire, dispel magic, or see invisibility.

Blurred Movement. Attack rolls against the quickling have disadvantage unless the quickling is incapacitated or restrained.

• Requires no bonus action.
• Being incapacitated (i.e. paralyzed, stunned, or otherwise unable to take an action) or restrained counters it.
• Does not require concentration.
• Is not countered by faerie fire, dispel magic or see invisibility.

I could see an argument being made that they're defensively comparable, but not that Blurred Movement is worse. Superior Invisibility covers you from more than attacks, but it's a lot more tenuous / can be disrupted more readily, whereas Blurred Movement only covers you from attacks, but is more reliable.

18 is 5 over 23, so only 2 steps.

Defensive CR is 1/2.

Offensive is 5 as noted.

Average is 2. Published as a 1.

Sure. But you're rounding 2.75 down to get 2.

What does the DMG have to say about rounding CR values? Top of page 275:

Average Challenge Rating. The monster's final challenge rating is the average of its defensive and offensive challenge ratings. Round the average up or down to the nearest challenge rating to determine your monster's final challenge rating. For example, if the creature's defensive challenge rating is 2 and its offensive rating is 3, its final rating is 3.

So I stand by my original evaluation. The quickling's CR is 3.

(a) as noted the DMG rules are an approximation of their internal ones, (b) this is off by 1 CR, (c) I asked for 5 for a reason. I have no doubt there are a few monsters with CR that is off. The least bad of the 5 worst you can find produces a reasonable bound on how bad 5e calculation is.

I agree, there's value in doing the maths for a broader cross-section of monsters, and I wish I had the time to do deep analysis of more monsters. Maybe I can tackle one or two more while I'm under quarantine -- I'll try to think of some that I have actual experience running which seemed different from the CR the book was telling me. The quickling came up in a game I ran in early 2017, and I think someone even asked about it on ENWorld, so I crunched some numbers.

For the quickling, a party of +5 to hit L2 PCs needs 1.5 hits to drop it. At 16 AC and disadvantage, that is 0.25 hits per attack, so 6 attacks to drop it. A party of 4 drops it after 1.5 rounds using at-will stuff.

(If someone has a non-dex save cantrip, faster).

It gets 5 swings (high initiative) for 36 total damage at +8 to hit. That is nasty. And lots of taps for near instant kills on downed foes. An ogre lasting 3 rounds does 39 at +6, and is a legit CR 2.

Yes, they can be nasty. My quickling encounter involved them carrying a magic lantern with a will-o-wisp inside through moorlands with treacherous sections that caused Small or Medium creatures to sink if you stayed on them too long. PCs were after the magic lantern. Quicklings zipped around without a concern for terrain and will-o-wisp magnified "attacking them while they're down" threat.

 

[Quickleaf]

@NotAYakk Aha! I remembered another one from earlier in my 5e experience -- the manticore, which is CR 3 in the MM.

It acted as a sort of "sin eater" figure for a cult, with its usual lurking spot being on a cliffside ledge where it looked down at repentant figures bearing tribute for its "wisdom" (which usually consisted of dishing out blackmail material on the supplicant's rivals who previously visited the manticore). The 3rd level PCs basically said "screw that noise" mid-conversation and attacked. I thought I played the manticore's tactics well enough, and it was accompanied by many cultists, but the manticore went down really hard in two rounds to focus fire. It didn't feel really threatening. IIRC the players had one or two encounters previously exploring the canyon environment, so they weren't fresh, but they weren't badly tapped either.

Looking at the maths now, I can see why...

Manticore, Defensive CR = 1/2
68 HP starts it at CR 1/2
AC 14 is about consistent with target AC 13 for CR 1/2, so no change

Manticore, Offensive CR = 3
DPR 22.5 (assuming tail spikes) starts it at CR 3
Attack bonus +5 is about consistent with target +6 for CR 3, so no change

Manticore, Total CR = (1/2 + 3) / 2 = 1.75
So its actual rounded-up CR is 2.

While I understand labeling it as CR 3 because flight is supposed to make it a suitable challenge for when spells like levitation and spider climb become available to 3rd level PCs, what I noticed was the big disparity between its poor defenses and decent offenses. With all the subclass features coming online at 3rd level, there's a spike in PC damage that the really weak defenses of the manticore couldn't keep pace with.

The CR being off by 1 probably wouldn't have been so noticeable in my play experience if it had a better Defensive CR.

 

[NotAYakk]

I wonder if they just didn't include multiattack for the quickling. Or an earlier draft dropped dex to damage on its additional attacks.

The manticore, flying is supposed to be big, and the PCs ambushed it on the ground.

 

[Quickleaf]

@NotAYakk I just stumbled upon another monster whose CR is not reflected by the DMG maths – the armanite (CR 7, MToF). This is one that I've never run in 5e so I can't speak on the personal experience front. But here's the breakdown of its maths according to the DMG...

Defensive CR = 3
HP = 84
eff AC = 16 + 2 magic resist = 18

Offensive CR = 7
DPR = [(27 * 2) Lightning Lance vs. two targets + (12+10+16) Multiattack + (12+10+16) Multiattack] / 3 = 43
attack +8 / spell save DC 15

Total CR = (3 + 7) / 2 = 5

Maths are telling me its actual CR is 5, versus a listed CR of 7.

This is from their most recent monster book! I was surprised, so I wondered if maybe assuming two targets for Lightning Lance is incorrect.

If a monster's damage output varies from round to round, calculate its damage output each round for the first three rounds of combat, and take the average. For example, a young white dragon has a Multiattack routine (one bite attack and two claw attacks) that deals an average of 37 damage each round, as well as a Breath Weapon that deals 45 damage, or 90 if it hits two targets (and it probably will). In the first three rounds of combat, the dragon will probably get to use its breath weapon once and its multiattack routine twice, so its average damage output for the first three rounds would be (90 + 37 + 37) + 3, or 54 damage (rounded down).

The designers assume that a 30-foot cone (~21 squares on a grid) hits 2 targets.

Lightning Lance's 60' x 10' line (24 squares on grid) is very comparable. On the off chance they assumed 3 targets – again, nothing in the DMG would suggest we should do this – but what if we assume 3 targets?

Defensive CR = 3
HP = 84
eff AC = 16 + 2 magic resist = 18

Offensive CR = 8
DPR = [(27 * 3) Lightning Lance vs. three targets + (12+10+16) Multiattack + (12+10+16) Multiattack] / 3 = 52
attack +8 / spell save DC 15

Total CR = (3 + 8) / 2 = 5.5

Still doesn't check out.

 

[NotAYakk]

It has resistance and immunity to 4 damage types. That is worth some HP, probably not enough.

Base attack is 38. Blast is 54 or 81. With 60' range you are likely to fire it off round one (unlike dragon, who should frighten). 1/3 round 2, 1/3 round 3, 1.66 lightning 1.34 melee.

Call it 2.5 targets with bolt for 68.

At (68*5+38*4)/9=(340+152)/9=(492)/9~=54 DPR, which boosts its offensive CR to 8. DC/ATK within 1 of target. Add 50% HP for significant resists and defensive CR hits 5+1=6.

Average is 7.

 

[Quickleaf]

Was working on converting a monster that reduces maximum HP, like an undead's Life Drain, and was wondering about the DMG saying that Life Drain has no impact on determining a monster's CR. So I ran quick numbers on the wight (CR 3) to double check, and found another discrepancy between what the DMG says and what the MM says.

Wight: Defensive CR 2
effective HP = 45 * 2 damage resistances for target CR 1-4 = 90
AC 14

Offensive CR 1
DPR = 5.5 life drain + 7.5 longsword = 13
+4 attack

Total CR = (2 + 1) / 2 = 1.5, rounding up to 2 (which is reasonable due to Life Drain)

Being off by a little isn't concerning, but what is a little concerning is how this effect can domino when you start to create stronger monsters which life drain for more HP. Sure, the PCs can cast greater restoration, but the value of a PC's action increases as they level, so that foregone action in combat is costing them more at 11th level than it did at 1st level. It sure looks to me like the designers are accounting for certain traits (like Life Drain) which the DMG says you don't need to account for.