I think that the deeper concern throughout D&D's balance is that for the most part CR isn't a good estimate for the power level of a monster. Before being able to tackle encounters as a whole, I think that we need a better basis for what CR means, and how deadly a given monster of that CR is.
The waters are further muddied by the designers themselves having said that the monster manual was made keeping in mind that the PCs have no magic items, which... usually isn't the case in many games I've played with. All of these together make any sort of encounter balancing nigh impossible if we want to grab monsters from the monster manual and run a given encounter.
First off, it's true that the MM was made without consideration for magic items. Magic items actually increase the PCs' "effective level" in a way that I plan to discuss in a future article in this series, which means they're actually punching above their weight in combat.
Now, usually, players don't have enough magic items to play at a full level above their own—usually only half a level or so—but if you push too far, you can absolutely "level up" your PCs through magic items alone.
CR itself is generally fine, in my experience. It's true that the encounter difficulty names are largely misleading or unhelpful, but the CR itself is generally pretty consistent. The sole exception comes in when you're dealing with glass cannons, such as spellcasters or dragons, because it becomes much easier for either side (players and monsters) to kick off a death spiral, as described in this article.
Two of the creatures that most get criticized under the CR system—shadows and intellect devourers—effectively break the mold not because they're particularly efficient at killing the PCs, but because they operate on a different resource system entirely, making it a lot easier for them to start a death spiral. I personally think that those types of creatures should have a "special" CR, rather than a normal one. Most monsters, though, don't fall into that category. For them, CR usually works pretty well.
I think even under that constraint, there are huge differences in the ability of monsters.
Off the top of my head, let's compare a quaggoth to a minotaur skeleton; the quaggoth has AC 13, 45 hit points, and two attacks at +5 that do 1d6+3 damage, with a special ability that comes online if it happens to have 10 or less hit points on its turn, which, if the party has a decent rogue or if the fighter took great weapon master, is rather unlikely to happen.
On the other hand, the skeleton has just one less point of AC but almost 50% more hit points at 67, along with one attack that does more than double what the quaggoth does on both hits (2d12+4) along with a special attack that's sure to go off at least once per combat that does 4d8+4 damage.
It's clear to see that the minotaur skeleton is much stronger than the quaggoth by almost all metrics, and that's just two monsters I randomly found that were CR 2 in the monster manual. Neither has super special secret attacks like the banshee, ghost, or intellect devourer, and both are meant to be "brutes" that don't do much apart from attack, and yet the difference between the two is staggering.
I'm certain I could find more examples, but I think this is enough to back up the assertion that the CR system as used by published monsters doesn't work very well, even ignoring that they were made without taking magic items into account.
The reason why the Minotaur Skeleton is CR 2 is twofold—because (1) the CR table has decided that an AC of 12 is sufficiently low as to create a negative defensive CR, and (2) because it has vulnerability to bludgeoning damage. I agree that the CR chart is somewhat oversensitive here, but the minotaur skeleton isn't as far above the quaggoth as the veteran statblock is, which is clearly and comfortably CR 3.
Would I put the minotaur skeleton closer to CR 2.5? Sure. But we don't have a CR 2.5; we have a CR 2.
I think that the deeper concern throughout D&D's balance is that for the most part CR isn't a good estimate for the power level of a monster. Before being able to tackle encounters as a whole, I think that we need a better basis for what CR means, and how deadly a given monster of that CR is.
The waters are further muddied by the designers themselves having said that the monster manual was made keeping in mind that the PCs have no magic items, which... usually isn't the case in many games I've played with. All of these together make any sort of encounter balancing nigh impossible if we want to grab monsters from the monster manual and run a given encounter.
I both agree and disagree.
First off, it's true that the MM was made without consideration for magic items. Magic items actually increase the PCs' "effective level" in a way that I plan to discuss in a future article in this series, which means they're actually punching above their weight in combat.
Now, usually, players don't have enough magic items to play at a full level above their own—usually only half a level or so—but if you push too far, you can absolutely "level up" your PCs through magic items alone.
CR itself is generally fine, in my experience. It's true that the encounter difficulty names are largely misleading or unhelpful, but the CR itself is generally pretty consistent. The sole exception comes in when you're dealing with glass cannons, such as spellcasters or dragons, because it becomes much easier for either side (players and monsters) to kick off a death spiral, as described in this article.
Two of the creatures that most get criticized under the CR system—shadows and intellect devourers—effectively break the mold not because they're particularly efficient at killing the PCs, but because they operate on a different resource system entirely, making it a lot easier for them to start a death spiral. I personally think that those types of creatures should have a "special" CR, rather than a normal one. Most monsters, though, don't fall into that category. For them, CR usually works pretty well.
I think even under that constraint, there are huge differences in the ability of monsters.
Off the top of my head, let's compare a quaggoth to a minotaur skeleton; the quaggoth has AC 13, 45 hit points, and two attacks at +5 that do 1d6+3 damage, with a special ability that comes online if it happens to have 10 or less hit points on its turn, which, if the party has a decent rogue or if the fighter took great weapon master, is rather unlikely to happen.
On the other hand, the skeleton has just one less point of AC but almost 50% more hit points at 67, along with one attack that does more than double what the quaggoth does on both hits (2d12+4) along with a special attack that's sure to go off at least once per combat that does 4d8+4 damage.
It's clear to see that the minotaur skeleton is much stronger than the quaggoth by almost all metrics, and that's just two monsters I randomly found that were CR 2 in the monster manual. Neither has super special secret attacks like the banshee, ghost, or intellect devourer, and both are meant to be "brutes" that don't do much apart from attack, and yet the difference between the two is staggering.
I'm certain I could find more examples, but I think this is enough to back up the assertion that the CR system as used by published monsters doesn't work very well, even ignoring that they were made without taking magic items into account.
The reason why the Minotaur Skeleton is CR 2 is twofold—because (1) the CR table has decided that an AC of 12 is sufficiently low as to create a negative defensive CR, and (2) because it has vulnerability to bludgeoning damage. I agree that the CR chart is somewhat oversensitive here, but the minotaur skeleton isn't as far above the quaggoth as the veteran statblock is, which is clearly and comfortably CR 3.
Would I put the minotaur skeleton closer to CR 2.5? Sure. But we don't have a CR 2.5; we have a CR 2.