I've seen a lot of people say that CR is useless, because of all the stuff that can't be accounted for. While D&D certainly does have a lot of factors that can't easily be put into a simple algorithm, I realized when I was looking at it that... a significant part of the problem is just that they did the math wrong.
The target for 5e combat appears to be that a fight against an even-CR foe will last 3 rounds and use up 1/6 of your daily resources, regardless of CR, and you probably have two such fights per short rest. If you look at actual scaling of monsters, it turns out that over the 1-20 range damage is about (CR+1)*5 per turn, hit points are about (CR+1)*15. This implies that, for example, they think an 8th level party has 3x the damage output of a second level party, and also has 3x the ability to absorb damage; a 14th level party is 5x at both.
Doing the math for damage output is hard, but doing it for damage capacity is much easier. Consider a party of 4: cleric(life), fighter(champion), rogue (thief), wizard (evoker). For simplicity, everyone starts with a 16 in their primary combat stat and a 14 in con, and I'm going to spend all cleric spell slots on healing.
At level 2, they have 68 combined hit points and 52 expected healing from hit dice, 22 from healing word, up to 30 from channel divinity (preserve life) and up to 21 from second wind (both are somewhat hard to use efficiently). Total 193
At level 8 (assumed ASIs: +4 to primary ability, fighter got Resilient(Wis), they have 236 combined hit points and 204 expected healing from hit dice, 38 from healing word, 204 from prayer of healing, 138 from mass healing word (3), 120 from mass healing word (4), up to 240 from channel divinity (preserve life), and up to 39 from second wind. Total 1,015.
At level 14 (assumed ASIs: cleric and wizard resilient(Con), rogue resilient (Con, Wis), fighter Resilient(Dex), Con+3, Wis+1) they have 432 combined hit points and 364 expected healing from hit dice, 38 from healing word, 204 from prayer of healing, 138 from mass healing word (3), 180 from mass healing word (4), 196 from mass cure wounds (5), 70 from heal (6), up to 642 from regenerate (unlikely to get the full amount, we'll credit it at (100)), up to 420 from channel divinity (preserve life), and up to 57 from second wind. Total 2199.
So the 8th level party is not 3x as durable as the second level party, it's more like 5x. The 14th level party is 11x. If we assume a CR 2 (15 dpr) is actually about right for the level 2 party, the equivalent encounter for an 8th level party is 79 dpr (CR 15), the equivalent for a 14th level party is 170 dpr (monster math changes above CR 20, so this is merely CR 24, not CR 33).
I tend to add a "handicap" character to my CR calculations, who is called Dwrmun the Dwarf. I make his level based on how powerful the party seems, with their magic combos and magic items and whatnot. For example, my current party has a good amount of magical loot right now, and are level 11, and they have Dwrmun trailing around with them at level 8, bumping up all the CR limits for the enemies so the fights are not too easy! And if they struggle too much, I will drop Dwrmun down to level 7. It's an easy way to incorporate it into existing CR based encounter calculators.
Now try it again without the healing. Combat healing is trash and should be avoided whenever possible.
If you assume no combat healing, the maximum damage that can be absorbed per combat is the total party hit points, multiplied by six encounters, or 408 at level 2, 1398 at level 8, and 2592 at level 14. As that is well over the healing total, leaving out combat healing has no effect. Also, you're wrong about combat healing being trash. Cure wounds is trash, but it's also not part of any of my figures.
Now try it again without the healing. Combat healing is trash and should be avoided whenever possible.
Tbh I'm slightly agree with Lunali. Combat healing in D&D can hinder your party's success. It takes away from potential damage and can become a game of catch-up. It can lead to complacency and other healing methods are often more effective. We should focus on dealing damage, avoid unnecessary risks, and use other healing methods for greater success when it comes to combat healing.
However, about the theory of "CR math breaks down at higher levels" I feel like you have a valid point here. As you presented a formula to estimate the expected damage output and hit points of a party based on their level. Then compare the calculated durability of a party against a given CR to the expected duration of a combat encounter. This can be a cool and useful tools for DMs when planning combat encounters. I understand the CR should be calculated correctly. But it's also important to understand that that CR is not the only factor that affects the difficulty of a fight.
Not sure how to phrase this but in my experience the party is more powerful than the sum of it's parts. E.g. if the designated tank goes down, the survivability of the rest of the party plummets (or the max DPS guy or whatnot, it's an example). The upshot of that is that the combat ability of a given party is not linear, I agree that it goes to 0 if all members are at 0 hp, but the level of combat ability likely has several plateaus and that drop from one to the next lower plateau, that transition, (when a party member goes down beyond potential combat healing) is steep if not vertical. In addition it probably also plays a role in which order the players go down.
All of that will change the equations you have set up and complicate them to no end.
Not saying that CR is great, the tool to end all tools, or that this calculation is bogus, I just think it's rather more complicated than the calculation implies
I didn't say combat healing was generally effective. In order to be effective, the actions gained (because characters don't go down, or downed characters get up) have to exceed the actions spent. Most often that means saving healing until someone goes down, though there are uncommon exceptions, generally involving very large amounts of healing.
However, that doesn't make a big difference in the calculations I was doing, because they were about damage recovered per day, and that could be from out of combat healing.
I've given up on CR completely with my 6 character, level 14 party. But having played up from level 1, I more or less know what they can handle on a full Long Rest in a single fight, or 3 fights, or 6. I homebrew pretty much every boss monster now. There's almost nothing outside ancient dragons in the official sources that can last more than a handful of turns.
My most recent 'boss' fight featured a mecha dragon with 1,300 hit points, 6 legendary actions, a five attack multiattack and a ton of special abilities. The fight lasted 5.5 hours and required a successful Divine Intervention to win. No fatalities but several K.O.s. We stick to RAW a lot but at heart I'm very much a Rule of Cool DM, so generic mathematics doesn't work all that well when the players are coming up with nuts ways to fight.
I've given up on CR completely with my 6 character, level 14 party. But having played up from level 1, I more or less know what they can handle on a full Long Rest in a single fight, or 3 fights, or 6. I homebrew pretty much every boss monster now. There's almost nothing outside ancient dragons in the official sources that can last more than a handful of turns.
To some degree that's working as intended; the 'average over 3 turns' method seems to think median combat lifespan for monsters should be 3 turns, so 'handful of turns' is about what they want. The problem is mostly that the math failed; monsters don't generally live three turns, and even if the do, they don't do enough damage.
Let's just assume you put a tarrasque against your party, in a situation where they actually have to engage it (no flying over it kiting). I expect median level 14 PCs to be something like attack +12/55 dpr, with advantage around half the time, so six of them should be putting out around 200 dpr against AC 25 and burn through 676 hp in three turns. The tarrasque's raw 232 dpr isn't going to be reduced a lot by AC unless there's some pretty unusual builds; call it -20%. Thus, in three turns the tarrasque does 696 base damage, reduced to 555 by armor. That's about 1/4 of the healing capacity (computed above) of a party of 4, or 1/6 of a party of six. If we assume a full day is probably expected to use around 2/3 of your resources, that Tarrasque should be 1/4 of your daily budget, or a Hard encounter.
My most recent 'boss' fight featured a mecha dragon with 1,300 hit points, 6 legendary actions, a five attack multiattack and a ton of special abilities. The fight lasted 5.5 hours and required a successful Divine Intervention to win. No fatalities. We stick to RAW a lot but at heart I'm very much a Rule of Cool DM, so generic mathematics doesn't work all that well when the players are coming up with nuts ways to fight.
Do your players actually like 5.5 hour fights? When I was younger and had more gaming time I remember a few fights that long, but even at the time it seemed excessive.
The ability to absorb damage also varies party to party. A Bear Totem Barbarian is all but sitting on 2x their on-page hp because Bear Totem. Any Druid with Conjure Animals or Conjure M.Elemental adds another hp pool on top of the party's pool just by standing around in tactically useful places. Ancients Paladin has their anti-magic damage Aura. Not to mention the easily combat-breakable AoE damage mitigation on the Twilight and Peace Clerics. And this isn't getting into how the PCs are positioned when combat starts, which can also make a big difference on how damage is distributed. IOW, I think there are more pieces in the puzzle even if you are just looking at average survivability based on level of party and # of PCs.
Actually, the repeat Tarrasque day sounds like an okay adventure concept. The Tarrasque shows up, the PCs kill it... and an hour later it resurrects (evolved, so it won't die the same way twice...). The PCs have to limit the damage it does while trying to figure out how to prevent it from resurrecting.
Cr is rubbish. Especially when you factor in random dice rolling. If your level 14 characters are killing a Tarrasque, it's probably being DM'd incorrectly. Especially when you take into account its legendary actions, resistances and so on.
I DMed for well over 20 years, before leaving the game for another 20 or so. The more I learn about this, and all these other newer rules-related headaches, the less I feel like taking on the Mantle of DM, ever again.
I DMed for well over 20 years, before leaving the game for another 20 or so. The more I learn about this, and all these other newer rules-related headaches, the less I feel like taking on the Mantle of DM, ever again.
None of these are newer rules-related headaches. Encounter balancing with CR in 5e is no worse than encounter balancing with CR in 3.x, or encounter balancing with no rules whatsoever in AD&D.
Agreed. Discard the CR and just use your experience to balance. Get on the horse again, your DND community need you.
That's easy to say if you have 20 years of experience as a DM. CR is still occasionally useful for newer DMs, which is why ways to improve the accuracy of rating encounters is generally a useful discussion to have even if the solutions are not easily arrived at.
Agreed. Discard the CR and just use your experience to balance. Get on the horse again, your DND community need you.
Rather than discard it, use it as intended, as a general guideline for how much damage a monster can take and deal. Doing exacting math with it or using monsters based on CR without reading them is going to run into massive problems. Using it to decide how old a dragon should be to be roughly the right challenge for the party on the other hand is rather reasonable.
Rather than discard it, use it as intended, as a general guideline for how much damage a monster can take and deal. Doing exacting math with it or using monsters based on CR without reading them is going to run into massive problems. Using it to decide how old a dragon should be to be roughly the right challenge for the party on the other hand is rather reasonable.
Well, it would be if the math worked right. As is, it's okay for gauging encounters against each other (encounter X was a reasonable challenge, so if I swap the monsters with different monsters of the same CR the new encounter will also be a reasonable challenge) but won't help much with designing the first encounter; your best bet there is "they handled X without any trouble, so let's make the next encounter harder".
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
I've seen a lot of people say that CR is useless, because of all the stuff that can't be accounted for. While D&D certainly does have a lot of factors that can't easily be put into a simple algorithm, I realized when I was looking at it that... a significant part of the problem is just that they did the math wrong.
The target for 5e combat appears to be that a fight against an even-CR foe will last 3 rounds and use up 1/6 of your daily resources, regardless of CR, and you probably have two such fights per short rest. If you look at actual scaling of monsters, it turns out that over the 1-20 range damage is about (CR+1)*5 per turn, hit points are about (CR+1)*15. This implies that, for example, they think an 8th level party has 3x the damage output of a second level party, and also has 3x the ability to absorb damage; a 14th level party is 5x at both.
Doing the math for damage output is hard, but doing it for damage capacity is much easier. Consider a party of 4: cleric(life), fighter(champion), rogue (thief), wizard (evoker). For simplicity, everyone starts with a 16 in their primary combat stat and a 14 in con, and I'm going to spend all cleric spell slots on healing.
At level 2, they have 68 combined hit points and 52 expected healing from hit dice, 22 from healing word, up to 30 from channel divinity (preserve life) and up to 21 from second wind (both are somewhat hard to use efficiently). Total 193
At level 8 (assumed ASIs: +4 to primary ability, fighter got Resilient(Wis), they have 236 combined hit points and 204 expected healing from hit dice, 38 from healing word, 204 from prayer of healing, 138 from mass healing word (3), 120 from mass healing word (4), up to 240 from channel divinity (preserve life), and up to 39 from second wind. Total 1,015.
At level 14 (assumed ASIs: cleric and wizard resilient(Con), rogue resilient (Con, Wis), fighter Resilient(Dex), Con+3, Wis+1) they have 432 combined hit points and 364 expected healing from hit dice, 38 from healing word, 204 from prayer of healing, 138 from mass healing word (3), 180 from mass healing word (4), 196 from mass cure wounds (5), 70 from heal (6), up to 642 from regenerate (unlikely to get the full amount, we'll credit it at (100)), up to 420 from channel divinity (preserve life), and up to 57 from second wind. Total 2199.
So the 8th level party is not 3x as durable as the second level party, it's more like 5x. The 14th level party is 11x. If we assume a CR 2 (15 dpr) is actually about right for the level 2 party, the equivalent encounter for an 8th level party is 79 dpr (CR 15), the equivalent for a 14th level party is 170 dpr (monster math changes above CR 20, so this is merely CR 24, not CR 33).
That's a good breakdown, thanks for that!
I tend to add a "handicap" character to my CR calculations, who is called Dwrmun the Dwarf. I make his level based on how powerful the party seems, with their magic combos and magic items and whatnot. For example, my current party has a good amount of magical loot right now, and are level 11, and they have Dwrmun trailing around with them at level 8, bumping up all the CR limits for the enemies so the fights are not too easy! And if they struggle too much, I will drop Dwrmun down to level 7. It's an easy way to incorporate it into existing CR based encounter calculators.
Make your Artificer work with any other class with 174 Multiclassing Feats for your Artificer Multiclass Character!
DM's Guild Releases on This Thread - latest release; the Harvest Sprite, a playable Jack-o-Lantern Race!
Or check them all out on DMs Guild!
DrivethruRPG Releases on This Thread - latest release: The College of Fisticuffs Bard!
I also dabble in art on here (my art thread)
Now try it again without the healing. Combat healing is trash and should be avoided whenever possible.
If you assume no combat healing, the maximum damage that can be absorbed per combat is the total party hit points, multiplied by six encounters, or 408 at level 2, 1398 at level 8, and 2592 at level 14. As that is well over the healing total, leaving out combat healing has no effect. Also, you're wrong about combat healing being trash. Cure wounds is trash, but it's also not part of any of my figures.
Tbh I'm slightly agree with Lunali. Combat healing in D&D can hinder your party's success. It takes away from potential damage and can become a game of catch-up. It can lead to complacency and other healing methods are often more effective. We should focus on dealing damage, avoid unnecessary risks, and use other healing methods for greater success when it comes to combat healing.
However, about the theory of "CR math breaks down at higher levels" I feel like you have a valid point here. As you presented a formula to estimate the expected damage output and hit points of a party based on their level. Then compare the calculated durability of a party against a given CR to the expected duration of a combat encounter. This can be a cool and useful tools for DMs when planning combat encounters. I understand the CR should be calculated correctly. But it's also important to understand that that CR is not the only factor that affects the difficulty of a fight.
Not sure how to phrase this but in my experience the party is more powerful than the sum of it's parts. E.g. if the designated tank goes down, the survivability of the rest of the party plummets (or the max DPS guy or whatnot, it's an example). The upshot of that is that the combat ability of a given party is not linear, I agree that it goes to 0 if all members are at 0 hp, but the level of combat ability likely has several plateaus and that drop from one to the next lower plateau, that transition, (when a party member goes down beyond potential combat healing) is steep if not vertical. In addition it probably also plays a role in which order the players go down.
All of that will change the equations you have set up and complicate them to no end.
Not saying that CR is great, the tool to end all tools, or that this calculation is bogus, I just think it's rather more complicated than the calculation implies
I didn't say combat healing was generally effective. In order to be effective, the actions gained (because characters don't go down, or downed characters get up) have to exceed the actions spent. Most often that means saving healing until someone goes down, though there are uncommon exceptions, generally involving very large amounts of healing.
However, that doesn't make a big difference in the calculations I was doing, because they were about damage recovered per day, and that could be from out of combat healing.
I've given up on CR completely with my 6 character, level 14 party. But having played up from level 1, I more or less know what they can handle on a full Long Rest in a single fight, or 3 fights, or 6. I homebrew pretty much every boss monster now. There's almost nothing outside ancient dragons in the official sources that can last more than a handful of turns.
My most recent 'boss' fight featured a mecha dragon with 1,300 hit points, 6 legendary actions, a five attack multiattack and a ton of special abilities. The fight lasted 5.5 hours and required a successful Divine Intervention to win. No fatalities but several K.O.s. We stick to RAW a lot but at heart I'm very much a Rule of Cool DM, so generic mathematics doesn't work all that well when the players are coming up with nuts ways to fight.
To some degree that's working as intended; the 'average over 3 turns' method seems to think median combat lifespan for monsters should be 3 turns, so 'handful of turns' is about what they want. The problem is mostly that the math failed; monsters don't generally live three turns, and even if the do, they don't do enough damage.
Let's just assume you put a tarrasque against your party, in a situation where they actually have to engage it (no flying over it kiting). I expect median level 14 PCs to be something like attack +12/55 dpr, with advantage around half the time, so six of them should be putting out around 200 dpr against AC 25 and burn through 676 hp in three turns. The tarrasque's raw 232 dpr isn't going to be reduced a lot by AC unless there's some pretty unusual builds; call it -20%. Thus, in three turns the tarrasque does 696 base damage, reduced to 555 by armor. That's about 1/4 of the healing capacity (computed above) of a party of 4, or 1/6 of a party of six. If we assume a full day is probably expected to use around 2/3 of your resources, that Tarrasque should be 1/4 of your daily budget, or a Hard encounter.
Bring on the four Tarrasque day, I guess?
Do your players actually like 5.5 hour fights? When I was younger and had more gaming time I remember a few fights that long, but even at the time it seemed excessive.
The ability to absorb damage also varies party to party. A Bear Totem Barbarian is all but sitting on 2x their on-page hp because Bear Totem. Any Druid with Conjure Animals or Conjure M.Elemental adds another hp pool on top of the party's pool just by standing around in tactically useful places. Ancients Paladin has their anti-magic damage Aura. Not to mention the easily combat-breakable AoE damage mitigation on the Twilight and Peace Clerics. And this isn't getting into how the PCs are positioned when combat starts, which can also make a big difference on how damage is distributed. IOW, I think there are more pieces in the puzzle even if you are just looking at average survivability based on level of party and # of PCs.
Actually, the repeat Tarrasque day sounds like an okay adventure concept. The Tarrasque shows up, the PCs kill it... and an hour later it resurrects (evolved, so it won't die the same way twice...). The PCs have to limit the damage it does while trying to figure out how to prevent it from resurrecting.
Cr is rubbish. Especially when you factor in random dice rolling. If your level 14 characters are killing a Tarrasque, it's probably being DM'd incorrectly. Especially when you take into account its legendary actions, resistances and so on.
I DMed for well over 20 years, before leaving the game for another 20 or so. The more I learn about this, and all these other newer rules-related headaches, the less I feel like taking on the Mantle of DM, ever again.
None of these are newer rules-related headaches. Encounter balancing with CR in 5e is no worse than encounter balancing with CR in 3.x, or encounter balancing with no rules whatsoever in AD&D.
I stand corrected.
Agreed. Discard the CR and just use your experience to balance. Get on the horse again, your DND community need you.
That's easy to say if you have 20 years of experience as a DM. CR is still occasionally useful for newer DMs, which is why ways to improve the accuracy of rating encounters is generally a useful discussion to have even if the solutions are not easily arrived at.
Rather than discard it, use it as intended, as a general guideline for how much damage a monster can take and deal. Doing exacting math with it or using monsters based on CR without reading them is going to run into massive problems. Using it to decide how old a dragon should be to be roughly the right challenge for the party on the other hand is rather reasonable.
Well, it would be if the math worked right. As is, it's okay for gauging encounters against each other (encounter X was a reasonable challenge, so if I swap the monsters with different monsters of the same CR the new encounter will also be a reasonable challenge) but won't help much with designing the first encounter; your best bet there is "they handled X without any trouble, so let's make the next encounter harder".